ORIGINAL ARTICLE

Advanced glycation end products induce moesin phosphorylationin murine retinal endothelium

Lingjun Wang • Qiaoqin Li • Jing Du •

Bo Chen • Qiang Li • Xuliang Huang •

Xiaohua Guo • Qiaobing Huang

Received: 12 January 2011 / Accepted: 2 February 2011 / Published online: 17 February 2011

� Springer-Verlag 2011

Abstract Increase in vascular permeability is the most

important pathological event during the development of

diabetic retinopathy. Deposition of advanced glycation end

products (AGEs) plays a crucial role in the process of

diabetes. This study was to investigate the role of moesin

and its underlying signal transduction in retinal vascular

hyper-permeability induced by AGE-modified mouse

serum albumin (AGE-MSA). Female C57BL/6 mice were

used to produce an AGE-treated model by intraperitoneal

administration of AGE-MSA for seven consecutive days.

The inner blood–retinal barrier was quantified by Evans

blue leakage assay. Endothelial F-actin cytoskeleton in

retinal vasculature was visualized by fluorescence probe

staining. The expression and phosphorylation of moesin in

retinal vessels were detected by RT–PCR and western

blotting. Further studies were performed to explore the

effects of Rho kinase (ROCK) and p38 MAPK pathway on

the involvement of moesin in AGE-induced retinal vascu-

lar hyper-permeability response. Treatment with AGE-

MSA significantly increased the permeability of the retinal

microvessels and induced the disorganization of F-actin in

retinal vascular endothelial cells. The threonine (T558)

phosphorylation of moesin in retinal vessels was enhanced

remarkably after AGE administration. The phosphorylation

of moesin was attenuated by inhibitions of ROCK and p38

MAPK, while this treatment also prevented the dysfunction

of inner blood–retinal barrier and the reorganization of

F-actin in retinal vascular endothelial cells. These results

demonstrate that moesin is involved in AGE-induced reti-

nal vascular endothelial dysfunction and the phosphoryla-

tion of moesin is triggered via ROCK and p38 MAPK

activation.

Keywords Advanced glycation end products �Blood–retinal barrier � Moesin � ROCK � p38MAPK

Introduction

As one of the most common microvascular complications of

diabetes, diabetic retinopathy (DR) is the leading cause of

visual loss, even blindness in working adult [1, 2]. The

pathogenesis of DR is complicated by the multi-involve-

ments of angiogenesis factor and inflammatory mediators

[3], characterizing with both the loss of integrity of blood–

retinal barrier function and the impairment in angiogenesis

[4]. The breakdown of blood–retinal barrier, eventually the

retinal vascular exudation are not only the important early

events in these angiogenic and inflammatory processes, but

also an adverse consequent disorder as macular edema as

well [1, 5, 6]. The blood–retinal barrier relies on the integrity

of endothelial function, as well as an intact extracellular

matrix [7, 8]. There has been an increasing focus on the

underlying molecular mechanisms involved in the develop-

ment of DR. It is proposed that the formation and accumu-

lation of advanced glycation end products (AGEs) in blood

are contributed to the disruption of vascular endothelial

barrier function during the development of DR [9, 10].

Resulting from the non-enzymatic reaction between

glucose and macromolecules, AGEs educe the pathogenic

effects by binding with receptor for AGE (RAGE) and

evoke the inflammatory activation through a series of

L. Wang � Q. Li � J. Du � B. Chen � Q. Li �X. Huang � X. Guo � Q. Huang (&)

Department of Pathophysiology,

Key Lab for Shock and Microcirculation Research,

Southern Medical University, 1023, Shatai Road,

Tonghe, 510515 Guangzhou, People’s Republic of China

e-mail: huangqb2000@yahoo.com

123

Acta Diabetol (2012) 49:47–55

DOI 10.1007/s00592-011-0267-z

signal transductions, such as, p21ras, mitogen-activated

protein kinases (MAPKs), and nuclear factor-j B [11, 12].

But the detail and dominant molecular pathways remain to

be elucidated. Our previous studies have demonstrated that

AGEs significantly disorganized endothelial F-actin cyto-

skeleton and VE-cadherin distribution and increased

endothelial monolayer permeability [13, 14]. Ezrin-radix-

in-moesin (ERM) proteins are known to work as linkers to

regulate actin-membrane interactions in a signal-dependent

manner. We have also shown that AGEs elicited a complex

signaling system which included the activations of Rho

kinase (ROCK) and p38a MAPK and the threonine phos-

phorylation of ERM proteins, especially moesin in human

dermal microvascular endothelial cells (HMVECs). The

activations of moesin and ERM are then linked with

F-actin, resulting in the reorganization of cytoskeleton and

the disruption of endothelial barrier function [15, 16].

By observing the Evans blue leakage and the arrange-

ment of F-actin in retinal vascular endothelial cells in an

AGE-stimulated mouse model, this present study is to

further verify the involvement of moesin phosphorylation

in situ and the effects of ROCK and p38 MAPK activations

in the development of AGE-induced blood–retinal barrier

dysfunction. The results indicate that ERM protein moesin

is involved in AGE-induced retinal vascular endothelial

dysfunction and the phosphorylation of moesin is triggered

via ROCK and p38 MAPK activations.

