compartments, Fzd6 and Vangl2
are found on distinct vesicular
compartments in vivo, indicating that
antero-posterior core PCP asymmetry
is preserved during mitosis. How is this
endosomal asymmetry maintained?
Devenport et al.  report that, in
cultured keratinocytes, mitotically
internalized Celsr1 colocalizes with
various endocytic markers, including
Rab5- and EEA1-positive early
endosomes, Rab11-positive recycling
endosomes, as well as caveolin .
This observation raises the possibility
that anterior and posterior cognate
PCP complexes could follow distinct
endocytic routes to be targeted to
distinct classes of endosomes, thereby
preventing the different PCP
complexes from mixing in mitosis.
During cytokinesis, Celsr1-positive
compartments are distributed in
a polarized manner at the anterior
and posterior poles of daughter cells
(Figure 1C). Strikingly, endosomal
vesicles are shown to interpret
antero-posterior cues independently
of mitotic spindle orientation. Stunning
mosaic experiments revealed that,
at this stage, polarisation of the
Celsr1-positive endosomes is dictated
in a cell-non-autonomous manner
by the interphasic neighboring
PCP-polarized cells . Whether
and how endosomes containing the
anterior or posterior PCP complex
selectively recognize and fuse with
the respective cognate anterior or
posterior cortex remains unknown.
It will also be interesting to understand
how polarized endosomal recycling
drives PCP re-establishment at the
boundaries of the two daughter cells
(Figure 1D). In addition, future studies
will assist our understanding of how
PCP complexes from neighboring
interphasic cells are maintained at the
boundaries of mitotic cells.
What are the underlying molecular
mechanisms and the biological
relevance of selective mitotic
internalization? Using a series of
domain swapping and point mutation
experiments, Devenport et al.  reveal
that a single juxtamembrane di-leucine
signal present in the cytoplasmic
domain of Celsr1 is necessary to
promote its mitotic internalization.
Importantly, in clones of cells
expressing the endocytic-defective
version of Celsr1, hair follicles are
no longer aligned along the
antero-posterior axis. Mutant cells
align one relative to the other,
a misorientation that is transmitted
in a dominant cell-non-autonomous
manner to adjacent wild-type cells .
These observations first strongly argue
that mitotic internalization of PCP
components is physiologically
important and second lead to the
proposal that mitotic uptake occurs
to prevent PCP signaling from the
rounded cell, therefore avoiding
disruption of PCP by aberrant
directional information. Is this
mechanism evolutionarily conserved?
Perhaps not, given that mitotic
internalization of PCP components has
not been reported in Drosophila [17,18]
and mitotic internalization motif of
Celsr1 is not conserved in dipters .
How then is PCP transmitted in
daughter cells inthefly?Clearly, further
investigation of mitotic endocytosis
of PCP components in model systems
will provide new and exciting insights
into how polarized trafficking allows
inheritance of PCP in tissues.
1. Adler, P.N. (2002). Planar signaling and
morphogenesis in Drosophila. Dev. Cell 2,
2. Lawrence, P.A., Struhl, G., and Casal, J. (2007).
Planar cell polarity: one or two pathways? Nat.
Rev. Genet. 8, 555–563.
3. Seifert, J.R., and Mlodzik, M. (2007). Frizzled/
PCP signalling: a conserved mechanism
regulating cell polarity and directed motility.
Nat. Rev. Genet. 8, 126–138.
4. Zallen, J.A. (2007). Planar polarity and tissue
morphogenesis. Cell 129, 1051–1063.
5. Strutt, H., and Strutt, D. (2009). Asymmetric
localisation of planar polarity proteins:
Mechanisms and consequences. Semin. Cell
Dev. Biol. 20, 957–963.
6. Devenport, D., and Fuchs, E. (2008). Planar
polarization in embryonic epidermis
of hair follicles. Nat. Cell Biol. 10, 1257–1268.
7. Simons, M., and Mlodzik, M. (2008). Planar cell
polarity signaling: from fly development to
human disease. Annu. Rev. Genet. 42, 517–540.
8. Devenport, D., Oristian, D., Heller, E., and
Fuchs, E. (2011). Mitotic internalization of
planar cell polarity proteins preserves tissue
polarity. Nat. Cell Biol. 13, 893–902.
