25 FEBRUARY 2022 • VOL 375 ISSUE 6583 817GRAPHIC: K. FRANKLIN/
SCIENCE
SCIENCE science.orgIndeed, DORAs are contraindicated in NT1
owing to an increased risk of cataplexy ( 8 ).
However, this leads to questions about the
mechanisms underlying disrupted sleep in
other pathologies.
Like aging, Alzheimer’s disease is com-
monly accompanied by disrupted sleep, as
well as hypocretin neuron loss ( 10 ) and in-
creased hypocretinergic tone ( 11 , 12 ). It will
thus be important to determine whether ag-
ing and Alzheimer’s disease share KCNQ2/3
channel-induced hypocretin neuron hyper-
activity as a common mechanism. NREM
sleep is important for the clearance of toxic,
misfolded brain peptides and proteins ( 13 )
such as b-amyloid (Alzheimer’s disease), tau
(tauopathies), and a-synuclein [Parkinson’s
disease ( 14 )]. Because these products can also
directly disrupt sleep, hypnotic-enhanced
NREM sleep may further enhance sleep qual-
ity and delay disease progression by reducing
their accumulation ( 14 , 15 ). However, clas-
sical hypnotics are unlikely to unlock such
benefits. These broad-spectrum inhibitors of
neuronal activity can induce cognitive com-
plaints and falls and are thus contraindicated
in such disorders and in older people in gen-
eral. DORAs, by contrast, have a better side-
effect profile in these domains ( 8 ). Together
with Li et al.’s evidence of improved recogni-
tion memory with flupirtine in aged mice,
selective targeting of hypocretinergic mecha-
nisms may prove superior in reaping the re-
wards of pharmacologically enhanced sleep
in older people in both health and disease.
The discovery by Li et al. reveals an age-
dependent mechanism underlying sleep
disruption and thus reopens the search for
new, targeted strategies to combat sleep
disturbances in older people, and future
research has the potential to expand to
neurodegenerative disorders. If translated
clinically, these findings pave the way for
the development of sleep medications that
specifically dampen hypocretin neuron hy-
peractivity to improve sleep and mental and
physical health in older people. jREFERE NCES AND NOTES
1. D. K o c e v s k a et al., Nat. Hum. Behav. 5 , 113 (2021).- B. A. Mander, J. R. Winer, M. P. Walker, Neuron 94 , 1 9 ( 2 0 1 7 ).
- C. J. Lowe et al., Neurosci. Biobehav. Rev. 80 , 586 (2017).
- D. J. Foley et al., Sleep 18 , 425 (1995).
- S.-B. Li et al., Science 375 , eabh3021 (2022).
- T. E. Scammell et al., Neuron 93 , 747 (2017).
- C. L. A. Bassetti et al., Nat. Rev. Neurol. 15 , 519 (2019).
- D. Hoyer et al., Br. J. Clin. Pharmacol. 86 , 244 (2020).
- N. J. Hunt et al. Neurobiol. Aging 36 , 292 (2015).
- J. Oh et al., Alzheimers Dement. 15 , 1253 (2019).
- Y. A. Dauvilliers et al., Front. Aging Neurosci. 6 , 119 (2014).
- C. Liguori et al., JAMA Neurol. 71 , 1498 (2014).
- M. Nedergaard, S. A. Goldman, Science 370 , 50 (2020).
- M. M. Morawska et al., Sci. Transl. Med. 13 , eabe7099
(2021). - B. P. Lucey, Neurobiol. Dis. 144 , 105031 (2020).
ACKNOWLEDGMENTS
L.H.J. has consulted for Eisai Co., Ltd.
10.1126/science.abo182 2GENETICSInferring human
evolutionary history
Unified genetic genealogy improves our understanding
of how humans evolved
By J a s m i n R e e s1,2 and Aida Andrés1,2G
enomes are invaluable tools for in-
ferring the demographic and adap-
tive history of human populations,
including migrations, population
splits, admixture, and genetic adap-
tations. Growing datasets of modern
and ancient genomes make this possible,
but their massive size comes with impor-
tant challenges, demanding techniques
that analyze immense amounts of data in
reasonable amounts of time while using as
much information as possible. Combining
genomes from different datasets poses per-
haps an even greater challenge, especially
when it comes to integrating ancient and
modern genomes. On page 836 of this issue,
Wohns et al. ( 1 ) report surmounting some
of these challenges to construct the largest
human genealogy to date, integrating mod-
ern and ancient genomes from multiple da-
tasets to infer key events in human history
together with their timings and geographi-
cal locations.
Phylogenetic trees are used to represent
the evolutionary or genetic relationshipsamong species. Similar trees can represent
the relationships among individuals within
a species, analogous to how family gene-
alogies represent relationships between
family members. However, because of the
effects of recombination, each locus has a
slightly different evolutionary history and
therefore a different tree. Tree-recording
methods ( 2 ) can infer these trees along the
genome, with each tree representing what
can be considered a nearly complete his-
tory of the locus. Consequently, tree record-
ing is superior to most classical methods,
which condense complicated evolutionary
patterns in relatively simple summary
statistics. By combining trees across the
genome, a theoretical genealogy can be
generated that embodies the genetic rela-
tionships among sampled individuals and
their inferred ancestors, within and across
populations. If the sampling of genomes
was sufficiently comprehensive, such gene-
alogy would in theory represent relation-
ships over the entire species, capturing the
genetic history of modern humans today.
Owing to a series of impressive theoret-
ical and computational advances, geneticYears agoAdding ancient DNAModern human samples Inferred ancestors
Ancient humansModern humans
Archaic humansHuman genealogy reconstruction and geographical inference
By building the tree from modern human and high-quality ancient genomes, one can infer ancestral human
relationships. Using additional ancient human samples to help infer the ages of alleles, Wohns et al. built a
unified genealogy that also includes the geographical location of inferred ancestors, which gives information
about populations and their migrations through human history.