Salmonella Degrades the Host Glycocalyx Leading to Altered Infection and Glycan Remodeling

Abstract

Complex glycans cover the gut epithelial surface to protect the cell from the environment. Invasive pathogens must breach the glycan layer before initiating infection. While glycan degradation is crucial for infection, this process is inadequately understood. Salmonella contains 47 glycosyl hydrolases (GHs) that may degrade the glycan. We hypothesized that keystone genes from the entire GH complement of Salmonella are required to degrade glycans to change infection. This study determined that GHs recognize the terminal monosaccharides (N-acetylneuraminic acid (Neu5Ac), galactose, mannose, and fucose) and significantly (p < 0.05) alter infection. During infection, Salmonella used its two GHs sialidase nanH and amylase malS for internalization by targeting different glycan structures. The host glycans were altered during Salmonella association via the induction of N-glycan biosynthesis pathways leading to modification of host glycans by increasing fucosylation and mannose content, while decreasing sialylation. Gene expression analysis indicated that the host cell responded by regulating more than 50 genes resulting in remodeled glycans in response to Salmonella treatment. This study established the glycan structures on colonic epithelial cells, determined that Salmonella required two keystone GHs for internalization, and left remodeled host glycans as a result of infection. These data indicate that microbial GHs are undiscovered virulence factors.

Full text

1
Scientific RepoRts | 6:29525 | DOI: 10.1038/srep29525
www.nature.com/scientificreports
Salmonella Degrades the Host
Glycocalyx Leading to Altered
Infection and Glycan Remodeling
Narine Arabyan1, Dayoung Park2, Soraya Foutouhi1, Allison M. Weis1, Bihua C. Huang1,
Cynthia C. Williams2, Prerak Desai1,†, Jigna Shah1,‡, Richard Jeannotte1,3,§, Nguyet Kong1,
Carlito B. Lebrilla2,4 & Bart C. Weimer1
Complex glycans cover the gut epithelial surface to protect the cell from the environment. Invasive
pathogens must breach the glycan layer before initiating infection. While glycan degradation is crucial
for infection, this process is inadequately understood. Salmonella contains 47 glycosyl hydrolases (GHs)
that may degrade the glycan. We hypothesized that keystone genes from the entire GH complement
of Salmonella are required to degrade glycans to change infection. This study determined that GHs
recognize the terminal monosaccharides (N-acetylneuraminic acid (Neu5Ac), galactose, mannose,
and fucose) and signicantly (p < 0.05) alter infection. During infection, Salmonella used its two GHs
sialidase nanH and amylase malS for internalization by targeting dierent glycan structures. The host
glycans were altered during Salmonella association via the induction of N-glycan biosynthesis pathways
leading to modication of host glycans by increasing fucosylation and mannose content, while
decreasing sialylation. Gene expression analysis indicated that the host cell responded by regulating
more than 50 genes resulting in remodeled glycans in response to Salmonella treatment. This study
established the glycan structures on colonic epithelial cells, determined that Salmonella required two
keystone GHs for internalization, and left remodeled host glycans as a result of infection. These data
indicate that microbial GHs are undiscovered virulence factors.
Epithelial cells in the human gastrointestinal tract (GIT) are covered with at least two glycan layers, composed of
multiple layers of glycoproteins (mucin) and complex oligosaccharides (glycocalyx) that protect cells from the
local environment and infection1. Mucin is the most distal layer of glycoproteins in the GIT lumen and is directly
exposed to the luminal microbiome. e glycocalyx layer is adjacent the epithelial membrane that is composed of
trans-membrane glycoproteins and glycolipids and are components of membrane lipid ras that extend from the
membrane, which are specic bacterial and viral receptors used for microbial invasion resulting in transduction
of extracellular signals into the cell2–5. Glycans represent the rst and crucial interface of the cell surface through
which microbes interact with the host immune system mediating recognition and communication processes;
thus, controlling immunological recognition, cell-cell adhesion, and pathogen binding. Glycans are mixture of
structures that include short chains of saccharides, with simple structures to highly branched complex oligosac-
charides that are complicated with an array of specic linkages between the monosaccharide residues leading to
a large diversity in arrangements during their synthesis to form higher order chemical structures1. In the case of
bacterial pathogens, this layer provides a barrier to physically exclude microorganisms from gaining access to
the epithelial membrane and the associated receptors used for infection and have been suggested to be used for
co-evolution with commensal bacteria2.
Bacterial interaction with the epithelial surface via the glycan has been recognized for many years with the
use of lectins to identify microbial activity6. Beyond system level interactions specic interactions with via fucose
1Department of Population Health and Reproduction, School of Veterinary Medicine, University of California,
Davis, CA 95616, USA. 2Department of Chemistry, University of California, Davis, CA 95616, USA. 3Universidad de
Tarapacá, Avenida General Velásquez N°1775, Arica, Chile. 4Department of Biochemistry and Molecular Medicine,
School of Medicine, University of California, Davis, CA 95616, USA. †Present address: Zoetis, 333 Portage Street,
Kalamazoo, MI 49007, USA. ‡Present address: MPI Research, 54943 North Main Street, Mattawan, MI 49071, USA.
§Present address: Department of Plant Sciences, College of Agricultural and Environmental Sciences, University of
California, Davis, CA 95616, USA. Correspondence and requests for materials should be addressed to B.C.W. (email:
bcweimer@ucdavis.edu)
Received: 09 February 2016
Accepted: 17 June 2016
Published: 08 July 2016
OPEN

www.nature.com/scientificreports/
2
Scientific RepoRts | 6:29525 | DOI: 10.1038/srep29525
and sialic acid have been implicated to regulate commensal interactions and provide sugar sources in complex
communities where some members may cleave the sugar for use by other community members6–10. e specic
linkages (i.e. α -1,2/3/4 and α -2,3/6 linkages), in addition to these specic sugars are also implicated in controlling
the microbial interaction11. e breath of the microbes that interact with mucin and the underlying glycocalyx
interactions is large and has been suggested to be a primary underpinning of co-evolution of an individual and
their microbiome11,12. In fact, recognition of the glycan as the primary interacting surface has also led to detailing
the very large and complex carbohydrate digestion enzymes in bacteria that are assembled in CAZymes13. Use of
the host glycan to provide nutrients that regulate bacterial infection and virulence is increasingly recognized as
an important characteristic for individual microbes to penetrate the mucin layers and subsequently gain access to
the cell membrane for association and invasion to progress into the disease state14,15.
Salmonella, and other invasive enteric pathogens, developed mechanisms to breach the protective glycan
layers of host cells, which permits access to the host membrane receptors, that results in invasive intracellular
infection. e diversity and complexity of the entire epithelial glycan barrier mandates that a breadth of enzy-
matic activities is required to degrade these diverse glycans to gain access of the cell membrane. Degradation of
these compounds relies in part on glycosyl hydrolases (GHs). GHs are diverse, widely distributed in bacteria,
yet poorly characterized enzymes, which hinder determining their specic role during infection dynamics16,17.