9. Jones, C., and Chen, P. (2007). Planar cell
polarity signaling in vertebrates. Bioessays 29,
10. Kibar, Z., Vogan, K.J., Groulx, N., Justice, M.J.,
Underhill, D.A., and Gros, P. (2001). Ltap,
a mammalian homolog of Drosophila
Strabismus/Van Gogh, is altered in the mouse
neural tube mutant Loop-tail. Nat. Genet. 28,
11. Curtin, J.A., Quint, E., Tsipouri, V., Arkell, R.M.,
Cattanach, B., Copp, A.J., Henderson, D.J.,
Spurr, N., Stanier, P., Fisher, E.M., et al. (2003).
Mutation of Celsr1 disrupts planar polarity of
inner ear hair cells and causes severe neural
tube defects in the mouse. Curr. Biol. 13,
12. Montcouquiol, M., Rachel, R.A., Lanford, P.J.,
Copeland, N.G., Jenkins, N.A., and Kelley, M.W.
(2003). Identification of Vangl2 and Scrb1 as
planar polarity genes in mammals. Nature 423,
13. Guo, N., Hawkins, C., and Nathans, J. (2004).
Frizzled6 controls hair patterning in mice.
Proc. Natl. Acad. Sci. USA 101, 9277–9281.
14. Strutt, H., Warrington, S.J., and Strutt, D.
(2011). Dynamics of core planar polarity
protein turnover and stable assembly into
discrete membrane subdomains. Dev. Cell 20,
15. Boucrot, E., and Kirchhausen, T. (2007).
Endosomal recycling controls plasma
membrane area during mitosis. Proc. Natl.
Acad. Sci. USA 104, 7939–7944.
16. Levayer, R., Pelissier-Monier, A., and Lecuit, T.
(2011). Spatial regulation of Dia and Myosin-II
by RhoGEF2 controls initiation of E-cadherin
endocytosis during epithelial morphogenesis.
Nat. Cell Biol. 13, 529–540.
17. Aigouy, B., Farhadifar, R., Staple, D.B.,
Sagner, A., Roper, J.C., Julicher, F., and
Eaton, S. (2010). Cell flow reorients the axis
of planar polarity in the wing epithelium of
Drosophila. Cell 142, 773–786.
18. Segalen, M., and Bellaiche, Y. (2009). Cell
division orientation and planar cell polarity
pathways. Semin. Cell Dev. Biol. 20,
CNRS, UMR 6061, Institut Ge ´ne ´tique
et De ´veloppement de Rennes, Universite ´
Rennes 1, UEB, IFR 140, F-35043 Rennes,
Protein Degradation: BAGging Up the
Cells efficiently uncover and degrade proteins that are misfolded. However, we
know very little about what cells do to protect themselves from mislocalized
proteins. A new study reveals a novel quality control pathway that recognizes
and degrades secretory pathway proteins that have failed to target to the
Tslil Ast and Maya Schuldiner*
Have you ever had the dubious
pleasure of finding groceries that
you’ve forgotten to place in the
refrigerator? Holding your breath and
looking away, the only thing left to do is
to promptly throw everything out. In
Current Biology Vol 21 No 18
a similar way, cells need to protect
themselves from ‘spoilt’ proteins that
have failed to make it to their final
cellular destination on time. One large
group of proteins that have to be
targeted to their cellular destination
consists of proteins that either reside in
the secretory pathway, are displayed
on the plasma membrane or are
secreted. Each and every one of these
proteins, which usually make up
manner to the central portal of the
secretory pathway — the endoplasmic
reticulum (ER). Once proteins have
made it into the ER, its lumen provides
a safe folding haven . Although
unsuccessful ER insertion is a
well-documented phenomenon, there
uncovering what happens to those
proteins that never make it into the ER.
