来客网

揭开脑的秘密:25年来中国内地学者第一次上《细胞》

声明:本文转载自 「新华网」, 或因排版与篇幅原因进行过编辑,内容未经本站独立核实,不代表本站立场、观点或建议。 如涉及版权问题,请联系我们,核实后立即删除。 [ 免责声明 ]

中国科学家进一步揭开脑的秘密(近25年来中国内地学者第一次在《细胞》杂志上发表论文)

新华网北京1月14日电(记者张小军)脑是世界上最奇妙的信息加工和传导装置。中国科学家最近在分子水平上确定了神经细胞极性的原理,为人类在更深层次上理解脑奠定了基础,也为治疗神经损伤提供了一种新的思路。

脑神经传导信息必须是有次序的,而实现这种有次序的信息传导是依靠神经细胞的特殊结构,即神经细胞的极性。所谓极性是指一个神经细胞通常有两种纤维,即树突和轴突,其中树突专门用来接收信号,轴突则只用来发送信号。科学家发现,每个神经元有多个树突,但轴突却只有一个。

由上海生命科学研究院神经科学研究所、中国科学院研究生院和北京生命科学研究所的科学家共同组成的研究小组在1月14日出版的美国《细胞》杂志上发表论文说,他们在体外培养海马神经元时发现,神经元一般先长出几个差不多的短突起,这些突起中的某一个会在发育后期迅速生长,而其他的突起长度维持不变。这个迅速生长的突起最终会成为轴突,而其他的突起则分化为树突。神经元是如何选出将成为轴突的突起,又是如何抑制其他突起生长的呢?

此篇论文的第一作者蒋辉介绍,他们进一步研究发现,一个称为GSK的蛋白激酶的活性对确定神经细胞的发育起着关键作用。轴突中的GSK活性比树突中的要低,而在未分化的神经元突起中,GSK活性比较均匀。研究人员提高GSK活性,就可以抑制轴突形成,而降低GSK活性就可以一反常规地在单个神经元上形成多个轴突。令人惊奇的是,在降低GSK活性的情况下,树突也可再变成轴突。研究人员还发现了调节GSK活性的上游分子。该论文通讯作者饶毅说,这项研究除了发现脑神经信息有次序传导的分子基础以外,还说明通过适当

的药物可以增加轴突数量,帮助神经损伤患者修复神经。

《细胞》杂志是国际生命科学的顶尖刊物,在科学界与《自然》、《科学》等期刊齐名。这是近25年来中国内地学者第一次在《细胞》杂志上发表论文。

《细胞》杂志发表中国脑神经研究成果

福康

光明日报 日期: 2005年1月14日

本报讯 1月14日出版的美国生命科学权威杂志《细胞》发表了中国科学家的一项研究成果,他们发现蛋白质的活性对确定神经细胞极性起关键作用,开拓了治疗神经损伤的新途径。

《细胞》杂志发表的论文第一作者是研究生蒋辉,还有研究生郭纬和梁新华,通讯作者是旅美中国学者饶毅教授。他们来自中科院上海生命科学研究院神经科学研究所、中科院研究生院和北京生命科学研究所等。

脑神经复杂的信息传导有赖神经细胞的特殊结构,神经细胞通常有两种结构接收讯号的树突、发送讯号的轴突。他们的研究揭示了神经细胞的这两个基本极性是如何形成的 一种称为GSK的蛋白激酶在发育过程中的分布有极性,在轴突中的活性比树突中要低。如果其活性太高,神经细胞会没有轴突,太低则会把树突变成轴突。它的活性由上游分子来调控,多个分子形成通路,控制着神经细胞的极性。

他们还发现,确定神经细胞极性的GSK还维持着极性。如果用药物改变GSK活性,就可以把树突变成轴突。这表明,可以通过药物来增加轴突数量,这可能成为神经元损伤修复和退行性神经系统疾病治疗的新突破口。

