1

Neuron, Volume 80

Supplemental Information

Accelerated Experience-Dependent Pruning

of Cortical Synapses in Ephrin-A2 Knockout Mice

Xinzhu Yu, Gordon Wang, Anthony Gilmore, Ada Xin Yee, Xiang Li, Tonghui Xu, Stephen J.

Smith, Lu Chen, and Yi Zuo

2

1. Supplemental Data

3

Figure S1. The gross brain anatomy of one-month-old ephrin-A2/A3 KOs is normal, related to

Figure 1.

(A, B) KO mice have comparable cortical mass (A) and cortical surface area (B) to age-matched wild-

type mice. Data are presented as mean ± SD.

(C) Coronal sections of wild-type and KO mice injected with DiI in the barrel cortex reveal normal

thalamocortical projections in KO mice. Cx, cortex; Hp, hippocampus; Th, thalamus; Hy, hypothalamus.

Scale bar, 500 μm.

(D) Cytochrome oxidase-stained sections of the barrel fields exhibit similar patterns between wild-type

and KO mice. Scale bar, 500 μm.

(E) Coronal sections of the motor (MC) and the barrel (BC) cortices of wild-type and KO mice, with

DAPI-labeled cell nuclei and endogenous YFP-labeled layer V neurons. Scale bar, 250 μm.

(F) Proportions of different cortical layers in the MC and the BC are comparable between wild-type and

KO mice.

4

Figure S2. Elimination of spines, but not filopodia, is significantly increased in ephrin-A2 KOs,

related to Figure 1.

(A) Percentages of spines eliminated in the motor cortex (MC), the sensory cortex (SC) and the barrel

cortex (BC) over 4 days in wild-type and ephrin-A2 KO mice at one month of age.

(B) Percentages of different spine types eliminated over 4 days in the motor cortex of wild-type and KO

mice.

(C) Proportional percentages of spines and filopodia in the motor cortex of wild-type and KO mice.

(D) Percentages of filopodia eliminated over 1 day in the motor cortex of wild-type and KO mice.

Data are presented as mean ± SD. *P<0.05, ***P<0.001.

5

6

Figure S3. Spine formation is not affected in ephrin-A2 KOs, related to Figure 2.

(A) Percentages of spine eliminated and formed over 4 days in the motor cortex (MC) of wild-type and

ephrin-A2 KO mice under control and sensory enriched conditions.

(B) Percentages of spines formed over 4 days in the barrel cortex (BC) of wild-type and ephrin-A2 KO

mice under different conditions.

Data are presented as mean ± SD. **P<0.01, ***P<0.001.

7

8

Figure S4. Array tomography reveals different synaptic localizations of ephrin-A3, ephrin-A2 and

EphA4 receptors in the mouse cortex, related to Figure 3.

(A) The labeling of ephrin-A2 and ephrin-A3 puncta is abundant in layer I of the barrel cortex in one-

month-old wild-type mice, but is completely absent in age-matched ephrin-A2/A3 KOs. The labeling of

presynaptic marker VGluT1 is comparable between wild-type and ephrin-A2/A3 KO mice. Scale bar, 10

μm.

(B) Left: volume rendering of 5 serial sections (70 nm each) through the entire cortical depth of the

barrel cortex reveals that ephrin-A2 (purple) and ephrin-A3 (yellow) are expressed throughout all

cortical layers in wild-type mice. Nuclei are stained with DAPI (cyan). Scale bar, 50 μm. Middle:

density plot of the two proteins reveals a consistently higher expression of ephrin-A2, compared to

ephrin-A3, throughout all cortical layers. Right: magnified images of the boxed regions in layer II/III,

layer IV and layer V from the left panel. Scale bar, 15 μm.

(C) Average numbers of ephrin-A3 puncta within 100 nm from the centers of different neuronal and

astrocytic constituents at synapses.

(D) Average numbers of ephrin-A2 puncta within 100 nm from the centers of various postsynaptic

markers and EphA4 at synapses.