To date there are no studies that dene the specic role of the diverse GHs found in Salmonella during infec-
tion16,17. In this study, we hypothesized that only a portion of the 21 dierent GH families that encompass 49
total carbohydrate-active enzymes in Salmonella may degrade glycans to gain access of the host membrane and
the microbe receptors used for invasion. Eilam et al. demonstrated that microbe’s glycan degradation potential is
associated with gut pathogens18. It was hypothesized in this study that unique and specic GHs in Salmonella are
required for invasive infection.
Here, we conducted a detailed study of how Salmonella degrades the glycocalyx layer of human colonic epi-
thelial cells (Caco-2) during in vitro infection and led to host glycan remodeling. We established that two GHs,
ΔnanH and ΔmalS, decrease invasion to non-invasive levels comparable to ΔinvA, suggesting that these GHs
may be new virulence factors. e N-glycome of Caco-2 cells during infection was proled and showed how
Salmonella used its glycan-degrading enzymes, nanH and malS, to degrade the glycocalyx layer. More surpris-
ingly, we identied that the host cell responded to microbial glycan degradation by modifying its own glycans
leading to decreased sialylation; however, increased fucosylation, higher-mannose, and more hybrid glycans.
is alteration in host glycans was due to induction of N-glycan biosynthesis pathways during Salmonella associ-
ation. is study demonstrates that a complex molecular interplay between epithelial cells and pathogens result
in altered association, but infection proceeds if bacteria have the appropriate GH compliment to overcome the
dynamic changes in the glycan.
Results
Glycan Degrading Enzymes Alter Host Membrane Access during Invasion of Salmonella. To
determine if glycans are involved during infection, we rst investigated if the removal of Neu5Ac (the outermost
monosaccharide on most N-glycans of animals) from Caco-2 cells aected Salmonella attachment. Caco-2 cells
were exposed to sialidase digestion to remove terminal Neu5Ac. As a control, Caco-2 cells were also subjected to
methyl-β -cyclodextrin (Mβ CD) treatment since Mβ CD inhibits caveolae-dependent endocytosis via disruption
of lipid ras19. Depletion of cell surface Neu5Ac led to a signicant (p < 0.05) reduction in Salmonella association
(Fig.1A), further implicating the importance of Neu5Ac in the adherence process. To determine if genes related
to metabolism of sugars released from glycan are involved during infection, nanT (STM3338; sialic acid trans-
porter), xylR (STM3662; xylose operon regulatory protein), and bax (STM3663; hypothetical protein similar to
ATP-binding protein) were deleted using homologous recombination20, characterized in their capacity to infect,
and for alteration in invasion in vitro with Caco-2 cells (Fig.1B). Deletion of these genes did not signicantly
change association compared to Salmonella enterica subsp. enterica serovar Typhimurium strain LT2, suggesting
that genes related to metabolism of sugars are not involved in additional invasion.
Deletion of each gene was veried using whole genome sequencing of the wild type and each deletion mutant
(Supplementary Figure S1B and C). Reference based assembly and de novo assembly showed that the genomes
were isogenic, that only the gene of interest was specically replaced with the chloramphenicol resistance gene,
and that the genome-to-genome distance was not signicantly dierent (p > 0.001). Taken together, these data
indicate that the gene deletions were adequate to delineate the multi-gene eect demonstrated in this work to
digest complex glycans to gain access to the host membrane.
To further test the hypothesis if glycan-degrading GHs alter host membrane access during infection, we ana-
lyzed the gene expression of the annotated GHs of Salmonella Typhimurium LT2 during in vitro infection of
Caco-2 cells (Supplementary Table S1). Knowing the most common terminal residues on glycans (i.e. Neu5Ac,
galactose, and mannose), GHs that can cleave the most common residues and linkages, along with gene expres-
sion analysis, we identied that sialidases (STM0928 and STM1252), galactosidase (STM4298), and amylases
(STM3664 and STM3537) likely play an important role during infection. To determine if these GHs are importand
for infection, these GHs in Salmonella Typhimurium LT2 were also deleted (Supplementary Figure S1A)
and used to characterize their ability to infect and alter invasion in vitro using dierentiated (i.e. polarized)
colonic epithelial cells (Caco-2 cells) (Fig.1B). ese enzymes targeted dierent glycans based on their termi-
nal sugar specicity, as seen with the dierences in the infectivity of each bacterial mutant (Fig.1B), further
implicating terminal sugar digestion could be important during infection so that the T3SS can gain access to
the membrane to initiate injection of eector molecules. α -Galactosidase (STM4298; melA) recognizes terminal
α -galactose molecules of glycans. Deletion of melA led to a similar invasion phenotype as the WT.
Salmonella LT2 contains two sialidases in its genome: nanH (STM0928) and CHPNeu (STM1252; Conserved
Hydrolase Putative Neu raminidase), both of which were deleted from the genome in this study. Genetic

www.nature.com/scientificreports/
3
Scientific RepoRts | 6:29525 | DOI: 10.1038/srep29525
comparison of these genes found that nanH has homology with genes from bacteria and parasites (Supplementary
Figure S2A). Gene homology for CHPNeu was broadly distributed among bacteria from the gut and environment
(Supplementary Figure S2B and S2C). In Salmonella there is very little DNA sequence homology between nanH
and CHPNeu; however, the domain structures were conserved, which allowed identication of function between
organisms to provide molecular markers to dene two dierent neuraminidases in Salmonella that are enzymat-
ically, and likely play dierent roles dierent during infection (Supplementary Figure S3). Nitrogenases among
microbial genomes also display wide sequence diversity, but retain conservation of enzymatically important
domains21. Deletion of nanH signicantly decreased invasion (p = 0.0059). e ΔCHPNeu mutant signicantly
increased invasion (p = 0.01).
e amylase genes (malS and glgX) are also divergent in their sequence, but both contain the required domains
for amylase activity (Supplementary Figure S4). Deletion of malS (STM3664; α -amylase) signicantly decreased
Salmonella invasion (p < 0.0001), resulting in the same level of invasion as ΔinvA (STM2896; needle complex
export protein for T3SS), which is considered non-pathogenic and decient for adhesion and invasion in vivo22.
Surprisingly, deletion of glgX (STM3537; glycogen debranching enzyme), which is an amylase-like enzyme with
broader hexose specicity, led to a signicant increase in invasion (p = 0.0002). Perhaps the deletion of CHPNeu
and glgX aected other genes involved in virulence, such as the nanH and malS. e invasion levels of nanH and
malS mutants were not signicantly dierent (p > 0.05) compared to those found for ΔinvA; hence, nanH and
malS may represent new virulence genes in Salmonella infections. ese observations led to the hypothesis that
Salmonella targets specic glycan structures during infection that enable the T3SS needle complex access to the
membrane since it is not long enough to penetrate the glycan structures on epithelial cells.
Salmonella Infection Led to Alteration in Host N-Glycome. We initiated the glycan denition by
determining the glycan structure and composition of uninfected Caco-2 cells to dene the N-glycome baseline
to allow comparison of the changes made by Salmonella during infection. is approach is of particular interest
because it extends the specicity beyond binding via lectin-like activities and can be directly linked to enzyme
Figure 1. Modulation of carbohydrate-degrading enzymes alter access during invasion of Salmonella W T.