Materials and methods

Materials

Mouse serum albumin (MSA, fraction V), D-glucose, and

Evans blue were purchased from Sigma (St. Louis, MO,

USA). Antibody recognizing phospho-moesin (T558) was

obtained from Santa Cruz Biotechnology (CA, USA). Anti-

bodies recognizing total moesin, total and phospho-p38

MAPK were purchased from CST (USA). ROCK I antibody

was purchased from Chemicon (USA) and phospho-ROCK I

antibody was from Upstate (USA). Rho kinase inhibitor

Y-27632 and p38 inhibitor SB203580 were acquired from

Calbiochem (San Diego, CA, USA). Rhodamine-phalloidin

was purchased from Molecular Probe (Carlsbad, CA, USA).

Complete protease inhibitor cocktail tablets and PhosSTOP

phosphatase inhibitor cocktail tablets were purchased from

Roche (Shanghai, China). Other chemicals were purchased

from Sigma (St. Louis, MO, USA) unless otherwise indicated.

Preparation of AGE-MSA

AGE-MSA was prepared according to the protocol of Hou

et al. [17]. Briefly, 3.5 mg/ml mouse serum albumin was

incubated in phosphate buffer solution (PBS, pH 7.4)with

100 mmol/l of D-glucose for 10 weeks at 37�C in a sterile

environment. Control albumin was incubated in the same

condition without glucose. The solutions were then

extensively dialyzed against PBS and concentrated with

Millipore 15-ml ultrafiltration column. Endotoxin was

removed by using a Pierce Endotoxin Removing Gel.

AGE-specific fluorescence was determined by using ratio

spectrofluorometry [15, 18, 19]. The AGE-MSA contained

72.50 arbitrary units of AGEs in per gram of protein, while

native albumin contained only 0.85 arbitrary units of AGEs

in per gram of protein.

Animals and anesthesia

Female C57/BL6 mice used in this experiment were

obtained from Laboratory Animal Center of Southern

China. The research protocols were approved by the Ani-

mal Care Committee of Southern Medical University of

China and performed in adherence to National Institute of

Health guidelines. The mice were anesthetized with an

intramuscular injection of 13.3% ethyl carbamate plus

0.5% chloralose (0.65 ml/kg) before all operational

manipulations.

AGE-MSA treatment of mice

Mice of 10–12 weeks old were randomly assigned to four

groups of equal size with n = 6 in each group. According

to the protocol of Stitt AW et al. [20], mice in control or

AGE-treated group were injected intraperitoneally (i.p.)

with either native mouse serum albumin (MSA, 10 mg/kg)

or AGE-MSA (10 mg/kg) daily for seven consecutive days.

SB203580 (1 mg/kg) or Y-27632 (5 mg/kg) were injected,

respectively, i.p. 30 min before every AGE-MSA admin-

istration in p38 or ROCK inhibition group.

Phosphorylations of moesin, p38 MAPK, and ROCK

in retinal microvascular cells

Retinas were carefully isolated, and stratum pigmenti retinae

was dissected and discarded for the collection of retinal

choriocapillary. Multiple retinas from the same group were

collected to obtain sufficient protein samples. Cells were

extracted by lysing and sonicating in lysis buffer with pro-

tease phosphatase inhibitors. Samples were subjected to

SDS–PAGE, and proteins were transferred to polyvinylidene

fluoride (PVDF) membranes. Blots were blocked and incu-

bated with 1:1,000 dilution of primary antibody of interest

overnight at 4�C on a rocker. After three washes for 5 min

each with TPBS, the blots were incubated with a 1:1,000

dilution of HRP-conjugated species-specific respective

secondary antibody (Dako Ltd, Ely, UK) for 1 h at room

48 Acta Diabetol (2012) 49:47–55

123

temperature. After three washes for 5 min each with TPBS,

protein bands were visualized by chemiluminescence.

Immunohistochemical analysis of moesin expression

in retina

Whole-mount of eyeball was conducted by fixing the

sample in 10% neutral-buffered formalin for 12–24 h at

room temperature and embedding in paraffin. Immuno-

histochemistry was carried out using standard techniques

by using anti-moesin or anti-phospho-moesin antibodies,

and biotin-free horseradish peroxidase (HRP) enzyme-

labeled polymer with 3,30-diaminobenzidine (EnVision

plus detection system, Dako Ltd, Ely, UK).

