Cell-intrinsic regulation of axonal morphogenesis by the Cdh1-APC target SnoN.
ABSTRACT Axonal growth is fundamental to the establishment of neuronal connectivity in the brain. However, the cell-intrinsic mechanisms that govern axonal morphogenesis remain to be elucidated. The ubiquitin ligase Cdh1-anaphase-promoting complex (Cdh1-APC) suppresses the growth of axons in postmitotic neurons. Here, we report that Cdh1-APC operates in the nucleus to inhibit axonal growth. We also identify the transcriptional corepressor SnoN as a key target of neuronal Cdh1-APC that promotes axonal growth. Cdh1 forms a physical complex with SnoN and stimulates the ubiquitin-dependent proteasomal degradation of SnoN in neurons. Knockdown of SnoN in neurons significantly reduces axonal growth and suppresses Cdh1 RNAi enhancement of axonal growth. In addition, SnoN knockdown in vivo suggests an essential function for SnoN in the development of granule neuron parallel fibers in the cerebellar cortex. These findings define Cdh1-APC and SnoN as components of a cell-intrinsic pathway that orchestrates axonal morphogenesis in a transcription-dependent manner in the mammalian brain.
- SourceAvailable from: John R Henley[show abstract] [hide abstract]
ABSTRACT: Pathfinding by growing axons in the developing or regenerating nervous system is guided by gradients of molecular guidance cues. The neuronal growth cone, located at the ends of axons, uses surface receptors to sense these cues and to transduce guidance information to cellular machinery that mediates growth and turning responses. Cytoplasmic Ca2+ signals have key roles in regulating this motility. Global growth cone Ca2+ signals can regulate cytoskeletal elements and membrane dynamics to control elongation, whereas Ca2+ signals localized to one side of the growth cone can cause asymmetric activation of effector enzymes to steer the growth cone. Modulating Ca2+ levels in the growth cone might overcome inhibitory signals that normally prevent regeneration in the central nervous system.Trends in Cell Biology 06/2004; 14(6):320-30. · 11.72 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: The anaphase-promoting complex (APC) is a multisubunit E3 ubiquitin ligase that targets key cell cycle regulatory proteins for degradation. Blockade of APC activity causes mitotic arrest. Recent evidence suggests that the APC may have roles outside the cell cycle. Several studies indicate that ubiquitin plays an important role in regulating synaptic strength. We previously showed that ubiquitin is directly conjugated to GLR-1, a C. elegans non-NMDA (N-methyl-D-aspartate) class glutamate receptor (GluR), resulting in its removal from synapses. By contrast, endocytosis of rodent AMPA GluRs is apparently regulated by ubiquitination of associated scaffolding proteins. Relatively little is known about the E3 ligases that mediate these effects. We examined the effects of perturbing APC function on postmitotic neurons in the nematode C. elegans. Temperature-sensitive mutations in APC subunits increased the abundance of GLR-1 in the ventral nerve cord. Mutations that block clathrin-mediated endocytosis blocked the effects of the APC mutations, suggesting that the APC regulates some aspect of GLR-1 recycling. Overexpression of ubiquitin decreased the density of GLR-1-containing synapses, and APC mutations blunted this effect. APC mutants had locomotion defects consistent with increased synaptic strength. This study defines a novel function for the APC in postmitotic neurons.Current Biology 12/2004; 14(22):2057-62. · 9.49 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Neurons extend long axons and highly branched dendrites, and our understanding of the essential regulators of these processes has advanced in recent years. In the past year, investigators have shown that transcriptional control, posttranslational degradation and signaling cascades may be master regulators of axon and dendrite elongation and branching. Thus, evidence is mounting for the importance of the intrinsic growth state of a neuron as a crucial determinant of its ability to grow, or to regenerate, axons and dendrites.Current Opinion in Neurobiology 11/2004; 14(5):551-7. · 7.34 Impact Factor
Neuron 50, 389–400, May 4, 2006 ª2006 Elsevier Inc.DOI 10.1016/j.neuron.2006.03.034
Cell-Intrinsic Regulation of Axonal Morphogenesis
by the Cdh1-APC Target SnoN
Judith Stegmu ¨ller,1Yoshiyuki Konishi,1,2
Mai Anh Huynh,1Zengqiang Yuan,1
Sara DiBacco,1and Azad Bonni1,*
1Department of Pathology
Harvard Medical School
77 Avenue Louis Pasteur
Boston, Massachusetts 02115
Axonal growth is fundamental to the establishment
of neuronal connectivity in the brain. However, the
cell-intrinsic mechanisms that govern axonal morpho-
genesis remain to be elucidated. The ubiquitin ligase
Cdh1-anaphase-promoting complex (Cdh1-APC) sup-
presses the growth of axons in postmitotic neurons.
Here, we report that Cdh1-APC operates inthe nucleus
tional corepressor SnoN as a key target of neuronal
Cdh1-APC that promotes axonal growth. Cdh1 forms
uitin-dependent proteasomal degradation of SnoN in
neurons. Knockdown of SnoN in neurons significantly
hancement of axonal growth. In addition, SnoN knock-
down in vivo suggests an essential function for SnoN
in the development of granule neuron parallel fibers in
the cerebellar cortex. These findings define Cdh1-APC
and SnoN as components of a cell-intrinsic pathway
tion-dependent manner in the mammalian brain.
The growth of axons is critical to the establishment of
neuronal connectivity and normal wiring of the develop-
ing nervous system. Studies of axonal development
have led to the identification of several families of poly-
peptide growth factors that regulate and guide the
growth of axons toward their targets (Dickson, 2002;
Tessier-Lavigne and Goodman, 1996). Axon growth
and guidance cues act on neurons via cell surface re-
ceptors that couple attractive and repulsive extrinsic
signals to the cytoskeletal machinery ofthe axon growth
cone (Dent and Gertler, 2003; Henley and Poo, 2004;
Huber et al., 2003; Luo et al., 1997).
Growing evidence suggests that cell-intrinsic mecha-
nisms also play a crucial role in the control of axonal
morphogenesis, but the nature of these mechanisms re-
mains incompletely understood (Goldberg, 2004). A few
transcription factors have been implicated in the control
of distinct aspects of axonal development including
axonogenesis and refinement of axonal projections
(Arlotta et al., 2005; Kania et al., 2000; Segawa et al.,
2001; Wang et al., 2002; Weimann et al., 1999). How-
trol axonal elongation and stabilization remain to be
In addition to transcriptional control, the ubiquitin-
proteasome machinery is emerging as a key cell-intrinsic
regulator of axonal development (Campbell and Holt,
2001; Konishi et al., 2004; Stegmuller and Bonni, 2005;
responses to chemotropic signals (Campbell and Holt,
2001). In addition, the ubiquitin-proteasome machinery
promotes the local degeneration of axon terminals dur-
ing developmentin Drosophila (Watts et al., 2003). While
ubiquitination of proteins within the growth cone has
been implicated in the regulation of axonal growth,
with few exceptions specific ubiquitin ligases and their
substrates in these processes remain to be identified
(Campbell and Holt, 2001; Watts et al., 2003).
We have found that the ubiquitin ligase Cdh1-
anaphase-promoting complex (Cdh1-APC) plays a criti-
cal role in the control of axonal growth and patterning in
the mammalian brain (Konishi et al., 2004). The finding
that Cdh1-APC controls axonal growth raises the major
question of how Cdh1-APC exerts this function in post-
mitotic neurons. Cdh1-APC is a multisubunit E3 ubiqui-
tin ligase that promotes the ubiquitination and conse-
quent degradation of B-type cyclins and other proteins
in dividing cells and thereby ensures the proper transi-
tions of the cell cycle (Harper et al., 2002; Page and
Hieter, 1999; Peters, 2002). The regulatory subunit
Cdh1 has the dual function of stimulating the APC ubiq-
uitin ligase activity and targeting APC to its substrates.