(A) Neu5Ac digestion. Methyl-β -cyclodextrin (control) disrupts lipid ras. Sialidase treatment led to reduction
in Salmonella association. Least Squares Means Dierences (LSD) was used for statistical analysis. Levels
not connected with the same letter are signicantly dierent, p < 0.05. (B) Salmonella WT knockout strains
characterized for the alteration in adhesion and invasion (A/I) using dierentiated Caco-2 cells 60 minutes post-
infection. White bars represent the CFU of Salmonella WT that adhered per Caco-2 cell. e gray bars represent
the CFU of Salmonella WT that invaded per Caco-2 cell. is was done in combination with transcriptional
proling (bottom panel) of Salmonella WT during infection of Caco-2 cells to gain insights about dierentially
expressed carbohydrate-degrading genes. Salmonella WT genes displaying changes in gene expression levels
during infection of Caco-2 cells. Colors indicate the expression of each gene induced (red) and repressed
(green). LSD was used for statistical analysis. Error bars indicate SEM between 3 biological replications,
* p < 0.05, * * p < 0.001, * * * p < 0.0001, Not Signicant (NS). e statistics shown at the top indicates the
statistical relevance in invasion levels of ΔinvA compared to the mutant strains.

www.nature.com/scientificreports/
4
Scientific RepoRts | 6:29525 | DOI: 10.1038/srep29525
activity and genes in Salmonella that are required to infect. e changes in host N-linked glycan prole are of par-
ticular interest as they comprise approximately 90% of the glycans present in eukaryotic cells and modication in
these types of glycans has been linked to multiple diseases23, but no specic structure has been directly linked to
infection even though specic sugars (i.e. sialic acid and fucose) are known to impact the microbiome and associ-
ation with the host. Uninfected Caco-2 cells contained 191 distinct glycan structures (Supplementary Figure S5A,
Supplementary Table S2). e Caco-2 N-glycans were rich in mannose, galactose, N-acetylglucosamine, fucose,
and N-acetylneuraminic acid (Neu5Ac). e relative abundance of specic structures changed signicantly
(p < 0.05) during Salmonella infection (Supplementary Figure S5B, Supplementary Table S2) when compared
to uninfected cells (Supplementary Figure S5A, Supplementary Table S2). Within 60 minutes of Salmonella
infection, the host glycome contained substantially dierent structures compared to uninfected Caco-2 cells
(Supplementary Figure S5B) that led to 185 total glycan structures. During infection 13 glycans completely
disappeared (Fig.2D, Supplementary Table S2) and seven glycans were appeared on the host surface (Fig.2F,
Supplementary Table S2). ere were also some glycans with intermediate abundance changes with decreases
(Fig.2A–C, Supplementary Table S2) while others increased (Fig.2E, Supplementary Table S2), yet other glycans
were similar to uninfected Caco-2 cells (Supplementary Table S2). Interestingly, the glycans resulting from infec-
tion were composed of high-mannose-containing structures and complex glycans that were polyfucosylated with
the terminal branches containing up to ve fucose molecules, while the majority of the glycans that disappeared
were polysialylated.
Host Cells Modify Their Own Glycan during Microbial Glycan Degradation. Using gene expres-
sion and canonical pathway analysis we examined the host response during infection and glycan degradation.
Signicantly dierentially (q < 0.05) regulated genes following Salmonella infection were marked by the induc-
tion of genes directly involved in glycan biosynthesis (Fig.3A, Supplementary Table S3). All mannosidases and
mannosyltransferases were all signicantly induced (q < 0.05) (Fig.3B) that was concomitantly observed in
the accumulation of mannose in the glycan structures using mass spectrometry described above. e fucosi-
dase, FUCA2, and all seven fucosyltransferases were signicantly induced (q < 0.05) (Fig.3C) that, again was
Figure 2. Host glycome is substantially altered during infection with Salmonella WT within 60 minutes.
e bars represent the relative abundance levels of each signicant glycan during infection with Salmonella.
(A) Decrease in complex-fucosylated glycans; (B) Decrease in high-mannose and hybrid glycans; (C) Decrease
in sialylated glycans; (D) Disappearance of sialylated glycans; (E) Accumulation of high-mannose and complex
glycans; (F) New complex-fucosylated and sialylated glycans. Error bars indicate SEM between 3 biological
replications, * p < 0.05, * * p < 0.001, * * * p < 0.0001.

www.nature.com/scientificreports/
5
Scientific RepoRts | 6:29525 | DOI: 10.1038/srep29525
concomitantly observed in the increase in fucose at the terminal residue in the cellular glycan. e Salmonella
genome does not contain fucosidases, so it is reasonable that these glycans accumulate via host production and
modication.
e overall Neu5Ac content of the glycan was a result of removal by Salmonella. We observed induction of
NEU1, NEU2, NEU3, and NEU4 in the host led us to presume that these genes were collectively responsible for
additional modication of the glycan. e four exo-α -sialidases in the host, each of which has a dierent substrate
specicity, were regulated independently with two induced and two unregulated, had median expression levels.
NEU2 and NEU4 were not regulated during infection (Fig.3D). NEU1, which is commonly located at the lyso-
some and NEU3, found on cell surfaces, were induced during infection (Fig.3D). e two sialidases in Salmonella
were induced during infection, suggesting that Neu5Ac would be released. e glycans were depleted in Neu5Ac
during infection, which may indicate the cooperative activity of the host and the pathogen resulted in a reduction
of the degree of sialylation from three to one (Fig.4).
Glycan Degrading Enzymes Modulated Glycans during Infection. To establish the denitive role
of the infection-associated glycan-degrading enzymes, we tested the glycan degradation capacity of Salmonella
enzymes determined during in vitro infection of Caco-2 cells with WT Salmonella, ΔmelA, ΔCHPNeu,
ΔnanH, ΔmalS, and ΔglgX. It was shown that GHs impacting invasive infection would be the same or less
than ΔinvA (Fig.1B). Bacterial degradation of the Caco-2 glycans is required for invasion. Glycan proles of
Figure 3. Host responds to microbial glycan degradation by modifying its own glycan. (A) Analysis of
host pathways involved during glycan degradation of Caco-2 cells following microbial association. Canonical
pathways whose biological functions were inuenced based on gene expression changes are shown (Fishers
exact test). Upregulated molecules in each pathway are represented as a percentage of the total canonical
pathway membership. (B–E) Networks display interactions between genes involved in mannose, fucose
and Neu5Ac ([D] sialidases and [E] sialyltransferases) metabolism, respectively, in Caco-2 cells treated for
60 minutes with Salmonella LT2. ST3GAL family sialyltransferases catalyzed the addition of Neu5Ac to a
terminal galactose of glycoconjugates in an α -2,3 linkage. e sialyltransferases in ST6GAL family transferred
alpha-2,6 linking Neu5Ac to galactose residues of N-glycans. e ST6GALNAc family sialyltransferases added
Neu5Ac to terminal N-acetylgalactosamine residues of glycoproteins and glycolipids, in an α -2,6 linkage. Lastly,
the ST8Sia family catalyzes the transfer of Neu5Ac in an alpha-2,8 linkage to other Neu5Ac residue present
in N- or O-glycans of Neural cells. Caco-2 up-regulation of enzymes in both gene networks is indicative of
microbial induced changes in host glycan biosynthesis. Gene induction is represented as a log ratio (q < 0.05)
and displayed in shades of red.

