meuhE‘

1..

5...
E

.lrl...: 15:...
1.7.2:. l.
.1: x

i1... 1.. .3.
3° ‘3'};
..- a 1 :1... .8. :.

91...
.4. .I.:_ I
.v.

1 x... s

39..) . .5 It
:3: in:

x .3

.1 .13.. 3., z
.4... a . .5;
.23.
v.

3...: 5.2.9...
an}; 1.
a.

1!..uv
‘1... it
3! 4r
..

..
a. .51
1.... 3... i
.
.23.... .. us .
.§.’}- 7.1. flaififinn‘!
,5 .. 3.4:...» ah
. _
a...

4... u... .1 0w...

.r.‘...: LI: 2
f 32......3. 1i.
5...! is...

$134 . a.»
u v. DID-a!

3:27
:1... .

x. x
.. rut-.7:
...a:r..y..l;z
1..., ... .

5

3‘5... v u

.1. i :11. (3.... «.2.
25.4.. .. ... l.

:I .

:53)
7.4.
t {.

 

 

ze$baz5é . -ww .. . . . .
. .: .12.: gfiufilflm . . . ... .. i .. .
£2. 1: . . . . .a. 5: . h;§fi&xfi.fi~u’

..
.r

as
Nil-""50
0“ ‘;

 

“CNN

Q
\.

LIBRARY
E‘fiichigan at.
University

-fi-r’"

 

 

This is to certify that the

dissertation entitled

Elevated Extracellular Glucose Elicits An
Osmotic Response In Osteoblasts Involving
Signaling And Transcription Factor Activation

presented by

Majd Zayzafoon

has been accepted towards fulfillment
of the requirements for

PhoD 0 degree in PhYSiOlogy

 

 

C’

Major professor

Date 3- ID @001

MS U is an Affirmative Action/Equal Opportunity Institution 0- 12771

 

PLACE IN RETURN Box to remove this checkout from your record.
To AVOID FINE return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 cycincmmpss-ms

ELEVATED EXTRACELLULAR GLUCOSE ELICITS AN OSMOTIC
RESPONSE IN OSTEOBLASTS INVOLVING SIGNALING AND
TRANSCRIPTION FACTOR ACTIVATION
By

Majd Zayzafoon

AN ABSTRACT OF A DISSERTATION
Submitted to
Michigan State University

in partial fulfillment of the requirement
for the degree of

DOCTOR OF PHILOSOPHY
Department of Physiology
2001

Professor Laura R. McCabe

ABSTRACT

ELEVATED EXTRACELLULAR GLUCOSE ELICITS AN OSMOTIC
RESPONSE IN OSTEOBLASTS INVOLva SIGNALING AND
TRANSCRIPTION FACTOR ACTIVATION
By

Majd Zayzafoon

Investigation of osmotic stress in mammalian tissues has focused principally on
kidney and the cellular responses to hyperosmolar conditions of 300 mOsm above
normal. Recently, many studies have focused on physiologically relevant extracellular
hyperosmolarity to increase our understanding of cellular osmotic response in diseases
such as uremia, hypernatremia and diabetes mellitus. Insulin dependent diabetes mellitus
(IDDM; type I) is a chronic disease stemming from little or no insulin production and
elevated blood glucose levels. Hyperglycemia causes cellular damage by many
mechanisms including osmotic Stress. IDDM is associated with an extensive list of late
complications involving nearly every tissue. One major concern to young and aging
diabetics is the association of IDDM with decreased bone mass and osteoporosis. These
conditions are associated with an increase in hip and vertebrae fracture which can be
debilitating and increase mortality.

Based on the literature, a major factor contributing to bone loss in diabetes is a
decrease in osteoblast function and ability to differentiate, we therefore hypothesize that
osteoblasts respond to increasing extracellular glucose concentration (as seen in diabetes)

through an osmotic response pathway which influences gene expression and cell

phenotype. We anticipate that protein kinases such as protein kinase C (PKC) and
mitogen activated protein kinases (MAPK) could be activated and result in changes in
AP-l and CRE transactivation. To experimentally test this hypothesis, four aspects of
osteoblast responsiveness were examined: 1) phenotype and gene expression, 2) signal
transduction, 3) transcription factor modulation and 4) in vivo vs in vitro effects.

Our findings, both in vitro and in vivo, are the first to demonstrate an increase in
expression of c-jun and collagen I as well as the alkaline phosphatase activity in response
to treatment with a physiologically relevant concentration of sugar (22 mM). In contrast,
extracellular mineralization is decreased under these conditions and is potentially the
result of osteoblast differentiation being halted at mid stage. The use of mannitol
treatment and glucose uptake studies supports the novel concept that this response is
osmotic in nature. The present studies also demonstrate that osteoblast responsiveness to
moderate osmotic stress leads to an increase in AP-l and CRE transactivation and DNA
binding marked by the increased presence of c-Jun and ATP-2 in the binding complex.
Inhibitor studies demonstrate that both PKC and p38 MAP kinase are involved in this
response.

Findings from these studies contribute to an understudied area of research, the
molecular and cellular pathways involved in diabetes induced bone loss. With an
increasing diabetic and elderly population this information is important for developing
counter measures to diabetes and possibly age related bone loss, perhaps by using drugs

directed at signaling pathways and transcription factors affected by elevated glucose.

This work is dedicated to my lovely wife Lamia, my son Alexander, and Abraham, the

new blessing we are expecting in December 2001.

I would also like to thank my parents, my two sisters Shaza and Nada, and my brother

Ahed for their constant encouragement and belief in me.

iv

ACKNOWLEDGMENTS

I would like to thank my committee members Drs. Laura R. McCabe, Patrick F.
Dillon, Donald B. Jump, Seth R. Hootman, Laurie K. McCauley, and William S.
Spielman. I am forever grateful to my mentor Dr. Laura McCabe for all the
encouragement, guidance and support she gave me throughout my graduate research
work. Furthermore, I would like to thank Drs Patrick Dillon, Donald Jump, and Seth
Hootman for giving me some of their valuable time, comments and great discussions.

I would like to thank all my laboratory friends for their support and help,
especially Shira Bassly, Brian Coates, Regina Irwin, Chris Ontiveros, Ruijie Shu and
Christine Stell. My special thanks as well for my friends in the physiology department,
Julia Busik, Nara Parameswaran, Maria Pino, Robert Reid, Panayiotis Vacratsis,

Last but not least, my thanks go to the department personnel, especially Sharon

Shaft for her continuous great help.

TABLE OF CONTENTS

Page
List of Figures ........................................................................................ ix
Key to Abbreviations ................................................................................. xi
1. Introduction ......................................................................................... 1
II. Literature Review ................................................................................. 5
1. Hyperosmolarity ............................................................................... 5
1.1 Osmotic Adaptation ...................................................................... 5
1.2 The effect of hyperosmolality .......................................................... 6
1.2.1 Osmotic stress and signal transduction ......................................... 6
1.2.2 The Kidney and Hypertonicity ................................................ 10
1.2.3 The Brain and Hypertonicity ................................................... 11
1.2.4 Blood vessels and Hypertonicity .............................................. 13
2. Bone as a tissue .............................................................................. 14
2.1 Bone Matrix and Mineral ......................................................... 14
2.2 Cellular organization within the bone matrix —the osteocyte ................ 15
2.3 The osteoblast and bone formation .............................................. 15
2.4 The osteoblast and bone resorption .............................................. 19
2.5 Bone remodeling ................................................................... 20
3. Influence of diabetes on bone ............................................................... 20
4. Diabetes and hyperosmolality .............................................................. 23
5. Transcription Factors ........................................................................ 24
5.1 AP-l family of transcription factors ................................................. 25
5.2 CREB/ATF family of transcription factors ......................................... 27
III. Extracellular Glucose Influences Osteoblast Differentiation
and c-jun Expression ........................................................................... 31
1. Abstract ....................................................................................... 31
2. Introduction ................................................................................... 32
3. Materials and Methods ...................................................................... 35
3.1 Cell culture system ...................................................................... 35
3.2 RNA analysis ............................................................................ 35
3.3 Statistical analysis ...................................................................... 36
4. Results ......................................................................................... 36
4.1 Elevated levels of extracellular sugar influences osteoblast gene expression..36
4.2 Elevated extracellular sugar levels stimulate c-jun expression ................... 37

vi

4.3 Inhibition of protein kinase C blocks glucose induction of c-jun
and collagen I expression .............................................................. 39
5. Discussion .................................................................................... 49

IV. Osteoblast Osmotic Response to Elevated Extracellular Glucose
in vitro and in vivo Involves AP~1 ........................................................... 53
1. Abstract .......................................................................................... 53
2. Introduction ...................................................................................... 54
3. Materials and Methods ......................................................................... 56
3.1 Cell Culture System ........................................................................ 56
3.2 Histology ..................................................................................... 57
3.3 RNA Analysis ............................................................................... 57
3.4 Glucose Media Measurement Assay ..................................................... 58
3.5 3-O-methylglucose Uptake Assay ........................................................ 58
3.6 Cell Growth and Proliferation Assay ..................................................... 59
3.7 In vivo analysis .............................................................................. 59
3.8 Transient transfection studies .............................................................. 60
3.9 Statistical analysis .......................................................................... 61
4. Results ............................................................................................ 61
4.1 Elevated levels of extracellular sugar suppress osteoblast mineralization
both in vitro and in vivo ................................................................... 61
4.2 Sugar treatment does not influence glucose uptake .................................... 62
4.3 c-jun is induced within one hour in response to elevated extracellular sugar. . ....63
4. 4 Glucose treatment does not affect growth of MC3T3-El cells ...................... 63
4.5 Elevation of extracellular glucose 1ncrease AP-l transactivation.. ........64
4. 6 AP-l dominant negative, A-FOS, blocks c-jun and Collagen 1 Induction ......... 64
4. 7 c-jun and COL 1 are induced within one hour in response
to elevated extracellular mannitol in vivo ............................................... 65
5. Discussion ......................................................................................... 75

V. p38 and ATF—Z involvement in osteoblast osmotic

Response to elevated extracellular glucose ................................................... 80

1. Abstract .......................................................................................... 80
2. Introduction ...................................................................................... 81
3. Materials and Methods ......................................................................... 84
3.1 Cell Culture System ........................................................................ 84

3.2 Whole cell protein extraction .............................................................. 84

3.3 Nuclear extract isolation ................................................................... 85

3.4 Western blot analysis ...................................................................... 85

3.5 EMSA ...................................... ' ................................................... 86

3.6 Measurement of CRE transactivation by transient transfection ...................... 86

3.7 Statistical analysis .......................................................................... 87

vii

4. Results ............................................................................................ 87
4.1 Elevation of extracellular sugar specifically enhances p38 phosphorylation ......87

4.2 Elevation of extracellular sugar enhances ATP-2 phosphorylation .................. 88
4.3 Elevation of extracellular glucose enhances p38 and ATP—2 phosphorylation in
time dependent manner ................................................................... 88
4.4 Elevation of extracellular glucose enhances p38 and ATF-2 phosphorylation in
concentration dependent manner .......................................................... 88
4.5 ATF-2 phosphorylation in response to elevation in
extracellular sugar is p38 dependent .................................................... 89
4.6 Increase in CRE DNA binding 1 hour after glucose or mannitol treatment ........ 89
4.7 Elevation of extracellular sugar increases CRE transactivation in
osteoblasts, a p38 dependent response ................................................... 9O
5. Discussion ...................................................................................... 102
VI. Summary and Conclusions ................................................................. 106
VII. Bibliography .................................................................................. 110

viii

LIST OF FIGURES

Figure Page
1. Illustration of the ERK, JNK and p38 MAPK kinase pathways ............................. 9
2. Osteoblast developmental sequence ........................................................... 18

3. Elevated levels of extracellular sugar influence osteoblast
gene expression 24 hours after treatment ...................................................... 40

4. Elevated levels of extracellular sugar influence osteoblast gene
expression one hour after treatment ............................................................ 41

5. Elevated extracellular sugar levels stimulate c-jun expression
one hour after treatment ......................................................................... 42

6. Glucose and mannitol treatment results in chronic elevation of c-jun expression. ....43
7. Induction of c-jun by glucose or mannitol is concentration dependent. . . . . . . . . . . ....45

8. Inhibition of protein kinase C blocks glucose induction of

c-jun and collagen I expression ............................................................... 47
9. Elevated levels of extracellular sugar suppress osteoblast mineralization ................ 66
10. Osmotic stress suppresses bone mineralization in mice ................................... 67
11. Sugar treatment does not influence glucose uptake ........................................ 68

12. Steady influx of glucose into MC3T3-E1 cells under hyperosmolar conditions ....... 69

13. c-jun is induced within one hour in response to elevated extracellular sugar .......... 70
14. Glucose treatment does not affect growth of MC3T3-E1 cells ........................... 71
15. Elevation in extracellular sugar increases AP-l transactivation .......................... 72
16. AP-l dominant negative, A-FOS, blocks c-jun and Collagen I Induction .............. 73
17. Osmotic stress in viva induces c-jun and COL I within one hour ........................ 74
18. Elevation of extracellular glucose enhances p38 phosphorylation ....................... 91

ix

l9. Elevation of extracellular glucose does not influence
ERK and JNK phosphorylation .............................................................. 93

20. Elevation of extracellular glucose enhances ATP-2 phosphorylation. . . . . . . . . . . . ...94

21. Elevation of extracellular glucose enhances p38 and ATP-2
phosphorylation in time dependent manner ................................................. 96

22. Elevation of extracellular glucose enhances p38 and ATP-2
phosphorylation in concentration dependent manner ...................................... 97

23. Extracellular glucose induction of ATP 2 phosphorylation
can be blocked by inhibiting p38 activity ................................................... 98

24. CRE binding increases 1 hour after elevation of
extracellular glucose or mannitol concentration ............................................ 99

25. Elevation of extracellular sugar increases CRE transactivation in osteoblasts,
p38 dependent response ...................................................................... 101

26. Illustrated model ..... _ ......................................................................... 109

AP-l
AQP

ATF
BGT
COL
CRE
CREB
CREM
ECF
ECM
ERK

Glut
ICF
ICAM
IDDM
J NK
LUC
MAPK
NAD
PKC

RVI
SMIT
TauT

KEY TO ABBREVIATIONS

Activating protein 1

Aquaporin

Aldose reductase

Activating transcription factor
Betain/y-amino-n-butydc acid transporter
Collagen

CAMP response element

cAMP response element binding proteins
CAMP response element modulator
Extracellular fluid

Extracellular matrix

Extracellular signal-regulated kinases
Fibronectin

Glucose transporter

Intracellular fluid

Intracellular adhesion molecule-1
Insulin dependent diabetes mellitus
c-Jun NHz-terminal kinases

Luciferase

Mitogen activated protein kinase
Nicotinamide adenine dinucleotide
Protein kinase C

Parathyroid hormone

Regulatory volume increase
Na+-dependent myo-inositol transporter
Taurine Transporter

xi

I. Introduction

Poorly controlled or untreated type I diabetes mellitus, which affects over 5
million people in the United States, is characterized by hyperglycemia and is associated
with decreased bone mass (Hui et al., 1985), impaired skeletal development (Levin et al.,
1976), increased fracture rate (Bouillon, 1991; Meyer et al., 1993), and osteoporosis
(Auwerx et al., 1988; Krakauer et al., 1995). With the longer life span of diabetics,
understanding the mechanisms responsible for bone loss is critical for developing
countermeasures to its deleterious effects.

Thomas et a1. (1996) have identified GLUTl and GLUT3 as the primary glucose
transporters in osteoblast like cells (Thomas et al., 1996 a; Thomas et al., 1996 b).
Therefore, osteoblasts have a limited capability to absorb glucose due to the lack of the
high Km glucose transporters. This means that at euglycemic state, when glucose is 3-5.5
mM, these transporters are fully saturated and have already reached their maximal
capacity to transport glucose. In poorly controlled or untreated type I diabetes mellitus,
blood glucose levels are constantly high. Consequently, this could place osteoblasts
under osmotic stress if glucose transporters at the cell membrane number remain constant
in number.

Cells can respond to different extracellular inputs from their environment by
activating protein kinases. These complex networks allow amplification of extracellular
signals. Activation of more than one selected pathway can mediate a diversity of cellular
responses. For example, osmotic stress can activate the PKC signaling pathway

(Schaffler et al., 2000) as well as mitogen-activated protein (MAP) kinase cascades.

Under physiologically relevant increases in extracellular sugar p38 kinase and PKC in
particular have been shown to be activated in different cell culture systems (Kreisbeg and
Kreisberg, 1995; Watts et al., 1998; Roger et al., 1999).

Both PKC and p38 pathways can potentially influence AP-l levels and
transactivation. Angel and Karin (1991) have shown that c-jun transactivation is
increased in response to an increase in PKC activity. Modulation of Fos and Jun family
member expression in bone by steroid hormones, overexpression, or targeted gene
ablation has implicated c-Fos and other Fos and Jun related proteins in the regulation of
bone tissue formation (Lian et al., 1991).

Activating transcription factor 2 (ATF-2) is a substrate of p38 and other stress-
activated MAP kinases. It is a DNA-binding protein that forms a homodimer or
heterodimer with c-Jun, binds to CRE and stimulates CRE dependent transcription of
genes (Kawasaki et al., 2000). Van Dam et a1. (1995) have shown that c-jun induction
can occur by ATF-2 activation through the phosphorylation of p38. Increased activation
and expression of transcription factors ultimately leads to gene modulation.

Based on our preliminary data and literature, we hypothesize that osteoblasts
respond to increasing extracellular glucose concentration (as seen in diabetes) through an
osmotic response pathway which influences gene expression and phenotype. Activation
of protein kinase C (PKC), modulation of the mitogen activated protein kinase (MAPK)
and changes in AP-l and CRE transcription factors are involved in these processes.

The aim of this thesis is to study the effect of moderate elevation of extracellular
glucose on osteoblasts, the bone forming cells. It is divided into six chapters. Chapter I

includes a short introduction and states the major aims and hypothesis. A review of the

relevant past and current literature in Chapter II summarizes our current understanding of
osmotic stress and the cellular response to hyperosmolarity. As this work is focused on
physiologically relevant hyperosmolarity as seen in diabetes, I reviewed the effect of
diabetes on bone and the role it plays as a hyperosmotic producing condition which leads
to the activation PKC and MAP kinase. The last part of this review summarizes the
current knowledge about osteoblasts and the importance of different transcription factors
in bone development.

Chapter III focuses on the effect of increased extracellular glucose on osteoblast
phenotype. This research project is designed to test the hypotheses that: (1) elevated
extracellular glucose influences gene expression in osteoblasts, (2) glucose mediates
AP-l expression, and (3) PKC activation is required for the modulation of AP-l and
collagen I. Chapter IV further addresses the role of AP-l in osteoblast responsiveness to
elevated extracellular glucose. We tested the hypotheses that: (1) osteoblasts elicit an
osmotic response to elevated extracellular glucose, (2) in viva studies are consistent with
in vitro data, and (3) the use of AP-l dominant negative abolishes both the AP-l
modulation and gene expression changes in response to elevated extracellular glucose.

Chapter V focuses on the mechanism of the osteoblast osmotic response to
elevated extracellular glucose. Based on the literature and our previous results this
section addresses the hypotheses that: (1) MAP kinase activation is part of the osteoblast
response to osmotic stress, and (2) downstream transcription factors of MAP kinase, such
as ATF-2, are activated. The final summary and the conclusions of this work are

summarized in Chapter VI.

Diabetes affects over 5 million people in the United States. Bone loss resulting
from diabetes is of major concern given the long-term survival of diabetic patients. Yet,
little is known in this area. Results from this project contribute to the understanding of
the cellular and molecular mechanisms regulating osteoblast phenotype in response to
diabetes. This information will aid in the development of drugs, directed at pathways
affected by elevated extracellular glucose levels, which can be used to increase bone
formation in diabetics and perhaps in the elderly. Such therapies can be used to reduce

and possibly prevent the detrimental effects of osteoporosis.

11. Literature Review

1. Hyperosmolarity

Water is the most abundant constituent in the body, comprising approximately 60
percent of body weight in men and 50 percent in women. Total body water is distributed
into two major compartments, 55 to 75 percent is intracellular [intracellular fluid (ICF)],
and 25 to 45 percent is extracellular [extracellular fluid (ECF)]. The ECF is further
subdivided into intravascular (plasma water) and extravascular (interstitial) space. Water
in the body is the fluid in which important salts and proteins are dissolved. The
concentration of these solutes or particles within a fluid is known as osmolality and is
expressed as milliosmoles per kilogram of water (mOsmol/kg H20). Water crosses cell
membranes to achieve osmotic equilibrium between extracellular and intracellular
compartments. Disparities in cellular membrane permeability and the presence of
transporters and active pumps allow extracellular and intracellular solutes to markedly
differ while still maintaining equal particle concentration and osmolality. Intracellular
particle concentration is normally constant while extracellular particle concentration can
significantly change. This causes water to move into or out of cells due to alterations in

extracellular osmolality resulting in a cellular volume change.