Substrate recognition by Cdh1 is dependent on specific
peptide motifs, including the D box and KEN box, that
are present in APC substrates (King et al., 1996; Pfleger
and Kirschner, 2000). The ubiquitin ligase activity of
Cdh1-APC is required for axon growth inhibition in
mammalian neurons (Konishi et al., 2004). Therefore, to
determine the mechanisms by which Cdh1-APC con-
trols axonal morphogenesis, it will be crucial to identify
the substrates of Cdh1-APC in neurons.
In this study, we report a mechanism by which Cdh1-
APC controls axonal morphogenesis in the mammalian
brain. We have found that Cdh1-APC operates in the nu-
cleus to inhibit the growth of axons. Using a candidate
approach, we have discovered that Cdh1-APC controls
axonal growth by inhibiting the transcriptional corepres-
sor SnoN. Cdh1 interacts with SnoN and thereby targets
SnoN for ubiquitin-dependent proteasomal degradation
in primary cerebellar granule neurons. Knockdown of
SnoN in primary neurons by RNAi significantly reduces
the growth of axons and completely suppresses the
ability of Cdh1 RNAi to enhance axonal growth. In addi-
tion, expression of a mutant SnoN protein that is resis-
tant to Cdh1-mediated degradation robustly stimulates
the growth of axons, an effect that is nonadditive with
Cdh1 RNAi-enhanced axonal growth. We also find that
the in vivo knockdown of SnoN in the developing rat
Institute of Life Science, 11 Minamiooya, Machida-shi, Tokyo
cerebellum profoundly impairs the development of gra-
nule neuron parallel fibers by inhibiting their elongation
and/orstabilization. Thesefindings uncoveran essential
function for the transcriptional corepressor SnoN in ax-
onal development in the mammalian brain. In addition,
our study suggests that,by acting inthe nucleus via reg-
ulation of SnoN-dependent programs of gene expres-
sion, Cdh1-APC may play a pivotal role in the control
of axonal morphogenesis.
Cdh1-APC Acts in the Nucleus to Control
To characterize the mechanism by which Cdh1-APC
controls axonal growth, we first determined the subcel-
lular site of action of Cdh1-APC in neurons. Cdh1, the
APC core protein Cdc27, and the associated Cdh1-
APC ubiquitin ligase activity in neurons are found pre-
dominantly but not exclusively in the nucleus in cerebel-
lar granule neurons (Konishi et al., 2004) (Figure S1A in
the Supplemental Data available with this article online).
Consistent with these results, immunohistochemical
analyses of the rat cerebellar cortex revealed Cdh1 im-
munoreactivity in the nucleus and cytoplasm of granule
neurons (Figure S1B). Cdh1-APC is thought to function
outside the nucleus in Drosophila neurons in the control
of synapse development (van Roessel et al., 2004).
Finally, ubiquitination of proteins within the growth
cone regulates axonal development (Campbell and
Holt, 2001; Watts et al., 2003). Together, these observa-
in the nucleus or cytoplasm in mammalian neurons to
control axonal growth.
We carried out structure-function analyses of Cdh1 in
the inhibition of axon growth. Since the overexpression
of Cdh1 in neurons failed to alter axon growth (data not
shown), we asked if an RNAi-resistant form of Cdh1
(Cdh1-Res) might rescue the Cdh1 knockdown-induced
ture-function analyses in this background. To perform
the rescue experiment, we constructed an expression
plasmid encoding a GFP-Cdh1 fusion protein using
wild-type Cdh1 cDNA and a plasmid encoding an
Immunoblotting revealed that Cdh1 RNAi induced the
knockdown of Cdh1, encoded by wild-type Cdh1
cDNA, but failed to effectively reduce the expression
of Cdh1-Res (Figure 1A). We next transfected cerebellar
granule neurons with the Cdh1 RNAi (U6/cdh1) or con-
trol U6 plasmid together with a DsRed expression plas-
mid (Figure 1B). In the background of Cdh1 RNAi, we
also expressed GFP-Cdh1 or GFP-Cdh1-Res. We found
that GFP-Cdh1-Res, but not GFP-Cdh1, reversed the
(Figure 1B). These experiments indicate that the axonal
phenotype upon Cdh1 RNAi is the result of specific
knockdown of Cdh1.
To determine if the localization of Cdh1 is of impor-
tance for Cdh1-APC function in neurons, we targeted
GFP-Cdh1-Res to the nucleus or the cytoplasm (Fig-
ure 1C). The GFP-Cdh1-Res expression plasmid was
modified by appending a nuclear exclusion sequence
(NES) or a nuclear localization sequence (NLS) to the
N terminus of Cdh1-Res (Yoneda et al., 1999). The
expression of GFP-NES-Cdh1-Res and GFP-NLS-
Cdh1-Res was assessed in both 293T cells and granule
neurons. The localization of GFP-NLS-Cdh1-Res was
predominantly nuclear in 293T cells and in primary neu-
rons. By contrast, GFP-NES-Cdh1-Res was excluded
from the nucleus (Figure 1C).
We next examined the ability of Cdh1-Rescue mu-
tants to reverse the increase in axonal length triggered
by Cdh1 knockdown (Figure 1D). Granule neurons
were transfected with the Cdh1 RNAi plasmid together
with the control GFP or one of the GFP-Cdh1-Res series
of plasmids and the DsRed expression vector. Neurons
were analyzed 3 days after transfection. Analysis of
axon length revealed that GFP-Cdh1-Res and GFP-
NLS-Cdh1-Res behaved in a very similar manner, re-
versing the Cdh1 RNAi-induced enhancement of axonal
growth (Figure 1D). In contrast, GFP-NES-Cdh1-Res
failed to reverse the Cdh1 RNAi axonal phenotype, and
axons remained at a comparable length as control
axons (Figure 1D). In control experiments, appending
the NES or NLS sequence to Cdh1 had little or no effect
on the ability of Cdh1 to interact with the APC core pro-
tein Cdc27 (Figure S2), suggesting that NES- or NLS-
tagged Cdh1 can still function as the regulatory subunit
of the APC. Together, these results suggest that the
localization of Cdh1 in the nucleus is required for Cdh1-
APC inhibition of axon growth.
The Cdh1-APC Target SnoN Promotes
The importance of the nuclear localization of Cdh1-APC
in the control of axonal growth suggests that the targets
of Cdh1-APC reside in the nucleus. Mounting evidence
points to a key role for transcriptional mechanisms in
the regulation of axonal growth (Goldberg, 2004; Steg-
muller and Bonni, 2005). Together, these observations
raised the possibility that Cdh1-APC might control axo-
nal growth at a transcriptional level.
To identify potential targets of neuronal Cdh1-APC
that might regulate axonal growth, we considered tar-
gets of APC in proliferating cells that control transcrip-
tion. Almost all APC substrates in dividing cells have
functions in anaphase and DNA replication (Harper
et al., 2002; Peters, 2002). Intriguingly, the transcrip-
tional corepressor SnoN represents a substrate of Cdh1-
APC in proliferating mammalian cells that directly
regulates transcription (Stroschein et al., 2001; Wan
et al., 2001). We asked if SnoN is expressed in neurons
and whether SnoN might control axonal growth down-
stream of Cdh1-APC.
We characterized the expression of SnoN in granule
neurons in the cerebellum. Immunoblotting of lysates
from primary cerebellar granule neurons with an anti-
body to SnoN revealed the expression of two alter-
natively spliced forms of SnoN (Figure 2A). Immuno-
precipitation of lysates from granule neurons followed
by immunoblotting with the SnoN antibody demon-
strated the specificity of the two bands representing
two SnoN isoforms (Figure 2A). SnoN was localized
predominantly in the nucleus in granule neurons as de-
termined by immunocytochemical analysis as well as
immunoblotting of fractionated lysates of granule neu-
rons (Figures 2B and 2C). SnoN localization remained
nuclear under different culture conditions, including in
the presence or absence of serum or membrane depo-
larization (data not shown).