Now, in a recent article published in
Nature, Hessa et al.  have elucidated
a novel mechanism by which
secretory proteins that have ‘missed
their train’ and have remained in the
cytosol are identified and sent off for
Translocation into the ER is an
intricate, multistep process that
involves protein targeting to the ER,
gating of the translocon, energy-driven
insertion of the protein through the ER
membrane, and achievement of the
correct orientation of the protein
relative to the membrane. Given the
complexity of this process, it is not
surprising that the efficiencies of ER
translocation differ from protein to
protein and that a substantial fraction
of many ER-bound proteins fail to
translocate. Such translocational
duds remain in the cytoplasm and
constitute a mislocalized protein (MLP)
pool [3,4]. Furthermore, even
proteins that are efficiently
translocated may have difficulties
with insertion during times of ER
stress, whereby translocational
attenuation is induced so as to not
overburden the ER with additional
folding requirements [5,6].
By their nature, MLPs can interfere
with cytoplasmic protein homeostasis.
Their exposed hydrophobic domains
(signal sequences as well as
transmembrane domains not shielded
by the membrane) may easily
aggregate, and therefore MLPs must
such proteins in an efficient and
timely manner can have severe
consequences. Onenotable exampleis
neurodegeneration resulting from the
accumulation of the inefficiently
translocated prion protein PrP .
Furthermore, there are several
examples of MLPs contributing to the
pathogenicity of Alzheimer’s disease
[8,9]. While the degradation of MLPs
has been shown to involve the
proteasome , the exact quality
control machinery that recognizes and
targets these proteins for degradation
has remained elusive.
It was previously thought that
misfolded proteins in the cytosol
would be recognized and degraded
by the general cytoplasmic quality
control machinery. Indeed, it has been
shown that translocation-incompetent
ER proteins can be targeted for
proteasomal degradation via the
which takes part in maintaining
general cytoplasmic folding
homeostasis [10,11]. However, these
studies were carried out on mutated or
abnormal proteins, which lacked
some of the determinants that are
unique to ER proteins — such as
transmembrane domains or signal
sequences that target the proteins to
the ER. It was therefore unclear
whether an additional, unique
molecular pathway existed to enable
recognition and proteasome targeting
Using an in vitro system, Hessa et al.
 reconstituted the translation
the cytoplasmic form of PrP, enabling
them to elucidate both the factors and
the signals that mediate the
degradation of MLPs. They elegantly
demonstrated that efficient
ubiquitination of the cytoplasmic PrP
was dependent both on the release of
the protein from the ribosome (without
having yet been inserted into the ER
membrane) and on the presence of
long linear hydrophobic stretches of
the protein. Specifically, PrP contains
two such hydrophobic domains — an
amino-terminal signal sequence,
which targets it for ER insertion, as
well as a carboxy-terminal sequence,
which facilitates the attachment of
a glycosyl phosphatidylinositol (GPI)
anchor. Deletion of both of these
sequences from PrP markedly reduced
its ubiquitination, while the addition of
these sequences or transmembrane
domains to GFP induced its
ubiquitination. These findings
indicated that it is not merely the
unfolded nature of the MLPs that
targets them for degradation, but
rather the presence of sequences that
are unique to proteins that must enter
the secretory pathway.
By carrying out crosslinking,
fractionation and reconstitution
assays, Hessa et al.  were able to
identify the machinery that carries
out the ubiquitination of cytoplasmic
PrP. Strikingly, they discovered that
MLPs that have been released from
the ribosome are bound by BAG6,
which maintains them in a
(Figure 1). BAG6, also known as BAT3
or Scythe, has been identified to take
part in the post-translational insertion
of tail-anchored (TA) proteins into the
ER [12,13] (Figure 1). Together with
TRC35 and UBL4A, BAG6 makes up
a tri-chaperone ribosome-associated
complex, which binds TA proteins
post-translationally, and loads them
onto the cytosolic ATPase TRC40,
which targets these proteins to the ER
[12,14–16]. However, in the case of
MLPs, BAG6 does not proceed to hand
off the proteins to TRC40, but rather
recruits the ubiquitination machinery
through its amino-terminal UBL
domain. Hessa et al.  demonstrated
that the E2 ubiquitin-conjugating
enzyme that takes part in the
ubiquitination reaction of MLPs is
UBCH5. However, the E3 ubiquitin
ligase that participates in this
ubiquitination remains so far unknown
This mechanism of MLP
identification and ubiquitination that
was initially uncovered in the in vitro
systems seems to hold true in vivo.