《细胞》杂志是国际权威刊物,对论文的要求很高,在学术界影响巨大。它曾在1980年发表过一篇中国大陆学者的论文,如今是25年后再次发表中国大陆学者论文,论文的全部研究都在中国进行。

----------------------------------------------------------------------

Cell, Vol 120, 123-135, 14 January 2005

Both the Establishment and the Maintenance of Neuronal Polarity Require Active Mechanisms

Critical Roles of GSK-3 and Its Upstream Regulators

Hui Jiang1, 2, Wei Guo1, 2, Xinhua Liang1 and Yi Rao, 1, 2, 3, ,

1 Institute of Neuroscience, Shanghai Institutes of Biological Sciences,

The Graduate School, Chinese Academy of Sciences, 320 Yueyang Road,

Shanghai 200031, China 2 National Institute of Biological Sciences,

Beijing, 7 Zhongguanchun Life Sciences Park, Beijing 102206, China

3 Department of Neurology, Northwestern University Feinberg School of

Medicine, 303 E. Chicago Avenue, Ward 10-185, Chicago, IL 60611, USA

Received 12 April 2004; Revised 8 September 2004; accepted 10 December

2004 Published: January 13, 2005 Available online 13 January 2005.

Abstract

Axon-dendrite polarity is a cardinal feature of neuronal morphology

essential for inxxxxation flow. Here we report a differential

distribution of GSK-3 activity in the axon versus the dendrites. A

constitutively active GSK-3 mutant inhibited axon xxxxation, whereas

multiple axons xxxxed from a single neuron when GSK-3 activity was

reduced by pharmacological inhibitors, a peptide inhibitor, or siRNAs.

An active mechanism for maintaining neuronal polarity was revealed by

the conversion of preexisting dendrites into axons upon GSK-3

inhibition. Biochemical and functional data show that the Akt kinase and

the PTEN phosphatase are upstream of GSK-3 in determining neuronal

polarity. Our results demonstrate that there are active mechanisms for

maintaining as well as establishing neuronal polarity, indicate that

GSK-3 relays signaling from Akt and PTEN to play critical roles in

neuronal polarity, and suggest that application of GSK-3 inhibitors

can be a novel approach to promote generation of new axons after

neural injuries.

Introduction

The morphological polarity of neurons with crucial functional

implications is an essential component of the neuron doctrine

established by Cajal and Ramon (1911). Typically, a neuron has a

single axon and multiple dendrites. Dendrites usually receive signals,

whereas the axon usually sends signals. Neuronal polarity is therefore

important for information processing and ensures unidirectional signal

flow in the nervous system or between the nervous and other systems.

Work in the past two decades has shown that, in addition to the

well-known morphological and functional polarities, there are also

molecular and organelle differences between the axons and the dendrites,

ranging from the preferential sorting of proteins and mRNAs to the

differential distribution of polyribosomes (Caceres et al., 1986;

Dotti and Banker 1987; Goslin et al. 1988; Kleiman et al. 1990 and

Steward 2002). How neuronal polarity xxxxs remains an interesting and

challenging question.

A well-established model for studying neuronal polarity is the pyramidal

neurons from the mammalian hippocampus (Craig and Banker, 1994).

Hippocampal neurons become polarized in successive steps (Banker and

Cowan 1977; Banker and Cowan 1979 and Dotti and Banker 1987). Shortly

after culturing, a neuron extends lamellipodia around the soma (stage

[st.] 1). It then extends several minor neurites in st. 2. At st. 3, one

neurite is significantly longer and becomes the axon, whereas the

others become dendrites. Experimental manipulations of neurite length

can reset neurite competition (Dotti and Banker, 1987). Exposure of a

neurite to extracellular adhesive molecules (Lein et al. 1992 and Esch

et al. 1999) or stimulation of a neurite by mechanical tension

promotes neurite growth and favors axon xxxxation (Lamoureux et al.,

2002). Cytoskeleton dynamics has been implicated in breaking symmetry

between stages 2 and 3 (Baas et al. 1988 and Ahmad et al. 1994).