www.nature.com/scientificreports/
6
Scientific RepoRts | 6:29525 | DOI: 10.1038/srep29525
the host cell aer infection showed similar composition, all of which contained increased mannose content.
During infection, the degrees of fucosylation and sialylation also shied to have higher fucosylation and less
sialylation (Fig.4A). e host cell glycan composition was not expected to be altered during infection with
ΔinvA since it contained all the glycan-degrading enzymes found in the WT. However, 11 dierent structures
were observed compared to the WT, while the fucose and Neu5Ac content remained unchanged. e host gly-
can composition remained unchanged when infected with ΔmelA with six glycan dierences compared to WT,
which may explain the similar invasion phenotype as compared to the WT (Fig.1B). Further examination of
the glycans led to identication of specic structures that uniquely linked to Salmonella GHs, specically siali-
dases and amylases, and regulated infection (Fig.4B–E). While sialidases are required to initiate glycan degra-
dation we also observed that other glycan-degrading enzymes aid to complete the degradation of these glycans
to invade. Glycans m/z = 2499.899 (Hex6HexNAc6Neu5Ac1), 2790.995 (Hex6HexNAc6Neu5Ac2), 3156.127
(Hex7HexNAc7Neu5Ac2), 2979.099 (Hex5HexNAc7Fuc3Neu5Ac1), 2239.798 (Hex6HexNAc4Fuc1Neu5Ac1), and
3026.089 (Hex6HexNAc5Fuc3Neu5Ac2) are sialylated glycans that were degraded during the infection with WT
(Fig.4C–E). Upon infection with mutant GHs, the most increase in relative abundance was seen during infection
with ΔmalS, ΔmalS, ΔCHPNeu, ΔmalS, ΔglgX, and ΔmalS, respectively for each of the glycans, which may sug-
gest that these GHs targeted and degraded those glycans (Fig.4C–E). Glycans m/z = 2167.777 (Hex7HexNAc5),
1722.645 (Hex3HexNAc6), 2953.047 (Hex7HexNAc6Neu5Ac2), 2776.020 (Hex5HexNAc6Fuc3Neu5Ac1), and
2556.946 (Hex4HexNAc5Fuc4Neu5Ac1) were accumulated (Fig.4B–D), suggesting Salmonella induces the host to
produce or increase the relative production of these glycans. Lastly, glycans m/z = 2688.004 (Hex5HexNAc7Fuc3)
and 2483.904 (Hex5HexNAc6Fuc1Neu5Ac1) stayed at similar levels throughout the infection (Fig.4B,D) suggest-
ing that Salmonella does not have the specicity to recognize these glycans to degrade. ese results suggest that
the GHs within a genome contribute to virulence. Perhaps GHs represent a new virulence mechanism that is con-
trolled by the host glycan structure and the gene content on the pathogen. is elucidates how glycan-degrading
enzymes play an important role by altering the host glycan proles during infection to gain access of the host
epithelial cells. Also identifying exact glycans during infection will lead to identifying pathways that Salmonella
uses to infect and invade the host.
Figure 4. Signature of glycan proles and how each glycan-degrading enzyme modulates the glycans
during infection. (A) Signature of glycan proles during infection. ere is an increase in abundances of high-
mannose and hybrid glycans. Also the glycans are switched from being highly sialylated to highly fucosylated
(the degree of fucosylation increases and the degree of sialylation decreases); (B–E) Glycans that are highly
regulated by Salmonella glycan-degrading enzymes during infection. e comparisons between CHPNeu and
nanH (Sialidases) and malS and glgX (Amylases) are shown. LSD was used for statistical analysis. Error bars
indicate SEM between 3 biological replications. Levels not connected with the same letter are signicantly
dierent. Statistical analysis was done within each structure.

www.nature.com/scientificreports/
7
Scientific RepoRts | 6:29525 | DOI: 10.1038/srep29525
Discussion
Use of mucin and the glycocalyx by the microbiome is widely recognized6. Degrading the glycocalyx layer dis-
rupts the host’s defensive barrier that provides access to the epithelial cell surface and receptors for binding, col-
onization, and invasion. Because Salmonella infections represent a persistent and major health challenge around
the world it is important to understand how to control Salmonella infections. is study examined the specic
enzymes used to degrade the glycan.
e sequence of events leading to Salmonella infection in the gut epithelial cells has been elucidated for the
T3SS22. However, prior to T3SS (invA) access to the membrane, Salmonella must overcome the protective gly-
cocalyx layer that coats the epithelium. In the gut Salmonella must rst attach to and second degrade the host’s
protective mucus and glycocalyx layer to gain access to the membrane and subsequently infect the underlying
tissue. ese results described here rst identied that depletion of Neu5Ac from Caco-2 cells led to decreases
in Salmonella association; therefore, this interaction lends insight into how Salmonella reaches the initial goals
necessary to establish infection by using Neu5Ac as a potential receptor for adherence process.
In order to breach the glycocalyx layer, pathogens have evolved mechanisms to degrade the glycocalyx, which
is common in many organisms6–10. Salmonella specically is equipped with 48 enzymes from 21 families of gly-
cosyl hydrolases to overcome this barrier. is study found expression of specic glycan-degrading enzymes in
Salmonella used to degrade the glycocalyx and alter the host glycans to mediate invasion, which were previously
unrecognized genes important to the virulence mechanism.