RT–PCR quantification of moesin mRNA expression

in retina

RT–PCR amplification was performed with moesin-specific

primers synthesized by Shanghai Yingjun Bio-tech Com-

pany. Moesin forward primer sequence was 50 cac tgt gct

gga gcc act aa 30; Moesin reverse primer sequence was

50 aac caa aag gaa tgc gtg tc; Mouse b-actin was used as

control with forward primer sequence as 50 tca tca cta ttg

gca acg agc 30 and reverse primer sequence as 50 aac agt ccg

cct aga agc ac30. Two microliters of RNA at the concen-

tration of 0.5 g/l was used as templates for cDNA synthesis

in the reverse transcriptase (RT) reaction. After an initial

4-min denaturation at 95�C, the cDNA was amplified for 30

cycles and then annealed at 58�C for moesin and b-actin. At

the end of the annealing process, 2 min of elongation phase

was followed by a single extension phase of 3 min at 72�C.

The PCR products were separated on 1.5% agarose gels,

and the PCR products are 227 base pairs for mouse moesin

and 399 base pairs for b-actin.

Quantification of inner blood–retinal barrier function

Evans blue leakage assay was applied to quantify the inner

blood–retinal barrier function according to Qaum’s and

Moore’s protocols [21–23]. Briefly, mouse was anesthe-

tized and femoral artery was cannulated. Evans blue was

injected through tail vein over 10 s at a dosage of 45 mg/kg,

and the uptake and distribution of the dye was confirmed by

immediately visible blue color in mouse. Twenty microli-

ters of blood was withdrawn from cannulated femoral artery

at 1, 15, 30, 45, and 60 min, respectively, to obtain the time-

average Evans blue levels, and the same amount of saline

was re-infused back to mouse each time. After the dye had

circulated for 60 min, the chest cavity was opened, and

kalium citricum was perfused through left ventricle while

right ventricle was cut open to allow the blood flushing out

for 2 min. Then, both eyes were enucleated and the retinas

were carefully dissected and thoroughly dried in a vacuum

dryer. Evans blue was extracted by incubating each retina in

65 ll formamide for 18 h at 70�C. The supernatant was

obtained by centrifuging in 7,000 rpm. The concentration of

Evans blue was measured by absorbance at 632 nm. The

amount of dye in the extracts was calculated from a standard

curve of Evans blue in formamide and normalized for retina

dry weight. Blood–retinal barrier breakdown was calculated

with the following equation, with results expressed as ll

plasma 9 g retinal dry wt-1 h-1.

Evans blue ðlgÞ=retina dry weight ðgÞTime average Evans blue concentration ðlg=llÞ � circulating time ðhÞ

Observation of endothelial F-actin cytoskeleton

in retinal vasculature

The retina in situ fluorescent staining was conducted at

mice according to Yu et al’s protocol [24]. The retina

vessels were perfused with a series of solutions via com-

mon carotid in the following order: carbogen-bubbled

sodium Krebs solution with papaverine (10-6 M) for

washing out the blood; 3% paraformaldehyde in 0.1 M

phosphate-buffered saline for fixation in 15 min; 0.1 M

phosphate-buffered saline for washing out; 0.1% v/v Tri-

ton-X-100 (Ajax Chemicals, Auburn NSW) in phosphate-

buffered saline for permeabilization in 15 min; 0.1 M

phosphate-buffered saline for washing out again; rhoda-

mine-phalloidin and 0.1 M phosphate-buffered saline for

fluorescence staining in 1 h. All the solutions were washed

out from cut-open external jugular vein. Through this

procedure, endothelial F-actin cytoskeleton in retinal vas-

culature was labeled with rhodamine-phalloidin. Then, the

eyeball was isolated and fixed in 4% paraformaldehyde.

The whole retina was then carefully dissected and mounted

on microwell dish. The organization of F-actin in retinal

microvascular endothelial cells was examined under con-

focal microscopy.

Statistics

Results are presented as the mean ± SD. Data were ana-

lyzed by using one-way ANOVA followed by post hoc

comparisons. P \ 0.05 was considered significant.

Result

The phosphorylation of moesin in retina tissue

from AGE-MSA-treated mice

Our previous study has demonstrated that moesin is

phosphorylated by AGEs and modulates endothelial per-

meability in HMVECs [10]. In this in situ report, we first

Acta Diabetol (2012) 49:47–55 49

123

explored the effect of AGE stimulation on retinal moesin

phosphorylation in an AGE-MSA-treated mouse model.

The data of western blot showed that there was an increase

in threonine 558 phosphorylation of moesin in retinal tissue

in AGE-MSA-treated mice (Fig. 1A), with P \ 0.05

compared with that in control mice (Fig. 1C). There was no

difference in total moesin expression between AGE-MSA-

treated and control groups (Fig. 1B).

The immunohistochemistry of retina demonstrated that

moesin and phosphorylated moesin were only detected in

retinal microvessel’s endothelial cells (Fig. 2), suggesting

the specific expression of moesin in endothelial cells. The

staining of phosphorylated moesin in vascular endothelia

was more remarkable in AGE-MSA-treated mouse than in

control mouse (Fig. 2C, D). RT–PCR experiment detected

no changed in moesin mRNA expression (data not shown).

These results indicated that AGEs stimulation induced the

moesin phosphorylation in mouse retinal endothelia,

without changing its mRNA and protein expressions.