Molecules affecting actin dynamics ( Bradke and Dotti, 1999; but see

also Ruthel and Hollenbeck, 2000) or microtubule (MT) dynamics such as

the collapsin response mediator protein-2 (CRMP-2) (Inagaki et al.

2001 and Fukata et al. 2002) regulate the xxxxation of axon-dendrite

polarity. The evolutionary conserved polarity complex composed of

Par3/Par6/aPKC was first found in C. elegans for their roles in

establishing the anterior-posterior polarity of the first blastomere.

Par3/Par6, as well as two small GTPases CDC42 and Rap1B, have recently

been implicated in the establishment of neuronal polarity in mammals

(Shi et al. 2003; Nishimura et al. 2004b and Schwamborn and Puschel

2004), though not in Drosophila (Rolls and Doe, 2004).

Glycogen synthase kinase-3 (GSK-3) is a multifunctional serine

(Ser)/threonine (Thr) kinase found ubiquitously in eukaryotes (Embi et

al. 1980 and Woodgett 1990). Biochemically, GSK-3 has a high basal

activity. Signaling pathways act by inhibiting GSK-3 activity.

Phosphorylation of Ser at position 9 in GSK-3 inhibits the activities of

GSK-3. GSK-3 is now known to be important in many biological processes,

ranging from the canonical Wnt signaling pathway, MT dynamics, to

astrocyte migration (Doble and Woodgett 2003 and Etienne-Manneville

and Hall 2003).

In the course of our work on axon guidance and neuronal migration (Li et

al. 1999; Wong et al. 2001 and Liu et al. 2004), we considered the

possibility that molecules implicated in directional guidance could be

involved in the development of neuronal polarity. We have obtained

evidence that GSK-3 played a regulatory role in establishing and

maintaining the axon-dendrite polarity. We further provide evidence that

Akt and PTEN function upstream of GSK-3, while the accompanying paper

provides evidence that GSK-3 directly inactivates CRMP-2 (Yoshimura et

al., 2004 [this issue of Cell]). Taken together, these findings indicate

that GSK-3 is a central regulator in a signaling pathway for neuronal

polarity.

Results

Polarized Distribution of a Phosphorylated xxxx of GSK-3 in

Hippocampal Neurons

We used hippocampal neurons to investigate a potential role of GSK-3

in neuronal polarity. Hippocampal explants were isolated from

embryonic day 18 (E18) rats and pyramidal neurons were cultured

(Banker and Cowan 1977; Dotti and Banker 1987 and Brewer et al. 1993).

To determine the distribution of GSK-3 in hippocampal neurons, we

double-labeled neurons with an antibody against total GSK-3 and

another one against GSK-3 phosphorylated at Ser 9 (referred to as pGSK-3

hereafter). GSK-3 was detected in all neurites, both in st. 2

nonpolarized neurons (Figure 1A, in green) and in st. 3 polarized

neurons (Figure 1B, in green). Interestingly, although pGSK-3 is present

in the tips of all neurites in st. 2 nonpolarized neurons ( Figure 1A,

in red and Figure 1C), there is more pGSK-3 present in the tips of

axons than that in the tips of dendrites in st. 3 polarized neurons (

Figure 1B, in red and Figure 1D). The ratio of pGSK-3/GSK-3 in the

tips of st. 3 axons was significantly higher than that in the tips of

st. 3 dendrites (Figure 1E).