Immunohistochemical analysis of the rat cerebellar
cortex revealed SnoN expression in both granule neu-
rons and Purkinje cells (Figure 2D). The pattern of
SnoN expression in the cerebellar cortex overlapped
that of Cdh1, which was also found to be expressed in
both granule neurons and Purkinje cells (Figure S1B).
In particular, both Cdh1 and SnoN were found in granule
neurons in the internal granule layer (IGL) (Figure 2D;
Figure S1B). SnoN expression in IGL neurons persisted
from P6 to P13 (Figure 2D), thus correlating temporally
withaxongrowth intheseneurons during braindevelop-
ment (Altman and Bayer, 1997).
To study the function of SnoN in granule neurons, we
used RNAi interference to acutely knock down SnoN
(Figure 3A). We used a SnoN RNAi plasmid (U6/snon)
that induces the knockdown of mouse SnoN (Sarker
et al., 2005). The SnoN hairpin RNAs (hpRNAs) target
a sequence in murine SnoN mRNA that is identical in
rat SnoN (XM 226979). Consistent with this observation,
granule neurons led to the efficient knockdown of
endogenous SnoN as determinedby immunocytochem-
ical analyses (Figure 3A). Whereas 70% of control
Figure 1. Cdh1-APC Acts in the Nucleus to Control Axonal Growth
(A) Schematic of GFP-Cdh1-Res and modified versions that contain the nuclear localization (NLS) or nuclear export (NES) sequences. Silent
mutations that render Cdh1-Res resistant to RNAi are indicated in red. (Right) Lysates of COS cells transfected with the U6 or U6/cdh1 plasmid
togetherwithaplasmid encoding FLAG-Cdh1orFLAG-Cdh1-ReswereimmunoblottedusingaFLAGor14-3-3 antibody. Cdh1RNAiinducedthe
knockdown of Cdh1 encoded by wild-type cDNA but failed to effectively induce knockdown of Cdh1-Res.
(B) Primary cerebellar granule neurons were transfected 8 hr after plating with the U6 or U6/cdh1 plasmid and the GFP, GFP-Cdh1 (wild-type), or
GFP-Cdh1-Res expression plasmid together with the DsRed and Bcl-xLexpression plasmid. Neurons were kept in media supplemented with
insulin. Three days later, cultures were subjected to immunocytochemistry using a polyclonal DsRed antibody. Total axonal length was
measured and shown as mean 6 SEM. Cdh1 knockdown in granule neurons significantly increased axon length as compared to control U6-
transfected neurons (p < 0.001, ANOVA); in the background of Cdh1 RNAi, GFP-Cdh1-Res but not GFP-Cdh1 (wild-type) significantly reduced
axon length when compared to Cdh1 knockdown neurons (p < 0.001, ANOVA). A total of 447 neurons were measured.
(C) 293T cells (left panels) and granule neurons (right panels) were transfected with the GFP, GFP-Cdh1-Res, GFP-NES-Cdh1-Res, or GFP-NLS-
Cdh1-Res expression plasmid. Cultures were subjected to immunocytochemistry using a monoclonal GFP antibody.
Res plasmid and the DsRed and Bcl-xLexpression plasmids. Neurons were analyzed as in (B). In the background of Cdh1 RNAi, total axonal
length of GFP-Cdh1-NES-Res- but not of GFP-Cdh1-NLS-Res-expressing neurons was significantly greater than in Cdh1Res-expressing
neurons (p < 0.001, ANOVA). A total of 519 neurons were measured.
SnoN Regulation of Axonal Growth
U6-transfected neurons displayed robust SnoN immu-
noreactivity, only 30% of SnoN hpRNA-expressing neu-
rons had SnoN immunoreactivity (Figure 3A). SnoN
knockdown in granule neurons led to a striking axonal
phenotype. Granule neurons in which SnoN RNAi was
triggered had significantly shorter axons than control
U6-transfected neurons (Figure 3B). The SnoN knock-
down and control U6-transfected neurons had compa-
rable expression of proteins enriched in postmitotic
granule neurons (Figure S3), suggesting that SnoN
RNAi did not alter the general differentiation state of
We next determined if the SnoN RNAi-induced axonal
phenotype is the result of specific knockdown of SnoN.
First, we used a plasmid encoding human SnoN
hpRNAs (U6/snon-h), which contain three nucleotide
mismatches with the rodent SnoN hpRNAs (Sarker
et al., 2005). While expression of SnoN hpRNAs in-
duced the efficient knockdown of endogenous SnoN
in the mouse neuronal cell line Neuro2A, human SnoN
hpRNAs failed to induce SnoN knockdown in these
cells (Figure S4). In granule neurons, expression of hu-
man SnoN hpRNAs failed to reduce axon length (Fig-
ure 3B). Second, we performed a rescue experiment
in the background of SnoN RNAi (Figure 3C). In these
experiments, we used a plasmid encoding human
SnoN harboring additional silent mutations in its
cDNA designed to render it resistant to RNAi (SnoN-
Res) (Sarker et al., 2005). We determined the effect of
SnoN-Res expression on axonal growth in the back-
ground of SnoN knockdown. We found that the expres-
sion of SnoN-Res triggered a significant increase in
axonal length, restoring total axonal length to 70% of
control U6-transfected granule neurons (Figure 3C).
These results indicate that the SnoN RNAi-triggered re-
duction in axonal length is the result of specific knock-
down of SnoN rather than off-target effects of SnoN
RNAi or nonspecific activation of the RNAi machinery.
Figure 2. SnoN Is Expressed in Granule Neurons in the Developing Cerebellum
(A) (Left) Lysates of granule neurons prepared from P6 rat pups and placed in culture for indicated days were immunoblotted using a polyclonal
SnoN antibody. (Right) Lysates of granule neurons were subjected to immunoprecipitation with the SnoN antibody or an antibody against actin
followed by immunoblotting with the SnoN antibody.
(B) Granule neurons were subjected to immunocytochemistry using the SnoN antibody and the DNA dye bisbenzimide (Hoechst 33258). SnoN
appeared to be predominantly in the nucleus in all granule neurons.
(C) Granule neurons were subjected to subcellular fractionation. The nuclear fraction (NF) and postnuclear supernatant (PNS) were immunoblot-
ted with the SnoN, SP1, or 14-3-3 antibody.
(D) Sagittal sections of cerebella from postnatal rat pups at indicated ages were subjected to immunohistochemistry using the SnoN antibody.
Cell nuclei were stained with the DNA dye bisbenzimide (Hoechst 33258). The external granule layer (EGL), molecular layer (ML), and internal
granule layer (IGL) are indicated. Asterisks indicate Purkinje cells. Scale bar, 100 mm.
Taken together, our results suggest that SnoN is re-
quired for the normal development of axons.
To assess the nature of SnoN function in the develop-
ment of axons, we transfected neurons with the SnoN
RNAi or control U6 plasmid at a time when they begin
to extend axons and measured the length of axons in
cohorts of neurons each day for 4 days, beginning 1
day after transfection (Figure 3D). In control U6-trans-
fected granule neurons, axons increased significantly
inlengthfromday 2today 5aftertransfection. However,
upon SnoN knockdown the axonal growth curve was
significantly reduced in slope (Figure 3D). Axons were
2-fold shorterinneurons with SnoNknockdown as com-
pared to control U6-transfected neurons at day 5 (Fig-
ure 3D). These results suggest that a key function of
the transcriptional corepressor SnoN in neurons is to
promote the growth of axons.
We also performed time-lapse analyses to ascertain
whether SnoN promotes axonal growth by stimulating
their elongation or by inhibiting axonal retraction. We
monitored individual neurons transfected with the
SnoN RNAi or control U6 plasmid (Figure 4A). While
the majority of control U6-transfected neurons robustly
extended axons over the entire period of observation,
SnoN knockdown almost completely inhibited axonal
extension (Figure 4B). However, SnoN RNAi did not
appear to induce retraction of axons. In short-term
changes and high motility of axonal tips in control
U6-transfected neurons, but the axonal tips in SnoN
knockdown neurons displayed little or no motility (data
not shown). Taken together, these results are consistent
elongation of axons in primary neurons.