Specifically, the cytoplasmic form of
PrP was stabilized in cultured cells that
overexpressed a version of BAG6 that
lacks its UBL domain. This
degradation was dependent on the
presence of hydrophobic stretches
on PrP, such as the signal sequence
and the GPI attachment sequence.
Moreover, these in vivo experiments
showed that BAG6 is a real
multitasker — even tagging TA
proteins for degradation in the
absence of TRC40 (Figure 1).
More generally, these findings
together with previous work [12,16]
suggest that BAG6 can bind terminal
hydrophobic stretches (the carboxyl
the case of signal-sequence-containing
proteins, the GPI attachment sequence
or the amino terminus) and act as
a triage factor. TA proteins are passed
on to TRC40 for ER insertion if
possible and targeted for degradation
when not, while MLPs are always
tagged for degradation. This dedicated
pathway appears to couple
translation and degradation. It ensures
failed to translocate, thereby
precluding their aggregation in the
cytosol as well as the futile recruitment
of cytoplasmic folding machinery.
Moreover, BAG6 may have additional
roles in maintaining the folding
homeostasis of the cytoplasm, as it has
been shown to mediate the
degradation of defective ribosomal
products , which make up about
a third of newly synthesized
Hessa et al.’s  work has answered
a fundamental cell biological question
regarding MLP degradation. Like
it now begs further queries, regarding
both the determinants and the
mechanisms that drive degradation.
For example, co-translationally
targeted proteins bind the signal
recognition particle (SRP) as they
emerge from the ribosome. How then
is the fraction of targeted (SRP-bound)
versus degraded (BAG6-bound)
proteins determined for any given
gene? This seems to be a regulated
and dynamic ratio, indicating that
there are trans-factors that modulate
this balance. In addition, how much
time is a protein given to be inserted
into the ER before it is targeted for
Finally, how is the handoff between
BAG6 and the degradation machinery
carried out? BAG6 has been shown
to bind to poly-ubiquitinated
proteins  — could this triage
protein also chaperone MLPs to the
Together, these findings have
opened an exciting and new field of
pre-insertional quality control. It seems
that life on the secretory export track is
rough — upon exiting the ribosomes,
polypeptides have to hurry and bind
the SRP or be translocated, otherwise
a clearance mechanism kicks in,
rapidly BAGging up the trash, and
ensuring that the cytosol is kept clean
and tidy (Figure 1).
1. Rapoport, T.A. (2007). Protein translocation
across the eukaryotic endoplasmic reticulum
and bacterial plasma membranes. Nature 450,
2. Hessa, T., Sharma, A., Mariappan, M.,
Eshleman, H.D., Gutierrez, E., and Hegde, R.S.
(2011). Protein targeting and degradation are
coupled for elimination of mislocalized
proteins. Nature 475, 394–397.
3. Levine, C.G., Mitra, D., Sharma, A., Smith, C.L.,
and Hegde, R.S. (2005). The efficiency of
protein compartmentalization into the
secretory pathway. Mol. Biol. Cell 16, 279–291.
4. Hegde, R.S., and Kang, S.-W. (2008). The
concept of translocational regulation. J. Cell
Biol. 182, 225–232.
5. Kang, S.-W., Rane, N.S., Kim, S.J.,
Garrison, J.L., Taunton, J., and Hegde, R.S.
(2006). Substrate-specific translocational
attenuation during ER stress defines a
pre-emptive quality control pathway. Cell 127,
6. Oyadomari, S., Yun, C., Fisher, E.A.,
Kreglinger, N., Kreibich, G., Oyadomari, M.,
Harding, H.P., Goodman, A.G., Harant, H.,
Garrison, J.L., et al. (2006). Cotranslocational
degradation protects the stressed endoplasmic
reticulum from protein overload. Cell 126,
7. Rane, N.S., Yonkovich, J.L., and Hegde, R.S.
(2004). Protection from cytosolic prion protein
toxicity by modulation of protein translocation.
EMBO J. 23, 4550–4559.
8. Devi, L., Prabhu, B.M., Galati, D.F.,
Avadhani, N.G., and
Anandatheerthavarada, H.K. (2006).
