6–10 month-old infants viewed faces
with dynamic eye gaze directed either
towards them or away from them.
Approximately 18 to 30 months later,
these children were clinically evaluated
for the presence of an ASD. Strikingly,
neural responses to dynamic eye gaze
shifts during the first year predicted
clinical outcomes at 36 months,
despite similar patterns of gaze as
measured by eye tracking. The authors
 conclude that ERP responses to eye
gaze in the first year of life reflect
developmental processes leading to
the later emergence of ASD.
As the field strives for earlier
methods of detecting autistic
development, these remarkable
findings offer hope for future clinical
practice, suggesting the possibility of
non-invasive, brain-based screening
methods that could detect differences
prior to behavioral emergence. ERPs
are collected with the same technology
used in hospitals around the world for
universal auditory screening of
newborns; the infrastructure might
already be in place to implement
population-based screening in an
affordable and highly efficient manner
. Of course, prior to realization of
such clinical benefits, it will be critical
to investigate the specificity of this
biomarker to autism, its presence in an
unselected, population-based sample,
and, most importantly, its viability in
individual patient data. Given historical
difficulty parsing heterogeneity in ASD,
these findings suggest the potential
power of systems neuroscience
approaches to identify meaningful
subtypes of ASD to inform treatment
and predicting outcome. We envision
a strategy of deep behavior and brain
phenotyping over longitudinal
development to offer a detailed profile
of brain–behavior performance for
a given individual for the purpose of
detection of atypical development,
subcategorization (for example, for
genetic analysis), treatment selection,
and prediction of treatment response.
There is an important historical
perspective to be noted here.
Elsabbagh and colleagues  focused
their analyses on the P1, N290 and
P400 components of the ERP signal,
components that are modulated in
a number of face perception tasks,
including tests of sensitivity to the
direction of eye gaze in infants as
young as four months . Critically,
experiments from a number of
laboratories around the world have
identified these ERP components in
infants as precursors of the
well-established face-sensitive N170
component in adults . Just over
sixteen years ago, in the Yale
Neuropsychology Laboratory, Gregory
McCarthy, Shlomo Bentin, and their
colleagues first described the ‘N170’:
while recording from scalp electrodes
in typically developing adult
volunteers, they discovered that
human faces and face parts (especially
the eyes) reliably evoked a negative
ERP at 172 ms (range 130–200 ms) that
they labeled the N170. This response
was absent from the ERPs elicited by
many other animate and inanimate
non-face objects, and was maximal
over occipitotemporal electrode sites.
This work, coupled with numerous
behavioral findings concerning
face-processing deficits in ASD, led
researchers (including an author of this
dispatch) to study the N170 in children,
adolescents, and adults with ASD .
This set of events, from the basic
science discovery of a neural signature
for face processing in the human brain
to its translation into a potential
biomarker for the emergence of ASD
represents the very finest in the
emerging field of translational
developmental social neuroscience.
1. American Psychiatric Association (2000).
Diagnostic and Statistical Manual of Mental
Disorders: DSM-IV-TR, 4th ed. (Washington,
District of Columbia: American Psychiatric
2. Kanner, L. (1943). Autistic disturbances of
affective contact. Nervous Child 2, 217–250.
3. Rogers, S.J. (2009). What are infant siblings
teaching us about autism in infancy? Autism
Res. 3, 125–137.
4. Elsabbagh, M., Mercure, E., Hudry, K.,
Chandler, S., Pasco, G., Charman, T., Pickles, A.,
Baron-Cohen, S., Bolton, P., Johnson, M.H., and
The BASIS Team (2012). Infant neural sensitivity
to dynamic eye gaze is associated with later
emerging autism. Curr. Biol. 22, 338–342.
5. McPartland, J.C., Coffman, M., and Pelphrey, K.
(2011). Recent advances in understanding the
neural bases of autism spectrum disorder. Curr.
Opin. Ped. 23, 628–632.
6. de Haan, M., Johnson, M.H., and Halit, H. (2003).
Development of face-sensitive event-related
potentials during infancy: a review. Int. J.
Psychophysiol. 51, 45–58.
7. Bentin, S., Allison, T., Puce, A., Perez, E., and
McCarthy, G. (1996). Electrophysiological
studies of face perception in humans. J. Cogn.
Neurosci. 8, 551–565.
8. McPartland, J., Dawson, G., Webb, S.J.,
Panagiotides, H., and Carver, L.J. (2004).
Event-related brain potentials reveal anomalies
in temporal processing of faces in autism
spectrum disorder. J. Child Psychol. Psych. 45,
Yale Child Study Center, Yale University
School of Medicine, 230 South Frontage
Road, New Haven, CT 06519, USA.