(E) Average numbers of EphA4 puncta within 100 nm from the centers of different neuronal and

astrocytic constituents at synapses.

(F) The ratio of phosphorylated EphA4 to total EphA4 protein expression in the cortex of wild-type and

ephrin-A2 KO mice. Data are presented as mean ± SEM.

9

10

Figure S5. Examinations of astrocytic morphology and the expression of glial glutamate

transporters in wild-type and KO mice, related to Figure 4.

(A) Bright-field images of S100 immunolabeling in the cortex of both wild-type and ephrin-A2 KO

mice. Inserts show individual astrocytes from boxed regions. Scale bar, 25 µm.

(B, C) Quantification of cell number (B) and cell body volume (C) of cortical astrocytes reveals no

significant difference between wild-type and ephrin-A2 KO mice.

(D) Single-plane confocal images of immunofluorescence staining for glial glutamate transporters

(GLAST and GLT-1) in the cortex of wild-type and KO mice reveal a lower intensity of

immunofluorescence in KO mice. Scale bar, 2 µm.

(E) Quantification of mRNA levels of GLAST and GLT-1 in the cortex of wild-type and KO mice at

one-month old of age.

(F, G) Ratio of GLAST and GLT-1 expression between deprived and control barrel cortices of the same

mice.

Data are presented as mean ± SEM. *P<0.05.

11

2. Supplemental Experimental Procedures

Examinations of Gross Brain Phenotypes

The brains of one-month old mice were dissected out and brain weights were measured. Dorsal view

images of brains were captured under a dissecting microscope (Olympus SZ61). The cortical areas were

quantified using ImageJ software (NIH).

Analysis of Cortical Laminations

Anesthetized mice were transcardially perfused with PBS followed by 4% paraformaldehyde (PFA).

Brains were dissected and postfixed in 4% PFA for 2-4 hours, then cryoprotected in 30% sucrose until

sectioning. Coronal sections (40 m) of the motor and the barrel cortices were incubated in 4'-6-

Diamidino-2-phenylindole (DAPI) solution (1:36,000) for 10 min at room temperature. The thickness of

different cortical layers was measured using ImageJ software and averaged for each group.

DiI Labeling

Mice (P14) were anesthetized and an incision was performed to expose the skull over the barrel cortex.

A focal craniotomy was created using a 25-gauge needle and a small amount of 10% 1,1’-dioctadecyl-

3,3,3’,3’-tetramethylindocarbocyanine (DiI) solution was injected using a pulled glass pipette and a

picospritzer (Parker Instruments). Two weeks after injection, mice were transcardially perfused with

PBS followed by 4% PFA. Coronal sections (150 m) were collected and images were acquired using

Keyence BZ-9000 microscope.

Cytochrome-oxidase Staining

Staining of barrel fields was performed as described previously (Triplett et al., 2012). Briefly, mice

(P28-P30) were decapitated and brains were dissected out. Cortices were carefully isolated and flattened

12

between two glass slides overnight in 4% PFA. Flattened cortices were then cryoprotected in 30%

sucrose for sectioning. Serial horizontal sections (80 m) were collected and incubated in cytochrome C

staining solution (5% sucrose, 0.03% cytochrome C, 0.02% catalase, 0.05% diaminobenzidine) for 4-5

hours at 37 C. Reactions were stopped with 0.1% sodium azide. Cortices were fixed in 4% PFA

overnight and mounted on gelatin-coated slides (LabScientific, Inc.). Images were acquired using

Keyence BZ-9000 microscope.

Spine Morphology Analysis

Based on their lengths and head diameters, spines were classified into four categories: mushroom,

stubby, thin and other spines (Harris et al., 1992). Percentages of spines in different categories

eliminated were normalized to total spines counted in the initial image.

Immunoprecipitation

Immunoprecipitation was performed using Pierce Classic IP kit (Thermo Scientific), according to

manufacturer’s directions. Cortical lysates (1 mg) were incubated with mouse anti-EphA4 antibody

(Inivitrogen) overnight at 4 C. Immunoprecipitates were then probed by immunoblotting with the

following primary antibodies: mouse anti-EphA4 receptor (1:1,000; Invitrogen) and anti-

phosphotyrosine clone 4G10 (1:1,000; Millipore). Horseradish peroxidase (HRP)-conjugated secondary

antibodies (1:5,000; Cell Signaling Technology) were used and signals were detected by luminol.