The effects of p38 and ROCK activations

on moesin phosphorylation

Our previous study suggested that p38 MAPK and ROCK

played important roles in AGE-induced moesin phosphor-

ylation in HMVECs [16]. We then further testified whether

p38 and ROCK are involved in moesin phosphorylation in

this AGE-MSA-treated mouse model. Displayed by the

higher level of phospho-p38 MAPK, the data demonstrated

that p38 MAPK was activated in retinal microvessels from

AGE-MSA-treated mice (Fig. 3A). While pretreatment of

specific p38 inhibitor SB203580 30 min before each AGE-

MSA application in mice attenuated the activation of p38

(Fig. 3A, B), threonine 558 phosphorylation of moesin was

also obviously down-regulated at the same time (Fig. 3A,

C), suggesting that the activation of p38 was involved in

AGE-MSA-induced moesin phosphorylation. The inhibi-

tion of p38 activation did not alter the total moesin protein

expression (Fig. 3).

The results also revealed that the phosphorylation level

of ROCK was elevated in retinal microvessels from AGE-

MSA-treated mice (Fig. 4A). This phosphorylation was

attenuated by pretreating the mice with ROCK inhibitor

Y-27632 30 min before each AGE-MSA application. The

threonine 558 phosphorylation of moesin was also signif-

icantly down-regulated by the inhibition of ROCK activa-

tion (Fig. 4C), with no changes in expression of total

moesin protein (Fig. 4B). These data indicated that the

activation of ROCK also participated in AGE-MSA-

induced moesin phosphorylation in mouse retina.

The disruption of blood–retinal barrier in AGE-MSA-

treated mouse

The measurement of blood–retinal barrier function showed

that AGE-MSA stimulation significantly weakened the

blood–retinal barrier while the leakage of Evans blue

increased from 1.938 ± 0.353 ll plasma 9 g retinal dry

wt-1 h-1 in control group to 5.438 ± 0.801 ll plasma 9 g

retinal dry wt-1 h-1 in AGE-MSA-treated mice (P \ 0.01).

0

0.2

0.4

0.6

0.8

1

Cha

nges

of

Moe

sin/

β-a

ctin

0

0.4

0.8

1.2

1.6

2

2.4

control AGE-MSAcontrol AGE-MSAFol

d in

crea

ses

in p

hos-

moe

sin *

A

B C

phos-moesin-Thr558

Control AGE-MSA

moesin

β-actin

Fig. 1 Increased moesin

phosphorylation in retinal tissue

from AGE-MSA-treated mice.

The inner blood–retinal barrier

layer tissues were dissected and

protein samples were collected

from retina. The expression of

moesin and the phosphorylation

of threonine 558 in moesin were

measured by western blot

analysis (A). The expression of

total moesin was calculated as

the ratio of b-actin (B). The

phosphorylation of moesin was

calculated as the ratio of total

moesin in the same group (C).

Data presented are the average

of three separate experiments

performed in duplicate

(mean ± SD). *P \ 0.05

compared with control; one-way

ANOVA

50 Acta Diabetol (2012) 49:47–55

123

The application of SB203580 or Y-27632 30 min before

every AGE-MSA injection remarkably attenuated this

increase in blood–retinal permeability, with the leakage of

Evans blue decreased to 2.576 ± 0.388 or 2.565 ± 0.528,

respectively (Fig. 5A). The results indicated that AGEs

induced a breakdown of the blood–retinal barrier and the

Fig. 2 Immunohistochemical analysis of paraffin-embedded mouse

retina with antibodies of moesin and phosphorylated moesin. The

protein expression was detected using a diaminobenzidine reagent

(brown, peroxidase). Moesin was located mainly in endothelial cells

of retinal microvessels both in control (A) or AGE-MSA-treated

mouse (B). The staining of phosphorylated moesin was more

remarkable in AGE-MSA-treated mouse (D) than in control

mouse (C)

phos-p38MAPK

moesin

β-actin

phos-moesin-Thr558

p38MAPK

AGEControl AGE+SB203580

0

0.5

1

1.5

2

2.5

control AGE AGE+SB203580

Fol

d in

crea

ses

in p

hos-

moe

sin

*

#

A

B C

0

0.4

0.8

1.2

1.6

2

control AGE AGE+SB203580

Fol

d in

crea

ses

in p

hos-

p38/

p38

*

#

Fig. 3 Effects of p38 inhibition

with SB203580 on p38 MAPK

and moesin phosphorylation in

retinal vascular endothelial

cells. The expression of p38,

phosphorylated p38, moesin and

phosphorylated moesin were

measured by western blot

analysis (A). The

phosphorylations of p38 or

moesin were calculated as the

ratio of total p38 (B) or moesin

in the same group (C). Data

presented are the average of

three separate experiments

performed in duplicate

(mean ± SD). *P \ 0.05 versus

control; and #P \ 0.05 versus

AGE; one-way ANOVA

Acta Diabetol (2012) 49:47–55 51

123

inhibition of p38 MAPK or ROCK activation preserved the

blood–retinal barrier function.