We next determined if SnoN function in axonal growth
is affected by extrinsic culture conditions. We assessed
the effect of SnoN knockdown in granule neurons that
were exposed to the growth factor insulin, serum, or se-
rum together with membrane depolarization. Under
each of these conditions, SnoN knockdown significantly
Figure 3. SnoN Promotes Axonal Growth in Cerebellar Granule Neurons
(A) Granule neurons were transfected with the SnoN RNAi (U6/snon) or control U6 plasmid together with an expression plasmid encoding far-
nesylated GFP. Three days later, cultures were subjected to immunocytochemical analysis using the GFP and SnoN antibodies. Arrowhead
points to a U6-transfected GFP-positive neuron that is also SnoN positive, and arrow points to a neuron transfected with the U6/snon plasmid
as indicated by GFP expression that is SnoN negative. Approximately 70% of the control U6-transfected neurons and only 30% of SnoN
hpRNAs-expressing neurons are SnoN positive. Scale bar, 20 mm.
(B) Granule neurons transfected with the control U6, U6/snon (mouse), or U6/snon-h (human) RNAi plasmid together with the GFP and Bcl-xL
expression plasmids and cultured in media supplemented with calf serum for 3 days were subjected to immunocytochemistry using the GFP
antibody and analyzed as in Figure 1B. Axon length in U6/snon (mouse)-expressing neurons is significantly reduced as compared to control
vector (pCMV5) and the GFP and Bcl-xLexpression plasmids were analyzed as in Figure 3B. In the background of SnoN knockdown in granule
neurons, expression of SnoN-Res significantly increased axonal length as compared to vector pCMV5-expressing neurons (p < 0.02, ANOVA). A
total of 241 neurons were measured.
(D) Granule neurons transfected with the SnoN RNAi or control U6 plasmid together with the GFP and Bcl-xLexpression plasmids were cultured
axonal length at 3 days and subsequent days as compared to control U6-transfected neurons (p < 0.005, ANOVA). A total of 716 neurons were
SnoN Regulation of Axonal Growth
reduced total axonal length in granule neurons (Figures
3B, 5A, and 5B). These results suggest that SnoN repre-
sents a cell-intrinsic regulator of axonal growth.
inaxonal growth inprimary dissociated cultures ofgran-
ule neurons, we next asked if SnoN promotes axonal
To address this question, we used cerebellar slice over-
lay assays. We plated P6 neurons that were transfected
with a plasmid encoding both SnoN hpRNAs and GFP
bicistronically (U6/snon-cmvGFP) or the control U6-
cmvGFP plasmid on top of cerebellar slices prepared
to immunohistochemistry. Granule neurons in which
SnoN RNAi was triggered had a significant reduction
in total axonal length as compared to the control U6
plasmid-transfected neurons (Figure 5C). These results
suggest that SnoN promotes axonal growth in granule
neurons in the tissue environment of the cerebellar
We also assessed if SnoN contributes to axonal
growth in neuronal cell types other than cerebellar gran-
ule neurons. We found that SnoN is also expressed in
cortical and hippocampal neurons (Figure 5D). Impor-
tantly, SnoN knockdown significantly reduced axonal
growth in cortical neurons (Figure 5E). These data
suggest that SnoN function in regulation of axonal mor-
SnoN Acts Downstream of Neuronal Cdh1-APC
in the Control of Axon Growth
To determine if Cdh1-APC and its substrate SnoN act in
a linear pathway to regulate axonal growth, we per-
formed epistatic analysis of the effects of Cdh1 and
SnoN knockdown on axonal length (Figure 6A). Primary
RNAi plasmid or both RNAi plasmids together. Cdh1
knockdown significantly stimulated an increase in total
axonal length, while SnoN knockdown significantly
inhibited axon growth compared to control neurons
Figure 4. SnoN Knockdown Impairs Axonal Elongation in Primary Granule Neurons
(A) Representative images of neurons transfected with the SnoN RNAi or control U6 plasmid. Starting at 2 DIV, images of neurons were taken
every 8 hr over a period of 48 hr. Axons of control U6-transfected neurons increased in length. In contrast, axons of SnoN knockdown neurons
did not grow or retract. Arrows indicate axons.
(B) (Left) Slope of axonal growth of individual neurons transfected with the SnoN RNAi or control U6 plasmid. (Right) Average slope of axonal
growth, presented as mean 6 SEM, was significantly higher in control U6-transfected neurons as compared to SnoN knockdown neurons
(p < 0.005, Student’s t test).
(Figure 6A). The simultaneous induction of Cdh1 and
SnoN RNAi led to an axonal phenotype that was identi-
cal to that of SnoN knockdown, resulting in significantly
shorter axons as compared to Cdh1 hpRNA-expressing
neurons or as compared to control U6-transfected neu-
rons (Figure 6A). These experiments indicate that SnoN
knockdown completely suppressed the Cdh1 RNAi-
induced axonal growth phenotype. These findings are
consistent with the interpretation that Cdh1-APC and
SnoN operate in a linear pathway, where SnoN acts
downstream of Cdh1-APC in the control of axon growth.
To corroborate our results suggesting that SnoN acts
downstream of Cdh1-APC in neurons, we carried out
structure-function analyses of SnoN (Figure 6B). SnoN
contains a conserved Cdh1 recognition destruction (D
box) peptide motif (Stroschein et al., 2001; Wan et al.,
2001), whose mutation renders SnoN resistant to APC-
mediated ubiquitination. We measured axonal length
in granule neurons in which we expressed wild-type
SnoN (SnoN WT) or a SnoN protein with a mutated D
box (SnoN DBM). Expression of SnoN in granule neu-
rons had little effect on axon length (Figure 6B). In con-
trast, neurons in which SnoN DBM was expressed had
a significant increase in total axonal length as compared
to SnoN-expressing neurons or control vector-trans-
fected neurons (Figure 6B). In experiments in which we
measured axonal length for several days after trans-
fection, SnoN DBM but not SnoN significantly enhan-
ced the rate of axonal growth as compared to control-
transfected neurons (Figure 6C). Together, these results
indicate that mutation of the Cdh1 recognition D box
in SnoN unmasks SnoN’s ability to promote axonal
We next tested the ability of SnoN DBM to promote
axonal growth in neurons in which Cdh1 RNAi was in-
down each enhanced axonal growth in granule neurons,
the combination of SnoN DBM overexpression and
Cdh1 knockdown did not result in an additive effect on
axonal length (Figure 6D). These findings corroborate
the results of the epistasis analyses in Figure 6A, sug-
gesting that Cdh1 and SnoN act in a shared pathway.
Taken together, our results support the conclusion
that Cdh1 controls axonal growth by inhibiting SnoN
function in neurons.
To characterize the mechanism by which neuronal
Cdh1-APC inhibits SnoN function, we first asked if
SnoN is regulated by the ubiquitin-dependent protea-
some machinery in granule neurons. Endogenous
SnoN levels increased in granule neurons treated with
the proteasome inhibitor lactacystin or MG132 as deter-
mined by immunoblotting (Figure 6E and data not
shown). In other experiments, endogenous SnoN was
found to be conjugated with ubiquitin in neurons as de-
termined by immunoblotting of immunoprecipitated
SnoN with antibodies to ubiquitin (Figure 6F). Together,
these results suggest that SnoN undergoes ubiquitin-
dependent proteasomal degradation in neurons.
Figure 5. SnoN Knockdown Impairs Axonal
Growth under DifferentEnvironmental Condi-
tions and in Distinct Populations of Neurons
(A and B) Granule neurons transfected with
the SnoN RNAi or control U6 plasmid to-
gether with the GFP and Bcl-xLexpression
plasmids were cultured in BME supple-
mented with insulin (A) or calf serum together
with membrane-depolarizing concentrations
of KCl (B) and were analyzed as in Figure 3B.
