SCIENCE science.orgpositioned in two locations of the CR and
one position of the NR, in entirely different
environments than in the IR. One rationale
for this moonlighting function, at least in
the CR, is the role of NUP93 in anchoring
the cytoplasmic NUP62-NUP88-NUP214
FG-containing complex, which is analogous
to the central NUP62-NUP58-NUP54 FG-
containing complex in the IR. Although only
one NUP62-NUP88-NUP214 complex is mod-
estly resolved in the protomer cryo-ET map,
stochiometric and steric arguments suggest
that there are two copies per protomer (or 16
copies per CR). Overall, the three studies of
human NPCs arrive at a reassuringly similar
conclusion about the positioning of NUPs.
Zhu et al. and Fontana et al. pursued a dif-
ferent approach. Oocytes from the African
clawed frog (Xenopus laevis) have nuclear
membranes packed with NPCs, which en-
ables single-particle reconstruction by use of
cryo–electron microscopy (cryo-EM). In cryo-
EM, many more particles are averaged com-
pared with that in cryo-ET, which boosts res-
olution. The studies of Zhu et al. and Fontana
et al. visualize a CR protomer at up to ~4
and ~7 Å, respectively. With well-resolved
secondary structure elements, and even res-
idue-level resolution for substantial parts of
the protomer, these studies enable detailed
analysis. The findings suggest that the core
scaffold of the vertebrate NPC is well con-
served. As previously determined, the nine-
membered core Y complex forms two concen-
tric, interwoven eight-membered rings ( 1 – 3 ).
The five NUP358 amino-terminal a-solenoid
structures per protomer are very similar in
both analyses, as are two NUP205 molecules.
Beyond that, the structures differ. Fontana et
al. resolve two NUP62-NUP88-NUP214 com-
plexes, as predicted by the cryo-ET–based
studies, but Zhu et al. do not. Instead, they
exclusively position two NUP93 molecules at-
tached to the CR. These differences may be
explained by different data processing strat-
egies, slightly different sample preparations,
or both. More likely, though, they point to a
larger issue, which is the intrinsic heteroge-
neity of NPCs.
A limitation of these studies is that to
achieve high resolution, many NPCs, and
specifically individual protomers, need to
be averaged. Given the high degree of struc-
tural redundancy among scaffold NUPs, it
is expected that NPCs are conformationally
and constitutionally heterogenous ( 3 , 11 ). A
recent structural analysis of NPCs from the
yeast Saccharomyces cerevisiae provided the
clearest example of this to date, in which dif-
ferent compositional classes of NPCs were
visualized ( 12 ). The more ambiguous parts
of the different reconstructions published in
this issue may point to inadvertent averag-
ing over nonidentical protomers, resulting in
partially distorted densities. This is difficult
to separate from flexibility, which commonly
reduces map contrast.
A prominent example of conformational
heterogeneity within the human NPC is the
recent discovery that the central channel
is substantially wider in intact cells when
compared with partially purified NPCs (57
versus 43 nm, respectively) ( 13 , 14 ). This con-
formational change is largely confined to the
IR and LR, not the two outer rings. It may
thus not be a coincidence that the CR is the
best-resolved NPC element, likely reflecting
higher homogeneity. How can the IR adopt
such different conformations? The consensus
from the studies of Mosalaganti et al. and
Petrovic et al. is that the protomers do not
structurally change but rather move as rigid
blocks, enabled by highly flexible linkers that
keep the protomers connected. This way, pe-
ripheral channels are established that enable
transport of embedded membrane proteins
from the outer to inner nuclear membrane,
and past the NPC.
Overall, the studies of Bley et al., Petrovic et
al., Mosalaganti et al., Zhu et al., and Fontana
et al. substantially advance the knowledge
of the assembly of NPCs, primarily in ver-
tebrates. NPCs can now be functionally and
structurally probed in unprecedented detail.