Mitosis: Short-Circuiting Spindle
The spindle checkpoint forms an intricate signaling circuit to sense unattached
kinetochores, to inhibit the anaphase-promoting complex/cyclosome (APC/C),
and to delay anaphase onset. Using clever genetic experiments in the budding
yeast, Lau and Murray define the endpoint of checkpoint signaling and provide
key mechanistic insights into checkpoint inhibition of APC/C.
Xuelian Luo1,* and Hongtao Yu1,2,*
The spindle checkpoint is a cell-cycle
surveillance system that guards
against chromosome missegregation
in mitosis and meiosis [1,2].
Dysregulation of the spindle
checkpoint can result in aneuploidy
and cancer predisposition. The
molecular dissection of this checkpoint
began two decades ago with yeast
genetic studies that identified Mad
(mitotic arrest deficient) and Bub
(budding uninhibited by benomyl)
proteins as key checkpoint
biochemical, and structural studies in
multiple organisms from yeast to man
then delineated a general framework of
how the spindle checkpoint
In this framework, checkpoint
proteins are recruited to kinetochores
Current Biology Vol 22 No 4
not properly attached to the mitotic
spindle in a hierarchical fashion. At the
kinase complexes: the Aurora
B-containing chromosome passenger
complex and the Bub1–Bub3 complex.
These kinases are required for the
proper recruitment of Mad3/
BubR1–Bub3 (BubR1 is the vertebrate
ortholog of Mad3) and Mps1, which in
turn are required for recruiting
Mad1–Mad2 and the checkpoint target
Cdc20 to the kinetochores. At the
kinetochores, Bub1 and Mps1 undergo
activation while Mad2 undergoes
conformational activation. The
activated checkpoint proteins promote
the formation of the mitotic checkpoint
complex (MCC) containing
Mad3/BubR1, Bub3,Mad2, and Cdc20,
which inhibits the activity of the
multisubunit ubiquitin ligase APC/C.
APC/C inhibition stabilizes its
substrates cyclin B and the separase
inhibitor securin, delaying
A key unresolved issue in spindle
checkpoint signaling is whether MCC
formation is sufficient to inhibit APC/C
in a physiological setting. Because
MCC could be detected in interphase
mammalian cells and in yeast cells
lacking functional kinetochores [6,7], it
has been proposed that an active
spindle checkpoint might additionally
modify MCC or APC/C or both in
the ability to inhibit APC/C. Using
clever genetic experiments in yeast,
as reported in a recent issue of
Current Biology, Lau and Murray 
now show that MCC formation is the
endpoint of checkpoint signaling.
Forced formation of this complex or
even a Mad2–Cdc20 sub-complex is
sufficient to block APC/C activation in
the absence of functional kinetochores
or other upstream checkpoint proteins.
Lau and Murray began their study by
testing the cellular effects of tethering
Mad2 and Mad3 (two MCC subunits)
either covalently through the
construction of Mad2–Mad3 fusion
proteins or non-covalently through
fusing each toleucine zippers knownto
form heterodimers . Both ways of
tethering Mad2 and Mad3 in cells
produced metaphase arrest, as
evidenced by the accumulation of
of the APC/C substrate Pds1 (the yeast
securin). The metaphase arrest exerted
by the Mad2–Mad3 fusion was still
observed in cells deficient of the core
kinetochore component Ndc10,
suggesting that the cellular effects of
the Mad2–Mad3 fusion were
independent of functional
kinetochores. The Mad2–Mad3 fusion
also arrested at metaphase cells
deleted of non-essential checkpoint
proteins, including Mad1, Mad2, Mad3,
Bub1, and Bub3. The metaphase arrest
caused by the Mad2–Mad3 fusion did
essential checkpoint kinases Ipl1 (the
yeast Aurora B) and Mps1, as chemical
inhibition of Mps1 and Ipl1 did not
affect Pds1 stabilization in cells
expressing the Mad2–Mad3 fusion.
These results strongly suggest that the
Mad2–Mad3 fusion directly binds to
Cdc20 and inhibits APC/C in cells. By
inference, the forced formation of the
Mad2–Mad3–Cdc20 complex in cells is
sufficient to inhibit APC/C, in a way that
is independent of upstream checkpoint
signaling. As Mad3 forms a constitutive
complex with Bub3, a fraction of the
tethered Mad2–Mad3–Cdc20 complex
likely contains Bub3 to form the intact
MCC. Because the Mad2–Mad3 fusion
still arrests Bub3-null cells at
metaphase, Bub3 in this complex is not
required for APC/C inhibition,
consistent with previous in vitro
biochemical studies .