The effects of AGE-MSA application on F-actin

cytoskeleton in retinal venule

ERM proteins, mostly moesin in endothelial cells, are

regarded as the linking molecules between plasma mem-

brane and actin cytoskeleton. When AGE treatment evoked

a significant phosphorylation of moesin in retinal endo-

thelial cells, it would be interesting to detect the alteration

of F-actin distribution in this same model. In control mice

retinal venules, F-actin appeared to be peripheral fibers at

outer area of endothelial cell near the cell–cell junctions,

formed a dense fiber bundle called peripheral actin rim

(PAR). The outline staining of vascular endothelium was

clear and continuous, showing the smooth grid structure of

endothelial cells along the prolate axis of the vessels

(Fig. 5B-a). Seven-day treatment of AGE-MSA to the mice

significantly blurred the staining line and induced the

depolymerization of PAR and broke down the grid struc-

ture (Fig. 5B-b). The inhibitions of p38 activation with

SB203580 or ROCK activation with Y-27632 helped to

restore the continuity of F-actin distribution in endothelial

cells (Fig. 5B-c, d). The result indicated that the treatment

of AGE-MSA affected the structure of F-actin and caused

the disarrangement of cytoskeleton. The activation of p38

MAPK or ROCK and the subsequent moesin phosphory-

lation might participate in the induction of F-actin

re-organization.

Discussion

Previously, in a human dermal microvascular endothelial

cells (HMVECs) model, we have confirmed the participa-

tion of moesin in AGE-induced endothelial response by

knockdown of moesin expression with siRNA. The inhibi-

tion of moesin expression could attenuate the formation of

F-actin stress fiber and the hyper-permeability response in

AGE-stimulated HMVECs. AGEs induced the phosphory-

lation of moesin and ERM proteins on the conserved thre-

onine residue in a time- and concentration-dependent

pattern. Moesin phosphorylation required AGE-induced

signaling pathways that include the binding of AGE-RAGE,

the activations of ROCK and p38 MAPK [15]. In this

present study, we demonstrate in an AGE-intervened ani-

mal model that threonine 558 phosphorylation of moesin in

retinal endothelial cells was remarkably increased and the

elevation of moesin phosphorylation could be attenuated by

inhibitions of p38 MAPK and ROCK activations. AGE

stimulation caused the disruption of blood–retinal barrier

integrity, and the increase in retinal vascular permeability

and the suppression of p38 MAPK and ROCK activations

helped to prevent these AGE-induced alterations.

To produce an animal model of AGE-evoked blood–

retinal barrier dysfunction, we adapted the mouse model

reported by Moore TC [22] in which AGE-MSA was

delivered intraperitoneally to the mouse. It has been

approved that treatment with AGE-MSA every day for

seven consecutive days did induce a statistically significant

increase in the amount of dye leakage from the retinal

phos-moesin-Thr558

moesin

β -actin

phos-ROCK

ROCK

AGEControl AGE+Y-27632A

B C

0

0.5

1

1.5

2

2.5

control AGE AGE+Y-27632

Fol

ds o

f ph

os-R

OC

K/R

OC

K

0

0.5

1

1.5

2

2.5

control AGE AGE+Y-27632

Fol

d in

crea

ses

in p

hos-

moe

sin

Fig. 4 Effects of ROCK

inhibition with Y-27632 on

ROCK and moesin

phosphorylation in retinal

vascular endothelial cells. The

expression of ROCK,

phosphorylated ROCK, moesin

and phosphorylated moesin

were measured by western blot

analysis (A). The

phosphorylations of ROCK or

moesin were calculated as the

ratio of total ROCK (B) or

moesin in the same group (C).

Data presented are the average

of three separate experiments

performed in duplicate

(mean ± SD). *P \ 0.05 versus

control; and #P \ 0.05 versus

AGE; one-way ANOVA

52 Acta Diabetol (2012) 49:47–55

123

vasculature in AGE-treated mice, indicating breakdown of

the blood–retinal barrier in AGE-infused mice. Our result

reaffirmed this method as an effective approach to

reproduce an in vivo AGE-treated model by demonstrating

a significant increase in Evans blue leakage from micro-

vessels to retinal tissues after 7 days of AGE-MSA

administration (Fig. 5A).

Yu et al. once reported that there was a disarrangement

of F-actin in diabetic rat retinal microvasculature [24].

Here, by in situ fixation and fluorescence probe staining in

endothelial cells, we first showed that there was a

remarkable alteration of F-actin distribution in this AGE-

treated mouse retinal microvessl (Fig. 5B). The disconti-

nuity of F-actin reflects the disruption of endothelial barrier

and could result in higher permeability [25].

It is suggested that ERM protein was involved in

inflammatory processes in several cell types [26–29]. As

the dominant ERM molecule expressed in endothelial cell,

moesin is more involved than the other ERM components

in endothelial functional regulation. Koss et al. have

demonstrated that tumor necrosis factor-a (TNF-a) induced

moesin phosphorylation and modulated permeability

increases in human pulmonary microvascular endothelial

cells [30]. We previously approved that moesin was

phosphorylated by AGEs and modulated endothelial per-

meability [15]. The development of diabetic retinopathy

has been directly linked to the accumulation of AGEs as

well as AGE-induced disruption of vascular endothelial

barrier function [9]. It is reasonable to presume that moesin

phosphorylation might play a role in pathogenesis of AGE-

induced diabetic retinopathy.

