inducible factor 1-a(HIF1-a) and production
of the antimicrobial peptide CRAMP (LL-37),
which together ensure transkingdom resis-
tance toC. albicans( 32 ). Overgrowth of the
yeastPichia kudriavzeviiin infant mice and
humans because of dysbiosis in early life is an
associated risk for allergic airway disease later
in life. Neonatal mice exposed toP. kudriavzevii
exhibit increased inflammatory responses as
they mature. Mice with the highest loads of
P. kudriavzeviihave the lowest fecal amounts
of bacterially derived SCFAs. Studies using in
vitro assays have reported that these SCFAs
inhibit the growth and morphology of asthma-
associatedP. kudriavzevii. In addition, mice
that have been administered a cocktail of
SCFAs (acetate, butyrate, and propionate) show
reduced colonization byP. kudriavzevii( 33 ).
Thus, although the mechanisms have yet to
be defined, these studies indicate that over-
growth of one fungal pathobiont stems from
a release of control by associated gut bacteria.
Likewise,Wallemia mellicola,acommensal
fungus of both humans and mice, expands
in the intestines of antibiotic-treated mice
but without any accompanying overgrowth
of the total fungal community. In antibiotic-
treated C57BL/6 mice or altered Schaedler
flora mice with a defined fungus-free bacte-
rial community, colonization byW. mellicola
aggravates asthma-like lung inflammation in
house dust mite antigen-induced allergic air-
way disease ( 34 ). Antibiotic therapy could be
an important risk factor for inflammatory dis-
ease triggered by fungal overgrowth because
of unrecognized, but potentially important,
mechanisms of interkingdom interactions be-
tween bacteria and fungi.
Macroparasites, including helminths and
protozoa, can also have complex, reciprocal
relationships with bacteria and viruses. The
protozoanTrichomonas vaginalisis the most
prevalent nonviral, sexually transmitted infec-
tion worldwide.T. vaginaliscolonization alters
the resident microbiota composition and leads
to urogenital dysbiosis. However,T. vaginalis
can itself be parasitized byT. vaginalisvirus,
which is found in ~50% of clinical isolates;
when the virus is present, it associates with a
greater severity of trichomoniasis in humans
because the virus itself induces a Toll-like re-
ceptor 3–driven proinflammatory response ( 35 ).
Another protozoan,Tritrichomonas musculis,
induces IL-18–driven TH1andTH17 responses
in mice, which modulates the outcome of
Salmonella typhimurium–induced enterocoli-
tis ( 36 ). Further work on coinfection of mice
with the helminthHeligmosomoides polygyrus
bakeriand West Nile virus showed increased
mortality, partially associated with altered gut
morphology and transit, as well as translocation
of gut bacteria into the spleen via blood ( 37 ).
However, infection of mice withH. p. bakeri
alone altered the intestinal microbiota and
increased SCFA production such that trans-
fer of theH. p. bakeri–modified microbiota
into antibiotic-treated or germ-free mice pro-
tected them against allergic asthma ( 38 ). A
recent study identified a previously unknown
bactericidal protein called small proline-rich
protein 2A (SPRR2A) that was induced in
intestinal secretory cells (goblet and Paneth
cells in mouse and goblet cells in human) upon
infection withH. polygyrus.IL-4andIL-13
elicited by the helminth infection strongly
induced SPRR2A expression. SPRR2A selec-
tively kills Gram-positive bacteria by disrupt-
ing their membranes, thus limiting bacterial
invasion of intestinal tissue during helminth
infection ( 39 ).
Bacteriophages are predominant in the virome
of the healthy gut. Phages are implicated inhorizontal gene transfer between bacterial
species, helping to shape the composition of
microbial assemblages and influencing mam-
malian immunity and physiology. Gut bacterio-
phages were shown to influence the dynamics
of bacterial populations in germ-free mice that
were colonized by a defined microbial com-
munity composed of 15 human symbionts ( 40 ).