EphA4 phosphorylation was calculated as the ratio of phosphorylated EphA4 to total EphA4 protein

expression.

13

Western Blots and Data Quantification

Mice were sacrificed by decapitation, and cortical tissues were immediately dissected and homogenized

in ice-cold HEPES buffer solution. After protein quantification, denatured lysates were

electrophoretically separated by 10% SDS-PAGE (10 g per lane) and transferred onto nitrocellulose

membrane. This was then probed at 4 C overnight with the following primary antibodies: rabbit anti-

GLAST (1:500; Abcam), guinea pig anti-GLT-1 (1:10,000; Millipore), mouse anti-GS (1:90,000; BD

Biosciences), rabbit anti-actin (1:500; Sigma-Aldrich) and mouse anti-tubulin (1:5,000; Millipore).

Horseradish peroxidase (HRP)-conjugated secondary antibodies (1:5,000; Cell Signaling Technology)

were used and detection was performed with luminol. Western blots were quantified using ImageJ

software. GLAST and GLT-1 levels were normalized to actin while GS level was normalized to tubulin.

Immunohistochemistry and Data Quantification

Brains were cryoprotected in 30% sucrose after fixation in 4% PFA for 2-4 hours. Serial 40 m sections

were collected and incubated with the following primary antibodies overnight at 4 C: rabbit anti-S100

(1:40,000; Dako), rabbit anti-GLAST (1:1,000; Abcam) and guinea pig anti-GLT-1 (1:5,000; Millipore).

Bright-field immunohistochemistry was followed by incubation with biotinylated secondary antibody

(1:400; Vector), avidin-biotin complex (ABC, Vector), and diaminobenzidine (Vector) for visualization.

Fluorescence immunohistochemistry was followed by incubation with Alexa Fluor 594-conjugated

secondary antibody (1:1,000; Invitrogen). Bright-field images were collected on Zeiss Axio Imager.M2,

using Axiovision software. Both numbers and cell body volumes of S100 -positive cells were obtained

using stereological image analysis software (StereoInvestigator, Microbrightfield). Confocal fluorescent

images were acquired using Leica SP5 confocal system.

14

Quantitative RT-PCR

Total RNA was isolated from the cortex of P28-P30 mice with Trizol (Invitrogen). 4 μg of total RNA

was reverse transcribed and amplified using the Super-Script III system (Invitrogen), and 5 ng cDNA

from this reaction was analyzed in triplicates using SYBR Green Supermix (Bio-Rad) and the Chrom4

system (Bio-Rad). The primers used were: GLT-1 forward, CCAAGCTGATGGTGGAGTTC; GLT-

1reverse, GTCCTTGATGGCGATGATCT; GLAST forward, GCCCTCCGACCGTATAAAAT;

GLAST reverse, GCCATTCCTGTGACGAGACT; GAPDH forward,

TGCCAAGTATGATGACATCAAGAAG; GAPDH reverse, TAGCCCAGGATGCCCTTTAGT. The

threshold cycle C(t) was determined for each sample using Biorad software Opticon 3.1. GAPDH was

used to normalize RNA content.

Cortical Slice Electrophysiology

Coronal slices (400 μm thick) of the somatosensory cortex were prepared from mice (P28-P30) in ice

cold sucrose cutting solution (75 mM NaCl, 2.5 mM KCl, 1.25 mM NaH2PO4, 26 mM NaHCO3, 25 mM

glucose, 75 mM sucrose, 2 mM MgSO4, and 0.5 mM CaCl2), and incubated in ACSF (119 mM NaCl,

2.5 mM KCl, 1.3 mM MgSO4, 1.0 mM NaH2PO4, 26.2 mM NaHCO3, 11 mM glucose, and 2.0 mM

