The ultimate benefit of urea nitrogen sal-
vage is nitrogen incorporation into host protein.
Using isotope ratio mass spectrometry, we found
that microbiome-intact squirrels incorporated
significant^15 N into liver and muscle protein,
but microbiome-depleted squirrels did not
(Fig. 4, B and C), thereby verifying that urea
nitrogen salvage is microbe-dependent and
benefits the host. Moreover,^15 N incorpora-
tion into liver and muscle protein was two to
three times as high in the late winter group
versus the summer group, indicating that urea
nitrogen salvage is most beneficial late in the
winter fast. We also found large amounts of
(^15) N incorporated into cecal content protein
(indicative of microbial protein; Fig. 4A) in
microbiome-intact squirrels, indicating nutri-
tional benefits for the gut microbiome. How-
ever, unlike in muscles and the liver, microbial
protein^15 N incorporation was highest in the
summer group (Fig. 4A), for which microbial
abundance was also highest (fig. S5).
Notably,^15 N incorporation into host protein
was highest when ureolytic activity was lowest
(i.e., late winter; Fig. 3D). This can be explained
by our observations at multiple steps (Fig. 1B):
First, urea transport capacity into the gut likely
increases during hibernation through elevated
UT-B abundance (Fig. 2B). Second, the micro-
biome becomes proportionally (though not ab-
solutely) more ureolytic in winter, evidenced
by elevated urease gene percentages (Fig. 3E)
and bacterial urease activity levels (fig. S4B)
that are higher than those predicted by the
breathd^13 C changes (Fig. 3D). Another possible
contributor is proportionally greater microbial
CO 2 fixation (e.g., acetogenesis) in winter than
in summer (fig. S6), thus reducing the quantity
of exhaled^13 CO 2. Third, higher bacterial death
rates in winter ( 10 ) could liberate relatively
higher amounts of bacterial contents for host
absorption during hibernation, including meta-
bolites containing urea nitrogen (Fig. 4A).
Salvaged urea nitrogen can potentially be
absorbed by the squirrel as amino acids or
ammonia. Though ceco-colonic amino acid
absorption occurs in rodents ( 11 ), our results
suggest that ammonia absorption and its
hepatic conversion to glutamine through gluta-
mine synthetase may be the key step that is
enhanced during hibernation. For example,
tissue abundance of^15 N-lysine (an essential
aminoacidthatmustbeabsorbed)wasaffected
only by season and not by microbiome pres-
ence. Additionally, numerous lines of evidence
support the absorption and hepatic conversion
of^15 N-ammonia, including elevated liver^15 N-
ammonia and^15 N-glutamine levels in early and
late winter (Fig. 4B), sustained liver glutamine
synthetase activity during hibernation (fig. S7),
and reduced urea cycle enzymes and inter-
mediates ( 12 , 13 ). This implies that during
hibernation, the condensation of ammonia
with glutamate to form glutamine is favored
over its conversion to urea. Indeed, in protein-
deficient rats, hepatic ammonia processing
shifts from urea toward glutamine produc-
tion ( 14 ), and in fasting hibernating arctic
ground squirrels, ammonia from muscle pro-
tein breakdown is directed away from the urea
cycle toward amino acid formation ( 13 ). De-
spite this implied shift in ammonia processing,
urea production continues during hibernation
(albeit at lower rates), with plasma concen-
trations lowest upon torpor reentry and highest
during interbout arousal (fig. S2) ( 7 ). Much of
this urea is likely transferred to the gut by urea
transporters, where it becomes available for
microbial ureolysis. Thus, nitrogen recycling
in hibernators involves both endogenous ( 13 )
and microbe-dependent mechanisms, under-
scoring the importance of maintaining nitro-
gen balance during hibernation.
Urea nitrogen salvage provides two major
benefits for hibernators: First, it augments
protein synthesis when dietary nitrogen is
absent and appears especially important late in
hibernation, just before the squirrel’s emergence
into its breeding season ( 2 , 15 ). By facilitating
protein synthesis and subsequent tissue func-
tion, this process may confer reproductive ad-
vantages. Second, as in water-deprived camels
( 16 ), urea nitrogen salvage may enhance water
conservation in water-deprived hibernators
by diverting urea away from the kidneys, thus
requiring less water for urine production.
462 28 JANUARY 2022•VOL 375 ISSUE 6579 science.orgSCIENCE
-1 0123
-30
-20
-10
0
Time (h)
Breath
13
C (‰)
Late winter
-1 0 1 2 3
-30
-20
-10
0
10
Time (h)
Breath
13
C (‰)
Early winter
-1 0123
-50
0
50
100
150
Time (h)
Breath
13
C (‰)
Summer
-100
0
100
200
300
AUC for breath
13
C a
b b
- Summer Early
winter
Late
winter
Summer Early Winter Late Winter
Taxon % UreaseGenes Taxon % UreaseGenes Taxon % UreaseGenes
Bacteroides 1.76% Alistipes 1.91% Alistipes 1.62%
Lactobacillus 1.15% Streptomyces 1.47% Streptomyces 1.47%
Alistipes 1.04% Bacteroides 1.47% Bacteroides 0.93%
Streptomyces 0.99% Pseudomonas 1.08% Flavonifractor 0.85%
Clostridiales 0.82% Flavonifractor 0.96% Pseudomonas 0.62%
Oscillibacter 0.71% Clostridiales 0.83% Oscillibacter 0.62%
Mycolicibacterium 0.38% Oscillibacter 0.77% Parabacteroides 0.54%
Pseudomonas 0.33% Parabacteroides 0.57% Intestinimonas 0.54%
Akkermansia 0.33% Akkermansia 0.38% Acidovorax 0.31%
Flavonifractor 0.27% Intestinimonas 0.38% Blautia 0.23%
ABC
DE
G
F
0.00
0.05
0.10
0.15
Urease genes in metagenome (%)
SummerEarly
winter
Late
winter
injection injection injection
Early Winter Late Winter
Early
Winter
Late
Winter
Early
Winter
Late
Winter
K03190 K01430 K01429 K01428 K03187 K03188 K03189
0
1
0
1
0
1
0
1
0
1
0
1
0
1
Summer
Fig. 3. Gut microbial ureolysis and metagenome.Mean breathd^13 C for (A) summer, (B) early winter, and
(C) late winter squirrels treated with^13 C,^15 N-urea. Horizontal and vertical error bars represent SEM, and
closed and open symbols represent squirrels with intact and depleted microbiomes, respectively. The arrows
indicate intraperitoneal^13 C,^15 N-urea injection. (D) Mean area-under-the-curve (AUC) values for data in (A)
to (C), where asterisks indicate significant differences between microbiome groups within a season, and
lowercase letters indicate significant seasonal differences among microbiome-intact groups [n= 5, (A) to
(D);n= 4 for early winter microbiome-depleted animals]. Seasonal effect on (E) the percentage of urease
genes in the bacterial metagenome, (F) urease operons including each geneÕs Kyoto Encyclopedia of Genes
and Genomes ontology identifier and seasonal relative abundance, and (G) the top 10 bacterial taxa from
taxonomic classification of urease genes as determined by sequence abundance.
RESEARCH | REPORTS
- Summer Early