Images of representative transfected neu-
rons in BME + insulin are shown. Scale bar,
50 mm. SnoN knockdown in granule neurons
reduced as compared to the corresponding
control U6-transfected neurons (p < 0.002
and p < 0.0002, respectively, Student’s t
test). A total of 110 and 165 neurons were
(C) Granule neurons transfected in sus-
pension with the U6/snon-cmvGFP RNAi or
control U6-cmvGFP plasmid together with
an expression plasmid encoding Bcl-xLwere
placed on top of cerebellar slices from P9
rat pups. Slices were fixed 3 days later, sub-
jected to immunohistochemistry with the
GFP antibody, and subjected to morphome-
try. Axonal length of U6/snon-expressing
neurons was significantly reduced as com-
pared to control U6-transfected neurons
(p < 0.002, Student’s t test; values indicate
mean 6 SEM). A total of 375 neurons were
(D)Hippocampal andcorticalneuronswereisolatedfromE18ratembryos. Neuronswereculturedfortheindicatedtimeperiod,andlysates were
subjected to immunoblotting with the SnoN antibody. Both hippocampal and cortical neurons express SnoN. Asterisks indicate nonspecific
(E) Axonal length was measured in cortical neurons transfected with the SnoN RNAi or control U6 plasmid together with the GFP and Bcl-XLex-
pression plasmids. Axon length was significantly reduced in neurons in which SnoN RNAi was induced as compared to control U6-transfected
neurons (p < 0.0001, Student’s t test; values indicate mean 6 SEM). A total of 194 neurons were measured. Scale bar, 200 mm.
SnoN Regulation of Axonal Growth
We next assessed the role of Cdh1-APC in the regula-
tion of SnoN protein turnover in neurons. In coimmuno-
to associate with endogenous SnoN (Figure 6G). To de-
termine if neuronal Cdh1-APC promotes SnoN degrada-
tion, we asked if Cdh1 knockdown increases the level of
SnoN protein in neurons. First, we established an assay
that accurately reflects increases in the amount of SnoN
protein in transfected neurons. We expressed a gene
encoding SnoN fused to renilla luciferase (Ren-SnoN)
in neurons, thus allowing us to use renilla luciferase
activity as a surrogate for the amount of SnoN in trans-
fected neurons (Figure 6H). Luciferase fusion proteins
have been successfully used to quantify protein turn-
over of APC substrates (Lukas et al., 1999). In our exper-
iments, a renilla fusion with the SnoN DBM mutant pro-
tein (Ren-SnoN DBM) had a significantly higher level of
renilla luciferase activity than renilla-SnoN, suggesting
that mutation of the Cdh1 recognition D box motif within
SnoN leads to increased levels of SnoN protein in neu-
rons (Figure 6H). Importantly, using the renilla fusion
assay, we found a significant increase in the level of
renilla-SnoN upon Cdh1 knockdown in granule neurons.
In contrast, Cdh1 knockdown had little effect on the
level of renilla-SnoN DBM activity (Figure 6I). These
results suggest that Cdh1 promotes the degradation of
SnoN protein. Taken together, our findings support the
conclusion that Cdh1-APC promotes the ubiquitination
and consequent proteasomal degradation of SnoN in
Figure 6. SnoN Acts Downstream of Cdh1-
APC in the Control of Axonal Growth
(A) Granule neurons transfected with the con-
trol U6, U6/cdh1, U6/snon plasmid, or both
U6/cdh1 and U6/snon RNAi plasmids to-
gether with the GFP and Bcl-xLexpression
plasmids were analyzed as in Figure 3B.
Knockdown of both Cdh1 and SnoN in gran-
ule neurons significantly reduced axonal
length as compared to control U6- and U6/
cdh1-transfected neurons, respectively (p <
0.0001 and p < 0.05 respectively, ANOVA).
A total of 521 neurons were measured.
(B) Granule neurons transfected with an ex-
pression plasmid encoding wild-type SnoN
(WT), mutant D box SnoN (DBM), or the con-
trol vector pCMV5 together with the GFP
and Bcl-xLexpression plasmids were ana-
lyzed as in Figure 3B. SnoN DBM expression
in granule neurons but not the expression
of SnoN WT significantly increased axonal
length as compared to control U6-trans-
fected neurons (p < 0.02, ANOVA). A total of
328 neurons were measured.
(C) Granule neurons transfected as in Fig-
ure 5B were analyzed at the indicated times
as in Figure 3B. Expression of SnoN DBM in
granule neurons significantly increased axo-
nal growth at day 4 as compared to SnoN ex-
pression or control U6-transfected neurons
(p < 0.001, ANOVA). A total of 423 neurons
(D) Granule neurons transfected with the
Cdh1 RNAi plasmid or SnoN DBM expres-
sion plasmid alone or together with their con-
trol vectors were analyzed as in Figure 3B.
Simultaneous knockdown of Cdh1 and ex-
pression of SnoN DBM did not result in addi-
tive axonal growth. A total of 314 neurons
(E) Lysates of granule neurons treated with 10 mm lactacystin or vehicle for 10 hr were subjected to immunoblotting with the SnoN or 14-3-3
(F) Lysates of granule neurons were subjected to immunoprecipitation with the HA (ctrl) or SnoN antibody followed by immunoblotting with an
antibody to ubiquitin.
(G) Lysates of granule neurons were immunoprecipitated with the HA (ctrl) and SnoN antibody followed by immunoblotting with the Cdh1 anti-
(H) Schematic of Renilla-SnoN WT and Renilla-SnoN DBM. Lysates of granule neurons transfected with Renilla-SnoN WT or Renilla-SnoN DBM
expression plasmid together with the SV40 firefly luciferase (pGL3 promoter) plasmid, the latter to serve as internal control for transfection
efficiency, were subjected to a Dual Luciferase assay (Promega). Renilla-SnoN DBM activity was significantly increased compared to SnoN
WT (p < 0.03, Student’s t test; n = 4; values indicate mean 6 SEM).
(I)Lysates of granuleneuronstransfected withthe Ren-SnoN WTorRen-SnoN DBMexpressionplasmid togetherwiththefirefly luciferase(pGL3
promoter) plasmid and a Cdh1 RNAi plasmid (pSUPER/cdh1) or its control vector (pSUPER). Renilla-SnoN activity was significantly increased
upon Cdh1 RNAi compared to control-transfected neurons (p < 0.02, ANOVA; n = 4; values indicate mean 6 SEM). Cdh1 knockdown had little or
no effect on activity of renilla-SnoN DBM.
SnoN Is Required for the Development of Granule
Neuron Parallel Fibers In Vivo
The identification of SnoN as a key Cdh1-APC target
protein that promotes axonal growth in primary granule
neurons led us to characterize the in vivo function of
SnoN in the cerebellar cortex. Since both gain-of-func-
tion and loss-of-function analyses implicated SnoN in
axonal growth in primary neurons, we decided to em-
ploy both approaches in the assessment of SnoN func-
tion in vivo. To do this, we used an electroporation
down in the cerebellar cortex in postnatal rat pups
(Figure 7A). In each case, we injected test plasmids
into the cerebellar cortex of P3 rat pups and subjected
these pups to electroporation. In both sets of experi-
ments, a GFP expression plasmid or cassette was in-
cluded. Five days after electroporation, the cerebellum
was isolated and coronal sections of the cerebellum
were subjected to immunohistochemistry with GFP
antibodies to allow the assessment of granule neuron
parallel fiber axons. GFP-positive granule neurons
were present in the cerebellar cortex of electroporated
animals. Most of the transfected GFP-positive granule
neurons were in the IGL, with a smaller population in
the EGL (Figures 7A–7C). In the gain-of-function analy-
ses, the SnoN DBM granule neurons appeared to have
normal parallel fibers of similar pattern as those in con-
trol pCMV5-transfected cerebella (Figure 7A). We were
unable to assess whether SnoN DBM stimulates axonal
growth in the cerebellar cortex, because it was not pos-
sible to measure the total length of individual parallel
fibers in vivo (Figure 7A).