Accumulation of amyloid precursor protein in
the mitochondrial import channels of human
Alzheimer’s disease brain is associated with
mitochondrial dysfunction. J. Neurosci. 26,
9. Huang, Y.D., Liu, X.Q., Wyss-Coray, T.,
Brecht, W.J., Sanan, D.A., and Mahley, R.W.
(2001). Apolipoprotein E fragments present in
Alzheimer’s disease brains induce
neurofibrillary tangle-like intracellular
inclusions in neurons. Proc. Natl. Acad. Sci.
USA 98, 8838–8843.
10. Lee, D.H., Sherman, M.Y., and Goldberg, A.L.
(1996). Involvement of the molecular chaperone
Ydj1 in the ubiquitin-dependent degradation of
short-lived and abnormal proteins in
Saccharomyces cerevisiae. Mol. Cell. Biol. 16,
11. Park, S.-H., Bolender, N., Eisele, F., Kostova, Z.,
Takeuchi, J., Coffino, P., and Wolf, D.H. (2007).
The cytoplasmic Hsp70 chaperone machinery
subjects misfolded and endoplasmic reticulum
import-incompetent proteins to degradation via
the ubiquitin-proteasome system. Mol. Biol.
Cell 18, 153–165.
12. Mariappan, M., Li, X., Stefanovic, S.,
Sharma, A., Mateja, A., Keenan, R.J., and
Hegde, R.S. (2010). A ribosome-associating
factor chaperones tail-anchored membrane
proteins. Nature 466, 1120–1124.
13. Leznicki, P., Clancy, A., Schwappach, B., and
High, S. (2010). Bat3 promotes the membrane
integration of tail-anchored proteins. J. Cell Sci.
insertion (TA proteins)
Figure 1. A schematic representation of the triage functions of BAG6 in protein translocation
Secretory pathway proteins are thought to enter the endoplasmic reticulum (ER) through three
routes: co- or post-translational translocation dependent on Sec61 (translocon) and Sec61-
independent post-translational translocation. Hessa et al.  show that translating ribosomes
bind both the signal recognition particle (SRP) and BAG6. If the signal sequence is recognized
by SRP, translational halt and co-translational translocation will occur. In the absence of SRP
binding, two types of protein bind BAG6 upon ribosomal release: first, tail-anchored (TA)
proteins require BAG6 to be loaded onto TRC40 for Sec61-independent post-translational
translocation. Second, secretory pathway proteins that have failed to insert into the ER will
bind BAG6. In this scenario, BAG6 recruits the ubiquitination machinery, thereby targeting
these mislocalized proteins for degradation. Some major open questions remain, such as
whether Sec61-dependent post-translational translocating proteins that do not bind SRP
utilize BAG6 either for insertion or for quality control, what is the identity of the E3 ubiquitin
ligase that ubiquitinates BAG6 substrates, and how is the ratio of ER-inserted versus
degraded protein determined for each gene.
Current Biology Vol 21 No 18
14. Schuldiner, M., Metz, J., Schmid, V., Denic, V.,
Rakwalska, M., Schmitt, H.D., Schwappach, B.,
and Weissman, J.S. (2008). The GET complex
mediates insertion of tail-anchored proteins
into the ER membrane. Cell 134, 634–645.
15. Stefanovic, S., and Hegde, R.S. (2007).
Identification of a targeting factor for
posttranslational membrane protein insertion
into the ER. Cell 128, 1147–1159.
16. Jonikas, M.C., Collins, S.R., Denic, V., Oh, E.,
Quan, E.M., Schmid, V., Weibezahn, J.,
Schwappach, B., Walter, P., Weissman, J.S.,
et al. (2009). Comprehensive characterization
of genes required for protein folding in the
endoplasmic reticulum. Science 323,
17. Minami, R., Hayakawa, A., Kagawa, H.,
Yanagi, Y., Yokosawa, H., and Kawahara, H.
(2010). BAG-6 is essential for selective
elimination of defective proteasomal
substrates. J. Cell Biol. 190, 637–650.
18. Schubert, U., Anton, L.C., Gibbs, J.,
Norbury, C.C., Yewdell, J.W., and Bennink, J.R.
(2000). Rapid degradation of a large fraction of
newly synthesized proteins by proteasomes.
Nature 404, 770–774.