1.1 Osmotic Adaptation
Many cells studied to date respond to induced volume changes by modulating
membrane transport and/or metabolic pathways that alter the concentration of

intracellular solutes. The first adaptive process occurring in response to extracellular

hypertonicity-induced cell shrinkage is a regulatory cell volume increase (RVI). The
RVI results from the stimulation of ion transporters such as the N a+-K+-2Cl' cotransporter
pathway (Lytle and Forbush, 1996) and Na+-H+ exchanger (Grinstein et al., 1986). These
transporters have been reviewed extensively in Parker, (1993). This immediate response
increases the intracellular ion content within minutes and partially restores the cellular
volume from the initial cell shrinkage (Parker, 1993; Sun et al., 1990). A second
adaptive mechanism occurs within hours to days and results in the induction of genes
encoding proteins involved in the accumulation of intracellular “compatible osmolytes”.
The general classes of organic osmolytes are sugars, polyols, amino acids, methylamines,
and urea. The accumulation of osmolytes is facilitated by the induction of specific
osmolyte transporters such as: betain/y—amino-n-butyric acid transporter (BGTI) for
betaine (Yamauchi et al., 1992); the Na+-dependent myo-inositol transporter (SMIT) for
inositol (Kwon et al., 1992); and the taurine transporter for the amino acid taurine
(Uchida et al., 1992). In some cases, enzymes that can produce osmolytes can be
increased such as aldose reductase (AR) which catalyzes the NADPH-mediated reduction
of aldehydes such as D-glucose to sorbitol (Bohren et al., 1989; Garcia-Perez et al.,
1989). These osmolytes increase intracellular osmolality and restore isotonicity (equal

intracellular and extracellular osmolality).

1.2 The effect of hyperosmolality
The normal plasma osmolality is about 300 mOsmol/kg H20 and is kept within a
narrow range by mechanisms capable of sensing a 1 to 2 percent change in tonicity.

However, a hyperosmolar blood milieu can commonly occur in a number of clinical

conditions such as diabetes mellitus, uremia, and hypematremia. The body and it’s
organs must adapt to this new environment. This adaptation response can lead to
pathological damage and altered organ function potentially as a result of signaling

pathway activation leading to modulation of gene expression.

1.2.1 Osmotic stress and signal transduction

Protein kinase C (PKC) and mitogen-activated protein (MAP) kinase cascades are
important intracellular signal-transduction pathways that have been demonstrated to be
activated in response to different extracellular stimuli including changes in osmolality.
PKC is regulated by two sequential mechanisms: 1) phosphorylation triggered by 3-
phosphoinositide-dependent kinase (PDK)-1 and 2) binding to the lipid second messenger
diacylglycerol. Each mechanism regulates the structure, subcellular localization, and
function of PKC (Dempsey et al., 2000). MAP kinases are serine/threonine kinases
activated via a cascade of kinases involving a sequential phosphorylation of two kinases
(MAP kinase kinase kinase and MAP kinase kinase), which activate a MAP kinase via a
dual phosphorylation on threonine and tyrosine residues (Kolch, 2000; Widmann et al.,
1999) (Figure.l). In mammalian cells, the MAP kinase family contains three major
subgroups responding to distinct extracellular stimuli: extracellular signal-regulated
kinases l and 2 (ERK); c-Jun NHz-terminal kinases (JNK, also known as stress-activated
protein kinases); and p38 kinases (also known as stress-activated protein kinases). The
ERK pathway is principally activated by growth factors, integrin-matrix interaction, and
hormones, while JNK and p38 are strongly activated by inflammatory cytokines,

ultraviolet light, and hypertonic stress (Chen et al., 1996;1garashietal., 1999).

Activation of specific signaling pathways in response to hyperosmolality can
mediate the required specific cellular adaptive response. Both PKC and p38 pathways
have been demonstrated to be co-activated in response to osmotic stress in several cell
systems. The target genes and importance of these pathways in response to osmotic
stress is not completely understood. However, a study by Schaffler et al., (2000)
demonstrates that inhibition of both PKC and p38 pathways can ameliorate
hyperglycemic and hyperosmotic induced vascular dysfunction in viva suggesting that
both pathways are involved in this pathology. Past studies examining the mechanisms of
cellular responsiveness to osmotic stress have focused principally on the kidney and the
responses to hyperosmolar conditions above 600 mOsm. More recently a variety of cell
systems have been shown to respond to physiologically relevant hyperosmolality as seen

in clinical conditions such as diabetes mellitus.

Ion Channels

  
 
  

      
 

 

 

“VIII-q!!!
J. : I, will:

IQ"

Slim-hutch . ‘

   
 

Am somycin

Thr223
m 52: 217/221

Thr202nyr204 l mummgj i

did:

UV. light
\ Genotoxrc Agents
1 MEKK 1

Tyr 59n] Ser 63 Set 73

  
   
 

   
 

Transcription

. Transcri pti on
Tran scn pt: on Factors

Transcription , . .
Factors

Osmotic stress

UV. light
Genotoxic Agents

p “nae. Amsomycin

l -.»
Ser189 1 Se 207
1 P r

Thr180/Tyr182 l

MAPKAP
Kinuso-Z

  
  
  
  
  

Thr71
Tyr 69l71
. Ser383

Transcn' pti on
Factors

Transcription

Fig. 1. Illustration of the ERK, JNK and p38 MAPK kinase pathways.

1.2.2 The Kidney and Hypertonicity

During diuresis and antidiuresis, the kidney medulla is exposed to large
fluctuations of interstitial osmolarity (Knepper, 1996) which challenge cell volume
constancy. Cells of the medullary thick ascending limb of Henle (MT AL) are of special
interest, since they are the major contributor to the generation of the renal cortico-
papillary osmotic gradient allowing urinary concentration in animals. Hyperosmolarity
studies in kidney have been conducted on renal medullary cells subjected to a very high
osmolar environment, 600 mOsm, consistent with the physiologic osmolarity that these
cells are exposed to in viva. Studies of cellular responses to high levels of hypertonicity
implicate the activation of all or some of the mitogen-activated protein (MAP) kinase
family members (Rosette and Karin, 1996; Galcheva-Gargova et al., 1994; Itoh et al.,
1994; Terada et al., 1994; Zhang and Cohen, 1996; Berl et al., 1997; Watts et al., 1998).
In MTAL cells Watts et al. (1998) demonstrated that hyperosmolality (600 mOsm)
results in the rapid (within 5 minutes) and sustained (up to 60 minutes) activation of p38
kinase. ERK was also activated in these cells peaking at 15 minutes. The lack of
responsiveness to treatment of MTAL cells with hyperosmolar levels of urea, a
compound that rapidly enters the cell before significant volume changes can occur,
demonstrates that cell shrinkage rather than extracellular hyperosmolality is critical for
the activation of both ERK and p38 kinases in response to high osmolar environment
(Roger et al., 1991). Activation of p38, in particular, is critical for regulatory volume
increases in medullary cells. Studies done by Sheikh-Hamad et al., (1998) using renal

medullary epithelial cells (MDCK) demonstrate that a specific p38 kinase inhibitor,

10

SB203580, can block the induction of betaine transporter mRNA in response to
hyperosmotic media (600 mOsm) and prevent regulatory volume increases.

Recent clinical trials have shown that the degree of hyperglycemia is an important
predictor of diabetic renal complications. Results of in viva studies have given
supportive evidence that increased ambient glucose concentrations exert an important
influence on extracellular matrix in renal diabetic pathology (Ziyadeh et al., 1993). As
diabetic nephropathy is characterized by the early appearance of glomerular hypertrophy,
many researchers have started to look into the effect of moderate osmolality increase on
other renal cells such as mesangial cells that are not usually exposed to high osmolar
environment of 600 mOsm. Elevating extracellular glucose or mannitol (a non-
metabolizable sugar) by 16-25 mM mimics the osmotic effect of diabetes mellitus,
marked by a 16-25 mOsm increase in blood osmolarity, results in increased mesangial
gene expression. In particular, extracellular matrix and attachment proteins expression
has been reported to be increased (Park et al., 2000; Nahman et al., 1992; Kreisbeg and
Kreisberg, 1995). The progressive accumulation of extracellular matrix (fibronectin,
larninin, and collagen IV) in the glomerular mesangium leads to glomerulosclerosis and
fibrosis (Mauer et al., 1994; Sharma et al., 1997; Ziyadeh et al., 1993). Both p38 and
PKC are implicated in these responses (Morrisey et al., 1999; Ingram et al., 1999;

Kreisbeg and Kreisberg, 1995).

1.2.3 The Brain and Hypertonicity

The kidney is not the only organ in the body which is sensitive to extracellular

osmotic fluctuation. The brain with its different cellular components is also affected.

11

Osmoreceptors sensitive to changes in the osmolality of circulating blood have been
located in the basal forebrain and their activation is responsible for the secretion of
vasopressin (Stricker, 1999). When exposed to a hyperosmotic medium, brain cells,
similar to many other cells in the body accumulate osmolytes and thus achieve osmotic
equilibrium with the medium while maintaining their volume. For example, when
astrocyte primary cultures are exposed to hyperosmolar medium they shrink rapidly and
then slowly regain their initial volume after several hours (Bitoun and Tappaz, 2000).
The mRNA levels of osmolyte transporters (taurine - TauT, myo-inositol - SMIT and
betaine - BGTl) showed significant and comparable increases after a four hour exposure
and return to near normal levels after 24 hours. When taurine was added to the
hyperosmotic medium, cell volume recovery was greatly accelerated and the osmo-
induced overexpression of TauT, SMIT and BGTl mRNA was prevented (Bitoun and
Tappaz, 2000). Taurine is an amino acid mainly known for its involvement in cell volume
regulation; it is one of the major inorganic osmolytes used by cells to compensate for
changes in extracellular osmolarity. In the supraoptic nucleus, taurine is highly
concentrated in astrocytes, and released in an osmodependent manner through volume-
sensitive anion channels (Saransaari et al., 2000). Via its agonist action on neuronal
glycine receptors, taurine is likely to contribute to the inhibition of neuronal activity
induced by hypotonic stimuli. This inhibitory influence would complement the intrinsic
osmosensitivity of supraoptic neurons, mediated by excitatory mechanoreceptors
activated under hypertonic conditions (Hussy et al., 2000). These observations
demonstrate the role taurine plays in the regulation of neural cell volume to that of whole

body fluid balance.

12

1.2.4 Blood vessels and Hypertonicity

The water in the blood and plasma is what constitute the intravascular
extracellular fluid. Therefore, a rise in the osmolality of the blood is considered an
extracellular hyperosmolality. This increase affects both cells circulating in the blood as
well as the cells lining the blood vessels. Similar to renal and brain cells, cell shrinkage
due to osmotic stress modulates leukocyte attachment and endothelial extracellular matrix
proteins through the activation of protein kinase pathways. ‘Hyperosmolality for
example, affects neutrophils partially by altering surface expression of adhesion
molecules, CD1 1b and L-selectin (Rizoli et al., 1998). Volume reduction in neutrophils
results in an increase in tyrosine phosphorylation and activation of p38. Similar to renal
medullary cells, the trigger for this response is cell shrinkage and not an increase in
osmolarity, ionic strength, or intracellular pH (Rizoli et al., 1999). The fact that cell
shrinkage is the triggering event for the osmotic response led Rizoli et a1. (2000) to
investigate if osmotically induced cytoskeletal changes might be related to this response.
They found that osmotic stress provoked a twofold increase in F-actin, induced the
formation of submembranous F-actin ring, and abolished depolymerization that normally
follows agonist-induced actin assembly. These results suggest that cytoskeletal
remodeling is a key component in the neutrophil response to hypertonicity. Monocytes
and macrophages in the blood are also targets for hyperosmolality. They use betaine and
myoinositol “compatible organic osmolytes” when exposed to osmotic stress.
Interestingly, p38 MAP kinase is found to be involved in the hyperosmolarity-induced

upregulation of the osmolyte transporters BGT-1 and SMIT (Denkert et al., 1998).

13

Vascular endothelial cells are not an exception when it comes to responding to
osmotic stress. Hypertonic treatment with physiologically relevant levels of glucose or
mannitol significantly increased the level of p38 activity. Other osmotic inducing agents
were used and the study concluded that different agents induce MAP kinase activation
differently. Urea did not affect the level of induction of the MAP kinase isoforms which
implies that cell shrinkage maybe an important component of hyperosmolality-induced
MAP kinase phosphorylation (Duzgun et al., 2000).

From all the above we can conclude that many cells in the body respond to
moderate increases in extracellular osmolarity as seen in diabetes where blood glucose
can be as high as 22mM. This response seems to be largely driven by a change in cell
volume “shrinkage” and leads to activation of various signal transduction pathways

resulting in changes in gene expression.

2. Bone as a tissue

Bone is a specialized connective tissue that makes up, together with cartilage, the
skeletal system. These tissues serve three functions: mechanical, protective, and
metabolic. In bone, as in all connective tissues, the fundamental constituents are the cells
and the extracellular matrix. The latter is particularly abundant in these tissues and is

composed of collagen fibers and noncollagenous proteins.

2.1 Bone Matrix and Mineral
Bone is formed by collagen fibers (type I, 90% of the total protein), usually

oriented in a preferential direction, noncollagenous proteins and a calcium phosphate

14

matrix termed hydroxyapatite. Spindle-or plate- shaped crystals of hydroxyapatite are
found on the collagen fibers, within them, and in the ground substance. They tend to be
oriented in the same direction as the collagen fibers. The ground substance is primarily
composed of glycoproteins and proteoglycans. These highly anionic complexes have a
high ion-binding capacity and are thought to play an important part in the calcification
process and the fixation of hydroxyapatite crystals to the collagen fibers.

The preferential orientation of the collagen fibers alternates in adult bone from
layer to layer, giving to this bone a typical lamellar structure, best seen under polarized
light or by electron microscopy. This fiber organization allows the highest density of
collagen per unit volume of tissue. The lamellae can be parallel to each other if deposited
along a flat surface (trabecular bone and periosteum), or concentric if deposited on a
surface surrounding a channel centered on a blood vessel (haversian system). However,
when bone is being formed very rapidly (during development and fracture healing, or in
tumors and some metabolic bone diseases), there is no preferential organization of the

collagen fibers; this type of bone is called woven bone as opposed to lamellar bone.

2.2 Cellular organization within the bone matrix —the osteocyte

The calcified bone matrix is not metabolically inert, and cells (osteocytes) are
found embedded deep within the bone in small osteocytic lacunae (25,000/mm3 of bone)
(Aarden et al., 1994). Osteocytes were once bone-forming cells, osteoblasts, that became
trapped in the bone matrix that they produced, which later became calcified. They
nevertheless express some specific membrane proteins. These cells have numerous and

long cell processes rich in microfilaments, which are in contact with cell processes from

15

other osteocytes or with processes from the cells lining the bone surface. These processes
are organized during the formation of the matrix and before its calcification; they form a
network of thin canaliculi permeating the entire bone matrix (Aarden et al., 1994).
Between the osteocyte’s plasma membrane and the bone matrix itself is the periosteocytic
space. This space exists both in the lacunae and in the canaliculi, and it is filled with

extracellular fluid.

2.3 The osteoblast and bone formation

The osteoblast is the bone-lining cell responsible for the production of the matrix
constituents (collagen and ground substance). It originates from a local mesenchymal
stem cell (bone marrow stromal stem cell or connective tissue mesenchymal stem cell).
These precursors, with the right stimulation, undergo proliferation and differentiate into
preosteoblasts and then into mature osteoblasts. Osteoblasts never appear or function
individually but are always found in clusters of cuboidal cells along bone surface (~100
to 400 cells per bone-forming site). At the light microscope level, the osteoblast is
characterized by a round nucleus at the base of the cell (opposite the bone surface), a
strongly basophilic cytoplasm, and prominent Golgi complex located between the
nucleus and the apex of the cell. Osteoblasts are always found lining the layer of bone
matrix that they are producing, before it is calcified (called, at this point, osteoid tissue).
Osteoid tissue exists because of a time lag between matrix formation and its subsequent
calcification (the osteoid maturation period), which is approximately 10 days (Pockwinse
et al., 1995). Behind the osteoblast can usually be found one or two layers of cells:

activated mesenchymal cells and preosteoblasts. At the ultrastructural level, the

16

osteoblast is characterized by (a) the presence of an extremely well-deve10ped rough
endoplasmic reticulum with dilated cisternae and a dense granular content, and (b) the
presence of large circular Golgi complex comprising multiple Golgi tasks (Ozawa et al.,
1994). Cytoplasmic processes on the secreting side of the cell extend deep into the
osteoid matrix and are in contact with the osteocyte processes in their canaliculi.
Junctional complexes (gap junctions) are often found between the osteoblasts (Ozawa et
al., 1994). The plasma membrane of the osteoblast is characteristically rich in alkaline
phosphatase (the concentration of which in the serum is used as an index of bone
formation) and has been shown to have receptors for parathyroid hormone but not for
calcitonin (Nijweide et al.,1986). Osteoblasts also express steroid receptors for estrogens
and vitamin D3 in their nuclei, as well as several adhesion molecules (integrins) and
receptors for cytokines. Toward the end of the secreting period, the osteoblast becomes
either a flat lining cell or an osteocyte.

In culture, like in viva, osteoblasts form a bone tissue-like organization by
undergoing three stages of development: proliferation, extracellular matrix maturation,
and extracellular matrix mineralization (Quarles et al., 1992; Owen et al., 1990). After 4-
5 weeks in culture, mineralized bone nodules are visible in the culture dish. During each
stage of development, specific subsets of genes are sequentially expressed or repressed
(see Figure 2), i.e.: proliferation — histone H4 and collagen I, extracellular matrix
maturation — alkaline phosphatase, and mineralization -— osteocalcin and osteopontin.
The regulation of gene expression in developing osteoblasts occurs predominantly at the
transcriptional level. This suggests that transcription factors play a major role in

regulating osteoblast phenotype and therefore bone formation.

17

Matrix
Proliferation Maturation Mineralization

 

1 Ha cou. Mind—— 5" be

9‘ MAXIMAL EXPRESSION

 

 

 

 

 

Figure 2. Osteoblast developmental sequence. Temporal expression of
osteoblast cell growth and phenotype related genes during 28 days of culture. Cells were
harvested on the indicated days and total RNA was analyzed by northern blot analysis. In
the proliferative period, histone (H4), and collagen I (COLL) are maximally expressed.
The peak level of alkaline phosphatase (AP) mRNA represents the ECM maturation
period. In the mineralized period, osteopontin (OP) and osteocalcin (OC) reach their
peak mRNA levels. A similar graph would also represent transcription rates, obtained

from nuclear run on assays, suggesting that transcription factors are important for this

orchestration of gene expression.

18

2.4 The osteoclast and bone resorption

The osteoclast is a giant multinucleated cell, containing four to 20 nuclei. It is
usually found in contact with a calcified bone surface and within a lacuna that is the
result of its own resorptive activity. It is possible to find up to four or five osteoclasts in
the same resorptive site, but there usually are only one or two (Roodman, 1996). Under
the light microscope, the nuclei appear to vary within the same cell: some are round and
euchromatic, and some are irregular in contour and heterochromatic, possibly reflecting
the asynchronous fusion of mononuclear precursors. The cytoplasm is “foamy” with
vacuoles. The zone of contact with the bone is characterized by the presence of a ruffled
border with dense patches on each side (the sealing zone)(Baron, 1993).

Characteristic features of osteoclast are the abundant Golgi complexes disposed
around each nucleus, the mitochondria, and the transport vesicles loaded with lysosomal
enzymes. The most prominent features of the osteoclast are the deep foldings of the
plasma membrane in the area facing the bone matrix. The ruffled border in the center is
surrounded by a ring of contractile proteins (sealing zone) that serve to attach the cell to
the bone surface, thus sealing off the subosteoclastic bone-resorbing compartment (Suda
et al., 1992). The attachment of the cell to the matrix is performed by integrin receptors,
which bind to specific sequences in matrix proteins. The plasma membrane in the ruffled
border area contains proteins that also found at the limiting membrane of lysosomes and
related organelles, and a specific type of electrogenic proton adenosine triphosphatase
(ATPase) involved in acidification. The basolateral plasma membrane of the osteoclast is
highly and specifically enriched in (Na+, K+) ATPase, HCOg'lCl' exchangers and NaVH"

exchangers (Baron, 1999).