While the microbe is degrading the host glycan, it is also very important to study the host response during
degradation. Eorts were made to improve our understanding of glycan degradation and alteration compexity
in its entirety via the use of high resolution mass spectrometry24 and gene expression proling. e host genes
that hydrolyze mannose, fucose, and Neu5Ac from the glycan; as well as sialyltransferases, were induced dur-
ing Salmonella infection (Fig.3); supporting the hypothesis that new glycan structures were produced de novo
(Fig.2F) during infection. Hooper et al. and Bry et al. also found in vivo that the host produced new glycans con-
taining fucosylation in the ileum that is microbe induced via α 1,2-fucosyltransferase transcripts6,7. Renement
of the glycan by the Salmonella during infection was also observed (Fig.4A). e shi to low Neu5Ac content is
supported by these changes in sialidases and sialyltransferases (Fig.3D,E, Supplementary Table S4). e accu-
mulation of fucosylated glycans is supported by the changes in host fucosidases and fucosyltransferases (Fig.3C)
aer infection, as previously observed6,7. Pickard et al. also observed that rapid fucosylation appears to be a pro-
tective mechanism that utilizes the host’s resources during host-microbe interactions during pathogen induced
stress25. ese observations conrm the host response to glycan modication during Salmonella LT2 association
and suggest a dynamic shi in the glycan is caused by Salmonella degradation coupled with the host remodeling
to produce a glycan of dierent structures as the infection progresses. is could be a protective mechanism for
the host during infection and suggests that the host is actively producing new glycans in response to Salmonella
infection. e mechanism dictating this phenomenon has yet to be elucidated, which requires additional exper-
imentation to uncover the direct eect of Salmonella on host glycan metabolism. Glycan biosynthesis is signi-
cantly aected by the disease states and distinct glycan structures could provide information about the specic
pathologic states of disease23. e host constantly remodels its glycans without altering its intrinsic function,
which alters microbiome association3 and modulates host immune surveillance methods26. e host may alters
its glycan composition on their cell surfaces to eliminate the expression of a terminal glycan structures in order
to limit pathogen binding. e host may discard non-critical glycans to allow its survival3. e loss of a particular
glycan may involve inactivation of one or more genes involved in glycan biosynthesis27 and can prevent recogni-
tion by pathogens using structure as a receptor. In an eort to alter its glycans to evade pathogens, the host may
create new glycan structures either by synthesis or modication.
By combining microbial genomics, glycan proling, and infection analyses, this study provides unprecedented
molecular details for the role of Salmonella GHs’ during infection of Caco-2 cells. Alterations in the degree of
branching, changes in the amount, linkage, and degrees of sialylation and fucosylation in N-glycans have been
reported as a consequence of diseases28,29, but this is the rst report to describe how Salmonella Typhimurium
degraded the glycans that resulted in an infection. Deleting GHs used for glycan degradation resulted in invasion
magnitude equal to those associated with T3SS in Salmonella, suggesting that these enzymes may be as important
as secretion systems. Dening the virulence characteristics and diversity of GHs in Salmonella can provide new
insights into the importance of glycan structures during host-pathogen interactions. is study further expands
our understanding of the infection characteristics and may lead to the host response in glycan modication as
mutualistic events between the host production and pathogen degradation. Some of the more encouraging pros-
pects of this research are the potential for new treatments for gastroenteritis caused by Salmonella. is work may
provide the basis of novel strategies to control enteric infections by targeting glycocalyx-degrading enzymes. e
complexity of the pan-genome of Salmonella, as well as the multiple methods of infection used by Salmonella30,31
demands the use of multi-omics to dene innovative targets to control infection. Use of antibiotics alone is not
meeting the needs to control this organism and oen leads to increased susceptibility to other pathogens32. Use
of glycan degradation will preempt Salmonella from gaining access of the membrane – a novel method to control
infection that may be of use in multiple pathogens in the gut and other epithelial surfaces.
Methods
Cell Culture. All cell in vitro experiments were done using colonic epithelial cells (Caco-2; ATCC HTB-37)
as described previous33–37. Detailed description of the cell growth conditions is available in Supplementary
Information.
Bacterial Strains and Growth Conditions. Salmonella enterica subsp. enterica serovar Typhimurium
strain LT2 (ATCC 700720; Salmonella WT) and the deletion mutants made in this study were used. All isolates

www.nature.com/scientificreports/
8
Scientific RepoRts | 6:29525 | DOI: 10.1038/srep29525
were grown in LB (Difco, BD) at 37 °C with shaking at 220 rpm for 14–16 hours before use. Each biological repli-
cate was done with a new vial of frozen stock aer thawing and growth twice as described above.
Gene Deletion. Bacterial gene deletions were done as described by Datsenko and Wanner20. Detailed
description of gene deletions is available in Supplementary Information.
Genomic sequencing and comparison. Each isolate was sequenced as described by Ludeke et al.38 as
dened by the methods of the 100 K Pathogen Genome Sequencing Project39–47. Abyss 1.5.2 was used to assemble
the paired end reads using k = 6448. Prokka was used for annotation49. Each genome was compared to the wild
type by determining the genome distance using Genome-to-Genome Distance Calculator (GGDC) (http://ggdc.
dsmz.de/distcalc2.php)50,51. Whole genome comparisons were done using Mauve under Progressive Mauve as
described by Darling et al.52,53. Contigs were reordered using the reorder contigs option in Mauve with default
parameters using Salmonella Typhimurium LT2 ATCC 700720 (accession number AE006468) as the reference
genome. All raw genome sequences generated in this study are available in the NCBI SRA as part of the 100 K
Pathogen Genome Project Bioproject Accession PRJNA186441. Accession numbers are listed in Table S5. Single
gene analyses were done by extracting sequences from each genome and using MUSCLE through Geneious
version 6.1.8 to align sequences54,55.
Bacterial Association Measurements. Bacterial association was determined using a modied gentam-
ycin protection assay56,57 aer adding each Salmonella treatment in a 96-well plate containing an MOI of 1:1000.
e assay was done aer incubation for 60 min at 37 °C with 5% CO2. Adhered bacteria were measured aer the
cell culture medium was aspirated and the host/microbe cell mixture was washed once with 200 μ l of 1X PBS
buer (pH 7). e host and associated microbes were lysed using 50 μ l of commercial Warnex lysis buer (AES
Chemunex) for use in qPCR assays to determine the absolute amount of host and bacteria57. All assays were done
in three biological replicates.
To calculate number of adhered bacteria, the mean of the number of invaded bacteria was subtracted from the
mean of the total number of host associated bacteria. e error for adhered bacteria was propagated using equa-
tion (Δ Z)2 = (Δ A)2 + (Δ B)2 where Δ Z is the standard error of mean (SEM) for adhered bacteria, Δ A is SEM for
total host associated bacteria and Δ B is SEM for invaded bacteria. One-way ANOVA with Tukey test was done to
nd signicant dierences across treatment’s and control’s group means.
Gene Expression during Infection. Samples were collected as previously described32,34,58,59. Briey, Caco-2
cells were cultured in T-75 asks (BD) and were serum starved 24 h before infection. Respective bacterial treat-
ments with Salmonella WT and the deletion mutants, at MOI of 1000, were used to infect epithelial cells as
described previously. All treatments were incubated at 37 °C with 5% CO2 for 60 min. 10 ml of TRIzol LS reagent
(Invitrogen, Carlsbad, CA) was added to the cells and mixed with pipette followed by centrifugation at 7,200 × g
for 5 min to pellet the host associated bacteria. TRIzol LS supernatant was stored in a clean tube and further pro-
cessed for RNA extraction from infected Caco-2 cells. e bacterial pellet was suspended in 2 ml of fresh TRIzol
LS, gently mixed and further processed for RNA extraction from host associated bacteria. e experiment was
done in two biological replicates.