Lau and Murray went on to show that
tethering Mad2 and Cdc20 with the
heterodimerizing leucine zippers was
sufficient to induce metaphase arrest
. When the Mad2-zipper protein was
overexpressed (driven by the GAL1
promoter), this metaphase arrest was
largely independent of Mad1, Mad2,
Mad3, Bub1, and Bub3, and did not
require the kinase activities of Mps1 or
Ipl1. Tethering Mad3 and Cdc20 with
the same approach did not produce
metaphase arrest, although it remained
to be tested whether tethering
Mad3 and Cdc20 compromised the
APC/CCdc20-inhibitory activity of
Mad3 in vitro. This caveat
notwithstanding, the sufficiency of the
tethered Mad2–Cdc20 complex to
produce metaphase arrest strongly
suggests that binding of Mad2 to
Cdc20 is a critical downstream event in
checkpoint signaling. Deletion of Mad3
did, however, diminish the extent of
metaphase arrest exerted by tethering
Mad2 and Cdc20 when Mad2-zipper
was overexpressed. If Mad2-zipper
were to be expressed at a lower level
with the native MAD2 promoter, the
metaphase arrest caused by tethering
Mad2 and Cdc20 might be more
dependent on Mad3. Therefore, their
results are consistent with the notion
that Mad2 and Mad3 cooperatively
The underlying reason for the
synergy between Mad2 and Mad3/
BubR1 in APC/C inhibition is not
B bR1 BubR1
Substrate binding blocked
Figure 1. Proposed mechanisms by which Mad2 and Mad3/BubR1 cooperate to inhibit
Mad2 alters the binding mode between APC/C and Cdc20, non-competitively and partially
inhibiting APC/CCdc20. Mad3/BubR1 competitively blocks substrate binding to APC/CCdc20
and partially inactivates it. Mad2 and Mad3/BubR1 mutually enhance each other’s binding
to Cdc20, engaging both mechanisms of APC/C inhibition and cooperatively inhibiting APC/C.
understood, but may stem from the
different mechanisms used by the
two proteins to inhibit APC/C
(Figure 1). Mad3/BubR1 contains
APC/CCdc20-binding motifs that are
commonly found in APC/C substrates
and competitively blocks substrate
binding to APC/CCdc20[10–12]. The
mechanism by which Mad2 inhibits
APC/CCdc20is unclear, but it does not
directly block substrate binding to
Cdc20. Instead, Mad2 may alter the
APC/C-binding mode of Cdc20 and
anchor Cdc20 to a site on APC/C that is
different from the catalytically
functional Cdc20-binding site [13,14].
The current evidence thus suggests
that Mad3/BubR1 and Mad2
cooperatively inhibit APC/CCdc20by
blocking substrate access and by
sequestering Cdc20 in a catalytically
compromised location on APC/C.
Tethering Mad2 to Cdc20 might be
sufficient to sequester Cdc20 away
from the active site in a process that
does not strictly require Mad3, thus
inhibiting APC/C and producing
The Lau and Murray study  also
sheds light on the mechanism of Mad2
conformational activation in cells.
Mad2 is an unusual two-state protein
with two native folds, termed
open/N1-Mad2 (O-Mad2) or
closed/N2-Mad2 (C-Mad2) [15,16].
O-Mad2 in the cytosol is recruited to
kinetochores through an O–C-Mad2
asymmetric dimerization event with
the kinetochore-bound Mad1–C-Mad2
core complex and is converted to an
active intermediate Mad2 (I-Mad2)
conformer, which then binds Cdc20 to
form the Cdc20–C-Mad2 complex
(Figure 2). Results from the tethering
experiments by Lau and Murray
support this Mad2 conformational
activation model. It has been proposed
that Cdc20–C-Mad2 can further recruit
and activate O-Mad2 to form more
copies of Cdc20–C-Mad2, thus
propagating checkpoint signal away
from the kinetochores  (Figure 2).
The Lau and Murray data are
inconsistent with this notion of
self-propagation by Cdc20–C-Mad2.
They showed that expression of
untagged Cdc20 driven by the native
CDC20 promoter relieved the
metaphase arrest induced by tethering
Mad2 to Cdc20. If the Cdc20-bound
C-Mad2 could further activate O-Mad2,
the tethered Cdc20–C-Mad2 complex
is expected toactivatethe endogenous
O-Mad2 in the cell, leading to the
inhibition of untagged Cdc20. The fact
that expression of untagged Cdc20
overcomes the metaphase arrest
induced by tethering Mad2 and Cdc20
thus argues against a role of
self-propagation by Cdc20–C-Mad2 in
In summary, by short-circuiting
spindle checkpoint signaling with
clever genetic manipulations in yeast,
Lau and Murray have defined the
endpoint of checkpoint signaling and
provided key insights into APC/C
inhibition by checkpoint proteins. This
approach will be useful in analyzing the
circuitry formed by upstream
checkpoint components in yeast.