19

2.5 Bone remodeling

Bone formation by osteoblasts and resorption by osteoclast occurs along the
surface of bone, mainly at the endosteal surface, and results in bone remodeling. This
activity is involved in bone growth and turnover and does not occur at random. Bone
formation and resorption are either part of the process of bone development and growth
or part of the turnover mechanism by which old bone is replaced by new bone. In the
normal adult skeleton, (i.e. after the period of development and growth), bone formation
occurs only where bone resorption has previously occurred (Marks and Hermey, 1996).
The sequence of events at the remodeling site involves osteoclast activation, bone
resorption and subsequently bone formation. During the intermediate phase between
resorption and formation (the reversal phase) a cement line is formed that marks the limit
of resorption and acts to cement together the old and the new bone. The duration of these
various phases has been measured. The complete remodeling cycle at each microscopic

site takes about 3 to 6 months.

3. Influence of diabetes on bone

Insulin dependent diabetes mellitus (IDDM; type I) is a chronic disease stemming
from little or no insulin production and elevated blood glucose levels. IDDM is
associated with an extensive list of late complications involving nearly every tissue. One
major concern to young and aging diabetics is the association of IDDM with osteoporosis
(Auwerx et al., 1988; Krakauer et al., 1995), decreased bone mass (Hui et al., 1985),
impaired skeletal development (Levin et al., 1976) and increased fracture rates (Bouillon,

1991; Meyer et al., 1993). Diabetic bone disease was first recognized at the beginning of

20

the twentieth century (Morrison and Bogan, 1927) and described as retardation of bone
development and bone atrophy in children with long-standing diabetes. Bemey, (1952)
also described the coexistence of diabetes and osteoporosis. Since then, diabetic bone and
mineral homeostasis have been studied in both humans and experimental animals. Studies
measuring bone loss by radiogrammetry and photon absorptometry demonstrate that both
male and female diabetics are vulnerable to this outcome (Levin et al., 1976).

A recent clinical study was done to evaluate the prevalence and severity of
osteopenia in patients with IDDM. It showed that 67% of the diabetic men and 57% of
the diabetic women suffered from osteopenia (a bone mineral density BMD 2 1 < 2.5
standard deviation of the population mean) of the femoral neck and/or lumbar spine
(Kemink et al., 2000). Fourteen percent of the male patients, but none of the female
patients, met the criteria for osteoporosis (BMD .>_ 2.5 SD). Analysis of bone formation
and resorption markers in the blood demonstrate that at least in male patients with IDDM,
osteopenia is the consequence of a lowered bone formation with a predominance of bone
resorption over formation (Kemink et al., 2000). These effects are even more
pronounced in diabetic rats. Verhaeghe et a1. (1990) showed that long standing diabetes
in BB rats results in severe low-tumover osteoporosis probably related to decreased
osteoblasts recruitment and/or function. When titanium alloy implants are placed on
normal and diabetic rat tibia for fourteen days, diabetic rats demonstrated significantly
less osseointegration than control (McCracken et al., 2000).

Several hypotheses have been proposed as possible pathogenic mechanisms of
diabetic osteopenia including primary disturbances in calcium absorption and secretion or

vitamin D metabolism (Raskin et al., 1978; Schneider and Schedl, 1972; Spencer et al.,

21

1980). However, histochemical studies suggest that osteoblasts are markedly affected by
diabetes. Specifically, osteoblast number/surface is reduced (cuboidal osteoblasts are
replaced by bone lining cells) and osteoid surfaces are markedly decreased with no
reduction in osteoclast number (Verhaeghe et al., 1990; Sasaki et al., 1991). TheSe
parameters do not correlate with bone loss resulting from low vitamin D or calcium
levels, and they do not occur in semistarved animals (Shires et al., 1980). Thus, diabetes
associated malnutrition and altered calcium and vitamin D metabolism cannot account for
the marked changes in bone.

Data from clinical, morphologic, and biochemical investigations has established
that hyperglycemia is a major metabolic consequence of diabetes and a major factor in
the development of chronic diabetic complications (Brownlee et al., 1984). In bone
however, most studies have focused on the effects of insulin. For example, insulin
treatment can stimulate bone mineralization in IDDM rats (Verhaeghe et al., 1992;
Hough et al., 1981).

Studies in type II non-insulin dependent diabetes mellitus also suggest a positive
effect of elevated insulin levels which are strong enough to overcome any negative
influences of other factors (Haffner and Bauer, 1993; Krakauer et al., 1995). However,
some animals still exhibit bone loss and increased fracture rate with NIDDM (Takeshita
et al., 1993) and increased bone loss with insulin treatment of IDDM (Waud et al., 1994).
Differences among these studies may result from difference in rat strain, differences in
overall daily blood glucose and insulin levels, and/or changes in other serum factors

(Taylor et al., 1987). These results suggest that modulation of the regulation of osteoblast

22

growth and differentiation is potentially a key component of the bone loss associated with

diabetes.

To directly address the role of glucose on osteoblast function, an in vitro culture
system can be used. Only a few studies have utilized this approach, but all have shown
an influence of glucose on osteoblasts. For example, when human osteoblasts (MG-63
cells) are exposed to elevated glucose levels, both parathyroid hormone (PTH) and
vitamin D responsiveness are impaired (Yoshida et al., 1995; Inaba et al., 1995). In
addition, Terada et a1. (1998) demonstrated that 7 day exposure of MG-63 cells to
elevated extracellular glucose levels results in attenuated responsiveness to IGF-l
treatment. These findings demonstrate that glucose can influence osteoblast
responsiveness to extracellular signals. However, the mechanism of these effects and
consequences of high glucose treatment on the regulation of osteoblast differentiation,

transcription factor levels, and gene expression remain unknown.

4. Diabetes and hyperosmolality

Diabetes is diagnosed by an elevated blood glucose level that is greater than 200
mg/dl (11 mM) at any random testing (Emancipator, 1999). Elevated extracellular
glucose has been shown to cause significant cellular effects including: 1) nonenzymatic
glycosylation of intracellular and extracellular proteins and DNA (Brownlee et al., 1984;
Bucala et al., 1984; Locatto et al., 1993; McCarthy et al., 1997; Katayama et al., 1996), 2)
modulation of cellular redox state (Hunt et al., 1990; Wolff and Dean, 1987), 3) changes

in cell metabolic pathways such as activation of the polyol pathway (Inaba et al., 1995;

23

Gabbay, 1973; Larkins and Dunlop, 1992), and 4) activation of protein kinase C (Lee et
al., 1994; Wolf et al., 1991; Craven et al., 1990; Ceolotto et al., 1999). Recently in rat
osteoblast like cells (UMR 106-01) GLUTl and GLUT3 glucose transporters have been
identified (Thomas et al., 1996 a; Thomas et al., 1996 b). GLUT 1 and GLUT 3 are basal
glucose transporters with a low Km, 1-2 mM for GLUTl and <1 mm for GLUT3 (Bell,
1991). The Km in this case is the glucose concentration at which transport is half
maximal. This means that at euglycemic state, when glucose is 3-5.5 mM, these
transporters are fully saturated and have already reached their maximal capacity to
transport glucose. In poorly controlled or untreated type I diabetes mellitus glucose level
is constantly high. Consequently, this could place osteoblasts under osmotic stress and
induce a combination of several mechanisms and signaling pathways, which contributes

to the glucose induced complications of osteoblast function in diabetes.

5. Transcription Factors

Downstream targets of protein kinase pathways include activation of transcription
factors and gene expression. For example, PKC has long been known to activate the
binding and transactivation at AP-l sites. This has been shown by the PKC activation
following TPA treatment leading to an increase in the AP-l consensus DNA binding on
the TPA response element (TRE). AP-l activation is involved in the regulation of the
transcription of different genes such as collagen I and osteocalcin (Katai et al., 1992;
McCabe 1996). MAP kinases have also been mostly attributed to the control of gene
transcription via phosphorylation of nuclear transcription factors such as ELK, Jun and

ATF-2. This eventually leads to gene modulation, which can play a part in the cellular

24

response to cell shape changes and osmotic stress. Hoffert et al. (2000) demonstrated that
aquaporin-S (AQPS), a water channel protein, is induced by hypertonic stress and this
induction requires activation of extracellular signal-regulated kinase (ERK). In diabetic
nephropathy, an excessive production of extracellular matrix (ECM) proteins by
glomerular mesangial cells occurs. Ishida et a1. (1999) demonstrates that MAP kinase
may contribute to the overproduction of fibronectin (FN) in mesangial cells. This
response was not only protein kinase mediated, but it also enhanced DNA-binding

activity of AP-l as well.

5.1 AP-l family of transcription factors

The nuclear proto-oncogenes, c-Fos and c-Jun, are the proto-typical members of
the AP-l family of transcription factors whose members also include: Fos B, Fra-l, Fra-
2, Jun B, and Jun D. All family members contain a leucine zipper and a basic region
required for dimer formation and DNA binding, respectively (Kouzarides and Ziff, 1988).
Fos and Jun proteins can form heterodimers, but only Jun proteins can form stable
homodimers (Angel and Karin, 1991). Dimerization is a prerequisite for DNA binding to
a consensus response element designated as an AP—l site (5’-TGAg/cTCA-3’).
Variations in core AP-l sequences or flanking sequences can affect DNA binding
affinities of specific dimer combinations (McCabe et al., 1996; Hadman et al., 1993;
Ryseck and Bravo, 1991). Compositional changes and post-translational modifications in
Fos and Jun dimer complexes can dramatically influence AP-l DNA binding activity and
transactivation potential (Gruda et al., 1994; Smeal et al., 1992). Therefore,

modifications in Fos and Jun family member proteins and their levels define AP-l dimer

25

subsets within the cell, and in turn could influence the transcription of genes with
appropriate target AP-l sites.

AP-l members are critical in the response of osteoblasts to changes in
extracellular stimuli. Modulation of Fos and Jun family member expression in bone by
steroid hormones, overexpression, or targeted gene ablation has implicated c-Fos and
other Fos and Jun related proteins in the regulation of bone tissue formation. For
example, vitamin D (Candeliere et al., 1991), PT H (Clohisy et al., 1992; Lee et al., 1994;
Koe et al., 1997), IGF (Merriman et al., 1990), BMP-2, (Ohta et al., 1992), or TGF-B
(Breen et al., 1994; Machwate et al., 1995) modulate c-fos and other fos/jun family
member expression and at the same time influence the growth and differentiation of
osteoblasts.

In osteoblasts, a significant change in the steady state levels (24 hours after
feeding) of Fos/Jun family members and AP-l DNA binding composition occurs during
the onset of differentiation (McCabe et al., 1996; McCabe et al., 1995). Specifically, Fra-
2 and Jun D are the predominant members present in differentiated osteoblasts at the
mRNA and protein level as well as in AP-l DNA binding complexes (McCabe et al.,
1996). This is in marked contrast to proliferating cells that express Fos B, Fra-l, Fra-2,
c-Jun, Jun B, and Jun D.

Although not well studied, AP-l-related sites are present in the promoters of
many developmentally regulated genes abundant in osteoblasts. This group includes
alkaline phosphatase, cyclin D1, collagen I, collagenase, osteopontin, and osteocalcin.
Binding ofivdifferent Fos and Jun dimers to unique AP-l sites with different flanking

sequences (yielding site selectivity of AP—l pairs) and overlapping with other

26

transcription factor elements provides multiple options for selectively influencing the
binding of factors necessary for basal and hormone stimulated gene expression.
Knowing that Fos/Jun expression is modulated during growth and differentiation, it is
possible to conceive of a system of gene regulation based on the presence of specific AP-

1 dimers in a cell and their affinity for different AP-l like sites within gene promoters.

5.2 CREB/ATF family of transcription factors

Cyclic AMP (cAMP)-responsive element (CREB) binding protein was originaly
identified as a target of the CAMP signaling pathway, but studies on activation of
immediate-early genes revealed that CREB is a target of other signaling pathways
activated by diverse array of stimuli (Sheng et al., 1990). During the characterization of
CREB, two other highly related products were identified: activating transcription factors
(ATF) (Lee et al., 1987) and CAMP response element modulator (CREM) (Foulkes et al.,
1991). Members of the CREB/ATF family bind to a consensus DNA sequence
(TGACGTCA), have a similar DNA binding domain and form selective heterodimers
with each other via the leucine zipper region. Although all CREB/ATF proteins share
similarity in their bZip domains; subgroups of proteins share additional similarity in other
regions. For example, ATF-1 (Hai et al., 1989), CREB (Gonzalez et al., 1989; Hoeffler et
al., 1988) and CREM (Foulkes et al., 1991) are similar in regions that contain the
phosphorylation sites. Similarly, ATF-2/CRE-BP1 (Maekawa et al., 1989) and ATF-3
(Gaire et al., 1990) share significant similarity in regions outside the bZip domain: the
first 100 N-terminal residues and the last 13 C-terminal residues. It is possible that

proteins within a given subgroup have closely related functions. Proteins between

27

subgroups, however, are completely different from each other outside the DNA binding
domain, indicating that they may interact with different proteins or ligands and have
different functions.

ATF-1 and CREB have been demonstrated to stimulate transcription in response
to cAMP and calcium influx (Gonzalez and Montminy, 1989; Liu et al., 1993; Rehfuss et
al., 1991), whereas ATF-2/CRE-BP1 has been demonstrated to stimulate transcription in
response to viral induction resulting in phosphorylation of residues in their amino-
terminal transactivation domain (Du et al., 1993). These sites are efficiently
phosphorylated by stress activated protein kinases (SAPKs). Phosphorylation of ATF-2
by SAPK in vitro at one or more sites outside the transactivation domain increases the
DNA binding activity of ATF-2 (Abdel-Hafiz et al., 1992). ATF-2 and c-Jun DNA
binding activity is highly regulated in response to stress. This complex will then target
various promoter sites, including the jun2TRE within the c-jun promoter (Morooka et al.,
1995).

ATF-3 on the other hand, is not an activator; it represses transcription when
bound to DNA (Chen et al., 1994). However, ATF-3 can heterodimerize with Jun
proteins, and ATF-3/Jun heterodimers have been demonstrated to activate transcription
(Hsu et al., 1992; Chu et al., 1994). Therefore, depending on the cellular context, ATF-3
may repress transcription as homodimers or activate transcription as heterodimers.

The role of CRE in the development of osteoblasts has not been fully
characterized yet. But increasing evidence points to the importance of CRE in the
development of bone and its response to different agents. Sakamoto et a1. (1998) reported

that in murine calvarial osteoblasts (MC3T3-E1) the amount of CREB protein is high

28

through all stages of osteoblast development and maximal in the proliferation stage. The
degree of CREB phosphorylation reached high levels in the proliferation stage and early
mineralization stage. In the early mineralization stage, most CREB bound to consensus
CRE DNA binding sites was phosphorylated, while both phosphorylated and
unphosphorylated CREB were bound in the proliferation stage. This leads to hypothesis
that the phosphorylation of CREB may regulate the expression of genes defining the
developmental sequence of MC3T3-El cells. ATF-2 mutant mice developed a defect in
endochondral ossification at epiphyseal plates similar to human hypochondroplasia
(Reimold et al., 1996).

Interestingly, the c-fas 5’ regulatory region has several imperfect CRE sequences
of which only one markedly causes CAMP-dependent c-fas promoter activation
(Berkowitz et al., 1989). Several in viva models have already identified Fos as an
important player in bone biology, where overexpression causes neoplasms and
collagenase producing bone tumors (Gack et al., 1994; Ruther et al., 1987) and Fos null
mice exhibit osteopetrosis and disorganized bone growth (Grigoriadis et al., 1994).
Pearman et al., (1996) showed that in normal osteoblasts, CREB participates in basal c-
fas expression through CRE. Upon PT H exposure, PKA phosphorylates and activates
CREB trans-activation functions. Increased Fos protein then participates in gene

regulation in bone.

In summary, diabetes is associated with bone loss and decreased osteoblast

function. Glucose levels are markedly increased in uncontrolled diabetes. The role of

glucose in altering osteoblast phenotype has not fully been examined, especially with

29

regard to changes in osteoblast gene expression and differentiation. Important
mechanisms by which glucose can influence osteoblast differentiation include altering
cell metabolism, extracellular osmolality, signal transduction pathways, and transcription
factor activities. Understanding the cellular and molecular mechanisms regulating
osteoblast phenotype in response to diabetes might aid in the development of drugs,
directed at pathways affected by elevated extracellular glucose levels, which can be used
to increase bone formation in diabetics and perhaps in the elderly. Such therapies can be

used to reduce and possibly prevent the detrimental effects of osteoporosis.

30

III. Extracellular Glucose Influences Osteoblast Differentiation

and c-jun Expression

(This chapter was published in Journal of Cellular Biochemistry 79(2): 301-310 2000)

1. Abstract

Insulin dependent diabetes mellitus, marked by high blood glucose levels and no
insulin secretion, is associated with decreased bone mass and increased fracture rates.
Analysis of bone histology suggests that osteoblast phenotype and function are
influenced by diabetes. To determine if elevated extracellular glucose levels could
directly influence osteoblast phenotype we treated mouse osteoblasts, MC3T3-El cells,
with 16.5 mM glucose and analyzed osteoblast gene expression. Collagen I mRNA levels
significantly increased while osteocalcin mRNA levels decreased twenty-four hours after
the addition of glucose. Expression of other genes, actin, osteopontin, and histone H4,
was unaffected. Effects on collagen I expression were seen as early as one hour after
treatment. c-Jun, an AP-l transcription factor involved in the regulation of osteoblast
gene expression and growth, was also modulated by glucose. Specifically, an increase in
c-jun expression was found at 1 hour and maintained for 24 hours following glucose
treatment. Treatment of osteoblasts with an equal concentration of mannitol completely
mimicked glucose treatment effects on collagen I and c-jun expression, demonstrating
that osmotic stress rather than glucose metabolism is responsible for the effects on
osteoblast gene expression and phenotype. Additional studies using staurosporine and
Ro-31-8220 demonstrate that protein kinase C is required for the glucose upregulation of

collagen I and c-jun. Taken together our results demonstrate that osteoblasts respond to

31

increasing extracellular glucose concentration through an osmotic response pathway that
is dependent upon protein kinase C activity and results in upregulation of c-jun and

modulation of collagen I and osteocalcin expression.

2. Introduction

Insulin dependent diabetes mellitus (IDDM) is a chronic disease stemming from
no insulin production and elevated blood glucose levels. IDDM manifest itself with
many well-known complications such as vascular abnormalities, retinopathy,
nephropathy, and neuropathy. Less well known, but of major concern, is the association
of IDDM with osteoporosis (Auwerx et al., 1988; Krakauer et al., 1995), decreased bone
mass (Hui et al., 1985), and increased fracture rates (Bouillon, 1991). Histochemical
studies demonstrate a decreased number early stage and mature osteoblasts in diabetic
bone (Verhaeghe et al., 1990; Sasaki et al., 1991). These changes cannot be accounted
for by altered calcium and vitamin D metabolism (Shires et al., 1980). These findings
suggest that modulation of the regulation of osteoblast growth and differentiation is
potentially a key component of the bone loss associated with diabetes.

Diabetes is diagnosed by an elevated blood glucose level that is greater than 200
mg/dL (11 mM) at any random testing (Emancipator et al., 1999). Elevated extracellular
glucose has been shown to cause significant cellular effects including: 1) nonenzymatic
glycosylation of intracellular and extracellular proteins and DNA (Brownlee et al., 1984;
Bucala et al., 1984; Locatto et al., 1993; McCarthy et al., 1997; Katayama etal., 1996), 2)
modulation of cellular redox state (Hunt et al., 1990; Wolf and Dean 1987), 3) changes in

cell metabolic pathways such as activation of the polyol pathway (Inaba et al., 1997;

32

Gabbay et al., 1973; Larkins and Dunlop 1992), and 4) activation of protein kinase C
(Lee et al., 1989; Wolf et al., 1991; Craven et al., 1990; Ceolotto et al., 1999). These
changes and perhaps other unidentified changes in signaling pathways and transcription
factor activities can directly affect cell growth, differentiation, and function. Only a few
studies have addressed the influence of glucose on osteoblast function. All suggest that
an elevation in extracellular glucose levels can influence osteoblast function. For
example, Katayama et al. (1996) have shown that culturing of osteoblasts on glycosylated
collagen I substrates can variably suppress alkaline phosphatase activity and osteocalcin
secretion. In addition to matrix effects, direct effects of glucose on osteoblast
responsiveness and gene expression have been suggested. When human osteoblasts
(MG-63 cells) are exposed to elevated glucose levels, parathyroid hormone (PTH),
vitamin D and insulin-like growth factor (IGF-l) responsiveness are impaired (Yoshida et
al., 1995; Inaba et al., 1995; Terada et al., 1998).