Bacterial RNA extraction and gene expression. Sample preparation for gene expression proling was performed
with RNA isolation, which was done using TRIzol LS reagent (Invitrogen) as described previously32,34. Total RNA
(10 μ g in 20 μ l) was reverse transcribed into cDNA with 6 μ g of random hexamers and 400 U of Superscriptase II
(Invitrogen) according to the manufacturer’s protocol. The reaction mixture was cleaned by using the
Qiaquick-PCR purification kit (Qiagen, Valencia, CA) according to the manufacturer’s instructions and as
described previously34. e puried single-strand cDNA was eluted from the columns twice with a total of 100 μ l
of nuclease free water (Ambion, Austin, TX). cDNA fragmentation was done using 0.6 U of DNaseI (Promega,
Madison, WI) per μ g of cDNA, according to the instructions. e fragmented 1 μ g of cDNA was labeled using 2 μ l
of GeneChip DNA Labeling reagent (Aymetrix, Santa Clara, CA) and 60 U of Terminal Transferase enzyme
(New England Biolabs, Ipswich, MA). e samples were denatured prior to hybridization, at 98 °C for 10 min
followed by snap cooling at 4 °C for 5 min.
Hybridization and normalization. Labeled cDNA was hybridized onto a custom made Aymetrix GeneChip
designed against all the annotated coding sequences of Salmonella LT2 ATCC 700720 (Salmonella WT)32,34,60. e
chips were hybridized and scanned at the Center for Integrated BioSystems (Utah State University, Logan, UT)
as per manufacturer’s protocols. Raw data (.cel les) was background corrected, quantile normalized and sum-
marized using MS-RMA32,61. e resultant normalized Log2 transformed intensity matrix was used for further
statistical analysis.
Caco-2 RNA extraction and gene expression. The TRIzol LS samples containing infected or non infected
Caco-2 cells were frozen (Liquid N2) and thawed (70 °C) twice. To 750 μ l of TRIzol LS sample, 250 μ l of water
was added. is was further processed for RNA extraction using manufacturer’s (TRIzol LS, Invitrogen) instruc-
tions. Synthesis of cDNA, biotin labeled cRNA, fragmentation and purication of cRNA were carried out using
one-cycle cDNA synthesis kit (Aymetrix, Santa Clara, CA).
Host hybridization and normalization. Labeled and fragmented cRNA (10 μ g) was hybridized onto the
Aymetrix HGU133Plus2 GeneChips as per manufacturer’s recommendations at the Center for Integrated
BioSystems (Utah State University, Logan, UT). Raw data (.cel les) was background corrected; quantile normal-
ized and summarized using RMA. RMA normalized data was then ltered through the PANP algorithm to make

www.nature.com/scientificreports/
9
Scientific RepoRts | 6:29525 | DOI: 10.1038/srep29525
presence-absence calls for each probe set. Probe sets that were called present in at least one of the samples were
included in further statistical analysis.
Statistical analysis for gene expression. Gene expression proles for Salmonella WT alone and in the presence
of the epithelial cells were obtained 60 min post infection. e data was analyzed as two class unpaired with T
statistic, using Signicance Analysis of Microarrays (SAM)62. All the genes were ranked based on the score from
SAM output. is pre-ordered ranked gene list was then used in Gene Set Enrichment Analysis soware (GSEA)
to detect the coordinate changes in the expression of groups of functionally related genes, upon respective treat-
ments. e gene sets were dened based on the annotations from Comprehensive Microbial Resource (CMR),
Cluster of Orthologous Groups of proteins (COGs), and Virulence Factors of pathogenic bacteria Data Base
(VFDB).
Glycan Degradation. Dierentiated Caco-2 cells in T-75 asks were infected with Salmonella WT and its
knockouts at MOI 1:1000 and incubated for 60 min at 37 °C, 5% CO2. Infection samples were washed three times
with ice cold 1X PBS to remove non-adherent bacteria and cellular debris. Cells were scraped from the ask with
cell scraper and were kept on ice until cell membrane extraction and N-glycan release as detailed in Supplemental
Information online. N-glycans were enriched and analyzed using Agilent HPLC-Chip-QTOF MS (Agilent, CA)
as detailed in the Supplementary Information. N-Glycans were identied by composition with a retrosynthetic
library using accurate mass according to mass tolerance, retention times, and abundance information24 and fur-
ther veried by tandem MS (Supplementary Figure S6).
Ingenuity Pathway Analysis. QIAGEN’S Ingenuity Pathway Analysis (IPA, QIAGEN Redwood City, www.
qiagen.com/ingenuity) soware was used to determine biological pathways associated with our gene expression
data (Ingenuity Systems, http://www.ingenuity.com, IPA release version summer 2014) similar to He et al.33.
Networks representing molecular interaction were constructed based on the IPA database. Determination of
pathway associations was determined through IPA (Fisher’s exact test).
References
1. Moran, A. P., Gupta, A. & Joshi, L. Sweet-tal: role of host glycosylation in bacterial pathogenesis of the gastrointestinal tract. Gut
60, 1412–1425, doi: 10.1136/gut.2010.212704 (2011).
2. McGucin, M. A., Linden, S. ., Sutton, P. & Florin, T. H. Mucin dynamics and enteric pathogens. Nature reviews. Microbiology 9,
265–278, doi: 10.1038/nrmicro2538 (2011).
3. Vari, A. Evolutionary forces shaping the Golgi glycosylation machinery: why cell surface glycans are universal to living cells. Cold
Spring Harbor perspectives in biology 3, doi: 10.1101/cshperspect.a005462 (2011).
4. Imai, M. & awaoa, Y. e role of receptor binding specicity in interspecies transmission of inuenza viruses. Current opinion in
virology 2, 160–167, doi: 10.1016/j.coviro.2012.03.003 (2012).
5. de Graaf, M. & Fouchier, . A. ole of receptor binding specicity in inuenza A virus transmission and pathogenesis. e EMBO
journal 33, 823–841, doi: 10.1002/embj.201387442 (2014).
6. Bry, L., Fal, P. G. & Gordon, J. L. Genetic engineering of carbohydrate biosynthetic pathways in transgenic mice demonstrates cell
cycle-associated regulation of glycoconjugate production in small intestinal epithelial cells. Proceedings of the National Academy of
Sciences of the United States of America 93, 1161–1166 (1996).
7. Hooper, L. V., Xu, J., Fal, P. G., Midtvedt, T. & Gordon, J. I. A molecular sensor that allows a gut commensal to control its nutrient
foundation in a competitive ecosystem. Proceedings of the National Academy of Sciences of the United States of America 96, 9833–9838
(1999).
8. Huang, Y. L., Chassard, C., Hausmann, M., von Itzstein, M. & Hennet, T. Sialic acid catabolism drives intestinal inammation and
microbial dysbiosis in mice. Nat Commun 6, 8141, doi: 10.1038/ncomms9141 (2015).
9. Ng, . M. et al. Microbiota-liberated host sugars facilitate post-antibiotic expansion of enteric pathogens. Nature 502, 96–99,
doi: 10.1038/nature12503 (2013).
10. Picard, J. M. & Chervonsy, A. V. Intestinal fucose as a mediator of host-microbe symbiosis. J Immunol 194, 5588–5593,
doi: 10.4049/jimmunol.1500395 (2015).