Although much of the approach by the differ-
ent researchers can be likened to solving a gi-
ant jigsaw puzzle, it is surprising that several
identical pieces fit into many different posi-
tions. It will be interesting to tease apart the
potential functional relevance of this unusual
binding behavior and to reveal NPC biology
at the level of detail comparable with that of
other central problems in cell biology. jR EFERENCES AND NOTES- B. Hampoelz et al., Annu. Rev. Biophys. 48 , 515 (2019).
- D. H. Lin, A. Hoelz, Annu. Rev. Biochem. 88 , 725 (2019).
- J. Fernandez-Martinez, M. P. Rout, Trends Biochem. Sci.
46 , 595 (2021). - C. E. Wing, H. Y. J. Fung, Y. M. Chook, Nat. Rev. Mol. Cell
Biol. 23 , 307 (2022). - C. J. Bley, A. Hoelz, Science 376 , eabm9129 (2022).
- S. Petrovic, A. Hoelz, Science 376 , eabm9798 (2022).
- S. Mosalaganti, M. Beck, Science 376 , eabm9506
(2022). - X. Zhu, Y. Shi, Science 376 , eabl8280 (2022).
- P. Fontana, H. Wu, Science 376 , eabm9326 (2022).
- H. B. Schmidt, D. Görlich, Trends Biochem. Sci. 41 , 46
(2016). - K. E. Knockenhauer, T. U. Schwartz, Cell 164 , 1162
(2016). - C. W. Akey et al., Cell 185 , 361 (2022).
- A. P. Schuller et al., Nature 598 , 667 (2021).
- C. E. Zimmerli et al., Science 374 , eabd9776 (2021).
ACKNOWLEDGMENTS
T. U.S. is supported by National Institutes of Health grant
R35-GM141834.
10.1126/science.abq4792METABOLISMAligning
mealtimes to
live longer
Calorie restriction, fasting,
and circadian rhythms sync
together for a long, healthy
life in mice
By Shaunak Deota and Satchidananda PandaC
alorie restriction (CR) involves
chronic reduction of energy intake by
20 to 40% without inducing malnutri-
tion ( 1 ). CR extends life span in mul-
tiple animal models and reduces the
risk of age-associated disorders, most
of which arise from metabolic dysfunction
and inflammation. However, extended daily
fasting or aligning daily meal timing to the
active period, even without reducing energy
intake, can also improve health and increase
life span in model organisms. On page 1192
of this issue, Acosta-Rodríguez et al. ( 2 ) re-
veal the specific contribution of fasting and
timing of calorie-reduced meals to the effi-
cacy of CR, as estimated by life-span exten-
sion in male mice.
In most rodent CR studies, the control
group is fed ad libitum (AL), whereas the
CR animals are fed a single meal per day
that contains ~20 to 40% fewer calories than
the AL group consumes. The CR animals eat
most of their daily ration within 2 hours ( 3 ).
Thus, CR studies inadvertently introduce
timed or time-restricted feeding and pro-
longed daily fasting, both of which can im-
prove health and delay aging independently
of CR ( 4 ). Disentangling the effect of CR, fast-
ing, and time of feeding on rodent life span is
not easy. Acosta-Rodríguez et al. developed a
system that automatically delivers a specific
quantity of food in a bolus, as in standard CR
studies, or in small meals at specific times ( 3 ).
All mice were housed under 12 hours of light
and dark to synchronize circadian rhythms
(internal body clock), and all cages were
equipped with wheels that measured volun-
tary wheel-running.
To assess the impact of calories alone on
life span, they split the CR ration (30% re-
duced calories relative to the AL group) intoThe Salk Institute for Biological Studies, La Jolla, CA, USA.
Email: [email protected]10 JUNE 2022 • VOL 376 ISSUE 6598 1159Newly resolved components of the human nuclear
pore complex include the luminal ring (orange),
cytoplasmic filaments (yellow), and lipid membrane
(white). The structure known in 2016 is in blue.