The AP-l family of transcription factors has been shown to play an important role
in the coupling of gene expression to changes in the environment (Morgan and Curran,
1991; Angel and Karin, 1991). This family includes c-Fos, Fos B, Fra-l, Fra-2, c-Jun Jun
B, and Jun D. FoszJun heterodimers and Juanun homodimers bind to a DNA consensus
response element designated as an AP-l site (5’-TGAg/cTCA-3’) (Angel et al., 1991).
Changes in the level of AP-l member expression and post-translational modifications
play an important role in regulating AP-l DNA binding and transactivation (McCabe et
al., 1996; Gruda et al., 1994; Smeal etal., 1992; Yamamoto et al., 1992).

Osteogenic hormones, bone strain and fracture, and Fos/Jun overexpression

stimulate both AP-l member expression and bone formation implicating this family of

33

transcription factors in the regulation of osteoblast growth and differentiation (McCabe et
al., 1996; Hadman et al., 1993; Candeliere et al., 1991; Clohisy et al., 1992; Koe et al.,
1997; Merriman et al., 1990; Breen et al.; 1994; Machwate et al., 1995; Ohta et al., 1991
& 1992; Mikuni-Takagaki Y,1999). This is not surprising seeing as osteoblasts must be
highly responsive to extracellular changes such as bone strain and calcium and nutrient
status within the body and must react immediately by modulating growth, differentiation,
and gene expression. Exactly how osteoblasts respond to the stress of increased
extracellular glucose is not known. Previous reports in other cell types suggest that
protein kinase C activity, an activator of AP-l expression, is elevated in response to
glucose treatment (Lee et al., 1989; Wolf et al., 1991; Craven et al., 1990; Ceolotto et al.,
1999). Furthermore, Kreisberg et a1. (1994) have shown that 1 hour after glucose
treatment mesangial cells have increased c-fas and c-jun expression, suggesting that AP-l
may play a role in cellular responsiveness to glucose. These findings led us to speculate
that diabetes associated changes in osteoblast phenotype could at least in part be the
result of altered AP-l family member expression which ultimately leads to changes in
gene expression. Our results support this hypothesis and demonstrate that osteoblasts are
sensitive to changes in extracellular glucose and respond uniquely through an osmotic
stress pathway involving protein kinase C and chronic induction of c-Jun expression and

altered osteoblast phenotype.

34

3. Materials and Methods

3.1 Cell culture system

MC3T3-El cells (Sudo et al., 1983) were plated at 100,000 cells per 100 mM dish
and fed every 2 days with alpha MEM (Gibco; Grand Island, NY) containing 5.5 mM
glucose (normal level) and supplemented with 10% fetal calf serum. The typical
concentration of insulin in the culture media is obtained from the serum component and
ranges between 1 and 5 picomole concentration, a level lower than physiologic (50-800
pM), but consistent with early development of IDDM. Eleven days after plating and 24
hours after the last feeding, glucose or mannitol (0.5M stock) were added directly to the
media in the tissue culture dish to yield the final concentrations of sugar noted for each
experiment (7.5 to 22 mM). For kinase inhibition studies, 30 min prior to addition of
sugar the cells were pre-treated with 50 nm of staurosporine or 200 nM of Ro-31-8220,
concentrations lower than those previously reported to inhibit protein kinase C (Nose and

Shibanuma, 1994; Beltman et al., 1996).

3.2 RNA analysis

Twenty four hours after feeding, glucose or mannitol (0.5 M stock) was added
directly to the media in the tissue culture dish to yield a final monosaccharide
concentration of 7.5 to 22 mM. Cells were washed and scraped in PBS at the indicated
times following treatment. Cells were centrifuged at 800 X g for 5 minutes and quick-
frozen in liquid nitrogen. Total RNA was extracted and analyzed by northern blot, as

previously described (McCabe et al., 1995; Chomczynski P and N Sacchi, 1987).

35

Northern blots were hybridized with random primed (Random primed DNA labeling kit,
Gibco) 32P—labeled complementary DNA probes specific for each AP-l family member
(generously provided by Dr. Rodrigo Bravo) and to cDNA probes for markers of
osteoblast differentiation (McCabe et al., 1995 & 1996, Lian and Stein, 1992).
Hybridzation signals were quantitated by phosphoimaging analysis and expressed relative

to 18S ribosomal subunit levels.

3.3 Statistical analysis

All statistical analyses were performed using Microsoft excel data analysis
program for t test analysis. Experiments were repeated at least three times unless
otherwise stated. The autoradiographs shown are of one representative experiment.

Values are expressed as a mean 1: SEM except where indicated.

4. Results

4.1 Elevated levels of extracellular sugar influences osteoblast gene expression

The use of an osteoblast cell line, MC3T3-El cells, which exhibits gene
expression in developmental manner similar to primary rat osteoblasts and bone in viva
(Quarles et al., 1992; Rodan et al., 1988), allowed us to examine modulation of one factor
associated with diabetes, glucose, while leaving insulin levels constant. Previously we
found that chronic treatment of osteoblasts with 16.5 mM glucose, a level seen in
untreated diabetics, results in decreased mineralization (Zayzafoon et al., 1998). To

determine if osteoblasts respond to acute glucose treatment, MC3T3-E1 cells were treated

36

(24 hours after feeding/media change) with 16.5 mM glucose which was directly added to
the tissue culture medium. Cells were harvested at 1 and 24 hours after treatment.
Within 24 hours of treatment gene expression was clearly modulated when compared to
control cultures which have 5.5 mM glucose (normal in viva and in vitro levels).
Specifically, the level of collagen I mRNA was significantly increased by greater than
two fold (Figure 3). In contrast, the level of osteocalcin mRNA was decreased to 30% of
control levels. Osteopontin, actin, and histone H4 expression was unaffected (Figure 3).
A two fold increase in collagen I expression was seen within one hour of treatment
(Figure 4) demonstrating that collagen I expression is highly and rapidly responsive to
elevation in extracellular glucose levels. In addition, the influence of extracellular
glucose on collagen I expression was detectable with as little as 4.5 mM glucose addition

(data not shown), a level that would be commonly seen in a diabetic patient.

4.2 Elevated extracellular sugar levels stimulate c-jun expression

Based on the role of AP-l in immediate response of cells to extracellular stimuli
(Morgan and Curran 1991; Angel and Karin 1991) and in the regulation of osteoblast
growth, differentiation, and gene expression (McCabe et al., 1996) we next examined
whether AP-l family member expression in osteoblasts was modulated by glucose
treatment. MC3T3-E1 cells were treated with 16.5 mM glucose for 1 hour. Northern
blot analysis demonstrates a greater than 3 fold induction of c-jun expression (Figure 5),
a rapid and transient induction in c-fas which is not always detectable by northern blot,
and a small but not statistically significant induction of fra-I (data not shown). RNA

levels of other AP-l members (jun B, jun D, fas B, fra-Z) did not change.

37

To examine the characteristics of the c-jun response a complete time course was
performed with osteoblasts being harvested at 0, 0.5, 1, 3, 8, and 24 hours following the
addition of 16.5 mM glucose. Figure 6 demonstrates that the induction of c-jun was
maximal at 05-1 hour and was maintained even 24 hours after the addition of glucose.
This response is unlike a serum response where c-jun levels rapidly decline after 1 hour
(McCabe et al., 1995). Thus, glucose treatment leads to long term changes in
transcription factor levels.

To determine if the effect on c-jun expression was concentration dependent
osteoblasts were treated with 2, 4.5, 9.5, and 16.5 mM glucose (final concentration equal
to 7.5, 10,15, and 22 mM glucose, respectively) and c-jun mRNA levels were examined 1
hour after treatment. Figure 7 demonstrates that as little as 4.5 mM of glucose added to
osteoblasts is enough to stimulate c-jun mRNA levels. This level of glucose (10 mM
final concentration; 180 mg/dL) as well as the addition of 9.5 mM glucose (15 mM final
concentration; 270 mg/dL) is often seen in patients who are diagnosed with diabetes.

Although the levels of glucose used in this study are not normally used to study
osmotic effects on cells, we still wanted to determine if any component of the osteoblast
response results from osmotic stress. Therefore osteoblasts were also treated with
mannitol, a non-absorbable sugar, to control for osmotic stress. Surprisingly, figures 3, 4,
5, 6 and 7 demonstrate that mannitol can completely mimic the effects seen with glucose.
The influence of mannitol on c-jun was seen with as little as 4.5 mM addition. These
results demonstrate that osmotic stress is the primary conveyor of the glucose—induced

changes in gene expression shown in figures 3, 4, 5, 6 and 7.

38

4.3 Inhibition of protein kinase C blocks glucose induction of c-jun and collagen I
expression

To begin to address the signaling pathways involved in the osteoblast response to
glucose we treated MC3T3-E1 cells with glucose in the presence of a protein kinase C
inhibitor, staurosporine. Northern blot analysis demonstrated that inhibition of protein
kinase C blocked the induction of collagen I and c-jun expression after 1 hour of glucose
treatment (Figure 8 A) while actin expression was not affected. Ro-31-8220, another
inhibitor of protein kinase C, also suppressed the influence of glucose on c-jun and
collagen 1 expression (Figure 8 B). Treatment with Ro-31-8220 alone had no effect on
basal collagen 1 expression but did enhance c-jun expression as has previously been
reported (Beltman et al., 1996). These findings further support an important role for

protein kinase C in the osteoblast response to osmotic stress.

39

 

 

H4 OP Actin Col 1 OC

 

24 hours

Figure 3. Elevated levels of extracellular sugar influence osteoblast gene expression
24 hours after treatment. Osteoblasts were fed every 2 days with media containing 5.5
mM glucose. On day 11 (24 hours after the last feeding) glucose (G, 16.5 mM) or
mannitol (M, 16.5 mM) were added directly to the tissue culture media of confluent
cultures of osteoblasts to yield a final concentration of 22 mM monosaccharide. Control
plates (C) were treated with PBS. Twenty-four hours after treatment the cells were
harvested for RNA. Autoradiographs of northern blot membranes hybidized with cDNAs
to histone H4 (H4), osteopontin (OP), actin, collagen I (COL I) and osteocalcin (0C) are
shown. Levels of mRNAs (as determined by phosphoimager analysis) for control
(white), 22 mM glucose (gray), or 22 mM mannitol (dark gray) treated cells were
expressed relative to ribosomal 18S subunit RNA levels and are graphed as a fold
increase relative to control levels set at 1. Values were obtained from 3-7 separate

experiments and are expressed +/- SE. *p<0.05

40

C G M
H434 11L.

Actin Q . i
COL 1 13 I I 0.5"

Fold increase

 

 

 

 

 

 

 

 

 

1 hour

Figure 4. Elevated levels of extracellular sugar influence osteoblast gene expression
one hour after treatment. Osteoblasts were fed every 2 days with media containing 5 .5
mM glucose. On day 11 (24 hours after feeding) glucose (G, 16.5 mM) or mannitol (M,
16.5 mM) were added directly to the tissue culture media of confluent cultures of
osteoblasts to yield a final concentration of 22 mM monosaccharide. Control plates (C)
were treated with PBS. One hour after treatment the cells were harvested for RNA.
Autoradiographs of northern blot membranes hybidized with cDNAs to histone H4 (H4),
actin, and collagen I (COL I) are shown. . Levels of mRNAs (as determined by
phosphoimager analysis) for control (white), 22 mM glucose (gray), or 22 mM mannitol
(dark gray) treated cells were expressed relative to ribosomal 18S subunit RNA levels
and are graphed as a fold increase relative to control levels set at 1. Values were obtained

from 3-7 separate experiments and are expressed +/- SE; *p<0.05

41

 

 

c G M jag 3i

‘2

0'1"“ it" E
2 2‘

IT"'TT7*.-f~~"‘ 9
185 W“* g 1"

 

 

 

 

 

 

 

Control Glucose

Figure 5. Elevated extracellular sugar levels stimulate c-jun expression one hour
after treatment. On day 11 osteoblasts were treated with 16.5 mM glucose or mannitol
to yield a final concentration of 22 mM. Cells were harvested 1 hour after addition of
sugar for RNA extraction. The autoradiograph was obtained a northern blot hybridized to
a cDNA specific for c-jun. The membrane contains RNA from control (C), 22 mM
glucose (G), or 5.5 mM glucose + 16.5 mM mannitol (M) treated osteoblasts. Levels of c-
jun mRNA (as determined by phosphoimager analysis) were expressed relative to
ribosomal 18S subunit RNA levels and graphed relative to control values set at 1: control
(white), 22 mM glucose (gray), or 22 mM mannitol (dark gray). Values were obtained

from 7 separate experiments and are expressed +/- SE; *p<0.01.

42

Figure 6. Glucose and mannitol treatment results in chronic elevation of c-jun
expression. Osteoblasts were fed every 2 days with media containing 5.5 mM glucose.
At day 11 osteoblasts were treated with 16.5 mM glucose or mannitol to yield a final
concentration of 22 mM. Cells were harvested at 0, 0.5, l, 3, 8, and 24 hours after the
addition of sugar for RNA extraction. A) A representative autoradiograph containing
RNA from glucose or mannitol treated osteoblasts hybridized to a cDNA specific for c-
jun or 18 S ribosomal subunit RNA. B) Levels of c-jun mRNA (as determined by
phosphoimager analysis) from glucose treated (gray) or mannitol treated (dark gray)
osteoblasts were expressed relative to ribosomal 18S subunit RNA levels and graphed as
a fold increase relative to basal control (0 hr) levels. Values were obtained from 3

separate experiments and are expressed +/- SE.

43

c-jun fold increase

Treatment (hour) 0 0-5 1 3 8 24

 

 

 

 

 

 

g c-Iunam --- i

3 . 1. z . t .-.». H.

<3 135 1““. w m m w m

g c-lun ‘3 a - - - -

C

r: .7 _

's“ 188 b 9...: on! w a“ tie
Glucose Mannitol

00.513824 00.51

Treatment time (hour)

3824

Figure 7. Induction of c-jun by glucose or mannitol is concentration dependent.
Osteoblasts were fed every 2 days with media containing 5.5 mM glucose. At day 11 (24
hours after feeding) osteoblasts were treated with additional glucose or mannitol to yield
a final sugar concentration of 7.5, 10, 15, or 22 mM. Cells were harvested 1 hour after
the addition of sugar and RNA extracted. A) A representative autoradiograph of a
northern blot containing RNA from glucose or mannitol treated osteoblasts hybridized to
a cDNA specific for c-jun or 18 S ribosomal subunit RNA. B) Levels of c-jun mRNA
(as determined by phosphoimager analysis) from glucose treated (gray) or mannitol
treated (dark gray) osteoblasts were expressed relative to ribosomal 188 subunit RNA
levels and graphed as a fold increase relative to basal control (5.5 mM) levels. Values

were obtained from 2 separate experiments.

45

Flnal sugar
concontratlon(mM) 5.5 7.5 10 15 22

c-lun e»; 5...» a - .

Glucose

18$ 11’ r«\ :5 £- “ "“' u“‘

g, c-lun M a it a)
‘E
‘6 ~! ! q.‘ '- ‘ 1v
E 188
2.5 .
Glucose Mannitol

c-jun fold increase

 

 

 

 

In In 0 In N In In
16 N F '- N 3 ,4 9 v-

22

Treatment concentration (mM)

46

Figure 8. Inhibition of protein kinase C blocks glucose induction of c-jun and
collagen I expression. Osteoblasts were fed every 2 days with media containing 5.5 mM
glucose. At day 11 (24 hours after feeding) osteoblasts were pretreated for 30 minutes
with 50uM staurosporine (A) or with IOOuM of Ro—3l-8220 (B). Cells were then treated
with 16.5 M glucose. After 1 hour cells were harvested and RNA extracted. A)
Representative autoradiographs of a northern blot containing RNA from control (5.5 mM
glucose), glucose (22 mM final sugar concentration) and glucose + staurosporine treated
cells. Levels of c-jun, collagen I (COLI), and actin mRNA were quantitated (by
phosphoimager analysis) from control (white), glucose treated (gray), or glucose +
staurosporine treated (dark gray) cells. Values (+/- SE) were expressed relative to 188
rRNA levels and expressed relative to control levels set at 1; n=3, *p<0.02. Values for
staurosporine alone treated cells did not differ from controls (data not shown). B)
Representative autoradiographs of a northern blot containing RNA from control (5.5 mM
sugar), glucose (22 mM final sugar concentration), glucose + Ro-31-8220, and Ro-31-
8220 alone treated cells. Levels of c-jun, collagen I (COLI), and actin mRNA were
quantitated from control (white), glucose treated (gray), glucose + Ro-3l-8220 treated
(dark gray), or Ro-31-8220 alone treated (diagonal lines) cells and expressed relative to

control cells; n=2.

47

 

 

 

 

 

 

 

 

Actin

c-jun

7////4

 

c-[un

Col I

Actln

 

CoLl Acfin

cjun

48

5. Discussion

Diabetes (IDDM) is marked by high glucose levels and is associated with bone
loss, decreased osteoblast number, and decreased osteoid surface suggesting that
osteoblast phenotype is altered. Few studies have examined the influence of increasing
extracellular glucose on osteoblast growth, differentiation, and gene expression. Our
results demonstrate that modulation of extracellular glucose concentration has an
immediate and significant effect on osteoblast gene expression twenty—four hours after
treatment: c—jun and collagen I expression is elevated while osteocalcin expression is
decreased. Treatment with mannitol, a nonabsorbable sugar, completely mimics the
effects of glucose hence the role of glucose metabolites can also be excluded and the role
of osmotic stress can be concluded as a primary inducer of the effects seen on gene
expression. This concept may be specific for osteoblasts which may be more sensitive to
this form of stress than other types of cells. Osteoblast sensitivity to osmotic stress is
demonstrated by the small addition of (4.5 — 16.5 mM) of either glucose or mannitol to
osteoblasts which results in increase in c-jun expression and in collagen I expression.
Elevation of extracellular glucose levels to 10-30 mM has been shown to influence gene
expression in other cells including mesangial cells (Ingram et al., 1999), monocytes
(Manduteanu et al., 1999) and fibroblasts (Benazzoug et al., 1998), potential mechanisms
causing these changes are thought to stem from metabolic effects although effects on
monocytes involve an osmotic component (Manduteanu et al., 1999).

Examination of glucose transporter expression by osteoblasts further supports a
role for increasing extracellular glucose concentration in causing osmotic stress. It is

known that osteoblasts do not express Glut 2 (the only high Km, 20mM, glucose

49

transporter; Waeber et al., 1994) but do express Glut l and Glut 3 glucose transporters
(Thomas et al., 1996 a & b) which are known to have low Kms (1-2 mM). Thus, entry of
glucose into the osteoblast is not significantly affected by increases in extracellular
glucose (Zayzafoon and McCabe, Manuscript submitted). Therefore, an elevation in
extracellular glucose should cause an osmotic stress which could be continuous or the
osteoblast may adapt by generating osmolytes or modulating water and ion channel
transport.

Increased expression of collagen I and other matrix mRNAs has been previously
reported for other cell types exposed to glucose including mesangial cells (Ziyadeh et al.,
1995) and fibroblasts (Han et al., 1999; Benazzoug et al., 1998). Consistent with our
results, the promoter of collagen I is known to contain an functional AP-l site (Chung et
al., 1996). Yet an elevation in collagen I expression is somewhat contradictory to
histological findings of decreased osteoid surface in diabetic animals (Verghaeghe et al.,
1990; Hough et al., 1981). One possibility is that collagen protein synthesis, processing,
or secretion is down regulated or altered by glucose treatment thereby negating the
elevation in mRNA expression. In addition our results are focused on the specific
modulation of extracellular glucose only, while in viva findings may involve the
additional effects of chronic exposure and the modulation of other factors including
insulin, IGF-1, and amylin; changes in these and other factors could lead to a more
complexly developed phenotype.

In contrast to the stimulation of collagen I expression, osteocalcin expression was
suppressed by elevation in extracellular glucose. The osteocalcin promoter contains

several AP-l sites which have been demonstrated to be functional. c-Jun exhibits binding

50

activity at these sites and when overexpressed can suppress osteocalcin transcription
(McCabe et al., 1996). This suppression is thought to result, at least in part, by the
competition with osteocalcin inducing transcription factors (such as the vitamin D
receptor) for overlapping promoter binding sites (Lian et al., 1991). Thus, suppression of
osteocalcin expression in response to increased glucose is consistent with the elevation in
c-jun expression which may be directly suppressing osteocalcin expression.

Activation of protein kinase C in diabetic tissue is a well-known outcome. It is
hypothesized that metabolism of glucose, through the pentose phosphate shunt or through
glycerolipid production, leads to an increase in diacylglycerol and subsequently protein
kinase C activation. Our studies demonstrate that PKC is also involved in a
nonmetabolic pathway involving osmotic stress. Specifically, induction of c-jun and
collagen I mRNA levels by either glucose or mannitol can be blocked by inhibitors of
protein kinase C, staurosporine and a more specific inhibitor Ro-31-8220.