CaCl2) for 30 min at 32°C and 1 h at room temperature before recording. Whole-cell recordings were

made from layer V pyramidal neurons morphologically identified under DIC. Synaptic currents were

evoked at 0.05 Hz using a concentric bipolar stimulating electrode placed in layer II/III of the same

whisker-barrel column, as visualized under bright-field microscope. To isolate NMDA receptor currents,

cells were voltage-clamped at -70 mV and perfused with a modified extracellular recording solution

lacking Mg2+

(119 mM NaCl, 2.5 mM KCl, 1.0 mM NaH2PO4, 26.2 mM NaHCO3, 11 mM glucose, 2.5

mM CaCl2, 300 mOsm), in the presence of 10 µM CNQX and 100 µM picrotoxin. To assess

concentration of cleft-glutamate, 250 μM L-APV was bath-applied in the same Mg2+

-lacking solution

15

with 10 µM CNQX and 100 µM picrotoxin, at a rate of 1 ml/min. In some recordings, the effect was

confirmed by wash-out of L-APV. All recordings were performed at room temperature, using a cesium-

based internal solution (either: 150 mM CeMeSO4, 1.3 mM MgCl2, 1 mM EGTA, 10 mM HEPES, and

0.1 mM CaCl2 or 122.5 mM Cs-Gluconate, 6.3 mM CsCl, 10 mM HEPES, 10 mM EGTA, 4 mM Mg-

ATP, 20 mM Na-phosphocreatine, and 0.3 mM NaGTP); cells where Ra exceeded 25 MΩ or changed

more than 25% over entire recording period were excluded from analysis. Analysis was done using

ClampFit. All recordings and analysis were performed by the experimenter blinded to the genotype of

the animals.

Statistics

P-values were calculated using the Student’s t-test. The numbers of mice analyzed were indicated in the

figure.

16

3. Supplemental Text

Elimination of Thin Spines Is Promoted in the Cortex of Ephrin-A2 KO Mice

It is generally believed that spine morphology is associated with synaptic strength and spine dynamics

(Alvarez and Sabatini, 2007; Hayashi and Majewska, 2005; Nimchinsky et al., 2002; Yuste and

Bonhoeffer, 2001). Previous studies have shown that thin spines are more susceptible to elimination than

mushroom and stubby spines (Holtmaat et al., 2005; Majewska et al., 2006). To determine whether

different types of spines were affected differently in ephrin-A2 KOs, we classified spines into four

categories based on their morphologies (i.e. mushroom, stubby, thin and other spines), and assessed

elimination rates of individual group in vivo. We found that while the proportional distribution of

different categories was unaltered in ephrin-A2 KOs, elimination of thin spines was selectively increased

compared with wild-type mice (Figure S2B). This result implies that the stability of a subset of synapses

is selectively affected in ephrin-A2 KOs.

The Ratio and the Dynamics of Filopodia Are Normal in Ephrin-A2 KO Mice

Dendrites of cortical neurons contain not only spines but also filopodia, which are long thin protrusions

without bulbous heads. Filopodia are highly dynamic compared with dendritic spines that the majority

of filopodia turned over on a daily basis (Xu et al., 2009; Zuo et al., 2005). We found that dendritic

protrusions in the cortex of both wild-type and ephrin-A2 KO mice were composed of similar

proportions of filopodia and dendritic spines. Specifically, 10.3±2.0% and 9.7±0.8% of the total

dendritic protrusions were filopodia in wild-type and KO mice, respectively (Figure S2C; P>0.3).

Furthermore, almost all the filopodia were eliminated within 1 day in wild-type mice and such dynamics

were unaltered in ephrin-A2 KOs (Figure S2D).

17

Ephrin-A2 Does not Colocalize with Postsynaptic Glutamate Receptors

To investigate the spatial relationship between ephrin-A2 and glutamate receptors at mouse cortical

synapses, we co-labeled ephrin-A2 with NR1 subunit of NMDA receptors and GluR2 subunit of AMPA

receptors on the postsynaptic membrane using array tomography. We found that the density of ephrin-

A2 puncta within 100 nm from the centers of NR1 and GluR2 was comparable to the density around

PSD95 (Figure S4D), suggesting that ephrin-A2 does not colocalize with postsynaptic glutamate

receptors.