(94K)

Figure 1. Distribution of GSK-3 and Phospho-GSK-3 (Ser 9) in Hippocampal

Neurons(A and B) Distribution of GSK-3 and pGSK-3 in st. 2 and st. 3

neurons, respectively. The scale bars are 20 m in this paper unless

specified otherwise. The small panels (scale bars of 5 m) on the right

part of (B) are higher magnification pictures of the tips of axons and

dendrites.(C and D) Fluorescence intensity after anti-pGSK-3

immunostaining of the st. 2 and st. 3 neurons.(E) pGSK-3 and GSK-3

levels in neurite tips of st. 3 neurons. Left panel: the average level

of pGSK-3 in dendrites was normalized as 1.0 ± 0.13 (n = 42); the

relative level in axons is 5.8 ± 0.38 (n = 16; p < 0.001).="" right="">

the average pGSK-3/GSK-3 ratio in dendrites was normalized as 1.0 ±

0.11 (n = 42), while that in axons was 2.4 ± 0.16 (n = 16, p <>

(F) Specificity of anti-pGSK-3 (Ser 9) antibody, which does not

recognize the GSK-3 S9A mutant transfected into HEK293 cells. IP:

immunoprecipitation. IB: immunoblotting.(G) Distribution of pGSK-3

(Tyr216) and pGSK-3 (Ser 9) in st. 3 neurons.

GSK-3 also has a tyrosine phosphorylation site at 216 (Tyr216) which

increases GSK-3 activity (Hughes et al. 1993 and Wang et al. 1994). No

differential distribution of GSK-3 Tyr216 was detected in polarized

hippocampal neurons (Figure 1G).

Inhibition of Axon xxxxation after Increasing GSK-3 Activity

To test for a functional role of GSK-3 in the xxxxation of neuronal

polarity, we used a constitutively active GSK-3 mutant S9A, in which Ser

9 was replaced with alanine. Dissociated hippocampal neurons were

transfected with a plasmid for the green fluorescent protein (GFP) or

cotransfected with a plasmid for GFP and another expressing GSK-3 S9A.

Neurons were cultured for 5 days before analysis of the axon-dendrite

polarity by the relative length of axons versus dendrites and by

staining with multiple specific antibodies. MAP2 is a somatodendritic

marker for all dendrites and the proximal part of axons (Caceres et al.,

1984; Figure 2A), whereas the Tau-1, the anti-Synapsin I, and the

anti-GAP43 antibodies recognize the distal part of the axon (Mandell and

Banker, 1996; Figure 2A).

(72K)

Figure 2. Inhibition of Axon xxxxation by a Constitutively Active xxxx

of GSK-3(A) GFP-transfected neurons showing expression of the dendrite

marker (MAP2) and axon markers (Tau-1, GAP43, and Synapsin I). MAP2 is

localized in the soma, the dendrites, and the proximal part of the axon,

while the axon markers are in the soma and the axon.(B) Neurons

transfected with GSK-3 S9A were stained with dendrite and axon markers.

GSK-3 S9A transfected neurons lacked the axon, and axonal markers

were mislocalized to all remaining neurites.(C) Quantification of

polarity defects. The p values of differences between control and

GSK-3 S9A were: <0.001 in="" the="" percent="" of="" neurons="" with="" a="" single="">

and <0.001 in="" those="" with="" no="" axons.="" numbers="" were="" discussed="" in="" the="">

(D) Neurite numbers per neuron. GFP alone (n = 56): 1.07 ± 0.07 axon

and 4.2 ± 0.22 dendrites. GSK-3 (n = 48): 1.04 ± 0.04 axon and 3.95 ±

0.27 dendrites. GSK-3 KD (n = 54): 1.04 ± 0.06 axon and 3.96 ± 0.24

dendrites. The numbers of axons and dendrites of neurons transfected

with GFP and GSK-3 S9A were counted in two categories: (S) was neurons

with single axons while (N) was neurons with no axons. Separate analysis

should give a more accurate account of the effect of GSK-3 S9A

because not all neurons positive for GFP were affected, most likely

due to variations in GFP and GSK-3 S9A cotransfection efficiencies in

different neurons.(E) The average axon length of GFP-transfected neurons

was normalized as 100%. The p values were <0.01 between="" the="" control="">

GSK-3 and <0.001 between="" the="" control="" and="" gsk-3="" s9a="" (s).(f)="">

average dendrite length of GFP-transfected neurons was normalized as

100%. The p value was <0.01 between="" the="" control="" and="" gsk-3="" s9a="">

Neurons were separately analyzed as into those with a single axon (S),

those with no axon (N), or those with multiple axons (M).