Given that protein kinase C is a major activator of AP-l expression and
transactivation, it is not surprising that we found an elevation in c-jun expression in
osteoblasts treated with increasing concentrations of extracellular glucose. In addition,
the c-jun promoter contains an AP-l site that allows for positive autoregulation. In
mesangial cell cultures, Kreisberg et al. (1994) also report an increase in c-jun mRNA
levels as well as c-fos mRN A levels at 0.5 and 2 hours after glucose treatment. Our time
course studies clearly demonstrate a long-term induction in c-jun mRNA expression
which is supported by an elevated c-jun signal at 24 hours in the study by Kreisberg et al.

(1994). In two experiments we were also able to detect an induction of c-fos mRNA

51

expression at 0.5 hours after glucose treatment. However this increase was not found in
subsequent experiments suggesting that it is highly rapid and transient.

Given that the expression pattern of c-Jun and collagen I is highest in proliferating
osteoblasts while osteocalcin is a marker of mature osteoblasts, our results suggests that
glucose treated osteoblasts may be “reverting” to a less differentiated state thereby
decreasing mineralization and ultimately bone mass in viva. Since both collagen I and
osteocalcin promoters contain AP—l sites which are respectively postively and negatively
responsive to c-Jun, it is highly possible that induction of c-jun by glucose is involved in
the effects on these markers of osteoblast phenotype.

Taken together our findings demonstrate that an additional effect of glucose,
osmotic stress, should be considered to contribute to the complications of diabetes with
regard to osteoblasts. Osteoblasts are highly sensitive to osmotic stress and respond
through a protein kinase C dependent pathway that results in an increase in c-jun and
collagen I expression and a decrease in osteocalcin expression. Future studies will
determine the precise role of c-Jun and the signaling pathways involved in the changes in

osteoblast gene expression.

52

IV. Osteoblast Osmotic Response to Elevated Extracellular Glucose

in vitro and in viva Involves AP-l
(Manuscript submitted to JBMR)

1. Abstract

Osteoblasts, the cells responsible for bone formation, express glucose transporters
with a low Km. In poorly controlled or untreated type I diabetes mellitus blood glucose
levels are constantly high. Consequently, this condition should place the osteoblast under
osmotic stress. Although controversial, several studies have demonstrated that diabetes
type I is associated with decreased bone mass. Here we demonstrate that long-term
culturing of MC3T3-El osteoblasts in elevated glucose or mannitol (final sugar
concentration equal to 22 mM) results in increased alkaline phosphatase activity and
decreased mineralization relative to euglycemic 5.5 mM cultures. Similar results were
obtained in viva following 7 daily mannitol injections over the calvaria.

Osteoblast growth was not affected by elevated glucose or mannitol treatment.
Short-term responsiveness, as determined by upregulation of c-jun and collagen I mRNA
levels was induced by a 1 hour treatment with 16.5 mM glucose, mannitol, or 3-0-
methyl-glucose. In combination with osteoblast uptake of sugar not differing between
5.5 and 22 mM glucose, these results further support that osteoblasts are exhibiting an
osmotic response. Similar acute responses were observed in viva. To understand the role
of AP-l in the response A-FOS, a dominant negative AP—l construct, was overexpressed
in osteoblasts and shown to block endogenous c-jun and collagen I mRNA induction by
elevated extracellular glucose. These findings demonstrate an important role for AP-l in

osteoblast osmotic responsiveness.

53

2. Introduction

Diabetic bone disease was recognized at the beginning of 20th century (Morrison
and Bogan, 1927). IDDM is associated with osteoporosis (Auwerx et al., 1988; Krakauer
et al., 1995), decreased bone mass (Hui et al., 1985; Levin et al., 1976) impaired skeletal
deve10pment (Levin et al., 1976) and increased fracture rates (Bouillon, 1991; Meyer et
al., 1993). A recent clinical study of patients with IDDM found that 67% of diabetic men
and 57% of diabetic women suffered from osteopenia (bone mineral density >1 but < 2.5
standard deviation (SD) of the population) of the femoral neck and/or lumbar spine
(Kemink et al., 2000). Osteoporosis (bone mineral density >2.5 SD) was diagnosed in
14% of the male patients. These effects are even more pronounced in diabetic rats
(Verhaeghe et al., 1990). Diabetic rats also exhibit significantly less osseointegration of
bone implants than control (McCracken et al., 2000). Histochemical studies demonstrate
that osteoblasts are directly influenced by diabetes. Specifically, osteoblast
number/surface is reduced (cuboidal osteoblasts are replaced by bone lining cells) and
osteoid surfaces are markedly decreased with no reduction in osteoclast number
(Verhaeghe et al., 1990; Sasaki et al., 1991). These changes do not correlate with bone
loss resulting from low vitamin D or calcium levels, and they do not occur in semistarved
animals (Shires et al., 1980).

Diabetes is diagnosed by an elevated blood glucose level that is greater than 200
mg/dl (11 mM) at any random testing (Emancipator, 1999). This level of glucose can
cause significant changes in cellular uptake and metabolism of glucose in cells, such as
those in the kidney, liver and beta islet cells of the pancreas that express GLUT2, a high

Km (~ 20 mM) glucose transporter. On the other hand, osteoblasts, the cells responsible

54

for bone formation, express only GLUTl and GLUT3 (Thomas et al., 1996 a; Thomas et
al., 1996 b), glucose transporters with a low Km, 1-2 mM and <1 mm, respectively (Bell,
1991). Therefore at euglycemic state, when glucose is 3-5.5 mM, these transporters have
reached their maximal capacity to transport glucose. In poorly controlled or untreated
type I diabetes mellitus glucose level is constantly high. Consequently, if the glucose
treatment does not alter the number of glucose transporters on the osteoblast cell surface
membrane, this condition should place the osteoblast under osmotic stress.

To form a bone tissue-like organization in vitro and in viva osteoblasts undergo
three stages of development: proliferation, extracellular matrix maturation, and
extracellular matrix mineralization (Quarles et al., 1992; Owen et al., 1990). During
extracellular matrix maturation alkaline phosphatase expression increases, while
osteocalcin (marker of the final stage of differentiation and matrix mineralization)
expression decreases. Regulation of gene expression during osteoblast differentiation
occurs at the level of transcription although it is clear that post-translational modifications
of transcription factors also play a critical role.

The AP-l family of transcription factors (Fos and Jun proteins) are critical in the
response of osteoblasts to changes in extracellular stimuli such as vitamin D (Candeliere
et al., 1991), PTH (Clohisy et al., 1992; Lee et al., 1994; Koe et al., 1997; McCauley et
al., 2001), IGF (Merriman et al., 1990), BMP—2, (Ohta et al., 1992), or TGF-B (Breen et
al., 1994; Machwate et al., 1995). All AP-l family members contain a leucine zipper
and a basic region required for dimer formation and DNA binding, respectively
(Kouzarides and Ziff, 1988). Dimerization is a prerequisite for DNA binding at AP-l

response elements (5’-TGAg/cTCA-3’). AP-l member levels as well as post-

55

translational modifications can influence DNA binding and transactivation (Gruda et al.,
1994; Smeal et al., 1992). Modulation of Fos and Jun family member expression in bone
by overexpression or targeted gene ablation further demonstrates a role for this family in
the regulation of osteoblast growth and differentiation (McCabe et al., 1996; Jochum et
al., 2001; Sabatakos et al., 2000; Grigoriadis et al., 1993; Sunters et al., 1998). '

The influence of glucose on osteoblast phenotype has not fully been examined,
especially with regard to changes in osteoblast gene expression and differentiation.
Important mechanisms by which glucose can influence osteoblast differentiation include
altering cell metabolism, extracellular osmolality, signal transduction pathways, and
transcription factor activities. Previously, we have shown that elevation of extracellular
glucose stimulates the expression of c-jun and collagen 1, genes that contain AP-l sites in
their promoters (Zayzafoon et al., 2000).. These findings suggest an important role for
AP-l in the osteoblast response to osmotic stress. Here we demonstrate that
hyperosmotic stress does not alter osteoblast growth, but chronic treatment does suppress
mineralization and is associated with increased alkaline phosphatase activity in vitro and
in viva. Furthermore, in vitro upregulation of c-jun and collagen I expression in response

to osmotic stress is dependent upon AP-l activity and also occurs in viva.

3. Materials and Methods

3.1 Cell Culture System
MC3T3-E1 cells (Sudo et al., 1983), subcloned for maximal alkaline phosphatase

staining and mineralization, were plated at 100,000 cells per 100 mM dish and fed every

56

2 days with alpha MEM (Gibco) containing 5.5 mM glucose (normal level) and
supplemented with 10% fetal calf serum. Eleven days after plating and 24 hours after the
last feeding, glucose, mannitol or 3-O-methylg1ucose (0.5M stock) were added directly to
the media in the tissue culture dish to yield the final concentrations of sugar noted for
each experiment (22 mM). For differentiation studies, upon confluency osteoblasts were
fed with media containing 25 ug/ml ascorbic acid and 2 mM beta-glycerol phosphate

(Wang et al., 1999).

3.2 Histology

Osteoblasts and dissected calvaria were fixed in 2% paraforrnaldhyde in
phosphate buffered saline for 10 minutes and stored at 4°C in 0.1 M cacodlyic buffer as
previously described (Breen et al., 1994). Alkaline phosphatase activity was visualized
by incubating cell layers or calvaria for 30 minutes at 37 °C with 0.5 mg/ml naphthol AS-
MX phosphate disodium salt and 2.8% (v/v) NN dimethyl forrnamide with 1 mg/ml Fast
Red TR salt (Sigma Chemical Co., St Louis, MO) in a 0.1 M Tris-HCl buffer, pH 8.4.
Mineral deposition was assessed by von Kossa staining of the cultures (30 minutes in 3%

AgNO3) (Clark, 1981).

3.3 RNA Analysis

Twenty-four hours after feeding, glucose, mannitol or 3-O-methylglucose (0.5 M
stock) was added directly to the media in the tissue culture dish to yield a final
monosaccharide concentration of 22 mM. Cells were washed and scraped in PBS at the

indicated times following treatment. Cells were centrifuged at 800 X g for 5 minutes and

57

quick-frozen in liquid nitrogen. Total RNA was extracted and analyzed by northern blot,
as previously described in (McCabe et al., 1995; Chomczynski and Sacchi, 1987).
Northern blots were hybridized with random primed (Random primed DNA labeling kit,
Gibco) 32P—labeled cDNA probes specific for c-jun and Collagen l (Zayzafoon et al.,
2000). Hybridization signals were quantitated by phosphoimaging analysis and

expressed relative to 18S ribosomal subunit levels.

3.4 Glucose Media Measurement Assay

MC3T3-E1 cells were treated with 5.5 mM glucose for 11 days. Twenty four
hours after the last feeding cells where treated with 16.5mM glucose or mannitol to yield
a final sugar concentration of 22 mM. Sample 10011.1 was taken from the media at a
different time point ranging from 0 to 24 hours. Glucose assay was done on the samples
to determine the concentration of glucose according to manufacture protocol (Reagents

for the quantitative determination of glucose, Sigma Diagnostics).

3.5 3-O-methylglucose Uptake Assay

The uptake of radiolabeled non-metabolized glucose analog [3H] 3-O-
methylglucose (specific activity of 60 Ci/mmole, Amersham Corp.) was measured as
previously described (Asano et al., 1992). Specifically, twenty—four hours after feeding,
glucose or mannitol (0.5 M stock) was added directly to the media in the tissue culture
dish to yield a final monosaccharide concentration of 22 mM. At the indicated times and
at 37°C, 0.411Ci or 1.6 uCi of [3H] 3-O-methylglucose was added to the 5.5 mM or 22mM

sugar treated plates, respectively, to maintain consistent 3-O-MG/sugar ratio. The glucose

58

uptake was halted after 10 seconds by adding 20 11M cytochalasin B followed by five

washes with chilled 1x PBS. Cells were then lysed and radioactivity counted.

3.6 Cell Growth and Proliferation Assay

Cell proliferation was determined by direct cell counting of viable cells and by
[3H] thymidine incorporation assay. For direct cell counting, MC3T3-El cells were
grown for different period of time as indicated in the study. Treatment duration varied
from day 1 until day 4, 4-8, 8-12, 12-14 from plating. At the end of the treatment, the
plates were washed 3 times with cold PBS and trypsinized for 1-3 minutes. Complete cell
detachment from plates was confirmed by observation under microscope. Osteoblasts
were resuspended in PBS and an aliquot was treated with 4% trypan blue. Viable cells
which excluded the dye were counted using a hemocytometer.

For [3H] thymidine incorporation, MC3T3-E1 cells were grown and treated as
above. At the end of the treatment, [3H] thymidine was added to the culture medium at
concentration of 5 uCi per ml. Cells were then incubated for 30 minutes at 37°C. Culture
medium was removed by aspiration and osteoblasts were washed twice with cold PBS.
Plates were kept on ice and 10% cold trichloroacetic acid (TCA) was added for 5
minutes. Ten % SDS was added to cells for 2 minutes after the removal of TCA (McCabe

et al., 1994).
3.7 In vivo analysis

Balb/c mice, 7 days old, were injected with 50 ul of PBS alone (vehicle) or

vehicle containing 22 mM mannitol subcutaneously over the calvaria. This method has

59

been previously used successfully to examine osteoblast response in viva to hormones
such as PTH (McCauley et al., 2001), FGF (Dunstan et al., 1999) and TGF-B (Fujimoto
et al., 1999). Calvaria were dissected and flash-frozen in liquid nitrogen one hour after
injection. Samples were subsequently processed for RNA analysis as described above.
For chronic studies, 7 day old Balb/c mice were injected daily as described above, for 7
days. Calvaria were dissected and processed for alkaline phosphatase and von Kassa

staining as described above.

3.8 Transient transfection studies

Osteoblasts, MC3T3-El cells, were plated at a concentration of 100,000 cells per
well of a 6 well dish. Twenty-four hours later cells were transfected with a luciferase
reporter plasmid driven by four tandem copies of the AP-l enhancer fused to TATA-like
promoter (PTAL) region from the Herpes simplex virus thymidine kinase (HSV-TK)
promoter (CLONTECH Laboratories); the same reporter vector without inserted
promoter; or an SV40 promoter-beta galactosidase reporter. Five hours after transfection
with the DNA-lipofectamine mix (Gibco), 1 mL of 20% FBS a-mem was added to the
cells. Fifteen hours later, media was aspirated and replaced with fresh 10% FBS a-MEM.
24 hours later, cells were treated for 3 hours with the indicated treatment in the study and
then harvested and lysed. Luciferase activity was read immediately using Promega
luciferase assay system and a luminometer. For AP-l dominant negative experiments, a
CMV driven A-FOS expression plasmid (Olive et al., 1997) or an empty CMV

expression vector were transfected into MC3T3-E1 cells as described above. Forty-eight

60

hours after transfection osteoblasts were treated with glucose, mannitol or PBS (control).

One hour after treatment RNA was extracted and analyzed as described above.

3.9 Statistical analysis

All statistical analyses were performed using Microsoft excel data analysis
program for t test analysis. Experiments were repeated at least three times unless
otherwise stated. The autoradiographs shown are of one representative experiment.

Values are expressed as a mean i SEM except where indicated.

4. Results

4.1 Elevated levels of extracellular sugar suppress osteoblast mineralization both in
vitro and in viva

Previously, we have shown that acute elevation of extracellular glucose can alter
MC3T3-E1 osteoblast gene expression. This suggests that long-term culturing of
osteoblasts under high glucose conditions could have significant phenotypic effects. To
test this, we fed osteoblasts every two days with media containing 5.5 mM glucose
(control) or treated with an additional 16.5 mM glucose or mannitol for 21 days.
Histochemistry demonstrated that staining of alkaline phosphatase activity, a marker of
the matrix maturation stage of osteoblast differentiation, was higher in 22 mM glucose
treated cultures compared to control (Figure 9). In contrast, mineralization as determined
by Von Kossa staining was lower in the sugar-treated cultures. This result is consistent

with the decreased bone formation seen clinically in diabetic patients (Kemink et al.,

61

2000). To determine if similar changes in osteoblast phenotype could be observed in
viva, 22 mM mannitol was injected subcutaneously over mouse calvaria daily for 7 days.
Calvaria were processed for histology. Consistent with our in vitro findings, Figure 10
shows that mannitol treated calvaria have increased alkaline phosphatase activity and
decreased mineralization compared to the saline injected controls. Similar results were
obtained when glucose was injected (data not shown). Given that mannitol is not
transported across the cell membrane nor is it metabolized, these findings suggest that
osmotic stress suppresses late stages of osteoblast differentiation associated with
mineralization, hence expression of genes associated with early stages of differentiation,

alkaline phosphatase, is maintained.

4.2 Sugar treatment does not influence glucose uptake

To determine if the increase in extracellular glucose stimulates glucose uptake, a
time course of 3-O-methyl-glucose uptake was examined in MC3T3-E1 cells. Osteoblasts
cultured in 5.5 mM glucose for 10 days were treated with 16.5 mM glucose or mannitol
to yield a final concentration of 22 mM sugar. Figure 11 demonstrates that elevation of
extracellular glucose is not accompanied by an increase in glucose uptake.
Correspondingly, measurements of glucose concentration in the medium over a twenty
four-hour period after treatment also did not demonstrate a significant difference in
glucose utilization among the various conditions (figure 12). These results are consistent
with constitutive levels of low Km glucose transporters present in the membranes of

osteoblasts.

62

4.3 c-jun is induced within one hour in response to osmotic stress

Next, we treated osteoblasts with different sugars to further determine if
absorption and/or metabolism are important for osteoblast responsiveness to increased
extracellular sugar. We have previously shown that c-jun mRNA levels are increased in
response to elevated extracellular glucose (Zayzafoon et al., 2000), therefore, we
examined the induction of this gene as a marker of responsiveness. Osteoblasts were
treated with 16.5 mM glucose (absorbed and metabolized), 3-O-methyl-glucose
(absorbed but not metabolized), or mannitol (not absorbed or metabolized) to yield a final
concentration of 22 mM. As shown in Figure 13, addition of any of the sugars resulted in
a significant increase in c-jun mRNA levels within one hour of treatment. Thus, sugar
absorption and metabolism are not required for osteoblast responsiveness to elevations in
extracellular glucose. This finding confirms our glucose uptake assay results and
demonstrates that the response is osmotic. It is also consistent with mannitol treatment
causing phenotypic changes in viva and in vitro that are similar to the changes seen in

vitro with chronic glucose treatment.

4.4 Glucose treatment does not affect growth of MC3T3-El cells

Elevation of c-jun expression can be associated with growth so we next examined
if increasing extracellular glucose was stimulating osteoblast proliferation. Osteoblasts
were incubated in high glucose or mannitol containing media (addition of 16.5 mM sugar
to obtain a final concentration of 22 mM sugar) for 4 days at different stages of the
growth and differentiation time course and cell counts and thymidine incorporation were

1
determined. As shown in Figure 14, elevation of extracellular sugar from day 1-4, 4-8, 8-

63

12, or 12-16 has no significant effect on osteoblast growth. Similarly, osteoblast number,
as determined by cell counts, did not differ during and following these treatments. (data

not shown).

4.5 Elevation of extracellular glucose increase AP-l transactivation

Alternatively, induction of c-jun could influence the expression of other genes not
associated with growth. For example, elevated extracellular glucose can increase
expression of collagen I in osteoblasts as we have previously shown (Zayzafoon et al.,
2000). To determine if AP-l transactivation is indeed increased in response to elevations
in extracellular glucose, osteoblasts were transfected with a luciferase reporter driven by
four tandem copies of AP-l. We demonstrate in figure 15 that elevation of extracellular
sugar increases AP-l transactivation over 3 fold. We have previously shown that PKC is
important for the c-jun induction (Zayzafoon et al., 2000), therefore, it is not surprising to
find an increase in AP-l transactivation as it is known that an increase in AP-l

transactivation is associated with the activation of PKC (Angel and Karin, 1991).

4.6 AP-l dominant negative, A-FOS, blocks c-jun and Collagen I Induction

The promoter of collagen I contains a functional AP-l site (Katai et al., 1992) and
therefore could be responsive to increased c-jun expression. To determine the importance
of AP-l in the induction of collagen I we utilized an AP—l dominant negative construct,
A-FOS (Olive et al., 1997), to inhibit AP-l binding activity. A-FOS has its basic region
replaced by an acidic domain; this modified domain binds to the basic region of its Jun

partner and therefore inhibits the AP-l complex from binding DNA. Compared to vector

transfected cells, A-FOS suppressed the glucose induced increase in endogenous collagen
I mRNA levels (Figure 16) consistent with collagen I being a downstream target of c-Jun.
Interestingly, A-FOS also suppressed induction of c-jun by elevated extracellular glucose.
This is consistent with the positive autoregulatory AP-l site located within the c-jun
promoter (Angel et al., 1988). Thus, AP-l binding and transactivation are important in

the osmotic upregulation of both c-jun and collagen I expression.