11. Tailford, L. E., Crost, E. H., avanaugh, D. & Juge, N. Mucin glycan foraging in the human gut microbiome. Front Genet 6, 81,
doi: 10.3389/fgene.2015.00081 (2015).
12. Bergstrom, . S. & Xia, L. Mucin-type O-glycans and their roles in intestinal homeostasis. Glycobiology 23, 1026–1037, doi: 10.1093/
glycob/cwt045 (2013).
13. El aoutari, A., Armougom, F., G ordon, J. I., aoult, D. & Henrissat, B. e abundance and variety of carbohydrate-active enzymes
in the human gut microbiota. Nature reviews. Microbiology 11, 497–504, doi: 10.1038/nrmicro3050 (2013).
14. Marcobal, A., Southwic, A. M., Earle, . A. & Sonnenburg, J. L. A rened palate: bacterial consumption of host glycans in the gut.
Glycobiology 23, 1038–1046, doi: 10.1093/glycob/cwt040 (2013).
15. Johansson, M. E. et al. e inner of the two Muc2 mucin-dependent mucus layers in colon is devoid of bacteria. Proceedings of the
National Academy of Sciences of the United States of America 105, 15064–15069, doi: 10.1073/pnas.0803124105 (2008).
16. Cantarel, B. L. et al. e Carbohydrate-Active EnZymes database (CAZy): an expert resource for Glycogenomics. Nucleic acids
research 37, D233–238, doi: 10.1093/nar/gn663 (2009).
17. Lombard, V., Golaconda amulu, H., Drula, E., Coutinho, P. M. & Henrissat, B. e carbohydrate-active enzymes database (CAZy)
in 2013. Nucleic acids research 42, D490-495, doi: 10.1093/nar/gt1178 (2014).
18. Eilam, O. et al. Glycan degradation (GlyDe) analysis predicts mammalian gut microbiota abundance and host diet-specic
adaptations. mBio 5, doi: 10.1128/mBio.01526-14 (2014).
19. Lim, J. S. et al. Caveolae-mediated entry of Salmonella typhimurium into senescent nonphagocytotic host cells. Aging cell 9,
243–251, doi: 10.1111/j.1474-9726.2010.00554.x (2010).
20. Datseno, . A. & Wanner, B. L. One-step inactivation of chromosomal genes in Escherichia coli -12 using PC products.
Proceedings of the National Academy of Sciences of the United States of America 97, 6640–6645, doi: 10.1073/pnas.120163297 (2000).
21. Dos Santos, P. C., Fang, Z., Mason, S. W., Setubal, J. C. & Dixon, . Distribution of nitrogen xation and nitrogenase-lie sequences
amongst microbial genomes. BMC genomics 13, 162, doi: 10.1186/1471-2164-13-162 (2012).
22. Smith, A. C. et al. A networ of ab GTPases controls phagosome maturation and is modulated by Salmonella enterica serovar
Typhimurium. e Journal of cell biology 176, 263–268, doi: 10.1083/jcb.200611056 (2007).
23. Chu, C. S. et al. Prole of native N-lined glycan structures from human serum using high performance liquid chromatography on
a microuidic chip and time-of-ight mass spectrometry. Proteomics 9, 1939–1951, doi: 10.1002/pmic.200800249 (2009).

www.nature.com/scientificreports/
10
Scientific RepoRts | 6:29525 | DOI: 10.1038/srep29525
24. ronewitter, S. . et al. e development of retrosynthetic glycan libraries to prole and classify the human serum N-lined
glycome. Proteomics 9, 2986–2994, doi: 10.1002/pmic.200800760 (2009).
25. Picard, J. M. et al. apid fucosylation of intestinal epithelium sustains host-commensal symbiosis in sicness. Nature 514, 638–641,
doi: 10.1038/nature13823 (2014).
26. irmiz, C. et al. A serum glycomics approach to breast cancer biomarers. Molecular & cellular proteomics: MCP 6, 43–55,
doi: 10.1074/mcp.M600171-MCP200 (2007).
27. Bishop, J. . & Gagneux, P. Evolution of carbohydrate antigens--microbial forces shaping host glycomes? Glycobiology 17, 23–34,
doi: 10.1093/glycob/cwm005 (2007).
28. An, H. J., ronewitter, S. ., de Leoz, M. L. & Lebrilla, C. B. Glycomics and disease marers. Current opinion in chemical biology 13,
601–607, doi: 10.1016/j.cbpa.2009.08.015 (2009).
29. Lebrilla, C. B. & An, H. J. e prospects of glycan biomarers for the diagnosis of diseases. Molecular bioSystems 5, 17–20, doi:
10.1039/b811781 (2009).
30. Jacobsen, A., Hendrisen, . S., Aaresturp, F. M., Ussery, D. W. & Friis, C. e Salmonella enterica pan-genome. Microbial ecology
62, 487–504, doi: 10.1007/s00248-011-9880-1 (2011).
31. Gordieno, E. N., azanov, M. D. & Gelfand, M. S. Evolution of pan-genomes of Escherichia coli, Shigella spp., and Salmonella
enterica. Journal of bacteriology 195, 2786–2792, doi: 10.1128/JB.02285-12 (2013).
32. Lai, Y. Y. et al. [A comparative study of uorescence in situ hybridization versus conventional cytogenetics in the detection of clonal
aberrations in myelodysplastic syndrome]. Zhonghua Yi Xue Za Zhi 93, 3175–3179 (2013).
33. He, X., Mishchu, D. O., Shah, J., Weimer, B. C. & Slupsy, C. M. Cross-tal between E. coli strains and a human colorectal
adenocarcinoma-derived cell line. Scientic reports 3, 3416, doi: 10.1038/srep03416 (2013).
34. Shah, J., Desai, P. T., Chen, D., Stevens, J. . & Weimer, B. C. Preadaptation to cold stress in Salmonella enterica serovar Typhimurium
increases survival during subsequent acid stress exposure. Applied and environmental microbiology 79, 7281–7289, doi: 10.1128/
AEM.02621-13 (2013).
35. Shah, J., Desai, P. T. & Weimer, B. C. Genetic mechanisms underlying the pathogenicity of cold-stressed Salmonella enterica serovar
typhimurium in cultured intestinal epithelial cells. Applied and environmental microbiology 80, 6943–6953, doi: 10.1128/
AEM.01994-14 (2014).
36. Ferruzza, S., ossi, C., Scarino, M. L. & Sambuy, Y. A protocol for dierentiation of human intestinal Caco-2 cells in asymmetric
serum-containing medium. Toxicology in vitro: an international journal published in association with BIBA 26, 1252–1255,
doi: 10.1016/j.tiv.2012.01.008 (2012).
37. Sambuy, Y. et al. e Caco-2 cell line as a model of the intestinal barrier: inuence of cell and culture-related factors on Caco-2 cell
functional characteristics. Cell biology and toxicology 21, 1–26, doi: 10.1007/s10565-005-0085-6 (2005).
38. Ludee, C. H., ong, N., Weimer, B. C., Fischer, M. & Jones, J. L. Complete Genome Sequences of a Clinical Isolate and an
Environmental Isolate of Vibrio parahaemolyticus. Genome Announc 3, doi: 10.1128/genomeA.00216-15 (2015).
