Science - USA (2022-01-28)

microbial ureolytic activity. Ureolysis was
greatest in summer squirrels (Fig. 3D and
fig. S4B), consistent with greater bacterial
abundance in summer versus hibernating
squirrels (fig. S5). Nevertheless, microbial ureolysis
continued throughout hibernation.
The metagenomes of hibernating squirrels
trended toward higher percentages of urease
genes than those of summer squirrels (Fig. 3E)
across the seven urease-related genes (Fig.
3F). This suggests that during hibernation, a
higher percentage of microbes have the poten-
tial to hydrolyze urea. Indeed,Alistipes—the
bacterial genus with the greatest detectable
genomic representation of urease genes in
early and late winter squirrels (Fig. 3G)—is
predominant during hibernation, with a six-

fold population increase between the summer

and late winter groups ( 9 ).

To benefit the host, microbial ureolytic ac-

tivity would need to provide nitrogenous com-

pounds such as amino acids and/or ammonia.

Using two-dimensional^1 H-^15 N NMR spectros-

copy, we found that more^15 N was incorporated

into the cecal content and liver metabolomes of

microbiome-intact compared with microbiome-

depleted squirrels (Fig. 4, A and B), a trend

that—with a few exceptions—also held for

specific metabolites such as ammonia, glutamine,

and alanine (Fig. 4, A and B).^15 N-metabolite

levels also varied seasonally. In cecum con-

tents, early and late winter metabolite levels

were generally lower than summer levels,

whereas in the liver, early and late winter

metabolite levels were generally higher than

summer levels and were typically highest in

the late winter group (Fig. 4, A and B). For

muscle, metabolite^15 N incorporation was gen-

erally unaffected by the presence of a micro-

biome (Fig. 4C), which could be due to the

timing of our^13 C,^15 N-urea dosing and tissue

sampling protocols. When tissues were sampled,

the^15 N-amino acids from the initial^13 C,^15 N-urea

dosemayhavealreadybeenincorporated

into muscle protein, thus explaining the

microbiome-dependent^15 N-protein results

(Fig. 4C), whereas the 3 hours between the

second dose and tissue sampling may have

been too brief for^15 N-metabolites to appear

in muscle. Rather, the muscle^15 N-metabolite

levels may represent microbiome-independent

background^15 N levels, which is consistent with

the equivalent^15 N-metabolite abundances in

the muscle of squirrels treated with labeled

and unlabeled urea and with intact and depleted

microbiomes (table S1).

SCIENCEscience.org 28 JANUARY 2022¥VOL 375 ISSUE 6579 461

A

B

0

50

100

150

Plasma urea concentration (mg dl

-1

)

Summer Early

Winter

Late

Winter

Season: P<0.0001

Microbiome: P=0.0002

0

100

200

300

400

500

UT-B abundance mg

-1

protein

Summer Late Winter

Season: P=0.0002

Microbiome: P<0.0001

Fig. 2. Plasma urea concentration and cecal urea

transporter (UT-B) expression.(A) Plasma urea

concentration in urea-treated squirrels with intact

(filled bars) and depleted (open bars) gut microbiomes

(n= 4 to 5 animals). (B) UT-B protein abundance per

milligram of cecal protein in cecum tissue (n= 12;

immunoblots in fig. S8). A subset of nonÐurea-treated

squirrels was used that lacked the early winter group.

Two-way analysis of variance (ANOVA) results are

shown in each panel.

Plasma urea

concentration:

UT-B abundance:

Microbial urease

activity:

CO 2 excretion:

Urea N incorporation

into host NH 3 :

% urease genes in

metagenome:

Urea N incorporation

into host AA:

Urea N incorporation

into host metabolome:

Urea-derived

acetogenesis:

Urea N incorporation

into host protein:

Urea nitrogen

salvage

process

Late Winter

relative to

Summer

Liver glutamine

synthetase activity:

AB

Fig. 1. Proposed mechanism for urea nitrogen salvage and relative changes during hibernation.
(A) Urea (yellow triangles), endogenously produced by the liver, is transported by epithelial urea transporters
(UT-B; brown squares) from blood into the cecal lumen where it is hydrolyzed into CO 2 and NH 3 by
urease-expressing gut microbes. CO 2 is excreted by the host and/or fixed by microbes. NH 3 is absorbed by
the host and converted into amino acids (AAs) and/or urea in the liver, or used by microbes to synthesize
AAs (circles) which are incorporated into the microbial proteome or potentially absorbed by the host through
ceco-colonic amino acid transporters (purple squares) ( 11 ). Ultimately, the AAs are used to synthesize
protein (blue ovals) in host tissues, recycling the urea nitrogen. (B) Arrows indicate how processes change
(increase, decrease, no change) in winter versus summer as revealed by this study. All changes are based on
statistically significant results except for percentage of urease genes in the metagenome (P= 0.083).

RESEARCH | REPORTS