In contrast to the gain-of-function analyses, the loss-
of-function analyses were informative. In these experi-
ments, we injected the U6/snon-cmvGFP RNAi or con-
trol U6/cmvGFP plasmid into the cerebellar cortex of
P3 rat pups and analyzed the cerebella 5 days later. Ex-
amination of the EGL neurons revealed that both control
U6-transfected and SnoN hpRNAs-expressing granule
ment (Figure 7B). However, analysis of the IGL granule
neurons revealed a striking SnoN knockdown-induced
phenotype in parallel fiber development (Figure 7C).
The control U6-transfected granule neurons residing in
Figure 7. SnoN Knockdown Impairs Granule
Neuron Parallel Fiber Development In Vivo
(A) The SnoN DBM expression plasmid or its
control pCMV5 vector together with the GFP
and Bcl-xL expression plasmids were in-
jected into the cerebellum of P3 rat pups.
Five days later at P8, cerebella were isolated
from rat pups, and 10 mm coronal sections of
the cerebella were subjected to immunohis-
tochemistry using the GFP antibody. No ap-
parent difference in parallel fiber patterning
in cerebella was detected in SnoN DBM-
expressing cerebella as compared to cere-
bella transfected with the control vector.
Arrows point to parallel fibers in the molecu-
lar layer; asterisks indicate granule neurons
in the internal granule layer.
(B) The U6/snon-cmvGFP RNAi or U6/
cmvGFP control plasmid together with the
Bcl-xLexpression plasmid were injected into
the cerebellum of P3 rat pups, and cerebella
were analyzed as in Figure 7A. Representa-
tive images of newly generated neurons that
extend axons in the EGL. Arrows point to
transfected newly generated neurons. Scale
bar, 50 mm.
(C) Coronal sections of cerebella from pups
were subjected to in vivo electroporation as
described in (B). The external granule layer
(EGL), molecular layer (ML), and internal
granule layer (IGL) are indicated. Arrowheads
indicate parallel fibers. Scale bars, 100 mm in
large panels and 50 mm in small panels.
(D) Quantification of parallel fibers. Trans-
fected granule neurons in the IGL were
counted in consecutive sections of the U6-
cmvGFP- or U6/snon-cmvGFP-transfected
cerebella. Axons were counted in the molec-
ular layer (see Experimental Procedures).
Graph indicates percentage of granule neu-
rons that were associated with parallel fibers.
Parallel fiber number in granule neurons is
significantly reduced upon SnoN knockdown
as compared to control U6-transfected neu-
rons (p < 0.001, Student’s t test; values in-
dicate mean 6 SEM). A total of 476 neurons
SnoN Regulation of Axonal Growth
the IGL displayed a normal anatomy including the typi-
cal T-shaped axons and association with parallel fibers
in the molecular layer (ML). Parallel fibers extended in
large numbers in and beyond the region of the trans-
fected IGL granule neurons (Figure 7C). By contrast, al-
though SnoN knockdown granule neurons were present
in the IGL in similar numbers as control U6-transfected
neurons, few of the SnoN knockdown neurons were
associated with parallel fibers in the ML (Figure 7C).
Wequantified the effectofSnoNknockdown onparal-
lel fibers in the cerebellar cortex. More than 90% of the
control U6-transfected IGL granule neurons were asso-
ciated with parallel fibers. In contrast, only 30% of the
SnoN knockdown IGL granule neurons were associated
with parallel fibers (Figure 7D). These remaining parallel
fibers in the neurons appeared underdeveloped and
failed to extend for long distances beyond the area of
the transfected IGL neurons (Figure 7C). These results
suggest that SnoN plays a key role in axonal growth
and/or stabilization of IGL granule neurons. Consistent
with this observation is the finding that SnoN is robustly
expressed in granule neurons in the IGL but not in the
EGL (Figure 2D). Taken together, our findings support
theconclusion thatSnoNspecifically promotesthe mor-
phogenesis of granule neuron parallel fibers in the cere-
bellar cortex in vivo at a developmental stage following
In this study, we have characterized a mechanism by
which the ubiquitin ligase Cdh1-APC controls axonal
morphogenesis in the mammalian brain. Our findings in-
dicate that Cdh1-APC operates in the nucleus to control
axonal growth in granule neurons of the developing
cerebellum. We have also identified the transcriptional
corepressor SnoN as a key target of neuronal Cdh1-
APC that promotes axonal growth. Cdh1 interacts with
SnoN and stimulates the ubiquitin-dependent degrada-
tion of SnoN in primary neurons. Consistent with these
results, epistasis analyses in neurons suggest that
SnoN acts downstream of Cdh1 in the control of axonal
electroporation points to an essential function of SnoN
in granule neuron parallel fiber development. Taken to-
gether, our findings suggest that Cdh1-APC and SnoN
form a cell-intrinsic pathway that orchestrates axonal
morphogenesis in the mammalian brain.
Our study uncovers a function for SnoN in neurons as
a potent promoter of axonal growth. In particular, our
findings in primary neurons and in vivo support the
view that SnoN promotes axonal elongation in the mam-
malian brain. The in vivo SnoN knockdown experiments
raise the possibility that SnoN may also actively inhibit
the destabilization of developing axons. In future stud-
ies, it will be interesting to determine how SnoN might
impact on both axonal elongation and stabilization.
Since SnoN is a member of the Ski/SnoN family of
transcriptional corepressors (Akiyoshi et al., 1999; He
et al., 2003; Luo et al., 1999; Nomura et al., 1999; Stro-
schein et al., 1999; Sun et al., 1999a, 1999b; Xu et al.,
2000), it will be important to identify the transcription
factors that act in concert with SnoN in neurons. Prior
toour study, SnoN function was unknown in the nervous
system. However, SnoN has been implicated as a proto-
oncogene in a number of tissues outside the nervous
system (Colmenares et al., 1991; Reed et al., 2001;
Zhang et al., 2003), owing to its function as a key nega-
tive regulator of signaling induced by transforming
growth factor-b (TGFb) (He et al., 2003; Luo et al.,
1999; Stroschein et al., 1999). SnoN physically interacts
with the TGFb-regulated transcription factors Smad2
and Smad3 and thereby represses Smad-dependent
gene expression (Akiyoshi et al., 1999; He et al., 2003;
Luo et al., 1999; Stroschein et al., 1999; Sun et al.,
1999a; Xu et al., 2000). Conversely, TGFb signaling
induces the Smad2- or Smad3-dependent recruitment
of SnoN to Cdh1, leading to Cdh1-APC-mediated ubiq-
uitination and consequent degradation of SnoN in prolif-
erating cells (Stroschein et al., 2001; Wan et al., 2001).
Thus, TGFb signaling triggers the release of a SnoN-
dependent repressor complex from Smad2 and Smad3,
leading to the activation of TGFb-dependent transcrip-
tion (Stroschein et al., 2001; Wan et al., 2001). In view
of these observations, it will be important to determine
if the Cdh1-APC-SnoN cell-intrinsic mechanism under-
lying axonal morphogenesis is subject to regulation by
TGFb signaling in neurons, and to identify components
of the TGFb-Smad signaling pathway that might control
SnoN-dependent axonal growth in the brain.
The finding that Cdh1-APC operates in the nucleus to
control axonal growth supports the view that Cdh1-APC
is a pivotal regulator of axonal development. This con-
clusion isbolstered bythe identification ofthe transcrip-
tional corepressor SnoN as a key downstream target of
Cdh1-APC in the control of axonal growth. These find-
ings suggest that Cdh1-APC controls the development
of axons by coordinating programs of gene expression
leading to specific effects on axonal morphogenesis.
Therefore, to gain insights into the mechanisms of
Cdh1-APC control of axonal growth, it will be important
in future studies to elucidate the programs of gene
expression that are regulated by SnoN in neurons.
substrates of Cdh1-APC in the control of axonal growth.