Department of Molecular Genetics,
Weizmann Institute of Science,
Rehovot 76100, Israel.
Affective Neuroscience: Tracing the
Trace of Fear
The trace of fear has been elusive and difficult to discern in the human brain.
Researchers have come up with a clever new way to track it down.
‘‘Even though I walk through the valley
of the shadow of death, I will fear no
evil’’ says David in the 23rdPsalm.
Had we scanned David’s brain while
walking through the dark valley of
death, what would we see? Could we
track his fears surface and crumble?
Bach et al.  would say, yes; last
month they reported in The Journal of
Neuroscience a novel way to identify
a stable but sparse fear memory trace
in the human amygdala.
Tracing the trace of fear in the human
brain is not an easy task. Researchers
ofthehuman brainhavelimited access.
They cannot insert electrodes, cut
slices, or inject toxic drugs, unless
a medical condition calls for it. Most
of what we know about the human
brain, we derive from animal studies.
From the middle of the 20thcentury
until this very day [2,3], the conclusions
from animal research are loud and
clear — the amygdala is critical for the
acquisition and expression of fear.
Drawan imaginarylinepassing through
your ear into the brain, and then
another line through your eye: the
amygdala roughly resides where these
has a mechanism to detect and predict
threats; this mechanism is so highly
conserved in evolution that we spot
it essentially everywhere, in a rat
and a human being alike.
The amygdala is in fact
a conglomerate of sub-nuclei, rather
than a cohesive brain region (Figure 1).
Although falling under the same
corporate structure, each of the
sub-nuclei engages in a completely
different business. The lateral nucleus,
for example, is where inputs into the
amygdala converge. The central
nucleus is the output station. In
between lies an island of inhibitory
neurons, the intercalated cells, which
transmit information within the
amygdala, and so forth [4–6].
Invasive techniques, such as
electrophysiological recording, allow
such detailed investigation. In humans,
instead, we use a non-invasive method
called functional magnetic resonance
Imaging by this method divides the
brain volume into spatial units called
voxels, which are analogous to pixels
but in a three-dimensional space.
Voxel sizes typically range between
one to three cubic millimeters. Each
voxel therefore contains millions of
neurons and it does not map onto any
naturally occurring layout of the brain.
The collective activation of those
millions of neurons sums up into one
data point. The fMRI method does
not, in fact, register direct neuronal
firing in each data point; it actually
reflects the impact of the collective
neural activity on nearby blood
vessels. To make things even blurrier,
each brain region contains hundreds
to thousands of voxels. The amygdala,
for example, hosts about 2000 voxels
in each side of the brain. Because of
issues of noise and statistical power,
fMRI studies usually report the
average activation from all these
voxels, or at least, from the few most
active ones. Studies using fMRI also
conventionally perform group analyses
and seldom examine individual
We are facing a double-edged sword
then. On the one hand, fMRI gives us
safe access into the human brain,
producing beautiful maps of the
evolvingactivation from everycorner of
the brain simultaneously. On the other
hand, it does so with poor resolution.
Ten years ago, Haxby et al.  came up
data. Instead of lumping neighboring
voxels together, they inspected
multiple voxels in parallel, to see what
kind of pattern they create. Think of
a group of sixth grade kids. According
to the traditional approach, you would
take the height of each kid, and end up
with a number representing the class’s
average height. Following Haxby’s
approach, you would ask all kids to
stay where are sitting and give each
one a flashlight that beams relatively to
their height: bright yellow to the tallest,
turning orange-red the shorter they are.
You would then look at the colorful
shiny pattern they create when they
all turn on their flashlights at the
Haxby et al.  showed how each
multi-voxel pattern of activity is in
fact a marker of a particular cognitive
state. When their participants
viewed different categories of images
(faces, houses, furniture, and so on),
each category produced a distinct
pattern in the visual cortex. What could
we do with such a technique? Mind
reading. Not to invade the privacy of
your thoughts, but rather as
a genuinely useful tool to understand
how the brain processes information. If
we decode a pattern reflecting an arm
movement in the brain of
a hemiplegic, for example, we could
feed it into a robotic arm to perform
the action that the hemiplegic
But Bach et al.  had other
intensions. They ingeniously proposed