The Mad2–Mad3 fusion and the
tethered Mad2–Cdc20 complex may
prove to be valuable tools for future
biochemical and structural studies on
1. Bharadwaj, R., and Yu, H. (2004). The spindle
checkpoint, aneuploidy, and cancer. Oncogene
2. Musacchio, A., and Salmon, E.D. (2007). The
spindle-assembly checkpoint in space and
time. Nat. Rev. Mol. Cell Biol. 8, 379–393.
3. Li, R., and Murray, A.W. (1991). Feedback
control of mitosis in budding yeast. Cell 66,
4. Hoyt, M.A., Totis, L., and Roberts, B.T. (1991).
S. cerevisiae genes required for cell cycle
arrest in response to loss of microtubule
function. Cell 66, 507–517.
5. Yu, H. (2007). Cdc20: a WD40 activator for a cell
cycle degradation machine. Mol. Cell 27, 3–16.
6. Sudakin, V., Chan, G.K., and Yen, T.J. (2001).
Checkpoint inhibition of the APC/C in HeLa
cells is mediated by a complex of BUBR1,
BUB3, CDC20, and MAD2. J. Cell Biol. 154,
7. Fraschini, R., Beretta, A., Sironi, L.,
Musacchio, A., Lucchini, G., and Piatti, S.
(2001). Bub3 interaction with Mad2, Mad3 and
Cdc20 is mediated by WD40 repeats and does
not require intact kinetochores. EMBO J. 20,
8. Lau, D.T., and Murray, A.W. (2012). Mad2 and
Mad3 cooperate to arrest budding yeast in
mitosis. Curr. Biol. 22, 180–190.
9. Tang, Z., Bharadwaj, R., Li, B., and Yu, H.
(2001). Mad2-independent inhibition of
APCCdc20 by the mitotic checkpoint protein
BubR1. Dev. Cell 1, 227–237.
10. King, E.M., van der Sar, S.J., and
Hardwick, K.G. (2007). Mad3 KEN boxes
mediate both Cdc20 and Mad3 turnover, and
are critical for the spindle checkpoint. PLoS
ONE 2, e342.
11. Burton, J.L., and Solomon, M.J. (2007). Mad3p,
a pseudosubstrate inhibitor of APCCdc20 in the
spindle assembly checkpoint. Genes Dev. 21,
12. Lara-Gonzalez, P., Scott, M.I., Diez, M., Sen, O.,
and Taylor, S.S. (2012). BubR1 blocks substrate
recruitment to the APC/C in a KEN-box-
dependent manner. J. Cell Sci. doi:10.1242/
13. Herzog, F., Primorac, I., Dube, P., Lenart, P.,
Sander, B., Mechtler, K., Stark, H., and
Peters, J.M. (2009). Structure of the
interacting with a mitotic checkpoint complex.
Science 323, 1477–1481.
14. Izawa, D., and Pines, J. (2011). How APC/
C-Cdc20 changes its substrate specificity in
mitosis. Nat. Cell Biol. 13, 223–233.
15. Luo, X., and Yu, H. (2008). Protein
metamorphosis: the two-state behavior of
Mad2. Structure 16, 1616–1625.
16. Mapelli, M., and Musacchio, A. (2007). MAD
contortions: conformational dimerization
boosts spindle checkpoint signaling. Curr.
Opin. Struct. Biol. 17, 716–725.
1Department of Pharmacology,2Howard
Hughes Medical Institute, University of Texas
Southwestern Medical Center, 6001 Forest
Park Road, Dallas, TX 75390, USA.
Self-propagation of Cdc20–C-Mad2 Self-propagation of Cdc20–C-Mad2
O Mad2O Mad2
Figure 2. Conformational activation of Mad2 in the spindle checkpoint.
The kinetochore-bound Mad1–C-Mad2 core complex recruits cytosolic O-Mad2, converts it to
the high-energy I-Mad2 conformer, and promotes the formation of the Cdc20–C-Mad2
complex. It has been suggested that Cdc20-bound C-Mad2 can further recruit and convert
O-Mad2 to I-Mad2, which can then form more Cdc20–C-Mad2 complexes (see shaded box).
This self-propagation of the Cdc20–C-Mad2 complex is not supported by the Lau and Murray
study , as a tethered Cdc20–C-Mad2 complex fails to induce untimely binding between the
endogenous O-Mad2 and Cdc20 proteins in yeast cells.
Current Biology Vol 22 No 4