Control hippocampal neurons (transfected with GFP alone) developed

normal polarity, with most of them having a single axon and multiple

dendrites and a small percentage of neurons with either no or multiple

axons (Figure 2C). GSK-3 S9A significantly inhibited neuronal polarity

(Figure 2C). Only 7.6% of control neurons has no axons, whereas

approximately 49% of neurons transfected with GSK-3 S9A has no axon (p

0.001), as exemplified by the neuron in Figure 2B that had no axon,

with MAP2 in the entire length of every neurite. Without an axon, the

axonal markers were mislocalized to all neurites (Figure 2B). These

neurites did not have recycling of synaptic vesicles ( Figure 4E), a

functional characteristic of axons.

GSK-3 S9A led to the absence of the axon but did not affect the number

of dendrites (Figure 2D). Dendrites were significantly longer in neurons

when axons were eliminated by GSK-3 S9A (Figure 2F), perhaps due to the

redistribution of membrane and cytoplasmic materials to the dendrites

in the absence of an axon.

Neither the wild-type GSK-3 nor a kinase-dead xxxx of GSK-3 affected

axon-dendrite polarity (Figures 2C and 2D). Wild-type GSK-3 reduced

the length of axons (Figure 2E) but did not affect the number or the

length of dendrites (Figure 2F).

xxxxation of Multiple Axons after Inhibition of GSK-3 Activity by

Pharmacological and Peptide Inhibitors and shRNAs

To further test for a role of GSK-3 in neuronal polarity, we used

pharmacological inhibitors of GSK-3: SB216763, SB415286 (Coghlan et al.,

2000), and lithium chloride (LiCl). Hippocampal neurons were treated

with each inhibitor or the control vehicle dimethyl sulfoxide (DMSO) and

were cultured for 5 days. Both SB216763 and SB415286 significantly

decreased the number of neurons with a single axon but increased the

number of neurons with multiple axons (Figures 3A and 3B). LiCl also

caused the xxxxation of multiple axons (Figure 3A), although not as

efficiently as the SB compounds (Figure 3B). The total number of

neurites per neuron was not significantly changed by SB216763, SB415286,

or LiCl: they increased the number of axons and decreased the number of

dendrites (Figure 3C), indicating the xxxxation of axons at the expense

of dendrites. Perhaps reflecting a limited amount of membrane and

cytoplasmic materials during neurite outgrowth, the length of axons

was shorter (Figure 3D) when multiple axons xxxxed (all cases marked

with (M) in Figure 3D). Shortening of axons was observed in all

experiments, described later in this paper, when multiple axons xxxxed.

The length of dendrites was not affected by the pharmacological

inhibitors (data not shown).

(72K)

Figure 3. xxxxation of Multiple Axons upon GSK-3 Inhibition by

Pharmacological and Peptide Inhibitors(A) Neurons cultured in the

presence of inhibitors (5 M SB216763, 10 M SB415286, or 1 mM LiCl)

were examined for expression of Tau-1 (in green) and MAP2 (in red).(B)

Quantification of polarity defects. The p values were: <0.001>

the control and SB216763, <0.001 between="" the="" control="" and="" sb415286,="">

<0.01 between="" the="" control="" and="" licl.(c)="" numbers="" of="" neurites="" per="">

(N): no axon. (S): single axon. (M): multiple axons. The p values of the

difference were: <0.001 between="" dmso="" and="" sb216763="" (m),=""><0.001>

DMSO and SB415286 (M), and <0.001 between="" dmso="" and="" licl="" (m).(d)="">

average axon length of control (DMSO) neurons was normalized as 100%.

