30 JANUARY 2015 • VOL 347 ISSUE 6221 477 SCIENCE sciencemag.org
lation of FRQ promotes its rapid degradation
through its association with a protein called
F-box and WD40 repeat–containing protein
1 (FWD-1). FWD-1 is the substrate-recruiting
subunit of an E3-ubiquitin ligase that supports FRQ ubiquitination and proteasomal
degradation. If degradation of FRQ is an essential step for its inactivation (as the prevailing model predicts), circadian rhythms
should be lost in Neurospora strains lacking
FWD-1. At first glance, this indeed seems to
be the case: In a mutant strain lacking FWD-
1, degradation of FRQ is severely impaired
and circadian rhythms of asexual spore formation (a convenient readout of the
Neurospora clock) are absent. However, Larrondo
et al. looked more closely at the molecular
clockwork using sensitive bioluminescence-based reporters and detected, surprisingly,
rather normal rhythms of FRQ expression. Thus, negative element turnover does
not seem to be a critical step for circadian
Why was degradation of the negative element thought to be required for circadian
rhythm generation? Probably because previous studies reported that the degree of FRQ
phosphorylation determines FRQ stability,
and FRQ stability determines circadian period. For example, in strains with long-period frq alleles (thus making the length of
a circadian cycle more than 24 hours), FRQ
protein is more stable, whereas in short-period strains, FRQ is less stable. Larrondo
et al. suggest that these are not cause-and-effect relations but just correlations. The authors could (pharmacologically) manipulate
the degree of FRQ phosphorylation without
affecting FRQ stability. In addition, more
stable FRQ does not always affect the circadian period. New FRQ can be synthesized
while old, hyperphosphorylated FRQ is still
around (3). It seems that the (
phosphorylation) quality of FRQ rather than its quantity
is important; it does not matter whether the
degradation of hyperphosphorylated FRQ
takes place because the circadian oscillator becomes blind toward old FRQ. This is
consistent with the inability of hyperphosphorylated FRQ to recruit CK1a (4) as well
as its weak binding to WCC (5), conditions
that make it an inefficient transcriptional
Can this circadian model be generalized
to other species? Indeed, several findings
from the fruit fly Drosophila melanogaster
and mammals hint at a less important role
of negative element destruction than previ-
ously assumed. In Drosophila, mutations in
the negative element Period (PER) that in-
terfere with binding to the E3-ligase subunit
Slimb (and thereby with PER degradation)
lead to hyperphosphorylated PER even at
times when new hypophosphorylated PER is
already synthesized (6). Moreover, constitu-
tive overexpression of Cryptochrome (CRY),
the strongest negative element of the mam-
malian oscillator, does not lead to long peri-
ods (although it does dampen the strength of
the rhythms) (7). Also, mice deficient for the
F-box/LRR-repeat protein 3 (FBXL3), an E3-
ligase subunit that is necessary for CRY pro-
tein degradation, display normal circadian
periods yet large amounts of CRY, but only
when a second clock gene (the nuclear hor-
mone receptor REV-ERBα, the inactivation
of which results in a rather normal period) is
also deleted (8).
Nevertheless, it is probably premature to
revise the prevailing clock model—at least
for animal clocks. Much evidence indicates
that perturbing the degradation of negative elements PER and CRY leads to drastic
period changes (1). Whether this is just a
correlation rather than cause-and-effect relation, as Larrondo et al. suggest, needs to
be carefully investigated. For example, it will
be necessary to unambiguously tease apart
which phosphorylation events regulate clock
speed and which events just target negative
elements for degradation without affecting
circadian period. This is likely a difficult endeavor given the dozens of phosphorylation
sites (and clusters) in FRQ, PER, and CRY
proteins (5, 9).
Our understanding of the circadian clock
appears to be moving toward a mechanism
of less evolved organisms—that is, posttranslationally controlled oscillators—but with
questions about the role of (theoretically
dispensable) degradation in its operation.
In cyanobacteria, a pure phosphorylation-based mechanism (a “phoscillator”) is at
the heart of the clock (10), and in mammals,
oscillators have been proposed (11). Individuals with extremely early chronotypes can
experience a 4- to 6-hour phase advance in
sleep-wake behavior, attributed to the mutation of a single phosphorylation site in a
clock constituent (9, 12). It may be that a
phosphorylation defect in a clock protein is
what keeps you tired in the morning. ■
1. S. A. Brown, E. Kowalska, R. Dallmann, Dev. Cell 22, 477
2. L. F. Larrondo, C. Olivares-Yañez, C. L. Baker, J. J. Loros, J. C.
Dunlap, Science 347, 1257277 (2015).
3. J. Cha, H. Yuan, Y. Liu, J. Biol. Chem. 286, 11469 (2011).
4. L. Lauinger, A. Diernfellner, S. Falk, M. Brunner, Nat.
Commun. 5, 3598 (2014).
5. C. L. Bakeretal . Mol.Cell 34, 354 (2009).
6. J. C. Chiuet al ., Genes Dev. 22, 1758 (2008).
7. H. R. Ueda etal ., Nat.Genet. 37, 187 (2005).
8. G. Shi etal ., Proc.Natl.Acad.Sci.U.S.A. 110, 4750 (2013).
9. K. Vanselow et al., Genes Dev. 20, 2660 (2006).
10. M.Nakajima et al., Science 308,414(2005).
11. J. S. O’Neill, A. B. Reddy, Nature 469, 498 (2011).
12. K.L. Toh et al., Science 291,1040(2001).
Regardless of cultural background and mathematical training, all humans have an intuitive sense of numerical magnitude (numerosity). We share with various nonhuman animals the ability to discriminate among different sets of quantities (1), but one aspect of
number processing is commonly assumed
to be uniquely human (2): the consistent
mapping of increasing quantities along
the horizontal extension of space—that is,
the construction of a mental number line.
On page 534 of this issue, Rugani et al. (3)
show that 3-day-old domestic chicks (see the
photo) associate small numerosities with the
left side, and large ones with the right side,
of a given space (see the figure, panel A). The
results show that newborn chicks can understand both relative and absolute quantities. They also suggest that the brain may be
prewired in how it relates numbers to space.
The authors first trained the chicks to find
food behind a central panel marked with five
dots. In the next test phase, the birds were
placed before two laterally placed panels.
When each panel showed two elements, the
birds searched behind the left of the two
panels; when each panel showed eight elements, they searched behind the right. In
Chicks with a
Far from bird-brained. Rugani et al. report elegant
experiments investigating the sense of numerical order
in 3-day-old domestic chicks.
By Peter Brugger
Chicks and humans map
numbers to space in a