Evidence for the existence of additional substrates of
Cdh1-APC that might be relevant in axonal morphogen-
esis comes from our analyses of overexpression of the
D box mutant SnoN (SnoN DBM) in the cerebellum.
Although we found in these in vivo experiments that it
was not possible to measure the anticipated gain-of-
function effect of SnoN DBM of increased axonal length,
we assessed whether overexpression of SnoN DBM
might induce a defect in axonal patterning, similar to
that with Cdh1 RNAi (Konishi et al., 2004). Interestingly,
we found little difference in parallel fiber patterning in
cerebella transfected with SnoN DBM as compared to
control vector (Figure 7A). These data suggest the inter-
esting possibility of additional substrates of neuronal
Cdh1-APC that might mediate the ability of Cdh1-APC
to control the patterning of axons.
Studies in Drosophila and Xenopus model systems
suggest that ubiquitin-dependent protein turnover
locally within axon growth terminals regulates distinct
responses and degeneration (Campbell and Holt, 2001;
Watts et al., 2003). Although we have found that the nu-
cleus is the prime site of action of Cdh1-APC in axonal
growth in mammalian neurons, the possibility remains
that Cdh1-APC might also contribute to the regulation
of axonal growth locally in the axon growth cone.
Cdh1-APC that resides outside the nucleus in mamma-
lian neurons might also be relevant to other functions,
including possibly synapse development and function,
as demonstrated in nematodes and flies (Juo and Ka-
plan, 2004; van Roessel et al., 2004). Thus, the localiza-
tion of Cdh1-APC in distinct subcellular compartments
in neurons might endow this ubiquitin ligase with the
ability to locally recruit its targets for degradation and
regulate diverse functions in the mammalian brain.
Cerebellar Granule Neuron Culture and Transfections
Granule neurons were prepared from isolated cerebella of P6 Long-
Evans rat pups as described (Shalizi et al., 2003). Neurons were
transfected either 8 hr after plating or at 2 days in vitro (DIV) with
a modified calcium phosphate method as described (Konishi et al.,
2002). To test if the calcium phosphate transfection method ensures
the efficient cotransfection of multiple plasmids, we cotransfected
expression plasmids encoding GFP and DsRed and found that
92% of DsRed-positive neurons also expressed GFP. To rule out
the possibility that the effects of RNAi or protein expression on axo-
nal length were due to any effect of these manipulations on cell
survival, the antiapoptotic protein Bcl-xLwas coexpressed in all
our experiments. The expression of Bcl-xLitself has little or no effect
on axonal length (Konishi et al., 2004).
Analysis of the morphology of axons of the cerebellar granule neu-
rons in vitro and in the slice overlay assay was carried out by captur-
ing images of the neurons in a blinded manner using a Nikon eclipse
TE2000 epifluorescence microscope. Measurements of axons were
performed using SPOT imaging software as described (Gaudilliere
et al., 2004).
In Vivo Electroporation
In vivo electroporation was performed as described (Konishi et al.,
2004). In brief, the U6-cmvGFP or U6/snon-cmvGFP plasmid to-
gether with the Bcl-xLexpression plasmid were diluted in PBS/
0.3% fast green (3–4 ml with 4 mg/ml of U6 plasmids and 1 mg/ml of
Bcl-xLplasmid) and injected into the cerebellum of P3 Sprague-
Dawley rat pups. The pups were subjected to five electric pulses
of 160 mV with 950 ms intervals. To analyze the parallel fibers in
the in vivo electroporation experiments, the identity of transfected
granule neurons in the EGL was confirmed based on the small size
of the nuclei, as determined by staining with the DNA dye bisbenz-
imide (Hoechst 33258) and MEF2 immunoreactivity, a marker of
granule neurons in the IGL. Transfected granule neurons were
counted in consecutive sections of individual cerebella. Parallel
fibers were counted in a restricted area of consecutive sections to
Slice Overlay Assay
Slice overlay assay was performed as described (Konishi et al.,
2004). In brief, cerebellar slices from P8 or P9 rat pups were pre-
pared using a McIllwain Tissue Chopper. Slices (400 mm) were cul-
tured on 0.4 mm membranes using the medium air interface method
(MEM; 25mM HEPES, 25% horse serum, 6.5 mg/ml D-glucose,1 ml/
100 ml PSG) for 24 hr at 36ºC/5% CO2before coculture with granule
neurons. Granule neurons were isolated from P6 rats as described
and transfected 3 hr later in suspension (2.5 3 10 cells/2 ml
U6 plasmid that also contained an expression cassette encoding
sion plasmid encoding Bcl-xL. Transfection reaction was terminated
by adding a large volume of DMEM. Cells were pelleted and then
plated on top of the cerebellar slices and cocultured for 3 days. Sli-
ces werethen subjected toimmunostaining using the GFP antibody.
Slice integrity was assessed using the DNA dye bisbenzimide
The Supplemental Data include four supplemental figures and can
be found with this article online at http://www.neuron.org/cgi/
We thank Shirin Bonni for providing the SnoN RNAi and SnoN
expression plasmids and for helpful discussions, Takahiko Matsuda
and Connie Cepko for sharing equipment, Jan-Michael Peters for
Cdh1 antibodies, Thijn Brummelkamp for the pSUPER Cdh1 RNAi
plasmid, Zhigang Xie for help with time-lapse analyses, Xuecai Ge
for help with experiments at an early stage of the study, and mem-
bers of the Bonni laboratory for critical reading of the manuscript.
Supported by an NIH grant (NS051255; A.B.), the Christopher Reeve
Foundation (A.B.), the Deutsche Forschungsgemeinschaft (J.S.),
and the Lefler Foundation (Z.Y.). A.B. is the recipient of a fellowship
from the Alfred P. Sloan Foundation, a Robert H. Ebert Clinical
Scholar Award from the Esther A. and Joseph Klingenstein Fund,
an EJLB Foundation award, and a Sidney Kimmel Foundation
Received: August 10, 2005
Revised: September 12, 2005
Accepted: March 27, 2006
Published: May 3, 2006
Akiyoshi, S., Inoue, H., Hanai, J., Kusanagi, K., Nemoto, N., Miya-
zono, K., and Kawabata, M. (1999). c-Ski acts as a transcriptional
co-repressor in transforming growth factor-b signaling through
interaction with smads. J. Biol. Chem. 274, 35269–35277.
Altman, J., and Bayer, S.A. (1997). Development of the Cerebellar
System: In Relation to Its Evolution, Structure, and Functions
(Boca Raton, FL: CRC Press).
Arlotta, P., Molyneaux, B.J., Chen, J., Inoue, J., Kominami, R., and
Macklis, J.D. (2005). Neuronal subtype-specific genes that control
corticospinal motor neuron development in vivo. Neuron 45, 207–
Campbell, D.S., and Holt, C.E. (2001). Chemotropic responses of
retinal growth cones mediated by rapid local protein synthesis and
degradation. Neuron 32, 1013–1026.
Colmenares, C., Sutrave, P., Hughes, S.H., and Stavnezer, E. (1991).
Activation of the c-ski oncogene by overexpression. J. Virol. 65,
Dent, E.W., and Gertler, F.B. (2003). Cytoskeletal dynamics and
transport in growth cone motility and axon guidance. Neuron 40,
Dickson, B.J. (2002). Molecular mechanisms of axon guidance. Sci-
ence 298, 1959–1964.
Gaudilliere, B., Konishi, Y., de la Iglesia, N., Yao, G., and Bonni, A.
(2004). ACaMKII-NeuroDsignaling pathway specifiesdendriticmor-
phogenesis. Neuron 41, 229–241.
Goldberg, J.L. (2004). Intrinsic neuronal regulation of axon and den-
drite growth. Curr. Opin. Neurobiol. 14, 551–557.
Harper, J.W., Burton, J.L., and Solomon, M.J. (2002). The anaphase-
promoting complex: it’s not just for mitosis any more. Genes Dev.
He, J., Tegen, S.B., Krawitz, A.R., Martin, G.S., and Luo, K. (2003).