Bacteriophages not only target specific bac-
teria but also affect nonhost bacterial species
through indirect effects on interbacterial in-
teractions. Fecal metabolomics has shown that
shifts in gut microbial communities caused by
phage predation can have a direct impact on
gut metabolites ( 41 ). Bacteriophages can also
interact directly with the mammalian immune
system because phage DNA activates inter-
ferong(IFN-g) responses through the Toll-like
receptor 9–dependent pathway. This inter-
action has been shown to exacerbate colitis.
Mucosal IFN-gresponses in ulcerative colitis
(UC) patients positively correlate with bacterio-
phage abundance ( 42 ).
One study has shown some early promise
in the use of bacteriophages to ameliorate
alcoholic liver disease.Cytolysin, an exotoxin
secreted byEnterococcus faecalis,isstrongly
correlated with severity of liver disease and
mortality in patients with alcoholic hepatitis.
Ethanol-induced gut barrier damage likely en-
ables translocation of cytolyticE. faecalisfrom
the intestine to the liver. In gnotobiotic mice
colonized with feces from patients with alco-
holic hepatitis, gavage of bacteriophages tar-
geting cytolyticE. faecaliseffectively decreases
cytolysin concentrations in the liver and reducesthe severity of ethanol-induced liver injury and
steatosis ( 43 ).Host-microbiota interaction through
metabolites and enzymes
In addition to direct and local communication
at host mucosal barriers, the microbiota pro-
duces metabolites that can communicate across
host mucosae to influence host physiology sys-
temically. Metabolites associated with various
organs and diseases have been cataloged and
used in functional screening (Fig. 2) ( 44 ). For
example, gut microbe–produced metabolites
that are derived from dietary tryptophan tar-
get the mammalian aryl hydrocarbon receptor
(AHR) to inhibit activation of the actin regu-
latory proteins MyoIIA and ezrin and thereby
locally strengthen intestinal barrier integrity
that is mediated by tight junctions and ad-
herens junctions ( 45 ). Tryptophan metabo-
lites also remotely function as AHR agonists
in astrocytes to mitigate central nervous sys-
tem inflammation by potentiating the type I
interferon response ( 46 ). Certain bacterial spe-
cies in the gut ferment nondigestible dietary
fibers into SCFAs—such as acetate, propionate,
and butyrate—to mediate host energy balance,
immune modulation, and mucosal barrier
function ( 47 ). Butyrate locally induces the
differentiation of Tregsin the colonic mucosa
to maintain gut immune homeostasis ( 48 )
while modulating asthma phenotypes in the
lung ( 49 ). These discoveries indicate that
SCFAs function in a cell type–dependent man-
ner and that local concentrations of SCFAs
are important.
Another clear example of the influence that
the microbiota can have on host metabolism has
emerged from an untargeted mass spectrometry–
based metabolomics analysis of plasma from
healthy patients and patients with cardiovas-
cular disease (Fig. 2). The disease group had
increased concentrations of trimethylamine
N-oxide (TMAO), which is produced in the liver
from trimethylamine (TMA), a microbiota-derived
metabolite generated from dietaryL-carnitine
and choline. TMAO alters host cholesterol and
steroid metabolism and increases the risk for
cardiovascular diseases when its concentra-
tions are increased in plasma ( 50 ). A proof-
of-concept study showed that inhibiting a major
microbial TMA-generating enzyme (CutC/D) by
a small molecule significantly reduced plasma
concentrations of TMAO and effectively reduced
the potential for thrombosis ( 51 ).
SimilartoTMAO,secondarybileacidsaregut
microbiota–liver axis metabolites. Bile acids are
a class of cholesterol derivatives synthesized in
the liver (as primary bile acids) that can be mod-
ified by gut microbes into secondary bile acids,
including deoxycholic acid (DCA). Bile acids
have been implicated in several host-microbiota
effects on host physiology. For example, at higher
concentrations, DCA binds one of its receptors,Lu et al., Science 376 , 950–955 (2022) 27 May 2022 4of6
“Dysbiosis triggered by antibiotic
treatment can allow the expansion
of pathobionts because of the loss
of commensal microbes...”