39. Jeannotte, . et al. Optimization of Covaris Settings for Shearing Bacterial Genomic DNA by Focused Ultrasonication and Analysis
Using Agilent 2200 TapeStation (2014).
40. Jeannotte, . et al. High-roughput Analysis of Foodborne Bacterial Genomic DNA Using Agilent 2200 TapeStation and Genomic
DNA ScreenTape System (2014).
41. ong, N. et al. Integrating the DNA Integrity Number (DIN) to Assess Genomic DNA (gDNA) Quality Control Using the Agilent
2200 TapeStation System (2014).
42. ong, N. et al. Quality Control of High-roughput Library Construction Pipeline for APA HTP Library Using an Agilent 2200
TapeStation (2014).
43. ong, N., Ng, W., elly, L. & Weimer, B. C. Quality Control of Library Construction Pipeline for PacBio SMTbell 10 b Library
Using an Agilent 2200 TapeStation System (2015).
44. ong, N., Ng, W., Lee, V., elly, L. & Weimer, B. C. Production and Analysis of High Molecular Weight Genomic DNA for NGS
Pipelines Using Agilent DNA Extraction it (p/n 200600) (2013).
45. ong, N. et al. Automated Library Construction Using APA Library Preparation its on the Agilent NGS Worstation Yields
High-Quality Libraries for Whole-Genome Sequencing on the Illumina Platform (2014).
46. ong, N. et al. Automation of PacBio SMTbell 10 b Template Preparation on an Agilent NGS Worstation (2014).
47. Miller, B. et al. A novel, single tube enzymatic fragmentation and library construction method enables fast turnaround times and
improved data qulaity for microbial whole-genome sequencing (2015).
48. Simpson, J. T. et al. ABySS: A parallel assembler for short read sequence data. Genome es 19, 1117–1123, doi: 10.1101/
Gr.089532.108 (2009).
49. Seemann, T. Proa: rapid proaryotic genome annotation. Bioinformatics 30, 2068–2069, doi: 10.1093/Bioinformatics/Btu153
(2014).
50. Auch, A. F., von Jan, M., len, H. P. & Goer, M. Digital DNA-DNA hybridization for microbial species delineation by means of
genome-to-genome sequence comparison. Stand Genomic Sci 2, 117–134, doi: 10.4056/Sigs.531120 (2010).
51. Meier-oltho, J. P., Auch, A. F., len, H. P. & Goer, M. Genome sequence-based species delimitation with condence intervals
and improved distance functions. Bmc Bioinformatics 14, doi: 10.1186/1471-2105-14-60 (2013).
52. Darling, A. C., Mau, B., Blattner, F. . & Perna, N. T. Mauve: multiple alignment of conserved genomic sequence with
rearrangements. Genome es 14, 1394–1403, doi: 10.1101/gr.2289704 (2004).
53. Darling, A. E., Mau, B. & Perna, N. T. progressiveMauve: multiple genome alignment with gene gain, loss and rearrangement. PloS
one 5, e11147, doi: 10.1371/journal.pone.0011147 (2010).
54. Edgar, . C. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic acids research 32, 1792–1797,
doi: 10.1093/nar/gh340 (2004).
55. earse, M. et al. Geneious Basic: an integrated and extendable destop soware platform for the organization and analysis of
sequence data. Bioinformatics 28, 1647–1649, doi: 10.1093/bioinformatics/bts199 (2012).
56. Elsinghorst, E. A. Measurement of invasion by gentamicin resistance. Methods in enzymology 236, 405–420 (1994).
57. Desai, P. T., Walsh, M. . & Weimer, B. C. Solid-phase capture of pathogenic bacteria by using gangliosides and detection with real-
time PC. Applied and environmental microbiology 74, 2254–2258, doi: 10.1128/AEM.02601-07 (2008).
58. Ganesan, B., Dobrowolsi, P. & Weimer, B. C. Identication of the leucine-to-2-methylbutyric acid catabolic pathway of Lactococcus
lactis. Applied and environmental microbiology 72, 4264–4273, doi: 10.1128/AEM.00448-06 (2006).
59. Xie, Y., Chou, L. S., Cutler, A. & Weimer, B. DNA Macroarray proling of Lactococcus lactis subsp. lactis IL1403 gene expression
during environmental stresses. Applied and environmental microbiology 70, 6738–6747, doi: 10.1128/AEM.70.11.6738-6747.2004
(2004).
60. Marcobal, A. et al. Bacteroides in the infant gut consume mil oligosaccharides via mucus-utilization pathways. Cell host & microbe
10, 507–514, doi: 10.1016/j.chom.2011.10.007 (2011).
61. Tusher, V. G., Tibshirani, . & Chu, G. Signicance analysis of microarrays applied to the ionizing radiation response. Proceedings
of the National Academy of Sciences of the United States of America 98, 5116–5121, doi: 10.1073/pnas.091062498 (2001).
62. Champine, P. J., Michaelson, J., Weimer, B. C., Welch, D. . & DeWald, D. B. Microarray analysis reveals potential mechanisms of
BMS1-mediated metastasis suppression. Clinical & experimental metastasis 24, 551–565, doi: 10.1007/s10585-007-9092-8 (2007).

www.nature.com/scientificreports/
11
Scientific RepoRts | 6:29525 | DOI: 10.1038/srep29525
Acknowledgements
Funding provided to BCW (NIH - 1R01HD065122-01A1; NIH - U24-DK097154; Agilent Technologies ought
Leader Award) and CBL (R01GM049077) are gratefully acknowledged.
Author Contributions
N.A. designed experiments, constructed deletion mutants, conducted bacterial infection and cell culture
experiments, analyzed MS data, glycan, and gene expression data, extracted DNA, analyzed genome sequences,
and wrote the manuscript; D.P., C.C.W. performed glycan assays and determined glycan structures; S.F. analyzed
the host gene expression; A.M.W. analyzed genome sequences; B.C.H. constructed DNA libraries for sequencing;
P.D., J.S. performed enzymatic glycan degradation, conducted gene expression experiments, and constructed
deletion mutants; R.J. helped with the analysis of MS data; N.K. constructed deletion mutants; C.B.L. reviewed
manuscript; B.C.W. conceived of the hypothesis, planned experiments, critically accessed the data, and wrote the
manuscript.
Additional Information
Supplementary information accompanies this paper at http://www.nature.com/srep
Competing nancial interests: e authors declare no competing nancial interests.
How to cite this article: Arabyan, N. et al. Salmonella Degrades the Host Glycocalyx Leading to Altered
Infection and Glycan Remodeling. Sci. Rep. 6, 29525; doi: 10.1038/srep29525 (2016).
is work is licensed under a Creative Commons Attribution 4.0 International License. e images
or other third party material in this article are included in the article’s Creative Commons license,
unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license,
users will need to obtain permission from the license holder to reproduce the material. To view a copy of this
license, visit http://creativecommons.org/licenses/by/4.0/