4.7 c-jun and COL I are induced within one hour in response to elevated
extracellular mannitol in viva

To examine acute responsiveness in viva, mice were given a single injection of 22
mM mannitol subcutaneously over the calvaria. One hour after injection RNA was
extracted from mannitol treated and saline injected controls. Figure 17 shows that both c-
jun and collagen I mRNA expression are induced by the mannitol injection while actin
expression is not affected. These results are again consistent with our in vitro response

data.

65

Control Glucose Mannitol
5.5 mM Glucose 22 mM Glucose 5.5 mM Glucose
+ 16.5 Mannitol

 

Figure 9. Elevated levels of extracellular sugar suppress osteoblast mineralization.
MC3T3-El cells were fed regularly with media containing 5.5 mM glucose (control,
left), 22 mM glucose (glucose treated, center) and 5.5 mM glucose +16.5 mM mannitol
(mannitol treated, right). Cells were treated beginning on day 4 and were fixed on day
28. Following fixation, samples were stained for alkaline phosphatase activity (red) and

mineralization (Black). Similar results were seen in three other experiments.

 

Control

Mannitol

 

 

 

 

 

Figure 10. Osmotic stress suppresses bone mineralization in mice. Mice 7 days old
were inject subcutaneously with 22 mM mannitol over the calvaria for 7 days. Calvaria
were then harvested and fixed. The specimens were then stained for alkaline phosphatase

activity (AP, red) and mineralization (VK, black).

67

.

u
in

._—_'4. _.. __..1._ ._ L_ ____.L __ _.i

H

 

n
In

Fold increase
(Relative to 0 hour)
at u

 

‘
_L

0.51

 

 

 

 

 

 

 

 

 

 

 

Ohou Smh loath 20mh 30mh 11101! Show More 24 hot:

"me after the addition of sugar

Figure 11. Sugar treatment does not influence glucose uptake. Osteoblasts were fed
every 2 days with media containing 5.5 mM glucose. On day 11 (24 hours after feeding)
glucose (G, 16.5 mM) or mannitol (M, 16.5 mM) were added directly to the tissue culture
media of confluent cultures of osteoblasts to yield a final concentration of 22 mM
monosaccharide. At the indicated times [3H] 3-O-methylglucose a non-metabolized
glucose analog was added. Glucose uptake was halted after 10 seconds by adding 20 11M
cytochalasin B followed by five washes with chilled 1x PBS. Cells were then lysed and
radioactivity counted. Data are the mean :t SD and are combined from two experiments,

triplicate determinations for each condition.

68

N
0|

N
D

d
01
1—““—"i 1'" ' *‘i “"_ '1

 
 

_-
O

Media’s glucose concentratlon (mM)
as

 

 

 

 

 

O
1

Figure 12. Steady influx of glucose into MC3T3-El cells under hyperosmolar
conditions. MC3T3-E1 cells were treated with 5.5 mM glucose for 11 days. At day 12,
16.5mM glucose or mannitol was added to the media. 100111 sample was taken from the
media at a different time point ranging from 0 to 24 hours. Glucose assay was done on
the samples to determine the concentration of glucose. Data are the mean :1: SD and are

combined from two experiments, triplicate determinations for each condition.

69

c-lun ——> iii i- ,1; Iii

 

 

 

 

 

 

 

 

 

\ \ \
rises
o c
0 G ‘9 O s
5..
or 1 *
* *
‘8 4i i-
92 i
i=3 :1-1
“C i
.9 2.:
.3 .
6, 11'
l
01; I J .
Control Glucose Mannitol 3—OMG

Figure 13. c-jun is induced within one hour in response to elevated extracellular
sugar. MC3T3-E1 cells were cultured in 5.5 mM glucose. On day 12 osteoblasts were
treated with glucose, mannitol or 3-O—Methyl-D-Glucose (3-OMG) to yield a final
concentration of 22 mM (actual levels were 5.5 mM glucose and 16.5 mM sugar). Sugars
were added directly to the media (24 hours after feeding) to prevent any influences from
changing the media. c-jun expression was analyzed by northern blot analysis using RNA
extracted from cells harvested at one hour after the addition of sugar to the media. Levels
of c-jun mRNA (as determined by phosphoimager analysis) were expressed relative to
ribosomal 18S subunit RNA levels and plotted relative to control values set at 1. Values

were obtained from 3 separate experiments and are expressed +/- SE; *p<0.05.

70

 

 

 

 

 

c. 12 C G M
.2
35 ‘
gg 03:
g8
-5 0.6
38
u°.§ 0.4
E
E ”2
o.
Day1-4 Day4-8 Day 912 Day12-16

Figure 14. Glucose treatment does not affect growth of MC3T3-E1 cells. MC3T3-E1
cells were treated with 5.5 mM glucose (C), 16.5 mM glucose (G), or 16.5 mM mannitol
(M) to yield a final concentration of 22 mM monosaccharide. Treatment duration was for
4 days and began on day 1,4,8 or 12 after plating. At the end of the treatment duration, 5
uCi/mL of [3H] thymidine was added for 30 minutes and thymidine incorporation was
measured. Data are the mean :1: SEM and are combined from three experiments, triplicate

determinations for each condition.

71

004501

AP-l Luciferase
Fold 1ncrease
to

 

.8
r
I

 

 

 

 

 

O

Control Glucose Mannintol

Figure 15. Elevation in extracellular sugar increases AP-l transactivation. MC3T3-
El cells were transfected with luciferase reporter plasmid driven by four tandem copies
of the AP-l. Forty-eight hours later, osteoblasts were treated for 3 hours with 16.5 mM
glucose or mannitol to yield a final concentration of 22 mM of monosaccharide. Cells
were then harvested and lysed for luciferase activity detection. Data are the mean :1: SE
and are combined from four experiments. Data are the mean :1: SEM and are combined

from three experiments, triplicate determinations for each condition. *p<0.05.

72

 

 

COL 1 fold increase

 

 

 

 

 

 

Control Citroen cm Gluco-

Vector A-FOS

 

Actin—i: DC.“

 

c-jm fold increase

 

 

 

Cellini 611100. Moi (loom
_

Vector A-F OS

 

 

Figure 16. AP-l dominant negative, A-FOS, blocks c-jun and Collagen 1 Induction.
MC3T3-E1 cells were transfected with CMV vector alone or CMV A-FOS expression
plasmid. Cells were treated with 16 mM glucose to yield a final concentration of 22 mM
of monosaccharide. One hour later cells were harvested for RNA extraction and c-jun and
COL 1 expression measured by northern blot analysis, representative autoradiographs are

shown. Data are the mean 1 SD and are combined from two experiments.

73

 

COL 1 fold Increase

 

 

 

 

 

Control Mannitol

&

u
1,_M . ___.L_.L__+ LL._._.._.L.CL_.._*_J

 

 

c-[un fold hereon
a to

 

 

 

 

 

Control Mamltol

Figure 17. Osmotic stress in viva induces c-jun and COL I within one hour. Balb/c
mice, 7 days old, were injected with 50 ul of PBS alone (vehicle) or vehicle containing 22
mM mannitol subcutaneously over the calvaria. Calvaria were dissected and flash-frozen
in liquid nitrogen one hour after injection. c-Jun and COL 1 expression was analyzed by
northern blot analysis using RNA extracted from specimens. Levels of c-jun mRNA (as
determined by phosphoimager analysis) were expressed relative to ribosomal 18S subunit
RNA levels and graphed relative to control values set at 1. Values were obtained from 3

separate experiments and are expressed +/- SE; *p<0.05.

74

5. Discussion

Elevated blood glucose levels greater than 200 mg/dl (11 mM) are the diagnostic
measure for diabetes. Elevations in extracellular glucose have been shown to influence a
variety of tissues including vascular, kidney, and hematopoietic cells. Here we show that
osteoblasts are also influenced by this condition. Long-term treatment of MC3T3-El
osteoblasts with 22 mM glucose (a physiologically relevant hyperosmolarity) results in
decreased mineralization. This is consistent with clinical data that suggest diabetes type I
is associated with decreased bone mass in humans (Auwerx et al., 1988; Krakauer et al.,
1995; Hui et al., 1985; Kernink et al., 2000) as well as in rodent models (Verhaeghe et al.,
1990; McCracken et al., 2000) and histochemical studies that suggest osteoblast
differentiation is suppressed (Verhaeghe et al., 1990). Mannitol treatment reproduced
results seen with glucose treatment demonstrating that osmotic stress is involved in the
response. We did notice some slight differences between mannitol and glucose
treatments such as mannitol being a better inhibitor of mineralization. This may stem
from mannitol being a purely osmotic effector. In addition, longterm treatment with
elevated glucose could have some metabolic effects that we did not detect in our short-
term uptake studies. Our chronic calvarial injections further demonstrate that the osmotic
response we are seeing is not simply an artifact of the in vitro culture system or of our
MC3T3-E1 cell line. Specifically, daily subcutaneous injections of 22 mM mannitol or
22 mM glucose over the calvaria suppress mineralization. This is the first in viva
evidence that osteoblast differentiation can be suppressed by osmotic stress.
Interestingly, both in vitro and in viva systems exhibited increased alkaline phosphatase

expression in response to elevated extracellular glucose treatment. This finding could

75

result from the progression of osteoblast differentiation being “blocked” in the matrix
maturation stage, marked by alkaline phosphatase expression (Owen et al., 1990), and
therefore osteoblasts are inhibited from entering the final stage of differentiation
associated with mineralization. This is consistent with our findings of suppressed
osteocalcin expression, a marker of fully differentiated osteoblasts, in response to glucose
treatment (Zayzafoon et al., 2000). Recently, Balint et al. have examined MC3T3-E1
phenotype in response to elevated glucose treatment (Balint et al., 2001). After 30 days of
treatment alkaline phosphatase activity was increased while calcium deposition was
decreased but nodule size and number was increased. Unlike our studies these changes
were not seen in mannitoltreated cultures. Differences between findings may result from
different MC3T3-E1 cell populations (Wang et al., 1999), as well as differences in the
time of analysis. Osteoblasts cultured for 30 days can exhibit different metabolic needs,
such as higher glucose utilization, and different responses compared to less confluent and
less differentiated cultures. In our studies, cells were fed daily after day 10 to try to
accurately maintain the level of glucose in the media.

Although differentiation was suppressed, we did not observe a change in
osteoblast growth in response to elevated glucose. We examined growth following 4 day
periods of glucose (22 mM) treatment (day 1-4, 4-8, 8-12, 12-16) to access if osteoblasts
at a particular stage of differentiation could be influenced to grow by elevated
extracellular glucose levels. In contrast, (Terada et al., 1998) found that osteoblast (MG-
63 cells) IGF-1 induced growth was decreased by elevated extracellular glucose whereas
Balint et al. (2001) found that glucose alone stimulated day 30 osteoblast growth.

Differences between these studies include sugar concentration, time point of analysis, and

76

cell types. For example, growth seen in day 30 osteoblast could result from an elevated
carbohydrate requirement of osteoblasts at this stage.

The characteristics of an osteoblast suggest that it should responds to elevated
extracellular sugar through an osmotic response. Principally, osteoblasts express two
glucose transporters GLUT] and GLUT 3 (Thomas et al., 1996 a; Thomas etal., 1996 b).
Each has a relatively low Km, 1-2 mM and <1 mm, respectively (Bell, .1991). Therefore
at euglycemic state, when glucose is 3-5.5 mM, these transporters have reached their
maximal capacity to transport glucose. When extracellular glucose levels are increased
it cannot enter the cell and be metabolized but rather it puts osteoblasts under osmotic
stress. This is supported by. our finding that glucose uptake is not modulated by elevating
extracellular glucose levels. Similarly, mesangial and endothelial cells that also express
GLUT 1 and GLUT 3 and not GLUT 2 have also been shown to exhibit an osmotic
response to small elevations in extracellular glucose (Kreisberg et al., 1994; Duzgun et
al., 2000). To definitively test this issue in osteoblasts we examine acute stimulation of
c-jun and collagen I expression in response to elevated glucose, mannitol (not absorbed),
or 3-O—methyl glucose (absorbed but not metabolized). Consistent with our previous
studies (Zayzafoon et al., 2000) all sugars stimulated c-jun and collagen I expression
demonstrating that indeed absorption and metabolism are not required for the response.

Small changes in extracellular osmolarity have been shown to influence other
cells. For example (Park et al., 2000) demonstrated that high glucose can induce
intercellular adhesion molecule-1 (ICAM-1) expression through an osmotic effect in rat
mesangial cells. Monocytes as well elicited an osmotic response due to elevation in

extracellular glucose involving ICAM-1 (Manduteanu et al., 1999). When fibroblasts

77

were cultured in physiologically relevant hyperosmolar medium (10-15 mM glucose)
both mRNA and protein levels of collagen III were increased. Fibronectin was also
increased (Benazzoug et al., 1998). These findings demonstrate that a moderate increase
in extracellular glucose can have a dramatic effect on gene expression in a variety of cells
including osteoblasts.

The induction of c-jun by osmotic stress is consistent with its role as an
immediate response gene and its activity being regulated by MAP kinase signaling
pathways that have been shown to be responsive to osmotic stress (Rosette and Karin,
1996; Duzgun et al., 2000; Sheikh-Hamad et al., 1998). Similar to our findings
(Kreisberg et al., 1994) demonstrated that c-jun was induced by 15 mM glucose in
mesangial cells. Upregulation of collagen I is consistent with it being an AP-l responsive
gene (Katai et al., 1992). Increases in extracellular matrix have been described in other
tissues in response to diabetes. Collagen l and other matrix mRNAs have been
previously reported to be elevated in response to extracellular glucose in mesangial cells
(Ziyadeh et al., 1995) and fibroblasts (Han et al., 1999; Benazzoug et al., 1998).
However an upregulation in osteoblast collagen I expression is not consistent with reports
of decreased osteoid surface in diabetic bone (Verhaeghe et al., 1990). One possibility is
that collagen RNA levels may not be directly related to matrix levels as a result of down
regulation of synthesis, processing or secretion. On the other hand some type of feed
back system may exist where collagen I expression is increased but it’s degradation is
increased. Collagenase-3 expression is also responsive to AP-l levels (Hess et al., 2001;
Winchester et al., 2000; Varghese et al., 2000; Rydziel et al., 2000) and could mask any

increase in collagen I matrix production.

78

Here we also demonstrate that our acute and chronic in vitro results are consistent
with the in viva responses and are not an artifact of the tissue culture system.
Subcutaneous injections above the calvaria have been previously used to examine
osteoblast response in viva to hormones such as PTH (McCauley et al., 2001), FGF
(Dunstan et al., 1999) and TGF-B (Fujimoto et al., 1999).

The use of a dominant negative AP-l construct, A-FOS demonstrates that osmotic
induction of collagen I and c-jun mRNA is dependent upon AP—l activation. These
findings are consistent with both gene promoters containing functional AP-l sites (Olive
et al., 1997). The AP-l site within the c-jun promoter has been shown to be involved in

the amplification of c-Jun levels (Angel et al., 1988).

79

V. p38 and ATF-2 involvement in osteoblast osmotic response to

elevated extracellular glucose

1. Abstract

Poorly controlled or untreated type I diabetes mellitus is characterized by
hyperglycemia and is associated with decreased bone mass and increased fracture rates.
Cells such as osteoblasts that lack the high Km glucose transporter (Glut2) have limited
capability to absorb glucose and consequently are placed under hyperosmolar stress.
Investigation of osmotic stress in mammalian tissues has focused principally on kidney
and cellular responses to hyperosmolar conditions above 600 mOsm. Recently, we have
shown that osteoblasts are sensitive to osmotic stress and respond to changes in
extracellular glucose and mannitol as little as 5 to 10 mM (180-270 mg/dL). The
osteoblast osmotic response is dependent upon protein kinase C activity and results in
upregulation of c-jun and modulation of collagen I and osteocalcin expression. To
determine if diabetic hyperosmolar stress plays a role in activating MAP kinases in
osteoblasts, we treated MC3T3-E1 osteoblasts (grown in 5.5 mM glucose) with 16 mM
glucose or mannitol for one hour. Immunoblots of phosphorylated p38 demonstrate
significant activation of p38 MAP kinase by both glucose and mannitol and no change in
p38 levels. Increased activity of p38 MAP kinase was not inhibited by staurosporine,
suggesting a PKC-independent pathway in this activation. This increase was time
dependent peaking at 20 minutes and staying detectable after 24 hours. It was also
concentration dependent, being detectable at concentration as low as 10 mM and

increasing gradually as total sugar concentration increased up to 22 mM. ATF-2

80

activation followed the same pattern as phospho p38. EMSA studies showed an increase
in CRE DNA binding 1 hour after hypertonic treatment with glucose or mannitol.
Supershift analysis demonstrated the involvement of ATF-2 in this binding.
Corresponding to EMSA results, CRE transactivation increased 3 hours after sugar
treatment. SB 203580, a p38 MAP kinase inhibitor, inhibited ATF-2 phosphorylation as
well as CRE transactivation. Therefore, we propose that increased p38 MAP kinase
activity and ATF-2 phosphorylation contribute to CRE activation. These findings suggest
that osmotic stress modulates osteoblast signaling pathways contributing to at least one

complication of diabetes, osteoporosis.

2. Introduction

Diabetes is diagnosed by an elevated blood glucose level that is greater than 200
mg/dl (11 mM) at any random testing (Emancipator, 1999). In poorly controlled or
untreated type I diabetes mellitus glucose level is constantly high. Consequently, this
places the osteoblasts under osmotic stress. Response to elevated extracellular glucose
differs greatly between cells. We have shown previously that elevation of extracellular
sugar elicits an osmotic response in osteoblasts (Zayzafoon et al., 2000; Zayzafoon and
McCabe, manuscript submitted).

Cells respond to different extracellular inputs from their environment by
activation of protein kinase cascades. These complex networks allow amplification of the
signal and can mediate the required diversity of cellular responses. Environmental cues
direct osteoblasts to proliferate and differentiate and MAP kinase pathways provide a key

link between the membrane bound receptors that receive these cues and changes in the

81

pattern of gene expression. The MAP kinase signaling pathways have been implicated in
playing a role in mediating the osmotic stress adaptive mechanisms (Roger et al., 1999;
Sheikh-Hamad et al., 1998; Rizoli et al., 1999; Rosette and Karin, 1996). The first
adaptive process occurring in response to extracellular hypertonicity-induced cell
shrinkage is a regulatory cell volume increase (RVI) which results from the stimulation
of ion transporters (Lytle and Forbush, 1996; Grinstein et al., 1986) and reviewed
extensively in (Parker, 1993). A second adaptive mechanism, in mammalian cells, is the
induction of genes encoding proteins involved in the accumulation of intracellular
“compatible osmolytes’ within hours and days (Yamauchi et al., 1992; Kwon et al., 1992;
Uchida et al., 1992; Bohren et al., 1989; Garcia-Perez et al., 1989). To date, the effects
of MAP kinases have been mostly attributed to the control of gene transcription via
phosphorylation of nuclear transcription. This eventually leads to gene modulation, which
plays part in the cellular response to osmotic stress (Hoffert et al., 2000; Ishida et al.,
1999; Schaffler et al., 2000).

IDDM is associated with an extensive list of late complications involving nearly
every tissue. One major concern to young and aging diabetics is the association of IDDM
with osteoporosis, decreased bone mass, impaired skeletal development and increased
fracture rates (Bouillon, 1991; Auwerx et al., 1988; Krakauer et al., 1995; Hui et al.,
1985; Levin et al., 1976; Meyer et al., 1993). Diabetic bone disease was first recognized
at the beginning of the twentieth century (Morrison and Bogan, 1927) and described as
retardation of bone development and bone atrophy in children with long-standing
diabetes. A recent clinical study (Kemink et al., 2000) showed a high prevalence of

osteopenia in patients with IDDM. It showed that 67% of the diabetic men and 57% of

82

the diabetic women suffered from osteopenia of the femoral neck and/or lumbar spine.
These effects are even more pronounced in diabetic rats. Verhaeghe et al., (1990) showed
that long standing diabetes in BB rats results in severe low-turnover osteoporosis
probably related to decreased osteoblasts recruitment and/or function.