The result of this present study on meosin phosphory-

lation status confirmed our hypothesis that AGE applica-

tion induced an increase in threonine 558 phosphorylation

of moesin in retinal endothelial cells (Fig. 1). The phos-

phorylation of moesin promoted the binding of F-actin, the

formation of long filopodia and the absence of spreading in

thrombin-activated human platelets [31]. Our previous data

have demonstrated that down-regulation of moesin

expression by siRNA in HMVECs prevented AGE-induced

formation of stress fiber, suggesting that moesin protein is

required in this F-actin rearrangement process [15]. It is

proposed that moesin not only functions as the linker of the

cytoskeleton to the membrane while controlling cell shape,

adhesion, and locomotion, but also plays an important role

in mediator-induced signaling transduction [32].

The total moesin expression was not changed in this

AGE-treated model, consistent with the result of unchan-

ged moesin mRNA level. It has been shown that TNF-aincreased the phosphorylation of ezrin/radixin/moesin

proteins without altering their expression in human pul-

monary microvascular endothelial cells [30]. Although the

protein sample in this study was from inner retina which

contained not only endothelia but also some pericytes,

Muller cells, and astrocytes [31], it is known that moesin

expresses mainly in endothelial cells [15, 33]. The

0

1

2

3

4

5

6

7

control AGE AGE+SB203580 AGE+Y-27632

Eva

ns B

lue

leak

age

(μμl p

lasm

a x

reti

nal g

/h)

*

# #

B

A

a b

c d

Fig. 5 Changes of blood–retinal barrier function and structure in

AGE-MSA-treated mice and the effects of p38 MAPK and ROCK

inhibition. Blood–retinal barrier function was analyzed by Evans blueleakage (A). While AGE-MSA injection increased the amount of dye

leakage from the retinal vasculature, statistically significant decreases

of leakage were noted in SB203580-pretreated and Y-27632-

pretreated mice. *P \ 0.01 for difference between AGE-MSA group

and control group; #P \ 0.01 for differences between SB203580 or

Y-27632 pretreated group and AGE-MSA group. Distribution of

F-actin cytoskeleton in endothelial cells of retinal microvessel was

conducted by in situ fluorescent staining (B). Continuous and smooth

grid structure (yellow arrow) of endothelial cells was showed in

normal retinal vessel (B-a); absence of F-actin was noticed in retina

from AGE-MSA-treated mouse (B-b), indicating by green arrow. The

inhibition of p38 with SB203580 or ROCK with Y-27632 restored the

continuity of F-actin distribution in endothelial cells (B-c, d) (yellowarrow)

Acta Diabetol (2012) 49:47–55 53

123

immunohistochemical data of retinas in this study clearly

showed that both moesin and phosphorylated moesin were

specifically displayed in retinal microvascular endothelial

cells, strengthening the endothelial specificity of moesin

expression in this AGE-treated model.

Acting like an inflammatory mediator, AGEs and

receptor for AGEs (RAGE) signaling exert complex effects

on the development of diabetic retinal disease via com-

plicated transduction pathways [34]. The kinases that

phosphorylate moesin and ERM proteins have been well

elucidated in several experiments and the participants

include Rho-ROCK [35], p38 MAPK [36], and PKC [30],

etc. We have showed in previous experiments that RhoA,

ROCK, and p38 MAPK were phosphorylated by AGEs

stimulation and the suppression of p38 or ROCK activation

by inhibitor SB203580 or Y-27632 could attenuate AGE-

triggered hyper-permeability responses in HMVECs [15,

16]. In present study, the administration of p38 MAPK

inhibitor SB203580 or ROCK inhibitor Y-27632 not only

decreased the relative phosphorylation of p38 or ROCK,

but also significantly suppressed the phosphorylation of

moesin (Figs. 3 and 4) in AGE-stimulated mouse retina.

These data convinced us that in AGE-affected mouse, p38

and ROCK activations also mediated the phosphorylation

of moesin. The down-regulation of p38 or ROCK activa-

tion was accompanied by much less Evans blue leakage

from retinal microvessels and the preservation of integrity

of F-actin distribution in retinal vascular endothelial cells

(Fig. 5), indicating the involvement of p38 and ROCK

pathways in this AGE-triggered blood-retinal barrier dys-

function. It is yet to show the direct relationship of moesin

phosphorylation with the morphological and functional

statutes in retinal endothelial cells. But our previous report

in the cellular model did approve the involvement of

moesin in regulating endothelial F-actin formation and

monolayer permeability by knockdown of moesin expres-

sion with siRNA. It would be more convincing to carry out

the experiment in a moesin knockout mice model, which

we are intended to conduct.