The p values were: <0.001 between="" dmso="" and="" sb216763="" (m),=""><>

between DMSO and SB415286 (M), and <0.001 between="" dmso="" and="" licl="">

(E) Axon and dendrite markers in neurons transfected with GFP and

GID5-6.(F) Quantification of polarity defects. p values were: <>

between GFP and GID5-6 (single axon), <0.001 between="" gfp="" and="">

(multiple axons), <0.001 between="" gid5-6="" and="" gid5-6lp="" (single="">

and <0.001 between="" gid="" 5-6="" and="" gid="" 5-6lp="" (multiple="" axons).(g)="">

neurons with single axons. (M): neurons with multiple axons. The p

values were <0.001 between="" the="" gfp="" control="" and="">

We then used GID5-6, a peptide inhibitor of GSK-3 derived from the GSK-3

interaction domain of axin (Hedgepeth et al., 1999). The control

GID5-6LP contained a single amino acid mutation rendering it unable to

interact with GSK-3 (Hedgepeth et al. 1999 and Zhang et al. 2003).

GID5-6 led to multiple axon xxxxation, as evidenced by multiple

markers ( Figures 3E and 3F and Supplemental Figure S1A at http://www.

cell.com/cgi/content/full/120/1/123/DC1/) and functional

characterization by vesicle recycling (Figure 4F). The total number of

neurites per neuron was not changed by GID5-6 (Figure 3G). Neurons

transfected with GID5-6LP developed normal polarity (Figure 3F).

(106K)

Figure 4. Effects of Specific siRNA Constructs on Hippocampal Neurons

and Functional Characterization of Axons(A) Localization of MAP2 in a

neuron transfected with GFP and GSK-3HP2.(B) Quantification of

polarity defects. The p values were: <0.001 between="" gsk-3hp1="" and="">

and <0.001 between="" gsk-3hp2="" and="" xash3hp.(c)="" the="" average="" gsk-3="">

in the neurites of control neurons (XASH3HP) was normalized as 100%. The

p values were: <0.001 between="" the="" control="" and="" gsk-3="" hp1="" or="" hp2.="">

single axon. (M): multiple axons.(D) A hippocampal neuron transfected

with GFP was cultured for 7 days before being loaded with FM4-64 in 45

mM K+ for 1 min. The neuron developed a single axon, which took up

FM4-64. Boxed region in the left panel is shown in higher

magnification in the lower right panels. The soma and dendrites had

little uptake of FM4-64 (upper right panels).(E) A GSK-3 S9A transfected

neuron took in little FM4-64 in their soma or any of the short

neurites.(F) A GID5-6 transfected neuron with multiple axons, which were

positive in FM4-64 uptake (boxed region in the left panel). The two

axons boxed in the left panel are shown in higher magnification in the

right panel: both were positive in FM dye uptake and release. Images

were taken at the indicated time points before and after stimulation

by 90 mM K+. The scale bars are 10 m.(G) Quantification of neurites,

with the axon defined as neurites positive in FM4-64 uptake, and

dendrites as neurites negative in dye uptake. The p values were: <>

001 between GFP and GID5-6 (M). (N): no axon. (M): multiple axons.

Finally, we used GSK-3HP1 and GSK-3HP2, two short hairpin RNA (shRNA)

constructs designed specifically against GSK-3 (Yu et al., 2003). The

control construct XASH3HP did not affect neuronal polarity ( Figures

4B and 4C, upper panels). Either GSK-3HP1 or GSK-3HP2 led to the

xxxxation of multiple axons (see Figures 4A and 4B) at the expense of

dendrites (Supplemental Figure S2A on the Cell website). When GSK-3

expression was examined (Figure 4C), reduction of GSK-3 expression

correlated with multiple axon xxxxation. Thus, GSK-3 expression was

significantly reduced in GFP-positive neurons that xxxxed multiple

axons, GSK-3HP1 (M) and GSK-3HP2 (M), but not significantly reduced in

those that had a single axon, GSK-3HP1 (S) and GSK-3HP2 (S).