Previously, we have shown that elevation of extracellular glucose stimulates the
expression of c-jun and collagen I, genes that contain AP-l sites in their promoters
(Zayzafoon et al., 2000). The use of PKC inhibitors as well as an AP—l dominant
negative (A-Fos) abolish this induction (Zayzafoon et al., 2000; Zayzafoon and McCabe,
manuscript submitted). Morooka et al., (1995) demonstrated that ATF-2 and c-Jun DNA
binding activity is highly regulated in response to stress. This complex then targets
various promoter sites, including the CAMP response element, CRE, (Cai et al., 2000).
Here we demonstrate that hyperosmotic stress caused by a physiologically relevant
increase in extracellular glucose results in the activation of p38 MAP kinase and its
downstream transcription factor, ATF-2. This activation is associated with an increase in
CRE DNA binding and its transactivation. SB 203580, p38 MAP kinase inhibitor, was
able to inhibit the phosphorylation of ATF-2 as well as CRE transactivation. These
results demonstrate that p38 MAP kinase and ATF-2 are involved in the osteoblast

response to moderate osmotic stress.

83

3. Materials and Methods

3.1 Cell Culture System

MC3T3-E1 cells (Sudo et al., 1983) were plated at 100,000 cells per 100 mM dish
and fed every 2 days with alpha MEM (Gibco; Grand Island, NY) containing 5.5 mM
glucose (normal level) and supplemented with 10% fetal calf serum. The typical
concentration of insulin in the culture media is obtained from the serum component and
ranges between 1 and 5 picomole concentration, a level lower than physiologic (50-800
pM), but consistent with early development of IDDM. Eleven days after plating and 24
hours after the last feeding, glucose or mannitol (0.5M stock) were added directly to the
media in the tissue culture dish to yield the final concentrations of sugar noted for each
experiment (7.5 to 22 mM). For kinase inhibition studies, 30 min prior to addition of
sugar the cells were pre-treated with 1011M of SB 203580, concentrations previously

reported for the inhibition of p38 activity (Nadkami et al., 1999; Volonte et al., 2001).

3.2 Whole cell protein extraction

MC3T3-E1 (1 100mm plate) were harvested for whole cell protein preparation.
Cells were washed with chilled 1 x PBS and left on ice 3 minutes. Cells were then
harvested and centrifuged at 800 X g for 5 minutes at 4C. Lysis buffer containing 50mM
Tris pH 7.4, 150 mM NaCl, lmM EDTA, 1%Triton X100 and 10% Glycerol was used to
lyse cells and extract protein. A cocktail of protease and phosphatase inhibitors was

added to the lysis buffer (lmM sodium ortho vanidate, 2 mM PMSF, 5 ug/ml aprotinin,

lmM EGTA, 10 mM NaF, 1 mM Na pyrophosphate and 0.1 mM Bglycerolphosphate).

84

Cells are then centrifuged at 14,000 rpm for 30 minutes at 4C. Protein concentration from

the supernatant was quantitated by the Biorad DC protein detection system.

3.3 Nuclear extract isolation

Osteoblasts (1 100mm plate) were harvested for nuclear extract preparation as
previously described (McCabe et al., 1996). Briefly, nuclei were isolated by detergent
lysis of the cells. Nuclei are then treated with a hypotonic solution followed by a 30-
minute incubation at 4C in a high salt (600 mM KCl) solution. Nuclei are then
centrifuged at 14,000 rpm and the supernatant represents the nuclear extract. All solutions
in this procedure contained a cocktail of protease inhibitors including: PMSF (5111M),
pepstatin (lug/ml), TPCK (70ug/ml), leupeptin (0.5ug/ml), spermidine (lmM) and
sperrnine (lmM) (McCabe et al., 1996). Protein concentration was quantitated by the
Biorad DC protein detection system, which can detect protein levels in the presence of

compounds that inhibit protein measurements in the Bradford method.

3.4 Western blot analysis

Fifty ug of whole cell extracts was loaded per lane on a mini-acrylamide gel.
Following electrophoreses, proteins were transferred to PVDF (Biorad) membranes using
a semi-dry transfer system (Owl). Protein transfer and size determinations were verified
using pre-stained protein markers. Blots were then blocked with 5% nonfat dry milk in
TTBS (McCabe etal., 1996) and were subsequently incubated with antibodies specific to

different MAP kinase proteins as well as transcription factors (Santa Cruz). Signals were

85

detected by using an HRP-conjugated secondary antibody and the enhanced

chemiluminescence (ECL, Amersham) detection kit.

3.5 EMSA

411g of nuclear extracts were incubated for 20 rrrinutes at room temperature with
32P-labeled oligonucleotide containing CRE sequence 5’ AGA GAT TGC CTG ACG
TCA GAG AGC TAG 3’. The DNA-protein complexes formed were separated from the

free probe on a native 5% gel. DNA binding activity was quantitated by phosphoimager

measurement of shifted bands (McCabe et al., 1996).

3.6 Measurement of CRE transactivation by transient transfection

Osteoblasts were transfected with a luciferase reporter plasmid driven by three
copies of the CRE—binding sequence fused to TATA-like promoter (PT AL) region from
the Herpes simplex virus thymidine kinase (HSV-TK) promoter (CLONTECH
Laboratories). MC3T3-E1 cells were plated at a concentration of 100,000 cells per well
of a 6 well dish. Twenty four hours after plating, a mix of 1 ug reporter plasmid and 6 111
of lipofectamine (Gibco) were prepared in Optimem (Gibco) and after 45 minutes added

to the osteoblasts. After 5 hours, 1 m of 20% FBS (x-mem was added to the cells. Fifteen

hours later, media is aspirated and a replaced by fresh 10% FBS 0t-mem. After 24 hours
cells were treated with the indicated treatment in the study and harvested and lysed three
hours later. Luciferase activity is read immediately using a luminometer. Additional

control for these studies include transfection of reporter vector without inserted promoter

86

elements and co—transfection with a CMV driven luciferase plasmid whose activity is not

affected by glucose (activity determined by Promega luciferase assay system).

3.7 Statistical analysis

All statistical analyses were performed using Microsoft excel data analysis program fort
test analysis. Experiments were repeated at least three times unless otherwise stated. The
autoradiographs shown are of one representative experiment. Values are expressed as a

mean 1' SEM except where indicated.

4. Results

4.1 Elevation of extracellular sugar specifically enhances p38 phosphorylation
Previously we demonstrated that elevation of extracellular glucose or mannitol
alters osteoblast gene expression. Given the important role of MAP kinase in the
regulation of transcription factor activities and gene expression, we examined the
influence of treating osteoblasts with 16.5 mM extracellular sugar on activation of MAP
kinase. Figure 18 and 19 demonstrate that total levels of p38, ERK, and JNK are
unaltered after 1 hour of sugar treatment. However, examination of active
phosphorylated MAP kinase forms show a 3-4 fold increase in phosphorylated p38 in
response to elevated extracellular glucose or mannitol. Activation of the p38 pathway
was specific since activation of JNK or ERK was not evident (figure 19). Under these
conditions, positive controls (TPA and 600mM glucose) demonstrated that these kinases

could be activated in osteoblasts.

87

4.2 Elevation of extracellular sugar enhances ATF-2 phosphorylation

ATF-2 is a known down stream target for p38 (Zhu and Lobie, 2000). To
determine if elevation of extracellular glucose or mannitol enhances ATF-2
phosphorylation, western blot analyses were performed using specific antibodies directed
against ATF-2 and phosphorylated ATF-2. Figure 20 demonstrates that while levels of
ATF-2 remain constant, phosphorylation of ATF-2, similar to p38, is clearly elevated 3-4

fold one hour after glucose treatment.

4.3 Elevation of extracellular glucose enhances p38 and ATF-2 phosphorylation in
time dependent manner

To examine the characteristics of the p38 and ATF-2 phosphorylation in response
to osmotic stress, a complete time course was performed with osteoblasts being harvested
at 0, 10, 15, 20, 30 and 60 minutes. Figure 21 A shows that the phosphorylation of p38 is
evident at 10 minutes, plateaus by 20 minutes and remains activated 60 minutes after
glucose treatment. Phosphorylation of ATF-2 is evident within 20 minutes after glucose
treatment; this is slower than the phosphorylation of p38 and is consistent with ATF-2 as
a downstream target of p38. The response after 24 hours was also examined. Although
p38 is considered to be a fast and transient activated kinase, p38 stayed activated even 24
hours after the onset of the osmotic stress (Figure 21 B). As expected, ATF-2 also

remained activated at 24 hours.

4.4 Elevation of extracellular glucose enhances p38 and ATF-2 phosphorylation in

concentration dependent manner

88

To determine if the effect on p38 and ATF-2 phosphorylation was concentration
dependent, osteoblasts were treated with 2, 4.5, 9.5 and 16.5 mM glucose (final
concentration equal to 7.5, 10, 15 and 22 mM glucose respectively) and phosphorylated
p38 and ATF-2 levels where examined 1 hour after treatment using western analysis.
Remarkably, figure 22 demonstrates that addition of as little as 4.5 mM of glucose is
enough to increase p38 phosphorylation in osteoblasts. This level of glucose (10 mM

final concentration; 180 mg/dL) is often seen in patients who are diagnosed with diabetes.

4.5 ATF-2 phosphorylation in response to elevation in extracellular sugar is p38
dependent

SB203580 (4-(fluorophenyl)-2-(4-methylsulfonylphenyl)-5-(4-pyridyl) imidazole
is a known inhibitor of p38 activity (Sheikh-Hamad et al., 1998). To determine the role
of p38 in the phosphorylation of ATF—2 in response to the extracellular sugar elevation
we treated osteoblasts with the 10 uM SB203580, 30 minutes prior to glucose treatment
and examined responsiveness. Figure 23 demonstrates that SB203580 pretreatment (a
concentration suitable for specifically inhibiting p—38 activity (Nadkami et al., 1999)

prior to the addition of 16.5 mM sugar inhibits ATF-2 phosphorylation.

4.6 Increase in CRE DNA binding 1 hour after glucose or mannitol treatment

To demonstrate that the increase in ATF-2 protein phosphorylation is‘associated
with an increase in CRE DNA binding, an electrophoretic mobility shift assay was
performed. Figure 24 A shows that indeed there is an increase in CRE DNA binding one

hour after glucose or mannitol treatment. To identify the involvement of ATF-2 in the

89

shifted DNA-protein complex, a supershift study was performed. Nuclear extracts from
glucose or mannitol treated osteoblasts were incubated with ATF-2 antibody and 32P
labeled CRE DNA consensus probe. Figure 24 B shows that ATF-2 was shifted from the
CRE DNA complex. This demonstrates that ATF-2 is a component of the increased CRE

DNA binding complex resulting from the elevated extracellular sugar.

4.7 Elevation of extracellular sugar increases CRE transactivation in osteoblasts, a
p38 dependent response

To determine if osmotic stress due to elevation of extracellular sugar can
functionally influence CRE transactivation, MC3T3-E1 cells were transfected with a
CRE-Luc or CMV-Luc reporter and treated with glucose or mannitol as described in the
methods section. Figures 25 demonstrates that increasing extracellular glucose or
mannitol enhances CRE transactivation 5 fold compared to the control cells (5.5 mM).
CMV-Luc did not show any increase demonstrating the specificity of the CRE induction.
To study the role of p38 in the transactivation of CRE, we treated osteoblasts with the 10
uM SB203580 30 minutes prior to glucose treatment and examined responsiveness.
Figure 25 demonstrates that p38 inhibition is enough to abolish the CRE transactivation

in response to elevated extracellular glucose.

90

Figure 18. Elevation of extracellular glucose enhances p38 phosphorylation, a PKC
independent response. A.) Osteoblasts cultured for 11 days were treated with 16.5 mM
glucose or mannitol to yield a final concentration of 22 mM. B.) Osteoblasts were
pretreated for 30 minutes with 50uM staurosporine (ST) before the sugar treatment. Cells
were harvested 1 hour after addition of sugar for whole cell protein extraction. Extracts
(50 pg per lane) were separated by 10% SDS-PAGE. Immunoblots were developed using
specific antibodies directed against p38 or phosphorylated p38. The blot is a
representative of three experiments and contains extracts from control 5.5 mM glucose
(C), 22 mM glucose (G) and 5.5 mM glucose + 16.5 mM mannitol (M) treated

osteoblasts. Data are the mean :tSEM and are combined from three experiments. *p<0.03.

91

A.

Phospho-p38——> «a. .- .

p38—9M
C G M

 

 

 

 

*
4 l
a *
g .
8 31
E
"C
6 2
U- .
1 i
i
o - : '
C G M
1 hour
B.
Phospho-p38 —> """'“‘ “'1" ”"'

ST ST+G ST+M

92

g 4———-— Phospho-ERKi
--—- —-- --- 4———— Phospho-ERKZ

,d—u-n ...... _. _ . 4——ERK1

C G MTPA

. .. . 4——Phospho-JNK2
11-. w «1...... “Wm 4—Phospho-JNK1

w _ man-r» WWI-‘1 g ‘ "—‘JNKZ
“I new w' w ‘———'JNK1

Figure 19. Elevation of extracellular glucose does not influence ERK and JN K
phosphorylation. Osteoblasts cultured for 11 days were treated with 16.5 mM glucose or
mannitol to yield a final concentration of 22 mM. Cells were harvested 1 hour after
addition of sugar for whole cell protein extraction. Extracts (50 ug per lane) were
separated by 10% SDS-PAGE. Immunoblots were developed using specific antibodies
directed against ERK 1-2, JN K 1-2, phosphorylated ERK 1-2 or phosphorylated JN K 1-2.
The blot is a representative of three experiments and contains extracts from control 5.5
mM glucose (C), 22 mM glucose (G) and 5.5 mM glucose + 16.5 mM mannitol (M)

treated osteoblasts.

93

Figure 20. Elevation of extracellular glucose enhances ATF-2 phosphorylation, a
PKC independent response. A.) Osteoblasts cultured for 11 days were treated with 16.5
mM glucose or mannitol to yield a final concentration of 22 mM. B.) Osteoblasts were
pretreated for 30 minutes with 50uM staurosporine (ST) before the sugar treatment. Cells
were harvested 1 hour after addition of sugar for whole cell protein extraction. Extracts
(50 pg per lane) were separated by 10% SDS-PAGE. Immunoblots were developed using
specific antibodies directed against ATF-2 or phosphorylated ATF-2. The blots are
representative of three separate experiments and contains extracts from control 5.5 mM
glucose (C), 22 mM glucose (G) and 5.5 mM glucose + 16.5 mM mannitol (M) treated
osteoblasts. Data are the mean 1 SEM and are combined from three experiments.

*p<0.05.

94

A.

Phospho-ATF 2 —. 7' - _.....—

C G M

ATF2—,----

5 *
*

g 4
8
5 3
.E ‘
E 2 I I
0
LL 1

1 [j

C G M
1 hour

B.

Phospho-ATF2——>i""-'" W

ST I “sad ST+M.

95

Phospho-p38 ——> _.....“

Phospho -ATF2 —>_ ' ‘ '~ “1..-
ATF 2 a..."

C 10 15 20 30 60 Minutes
Glucose22mM

B.

Phospho-p38 —-+ --—-- ----

 

   

Phospho -ATF 2 ——>--- ’ , .. g.
C G M 24 hours

Figure 21. Elevation of extracellular glucose enhances p38 and ATF-2
phosphorylation in time dependent manner. Osteoblasts cultured for 11 days were
treated with 16.5 mM glucose or mannitol to yield a final concentration of 22 mM. Cells
were harvested at the indicated times from 0-60 minutes (A) and 24 hours (B) for whole
cell protein extraction. Extracts (50 pg per lane) were separated by 10% SDS—PAGE.
Irnmunoblots were developed using antibodies specific for ATF-2, phosphorylated p38 or
ATF-2. The blots are representative of two separate experiments and contain extracts
from control 5.5 mM glucose (C), 22 mM glucose (G) and 5.5 mM glucose + 16.5 mM

mannitol (M) treated osteoblasts.

96

Phospho—p38 ———+ .. -......

Phospho -ATF 2——2 - -. _... .11

AFT—fl“..-
C 7.5 10 15 22 mM Glucose

 

1 hour

Figure 22. Elevation of extracellular glucose enhances p38 and ATF-2
phosphorylation in concentration dependent manner. Osteoblasts cultured for 11 days
were treated with 16.5 mM glucose to yield a final concentration of 22 mM. Cells were
harvested at the indicated time for whole cell protein extraction. Extracts (50 pg per lane)
were separated by 10% SDS-PAGE. Irnmunoblots were developed using antibodies
specific for ATF-2, phosphorylated p38 or ATF-2. The blots are representative of two
separate experiments and contains extracts from control 5.5 mM glucose (C), 7.5, 10, 15,

and 22 mM glucose treated osteoblasts.

97

Phospho-ATF 2 —> a... ...fi .1- ..— .1
i

ATF2 —* --‘fifi-

CGMC GM
88203580 - - - + + +

Figure 23. Extracellular glucose induction of ATF- 2 phosphorylation can be blocked
by inhibiting p38 activity. Osteoblasts were fed every 2 days with media containing 5.5
mM glucose. On day 11 (24 hours after feeding) osteoblasts were pretreated for 30
minutes with 10 11M SB203580. Cells were then treated for 1 hour with 16.5 M glucose
(G) or 16.5 mannitol (M) to yield a final concentration of 22 mM. Extracts (50 pg per
lane) were separated by 10% SDS-PAGE. Immunoblots were developed using antibodies
specific for phosphorylated ATF-2. The blots are representative of two separate
experiments and contains extracts from control 5.5 mM glucose (C), 22 mM glucose (G)

and 5.5 mM glucose + 16.5 mM mannitol (M) treated osteoblasts.

98

Figure 24. (A.) CRE binding increases 1 hour after elevation of extracellular glucose
or mannitol concentration. MC3T3-El cells were treated with 16.5 mM glucose (G) or
mannitol (M) to yield a final concentration of 22 mM. Control cells (C) were untreated.
Nuclear extracts were prepared and analyzed as described in methods. 41tg of nuclear
extracts were mixed with a 32P-labeled oligonucleotide containing a consensus CRE
sequence 5’ AGA GAT TGC CTG ACG TCA GAG AGC TAG 3’ and EMSA was
performed. The DNA-protein complexes were separated by 5% polyacrylamide gel. The
auto radiographs are representative of 3 separate experiments.

(8.) Increase ATF-2 binding on CRE DNA element in response elevation in
extracellular sugar. 411g of nuclear extracts were incubated overnight with a 32P—labeled
oligonucleotide containing CRE sequence 5’ AGA GAT TGC CTG ACG TCA GAG
AGC TAG 3’ and in the absence (-) or presence (+) of ATF-2 antibody. DNA-protein
complexes were separated by 5% polyacrylamide gel electropheresis and visualized by

autoradiography.EMSA was performed. n=2.

99

A. 1 hour
C G M

CRE—>11? I?

B- ATF-2: - +- + -+

Extract: C C G G M M

ATF-2
super shift

 

 

 

 

0»:

91

“l

a 7.
g 1
fig Bi
"9E s?
is 3‘.

G

in 21
11.

01+

(0‘

0°65

 

Figure 25. Elevation of extracellular sugar increases CRE transactivation in
osteoblasts, p38 dependent response. MC3T3-El cells were transfected with a CRE-
Luc reporter construct. 48 hours after transfection, 16.5 mM glucose or mannitol was
added to the plates to yield a final concentration of 22 mM. Reporter activity was
measured 3 hours after treatment. SB203580 (10 um) was used to inhibit p38 activity.
Osteoblasts were pretreated for 30 minutes with 10 um SB203580 before adding the
sugar. Data are the mean :1: SD of two separate experiments containing triplicate

determinations for each condition.

101

5. Discussion

We have shown previously that osteoblasts respond to an elevation in
extracellular glucose through an osmotic response pathway (Zayzafoon et al., 2000).
Protein kinases are known to be activated in response to different environmental
challenges. We have shown that moderate hyperosmolarity as seen in diabetes affects
osteoblasts by increasing c-jun expression through a PKC dependent pathway resulting in
gene modulation manifested by increase in collagen I and decrease osteocalcin
expression. Here, we show that physiologically relevant hypertonicity increases the
phosphorylation of p38 MAP kinase after 1 hour of exposure to 22mM glucose or
mannitol. This level of hypertonicity failed to activate the other MAP kinases (ERK and
JNK) in MC3T3-E1 cells. The specific activation of p38 in response to osmotic stress has
been reported before in other tissues (Rizoli et al., 1999; Denkert et al., 1998; Nadkarni et
al., 1999). Similar to our results, Duzgun et al., (2000) demonstrate in endothelial cells
that p38 phosphorylation is fast and robust in response to a 50 mM increase in glucose or
mannitol. Unlike ERK and JNK, p38 induction by osmotic shock occurred as early as 2
minutes and this time point coincides with the activation of the volume regulatory
proteins (O’Donnell etal., 1995; ODonnell, 1993). This observation is consistent with the
idea that p38 activation is directly related to the initial cell response to osmotic shock. In
this regard, HOG 1, the homologue of mammalian cell p38 in yeast, has been previously
demonstrated to be critical element of yeast adaptation to osmotic stress (Takekawa et al.,
1997). HOG 1 defective yeast do not proliferate under hyperosmotic condition (Brewster

and Gustin, 1994).