The transforming activity of Ski and SnoN is dependent on their abil-
ity to repress the activity of Smad proteins. J. Biol. Chem. 278,
Henley, J., and Poo, M.M. (2004). Guiding neuronal growth cones
using Ca2+ signals. Trends Cell Biol. 14, 320–330.
Huber, A.B., Kolodkin, A.L., Ginty, D.D., and Cloutier, J.F. (2003).
Signaling at the growth cone: ligand-receptor complexes and the
SnoN Regulation of Axonal Growth
control of axon growth and guidance. Annu. Rev. Neurosci. 26, 509–
Juo, P., and Kaplan, J.M. (2004). The anaphase-promoting complex
regulates the abundance of GLR-1 glutamate receptors in the ven-
tral nerve cord of C. elegans. Curr. Biol. 14, 2057–2062.
Kania, A., Johnson, R.L., and Jessell, T.M. (2000). Coordinate roles
for LIM homeobox genes in directing the dorsoventral trajectory of
motor axons in the vertebrate limb. Cell 102, 161–173.
King, R.W., Glotzer, M., and Kirschner, M.W. (1996). Mutagenic anal-
ysis of the destruction signal of mitotic cyclins and structural char-
acterization of ubiquitinated intermediates. Mol. Biol. Cell 7, 1343–
Konishi, Y., Lehtinen, M., Donovan, N., and Bonni, A. (2002). Cdc2
phosphorylation of BAD links the cell cycle to the cell death machin-
ery. Mol. Cell 9, 1005–1016.
Konishi, Y., Stegmuller, J., Matsuda, T., Bonni, S., and Bonni, A.
(2004). Cdh1-APC controls axonal growth and patterning in the
mammalian brain. Science 303, 1026–1030.
Lukas, C., Sorensen, C.S., Kramer, E., Santoni-Rugiu, E., Lindeneg,
C., Peters, J.M., Bartek, J., and Lukas, J. (1999). Accumulation of
cyclin B1 requires E2F and cyclin-A-dependent rearrangement of
the anaphase-promoting complex. Nature 401, 815–818.
Luo, L., Jan, L.Y., and Jan, Y.N. (1997). Rho family GTP-binding pro-
teins in growth cone signalling. Curr. Opin. Neurobiol. 7, 81–86.
Luo, K., Stroschein, S.L., Wang, W., Chen, D., Martens, E., Zhou, S.,
and Zhou, Q. (1999). The Ski oncoprotein interacts with the Smad
proteins to repress TGFb signaling. Genes Dev. 13, 2196–2206.
Nomura, T.,Khan,M.M.,Kaul,S.C., Dong,H.D., Wadhwa,R.,Colme-
nares, C., Kohno, I., and Ishii, S. (1999). Ski is a component of the
histone deacetylase complex required for transcriptional repression
by Mad and thyroid hormone receptor. Genes Dev. 13, 412–423.
Page, A.M., and Hieter, P. (1999). The anaphase-promoting com-
plex: new subunits and regulators. Annu. Rev. Biochem. 68, 583–
Peters, J.M. (2002). The anaphase-promoting complex: proteolysis
in mitosis and beyond. Mol. Cell 9, 931–943.
Pfleger, C.M., and Kirschner, M.W. (2000). The KEN box: an APC
recognition signal distinct from the D box targeted by Cdh1. Genes
Dev. 14, 655–665.
Reed, J.A., Bales, E., Xu, W., Okan, N.A., Bandyopadhyay, D., and
Medrano, E.E. (2001). Cytoplasmic localization of the oncogenic
protein Ski in human cutaneous melanomas in vivo: functional impli-
cations for transforming growth factor beta signaling. Cancer Res.
Sarker, K.P., Wilson, S.M., and Bonni, S. (2005). SnoN is a cell type-
specific mediator of transforming growth factor-beta responses.
J. Biol. Chem. 280, 13037–13046.
Segawa, H., Miyashita, T., Hirate, Y., Higashijima, S., Chino, N., Uye-
mura, K., Kikuchi, Y., and Okamoto, H. (2001). Functional repression
of Islet-2 by disruption of complex with Ldb impairs peripheral axo-
nal outgrowth in embryonic zebrafish. Neuron 30, 423–436.
Shalizi, A., Lehtinen, M., Gaudilliere, B., Donovan, N., Han, J.,
Konishi, Y., and Bonni, A. (2003). Characterization of a neurotrophin
signaling mechanism that mediates neuron survival in a temporally
specific pattern. J. Neurosci. 23, 7326–7336.
Stegmuller, J., and Bonni, A. (2005). Moving past proliferation: new
roles for Cdh1-APC in postmitotic neurons. Trends Neurosci. 28,
Stroschein, S.L., Wang, W., Zhou, S., Zhou, Q., and Luo, K. (1999).
Negative feedback regulation of TGF-b signaling by the SnoN onco-
protein. Science 286, 771–774.
Stroschein, S.L., Bonni, S., Wrana, J.L., and Luo, K. (2001). Smad3
recruits the anaphase-promoting complex for ubiquitination and
degradation of SnoN. Genes Dev. 15, 2822–2836.
Sun, Y., Liu, X., Eaton, E.N., Lane, W.S., Lodish, H.F., and Weinberg,
R.A. (1999a). Interaction of the Ski oncoprotein with Smad3 regu-
lates TGF-b signaling. Mol. Cell 4, 499–509.
Sun, Y., Liu, X., Ng-Eaton, E., Lodish, H.F., and Weinberg, R.A.
(1999b). SnoN and Ski protooncoproteins are rapidly degraded in
response to transforming growth factor beta signaling. Proc. Natl.
Acad. Sci. USA 96, 12442–12447.
Tessier-Lavigne, M., and Goodman, C.S. (1996). The molecular biol-
ogy of axon guidance. Science 274, 1123–1133.
van Roessel, P., Elliott, D.A., Robinson, I.M., Prokop, A., and Brand,
A.H. (2004). Independent regulation of synaptic size and activity by
the anaphase-promoting complex. Cell 119, 707–718.
Wan, Y., Liu, X., and Kirschner, M.W. (2001). The anaphase-promot-
ing complex mediates TGF-b signaling by targeting SnoN for
destruction. Mol. Cell 8, 1027–1039.
Wang, S.W., Mu, X., Bowers, W.J., Kim, D.S., Plas, D.J., Crair, M.C.,
Federoff, H.J., Gan, L., and Klein, W.H. (2002). Brn3b/Brn3c double
knockout mice reveal an unsuspected role for Brn3c in retinal gan-
glion cell axon outgrowth. Development 129, 467–477.
Watts, R.J., Hoopfer, E.D., and Luo, L. (2003). Axon pruning during
Drosophila metamorphosis: evidence for local degeneration and
requirement of the ubiquitin-proteasome system. Neuron 38, 871–
Weimann, J.M., Zhang, Y.A., Levin, M.E., Devine, W.P., Brulet, P.,
and McConnell, S.K. (1999). Cortical neurons require Otx1 for the re-
finement of exuberant axonal projections to subcortical targets.
Neuron 24, 819–831.
Xu, W., Angelis, K., Danielpour, D., Haddad, M.M., Bischof, O., Cam-
pisi, J., Stavnezer, E., and Medrano, E.E. (2000). Ski acts as a co-
repressor with Smad2 and Smad3 to regulate the response to type
beta transforming growth factor. Proc. Natl. Acad. Sci. USA 97,
Yoneda, Y., Hieda, M., Nagoshi, E., and Miyamoto, Y. (1999). Nucle-
ocytoplasmic protein transport and recycling of Ran. Cell Struct.
Funct. 24, 425–433.
Zhang, F., Lundin, M., Ristimaki, A., Heikkila, P., Lundin, J., Isola, J.,
Joensuu, H., and Laiho, M. (2003). Ski-related novel protein N
(SnoN), a negative controller of transforming growth factor-beta
signaling, is a prognostic marker in estrogen receptor-positive
breast carcinomas. Cancer Res. 63, 5005–5010.