Functionality of Multiple Axons Induced by GSK-3 Inhibition

Although multiple axons are usually recognized by the relative length

and expression of molecular markers, it is important to examine

whether multiple axons are functional. A basic feature of functional

axons is that they have active synaptic vesicle recycling which can be

monitored by visualizing the FM4-64 dye uptake and release (Ryan et al.,

1993). The FM dye was taken into the axon of control hippocampal

neurons after stimulation by potassium (K+) (Figure 4D), whereas

little FM4-64 was taken into the somatodendritic region (Figure 4D). The

GSK-3 inhibitor GID5-6 caused the xxxxation of multiple axons

positive for FM4-64 uptake after stimulation by 45 mM K+ for 1 min

(Figures 4F and 4G and Supplemental Figure S2B on the Cell website),

whereas the dendrites did not take up the dye (Supplemental Figure S2B

on the Cell website and Figure 4G). Vesicle recycling was observed:

after the initial FM dye loading, treatment with 90 mM K+ depleted it

from two axons of a single GID5-6 transfected neuron (Figure 4F).

Similar results were obtained with axons after GSK-3HP2 transfection

(data not shown). These results indicate that multiple axons induced

by GSK-3 inhibition are functional in vesicle recycling.

We also examined FM dye uptake by neurites from neurons transfected with

GSK-3 S9A with the same assay and found that neurites remaining after

GSK-3 S9A transfection were negative in uptake (Figures 4E and 4G),

confirming the earlier conclusion that GSK-3 S9A eliminated axons.

Role of GSK-3 in Establishing Neuronal Polarity

Establishment of neuronal polarity can be studied by examining the

effect of manipulations before or up to st. 3, because most neurons

reached st. 3 with axon-dendrite polarity established after culturing

for 48 hr under the present conditions (Brewer et al., 1993). To study

the role of GSK-3 in establishing neuronal polarity, neurons were

treated with SB415286 for the first 24 hr and then removed. Neurons were

cultured until day 6 (144 hr) before polarity examination. SB415286

caused the xxxxation of multiple axons even when it was applied only for

the first 24 hr (Figures 5A and 5B), indicating a role for GSK-3 in

establishing axon-dendrite polarity.

(79K)

Figure 5. Roles of GSK-3 in Establishing and Maintaining Neuronal

Polarity(A) Neurons were treated with SB415286 from 0–24 hr, 72–144

hr, or 0–144 hr before being fixed for immunocytochemistry at day 6

(144 hr) with Tau-1 (in green) and MAP2 (in red).(B) Quantification of

polarity defects. The p values were: <0.001 between="" dmso="" and="">

for 0–24, or 72–144, or 0–144 hr.(C) Conversion of a preexisting

dendrite into an axon by SB415286. The left panel shows a neuron on

day 3, with one presumptive axon (a long neurite marked by a white

arrow) and several dendrites (the tip of two dendrites marked by

arrowheads). The middle and right panels show the same neuron on day 6.

The long neurite indicated by the white arrow was indeed an axon (as

shown by Tau-1 staining). One of the short neurites (indicated by the

yellow arrowhead) in the leftmost panel on day 3 had grown by day 6 into

an axon in the right panel. The red arrow points to the tip of the

new axon.(D) Neurite dynamics. St. 2 neurons were imaged by time lapse

microscopy at an interval of 10 min for 2 hr. The % of time spent by

each neurite on retraction, pause, or growth was recorded. The p

values of differences between control and SB415286 were: retraction,

<0.001; pause,=""><0.05; and="" growth,=""><0.05.(e) dynamics="" of="" neurites="">

a representative neuron treated with DMSO (upper panel) or SB415286

(lower panel).

评论 (0)