Acknowledgments This work was supported by General Program

from Natural Science Foundation of China (30771028, 30971201);

Program for Changjiang Scholars and Innovative Research Team in

University (PCSIRT, No. IRT0731); and National Key Foundation for

Basic Science Research of China (G2005CB522601).

References

1. Williams R, Airey M, Baxter H, Forrester J, Kennedy-Martin T,

Girach A (2004) Epidemiology of diabetic retinopathy and

macular oedema: a systematic review. Eye 18:963–983

2. Bloomgarden ZT (2007) Screening for and managing diabetic

retinopathy: current approaches. Am J Health Syst Pharm

64:S8–S14

3. King GL (2008) The role of inflammatory cytokines in diabetes

and its complications. J Periodontol 79:1527–1534

4. Menghini R, Casagrande V, Cardellini M, Martelli E, Terrinoni

A, Amati F, Vasa-Nicotera M, Ippoliti A, Novelli G, Melino G,

Lauro R, Federici M (2009) MicroRNA 217 modulates endo-

thelial cell senescence via silent information regulator 1. Circu-

lation 120:1524–1532

5. Sander B, Hamann P, Larsen M (2008) A 5-year follow-up of

photocoagulation in diabetic macular edema: the prognostic value

of vascular leakage for visual loss. Graefes Arch Clin Exp

Ophthalmol 246:1535–1539

6. Pfister F, Wang Y, Schreiter K, vom Hagen F, Altvater K,

Hoffmann S, Deutsch U, Hammes HP, Feng Y (2010) Retinal

overexpression of angiopoietin-2 mimics diabetic retinopathy and

enhances vascular damages in hyperglycemia. Acta Diabetol

47:59–64

7. Tarallo S, Beltramo E, Berrone E, Dentelli P, Porta M (2010)

Effects of high glucose and thiamine on the balance between

matrix metalloproteinases and their tissue inhibitors in vascular

cells. Acta Diabetol 47:105–111

8. Cardellini M, Menghini R, Martelli E, Casagrande V, Marino A,

Rizza S, Porzio O, Mauriello A, Solini A, Ippoliti A, Lauro R,

Folli F, Federici M (2009) TIMP3 is reduced in atherosclerotic

plaques from subjects with type 2 diabetes and increased by

SirT1. Diabetes 58:2396–2401

9. Yamagishi S, Ueda S, Matsui T, Nakamura K, Okuda S (2008)

Role of advanced glycation end products (AGEs) and oxidative

stress in diabetic retinopathy. Curr Pharm Des 14:962–968

10. Federici M, Lauro R (2005) Review article: diabetes and ath-

erosclerosis–running on a common road. Aliment Pharmacol

Ther 22 (Suppl 2):11–15

11. Barile GR, Schmidt AM (2007) RAGE and its ligands in retinal

disease. Curr Mol Med 7:758–765

12. Naka Y, Bucciarelli LG, Wendt T, Lee LK, Rong LL, Ramasamy

R, Yan SF, Schmidt AM (2004) RAGE axis: animal models and

novel insights into the vascular complications of diabetes. Arte-

rioscler Thromb Vasc Biol 24:1342–1349

13. Guo XH, Huang QB, Chen B, Wang ShY, Hou FF, Fu N (2005)

Mechanism of advanced glycation end products-induced hyper-

permeability in endothelial cells. Acta Physiol Sino 57:205–210

14. Wang ZhH, Guo XH, Liu XL, Li Q, Wang JP, Wang LJ, Huang

QB (2008) The morphological changes of vascular endothelial

cadherin in human umbilical vein endothelial cells induced by

advanced glycation end products. Chin J Arterioscler 16:505–509

15. Guo XH, Wang LJ, Chen B, Li Q, Wang J, Zhao M, Wu W, Zhu

P, Huang X, Huang Q (2009) ERM protein moesin is phos-

phorylated by advanced glycation end products and modulate

vascular permeability. Am J Physiol Heart Circ Physiol

297:H238–H246

16. Wang JP, Guo XH, Wang LJ, Li Q, Chen B, Wu W, Huang X,

Huang Q (2009) Effects of Rho/Rock signal pathway on AGEs

induced morphological and functional changes in HMVECs-1.