Ephrin-A2 Colocalizes with Neuronal EphA4 Receptors at Cortical Synapses

EphA4 receptors have been demonstrated to interact with ephrin-A3 in the hippocampus (Murai et al.,

2003), and electron microscopic examination has revealed a perisynaptic location of EphA4 in the

cortex (Bouvier et al., 2008). Consistently, our AT analyses revealed that EphA4 receptors colocalized

with most pre- and post-synaptic neuronal markers but not astrocytic markers (Figure S4E). In addition,

the number of ephrin-A2 puncta within 100 nm from the centers of EphA4 receptors was much higher

than those from the centers of neuronal markers (Figure S4D), suggesting a potential interaction

between astrocytic ephrin-A2 and neuronal EphA4 receptors in the mouse cortex. However, the

phosphorylation level of EphA4 receptors in the cortex of ephrin-A2 KOs was not significantly different

from wild-type mice (Figure S4F), indicating that the phenotype we observed in ephrin-A2 KOs is

unlikely to be mediated by altered EphA4 signaling.

18

4. Supplemental References

Alvarez, V.A., and Sabatini, B.L. (2007). Anatomical and physiological plasticity of dendritic

spines. Annu Rev Neurosci 30, 79-97.

Bouvier, D., Corera, A.T., Tremblay, M.E., Riad, M., Chagnon, M., Murai, K.K., Pasquale, E.B.,

Fon, E.A., and Doucet, G. (2008). Pre-synaptic and post-synaptic localization of EphA4 and

EphB2 in adult mouse forebrain. Journal of neurochemistry 106, 682-695.

Harris, K.M., Jensen, F.E., and Tsao, B. (1992). Three-dimensional structure of dendritic spines

and synapses in rat hippocampus (CA1) at postnatal day 15 and adult ages: implications for the

maturation of synaptic physiology and long-term potentiation. J Neurosci 12, 2685-2705.

Hayashi, Y., and Majewska, A.K. (2005). Dendritic spine geometry: functional implication and

regulation. Neuron 46, 529-532.

Holtmaat, A.J., Trachtenberg, J.T., Wilbrecht, L., Shepherd, G.M., Zhang, X., Knott, G.W., and

Svoboda, K. (2005). Transient and persistent dendritic spines in the neocortex in vivo. Neuron 45,

279-291.

Majewska, A.K., Newton, J.R., and Sur, M. (2006). Remodeling of synaptic structure in sensory

cortical areas in vivo. J Neurosci 26, 3021-3029.

Murai, K.K., Nguyen, L.N., Irie, F., Yamaguchi, Y., and Pasquale, E.B. (2003). Control of

hippocampal dendritic spine morphology through ephrin-A3/EphA4 signaling. Nature

neuroscience 6, 153-160.

Nimchinsky, E.A., Sabatini, B.L., and Svoboda, K. (2002). Structure and function of dendritic

spines. Annual review of physiology 64, 313-353.

Triplett, J.W., Phan, A., Yamada, J., and Feldheim, D.A. (2012). Alignment of multimodal sensory

input in the superior colliculus through a gradient-matching mechanism. J Neurosci 32, 5264-5271.

Xu, T., Yu, X., Perlik, A.J., Tobin, W.F., Zweig, J.A., Tennant, K., Jones, T., and Zuo, Y. (2009).

Rapid formation and selective stabilization of synapses for enduring motor memories. Nature 462,

915-919.

Yuste, R., and Bonhoeffer, T. (2001). Morphological changes in dendritic spines associated with

long-term synaptic plasticity. Annual review of neuroscience 24, 1071-1089.

Zuo, Y., Lin, A., Chang, P., and Gan, W.B. (2005). Development of long-term dendritic spine

stability in diverse regions of cerebral cortex. Neuron 46, 181-189.