Acta Physiol Sinica 61:132–138

17. Hou FF, Miyata T, Boyce J, Yuan Q, Chertow GM, Kay J,

Schmidt AM, Owen WF (2001) Beta(2)-Microglobulin modified

with advanced glycation end products delays monocyte apopto-

sis. Kidney Int 59:990–1002

18. Gopalkrishnapillai B, Nadanathangam V, Karmakar N, Anand S,

Misra A (2003) Evaluation of autofluorescent property of

hemoglobin-advanced glycation end product as a long-term gly-

cemic index of diabetes. Diabetes 52:1041–1046

19. Sattarahmady N, Khodagholi F, Moosavi-Movahedi AA, Heli H,

Hakimelahi GH (2007) Alginate as an antiglycating agent for

human serum albumin. Int J Biol Macromol 41:180–184

20. Stitt AW, McGoldrick C, Rice-McCaldin A, McCance DR, Glenn

JV, Hsu DK, Liu FT, Thorpe SR, Gardiner TA (2005) Impaired

54 Acta Diabetol (2012) 49:47–55

123

retinal angiogenesis in diabetes: role of advanced glycation end

products and galectin-3. Diabetes 54:785–794

21. Qaum T, Xu Q, Joussen AM, Clemens MW, Qin W, Miyamoto

K, Hassessian H, Wiegand SJ, Rudge J, Yancopoulos GD,

Adamis AP (2001) VEGF-initiated blood-retinal barrier break-

down in early diabetes. Invest Ophthalmol Vis Sci 42:2408–2413

22. Moore TC, Moore JE, Kaji Y (2003) The role of advanced gly-

cation end products in retinal microvascular leukostasis. Invest

Ophthalmol Vis Sci 44:4457–4464

23. Hughes JM, Kuiper EJ, Klaassen I, Canning P, Stitt AW, Van

Bezu J, Schalkwijk CG, Van Noorden CJ, Schlingemann RO

(2007) Advanced glycation end products cause increased CCN

family and extracellular matrix gene expression in the diabetic

rodent retina. Diabetologia 50:1089–1098

24. Yu PK, Yu DY, Cringle SJ, Su EN (2005) Endothelial F-actin

cytoskeleton in the retinal vasculature of normal and diabetic rats.

Curr Eye Res 30:279–290

25. Rist RJ, Romero IA, Chan MW, Couraud PO, Roux F, Abbott NJ

(1997) F-actin cytoskeleton and sucrose permeability of im-

mortalised rat brain microvascular endothelial cell monolayers:

effects of cyclic AMP and astrocytic factors. Brain Res

768:10–18

26. Masumoto J, Sagara J, Hayama M, Hidaka E, Katsuyama T,

Taniguchi S (1998) Differential expression of moesin in cells of

hematopoietic lineage and lymphatic systems. Histochem Cell

Biol 110:33–41

27. Iontcheva I, Amar S, Zawawi KH, Kantarci A, Van Dyke TE

(2004) Role for moesin in lipopolysaccharide-stimulated signal

transduction. Infect Immun 72:2312–2320

28. Hiroki J, Shimokawa H, Higashi M, Morikawa K, Kandabashi T,

Kawamura N, Kubota T, Ichiki T, Amano M, Kaibuchi K,

Takeshita A (2004) Inflammatory stimuli upregulate Rho-kinase

in human coronary vascular smooth muscle cells. J Mol Cell

Cardiol 37:537–546

29. Killock DJ, Parsons M, Zarrouk M, Ameer-Beg SM, Ridley AJ,

Haskard DO, Zvelebil M, Ivetic A (2009) In vitro and in vivo

characterization of molecular interactions between calmodulin,

Ezrin/Radixin/Moesin, and L-selectin. J Biol Chem 284:8833–

8845

30. Koss M, Pfeiffer GR II, Wang Y, Thomas ST, Yerukhimovich M,

Gaarde WA, Doerschuk CM, Wang Q (2006) Ezrin/radixin/

moesin proteins are phosphorylated by TNF-alpha and modulate

permeability increases in human pulmonary microvascular

endothelial cells. J Immunol 176:1218–1227

31. Nakamura F, Amieva MR, Furthmayr H (1995) Phosphorylation

of threonine 558 in the carboxyl-terminal actin-binding domain

of moesin by thrombin activation of human platelets. J Biol Chem

270:31377–31385

32. Kaur C, Foulds WS, Ling EA (2008) Blood-retinal barrier in

hypoxic ischaemic conditions: basic concepts, clinical features

and management. Prog Retin Eye Res 27:622–647

33. Berryman M, Franck Z, Bretscher A (1993) Ezrin is concentrated

in the apical microvilli of a wide variety of epithelial cells

whereas moesin is found primarily in endothelial cells. J Cell Sci

105:1025–1043

34. Pachydaki SI, Tari SR, Lee SE, Ma W, Tseng JJ, Sosunov AA,

Cataldergirmen G, Scarmeas N, Caspersen C, Chang S, Schiff

WM, Schmidt AM, Barile GR (2006) Upregulation of RAGE and

its ligands in proliferative retinal disease. Exp Eye Res

82:807–815

35. Matsui T, Maeda M, Doi Y, Yonemura S, Amano M, Kaibuchi K,

Tsukita S, Tsukita S (1998) Rho-kinase phosphorylates COOH-

terminal threonines of ezrin/radixin/moesin (ERM) proteins and

regulates their head-to-tail association. J Cell Biol 140:647–657

36. Lan M, Kojima T, Murata M, Osanai M, Takano K, Chiba H,

Sawada N (2006) Phosphorylation of ezrin enhances microvillus

length via a p38 MAP-kinase pathway in an immortalized mouse

hepatic cell line. Exp Cell Res 312:111–120

Acta Diabetol (2012) 49:47–55 55

123