102

To date, the effects of MAP kinases have been mostly attributed to the control of
gene transcription via phosphorylation of nuclear transcription. This eventually leads to
gene modulation, which plays a part in the cellular response to cell shape change and
osmotic stress (Hoffert et al., 2000; Ishida et al., 1999; Schaffler et al., 2000). We have
already shown that AP-l is a crucial component involving increase expression of c-jun
and modulation of collagen I in osteoblast responsiveness to osmotic stress (Zayzafoon
and McCabe, manuscript submitted). Here we show that ATF-2 phosphorylation
increases dramatically after one hour treatment with 22 mM glucose or mannitol. ATF-2
phosphorylation is known to increase in response to cellular stresses such as short-
wavelength UV (Wilhelm et al., 1995; van Dam et al., 1995) and _cellular reperfusion
after ischemia (Morooka et al., 1995). ATF-2 target genes include important bone
regulatory genes such as tumor necrosis factor 01 (TNF or) (Tsai et al., 1996),
transforming growth factor [3 (Kim et al., 1992), cyclin A (Shimizu et al., 1998) and c-jun
(van Dam et al., 1993). These genes are also known to play important roles in the stress
response, cell growth and differentiation, and immune response, therefore it is not
surprising that osteoblasts respond to osmotic stress by increasing ATF-2
phosphorylation as we show in Figure 20. Although the role of ATF-2 in controlling
osteoblast growth and differentiation is still largely unknown, it is known that both
parathyroid hormone (PTH) and prostaglandin E2 (PGE2) influence the growth and
differentiation of osteoblasts mediating CRE (Pearman et al., 1996; Thomas et al., 1996).
Phosphorylation of amino acid residues Thr-69 and Thr-71 on ATF-2 has been
demonstrated to increase transcriptional activation of ATF-2 (Abdel-Hafiz et al., 1992;

van Dam et al., 1995; Tsai et al., 1996). p38 MAP kinase has been shown to

103

phosphorylate these sites (Raingeaud et al., 1996). Therefore, we expected that the
characteristics of ATF-2 phosphorylation to be similar to that of phosphorylated p38. Our
results demonstrate that p38 is phosphorylated within 10 minutes after treatment with 22
mM glucose, but there is a lag period of 10 minutes before ATF-2 is phosphorylated.
This is consistent with ATF-2 being a downstream target to p38. But what we found to be
more interesting was the persistence of phosphorylation of both p38 and ATF-2 after 24
hours. It is known that the ubiquitination of ATF-2 as well as c-Jun is phosphorylation
dependent and the early increase in ATF-2 phosphorylation should increase the
degradation of this protein and make it unlikely to be phosphorylated by 24 hours (Fuchs
and Ronai, 1999). However, it is possible that p38 stress activating protein kinase
mediated phosphorylation may stabilize ATF-2, as has been reported in the case with c-
Jun (Fuchs et al., 1996; Musti et al., 1997). Phosphorylation of both p38 and ATF-2
when extracellular glucose is between 10 and 15 mM is very interesting because diabetes
is diagnosed when glucose levels are between 10‘ and 15 mM. This demonstrates that
osteoblasts are very sensitive to fluctuation in extracellular osmolarity. They elicit an
osmotic response as soon as osmolarity has passed a certain threshold which is around 10
mM.

88203580 is a known specific inhibitor of p38 activity at a low concentration of
and 10 11M and is widely used for this purpose (Nadkami et al., 1999). Using this
inhibitor demonstrates that the hypertonicity-induced ATF-2 phosphorylation is p38
dependent. Consistent with our results Nadkami et al., (1999) showed in hepatic cells
(Hep2G) that the hypertonic induction of aldose reductase mRNA as well as osmotic

response element (ORE)—driven transactivation are both inhibited by SB203580,

104

suggesting that p38 MAP kinase mediates the activation of the transcription factors
necessary for ORE activation.

Next, we show that hypertonicity increased CRE DNA binding and supershift
study shows that ATF-2 is an important part of the complex. In addition, we demonstrate
a five fold increase in CRE-Luc activation in response to glucose or mannitol treatment.
This activation was shown to be p38 dependent as it was inhibited by the use of
SB203580. Similarly, in fibroblasts exposed to stress (heat shock stress in this case) CRE
binding and transactivation were increased (Van dam et al., 1992).

Our data indicate that osteoblast are sensitive to small changes in extracellular
osmolality and that they activate a p38 MAP kinase pathway. This activation leads to
increase in ATF-2 transcription factor phosphorylation. CRE DNA binding and
transactivation were also increased in response to hyperosmolarity. These findings
suggest that osmotic stress modulates osteoblast signaling pathways contributing to at

least one complication of diabetes, osteoporosis.

105

VI. Summary and Conclusions

Osteopenia and osteoporosis are associated with diabetes. Bone mineral density in
patients with both type 1 and type II diabetes has been reported to be decreased by greater
than 10% compared with gender- and age- matched health persons (Levin et al., 1976;
Mathiassen et al., 1990; McNair et al., 1988). The objective of this thesis is to understand
the mechanisms by which elevation of extracellular glucose can suppress osteoblast
function.

We have shown that elevation in extracellular glucose was able to acutely
increase the expression of c-jun and collagen I while decreasing osteocalcin expression.
These results were consistent with a suppression of osteoblast differentiated phenotype.
Interestingly, we demonstrate that mannitol can completely mimic the effect of glucose
which leads us to hypothesize that the results we are seeing are due to an osmotic
response elicited by the osteoblast in response to increased extracellular osmolality. In
the process of proving our hypothesis, we have shown that PKC and AP-l are important
for the induction of c-jun and col I expression. In viva results support our in vitro findings
and solidified the hypothesis that osmotic stress is responsible for the decreased
mineralization.

Furthermore, we demonstrate that another leucine zipper transcription factor
family is involved in the osteoblast osmotic response, the CREB/ATF family. We
demonstrate that physiologically relevant hypertonicity increases CRE binding and
transactivation. ATF-2, a family member of the CRE/ATF family, is present in CRE

binding complex and its phosphorylation is noticeably increased by glucose and mannitol

106

treatment. p38 MAP kinase was also involved and ATF-2 phosphorylation and CRE
transactivation are p38 dependent based on experiments using SB203580 a p38 activity
inhibitor.

The combination of our results can be used to hypothesize a novel mechanism
employed by osteoblasts in response to physiologically relevant hypertonicity. We have
previously shown that hyperosmotic stress induces c-jun. The major regulators of c-jun
expression are ATF-2 and c-Jun (Van Dam et al., 1995). The promoter for c-jun contains
two AP-l sites; a proximal site (TGAGTCA) and a distal site called jun2TRE
(T'I‘ACCTCA) (Angel et al., 1988). In vitro binding studies reveal that, in contrast to the
consensus AP-l site which is preferentially targeted by dimers composed of Jun and Fos
families, the jun2TRE binds heterodimers composed of c-Jun and ATF (-like) proteins.
Our data indicate that osmotic stress, indeed, causes an increase in the expression of c-jun
and it is possible that the two AP-l sites on the c-jun promoter are involved through an
initial c-Jun: c-Jun complex binding on the AP-l consensus followed by c-Jun: ATF-2
heterodimer binding on the jun2TRE. The resulting increase in c-jun expression could
ultimately participate in blocking the function and maturation of osteoblasts by changing
their phenotype (increasing COL l and decreasing osteocalcin and perhaps modulating
expression of other important genes).

This work provides a novel mechanism in which osteoblast, the bone forming
cells, respond to increased extracellular osmolality by activating selected signaling
pathways and transcription factors. This leads to the modulation of osteoblast gene
expression, which can play a role in altering cell phenotype. This data can be used to

increase our understanding of not only diabetic bone disease but also the importance of

107

osteoblast cell shape in proliferation and differentiation. Understanding the cellular and
molecular mechanisms regulating osteoblast phenotype in response to diabetes might aid
in the development of drugs, directed at pathways affected by elevated extracellular
glucose levels, which can be used to increase bone formation in diabetics and perhaps in
the elderly. Such therapies can be used to reduce and possibly prevent the detrimental

effects of osteoporosis.

108

$85 a... a 388.... is: 2: «assessed .82: < .8 we

secede—E I? E

«ma—aooumc c282 /

 

109

..oooo...o.o...........o....o..ooo..oo...l« s
:3".....Iz..v...:.:::.:::::.VOR

....................2..,...‘......._..,....L.
co.

00......0......goooouooooo o...ooooo.o..

<
28.—egg
8.3.25 111882“;

 

 

 

«£505 an 3:5... 8.”

$2: “5ng

VII. Bibliography

Abdel-Hafiz HA, Heasley LE, Kyriakis JM, Avruch J, Kroll DJ, Johnson GL and
Hoeffler JP. (1992) Activating transcription factor-2 DNA-binding activity is stimulated
by phosphorylation catalyzed by p42 and p54 microtubule-associated protein kinases.

Mal Endacrinal 6(12): 2079-2089.

Angel P, Hattori K, Smeal T and Karin M. (1988) The jun proto-oncogene is
positively autoregulated by its product, Jun/AP-l. Cell 55(5): 875-885.

Angel P and Karin M. (1991) The role of Jun, Fos and the AP-l complex in cell-
proliferation and transformation. Biachim Biophys Acta 1072(2-3): 129-157

Asano T, Takata K, Katagiri H, Tsukuda K, Lin JL, Ishihara H, Inukai K, Hirano
H, Yazaki Y and Oka Y. (1992) Domains responsible for the differential targeting of
glucose transporter isoforms. J Biol Chem 267(27): 19636-19641.

Auwerx J, Dequeker J, Bouillon R, Geusens P and Nijs J. (1988) Mineral metabolism
and bone mass at peripheral and axial skeleton in diabetes mellitus. Diabetes 37(1): 8-12

Balint E, Szabo P, Marshall CF and Sprague SM. (2001) Glucose-induced inhibition
of in vitro bone mineralization. Bane 28(1): 21-28.

7. Bell GI. (1991) Lilly lecture 1990. Molecular defects in diabetes mellitus. Diabetes
40(4): 413-422

Benazzoug Y, Borchiellini C, Labat-Robert J, Robert L and Kern P. (1998) Effect of
high-glucose concentrations on the expression of collagens and fibronectin by fibroblasts
in culture. Exp Gerantal 33(5): 445-455

Berkowitz LA, Riabowol KT and Gilman M2. (1989) Multiple sequence elements of a
single functional class are required for cyclic AMP responsiveness of the mouse c-fos
promoter. Mal Cell Biol 9(10): 4272-4281.

Berl T, Siriwardana G, A0 L, Butterfield LM and Heasley LE. (1997) Multiple
mitogen-activated protein kinases are regulated by hyperosmolality in mouse IMCD
cells. Am J Physiol 272(3 Pt 2): F305-311.

Berney PW. (1952) Osteoporosis and diabetes mellitus. J Iowa Med Soc 42: 10-12
Bitoun M and Tappaz M. (2000) Gene expression of the transporters and biosynthetic

enzymes of the osmolytes in astrocyte primary cultures exposed to hyperosmotic
conditions. Glia 32(2): 165-176.

110

Bohren KM, Bullock B, Wermuth B and Gabbay KH. (1989) The aldo-keto reductase
superfarnily. cDNAs and deduced amino acid sequences of human aldehyde and aldose
reductases. J Biol Chem 264(16): 9547-9551.

Bouillon R. (1991) Diabetic bone disease [editorial]. Calcif Tissue Int 49(3): 155-160

Breen EC, Ignotz RA, McCabe L, Stein JL, Stein GS and Lian JB. (1994) TGF beta
alters growth and differentiation related gene expression in proliferating osteoblasts in

vitro, preventing development of the mature bone phenotype. J Cell Physiol 160(2): 323-
335

Brewster JL and Gustin MC. (1994) Positioning of cell growth and division after
osmotic stress requires a MAP kinase pathway. Yeast 10(4): 425-439.

Brownlee M, Vlassara H and Cerami A. (1984) Nonenzymatic glycosylation and the
pathogenesis of diabetic complications. Ann Intern Med 101(4): 527-537

Bucala R, Model P and Cerami A. (1984) Modification of DNA by reducing sugars: a
possible mechanism for nucleic acid aging and age-related dysfunction in gene
expression. Prac Natl Acad Sci U S A 81(1): 105-109

Cai Y, Zhang C, Nawa T, Aso T, Tanaka M, Oshiro S, Ichijo H and Kitajima S.
(2000) Homocysteine-responsive ATF-3 gene expression in human vascular endothelial
cells: activation of c-Jun NH(2)-terminal kinase and promoter response element. Blood
96(6): 2140-2148.

Candeliere GA, Prud’homme J and St-Arnaud R. (1991) Differential stimulation of
fos and jun family members by calcitriol in osteoblastic cells. Mol Endacrinal 5(12):
1780-1788

Ceolotto G, Gallo A, Miola M, Sartori M, Trevisan R, De] Prato S, Semplicini A and
Avogaro A. (1999) Protein kinase C activity is acutely regulated by plasma glucose
concentration in human monocytes in vivo. Diabetes 48(6): 1316-1322

Chen BP, Liang G, Whelan J and Hal T. (1994) ATF-3 and ATF-3 delta Zip.
Transcriptional repression versus activation by alternatively spliced isoforms. J Biol
Chem 269(22): 15819-15826.

Chen YR, Wang X, Templeton D, Davis RJ and Tan TH. (1996) The role of c-Jun N-
terminal kinase (JNK) in apoptosis induced by ultraviolet C and gamma radiation.
Duration of JNK activation may determine cell death and proliferation. J Biol Chem
271(50): 31929-31936

Chomczynski P and Sacchi N. (1987) Single-step method of RNA isolation by acid
guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162(1): 156-159

111

Chu HM, Tan Y, Kobierski LA, Balsam LB and Comb MJ. (1994) Activating
transcription factor-3 stimulates 3’,5’-cyclic adenosine monophosphate-dependent gene
expression. Mol Endocrinol 8(1): 59-68.

Clark G. (1981) Miscellaneous stains. In Clark G (ed) Stain Procedures, ed 4. Williams
and Wilkins, Baltimore, p187

Clohisy J C, Scott DK, Brakenhoff KD, Quinn CO and Partridge NC. (1992)
Parathyroid hormone induces c-fos and c-jun messenger RNA in rat osteoblastic cells.
Mol Endacrinal 6(11): 1834-1842

Craven PA, Davidson CM and DeRubertis FR. (1990) Increase in diacylglycerol mass
in isolated glomeruli by glucose from de novo synthesis of glycerolipids. Diabetes 39(6):
667-674

Denkert C, Warskulat U, Hensel F and Haussinger D. (1998) Osmolyte strategy in
human monocytes and macrophages: involvement of p38MAPK in hyperosmotic

induction of betaine and myoinositol transporters. Arch Biochem Biophys 354(1): 172-
180.

Du W, Thanos D and Maniatis T. (1993) Mechanisms of transcriptional synergism
between distinct virus-inducible enhancer elements. Cell 74(5): 887-898.

Dunstan CR, Boyce R, Boyce BF, Garrett IR, Izbicka E, Burgess WH and Mundy
GR. (1999) Systemic administration of acidic fibroblast growth factor (FGF-1) prevents
bone loss and increases new bone formation in ovariectomized rats. J Bone Miner Res
14(6): 953-959.

Duzgun SA, Rasque H, Kito H, Azuma N, Li W, Basson MD, Gahtan V, Dudrick SJ
and Sumpio BE. (2000) Mitogen-activated protein phosphorylation in endothelial cells
exposed to hyperosmolar conditions. J Cell Biochem 76(4): 567-571.

Emancipator K. (1999) Laboratory diagnosis and monitoring of diabetes mellitus. Am J
Clin Pathol 112(5): 665-674

Foulkes NS, Borrelli E and Sassone-Corsi P. (1991) CREM gene: use of alternative
DNA-binding domains generates multiple antagonists of CAMP-induced transcription.
Cell 64(4): 739-749.

Fuchs SY, Dolan L, Davis RJ and Ronai Z. (1996) Phosphorylation-dependent
targeting of c-Jun ubiquitination by Jun N-kinase. Oncogene 13(7): 1531-1535.

Fuchs SY and Ronai Z. (1999) Ubiquitination and degradation of ATF-2 are
dimerization dependent. Mal Cell Biol 19(5): 3289-3298.

112

 

Fujimoto R, Tanizawa T, Nishida S, Yamamoto N, Soshi S, Endo N and Takahashi
HE. (1999) Local effects of transforming growth factor-betal on rat calvaria: changes
depending on the dose and the injection site. J Bone Miner Metab 17(1): 11-17

Gabbay KH. (1973) The sorbitol pathway and the complications of diabetes. N Engl J
Med 288(16): 831-836

Gack S, Vallon R, Schaper J, Ruther U and Angel P. (1994) Phenotypic alterations in
fos-transgenic mice correlate with changes in Fos/Jun-dependent collagenase type I

expression. Regulation of mouse metalloproteinases by carcinogens, tumor promoters,
CAMP, and Fos oncoprotein. J Biol Chem 269(14): 10363-10369.

Gaire M, Chatton B and Kedinger C. (1990) Isolation and characterization of two
novel, closely related ATF cDNA clones from HeLa cells. Nucleic Acids Res 18(12):
3467-3473.

Galcheva-Gargova Z, Derijard B, Wu [H and Davis RJ. (1994) An osmosensing
signal transduction pathway in mammalian cells. Science 265(5173): 806-808.

Garcia-Perez A, Martin B, Murphy HR, Uchida S, Murer H, Cowley BD, J r.,
Handler JS and Burg MB. (1989) Molecular cloning of cDNA coding for kidney aldose
reductase. Regulation of specific mRNA accumulation by NaCl-mediated osmotic stress.
J Biol Chem 264(28): 16815-16821.

Gonzalez GA and Montminy MR. (1989) Cyclic AMP stimulates somatostatin gene
transcription by phosphorylation of CREB at serine 133. Cell 59(4): 675-680.

Gonzalez GA, Yamamoto K, Fischer WH, Karr D, Menzel P, Biggs W, 3rd, Vale
WW and Montminy MR. (1989) A cluster of phosphorylation sites on the cyclic AMP-
regulated nuclear factor CREB predicted by its sequence. Nature 337(6209): 749-752.

Grigoriadis AE, Schellander K, Wang ZQ and Wagner EF. (1993) Osteoblasts are
target cells for transformation in c-fos transgenic mice. J Cell Biol 122(3): 685-701.

Grigoriadis AE, Wang ZQ, Cecchini MG, Hofstetter W, Felix R, Fleisch HA and
Wagner EF. (1994) c-Fos: a key regulator of osteoclast-macrophage lineage
determination and bone remodeling. Science 266(5184): 443-448.

Grinstein S, Goetz-Smith JD, Stewart D, Beresford BJ and Mellors A. (1986) Protein
phosphorylation during activation of Na+/H+ exchange by phorbol esters and by osmotic
shrinking. Possible relation to cell pH and volume regulation. J Biol Chem 261(17):
8009-8016.

Gruda MC, Kovary K, Metz R and Bravo R. (1994) Regulation of Fra-l and Fra-2

phosphorylation differs during the cell cycle of fibroblasts and phosphorylation in vitro
by MAP kinase affects DNA binding activity. Oncogene 9(9): 2537-2547

113

Hadman M, Loo M and Bos TJ. (1993) In vivo viral and cellular Jun complexes exhibit
differential interaction with a number of in vitro generated ’AP-l- and CREB-like’ target
sequences. Oncogene 8(7): 1895-1903

Haffner SM and Bauer RL. (1993) The association of obesity and glucose and insulin
concentrations with bone density in premenopausal and postmenopausal women.
Metabolism 42(6): 735-738

Hai TW, Liu F, Coukos WJ and Green MR. (1989) Transcription factor ATF cDNA
clones: an extensive family of leucine zipper proteins able to selectively form DNA-
binding heterodimers. Genes Dev 3(12B): 2083-2090.

Han DC, Isono M, Hoffman BB and Ziyadeh FN. (1999) High glucose stimulates
proliferation and collagen type 1 synthesis in renal cortical fibroblasts: mediation by
autocrine activation of TGF-beta. J Am Soc Nephrol 10(9): 1891-1899.

Hess J, Porte D, Munz C and Angel P. (2001) AP-l and bea/Runt Physically Interact
and Regulate Parathyroid Hormone-dependent MMP13 Expression in Osteoblasts
through a New Osteoblast-specific Element 2/AP-1 Composite Element. J Biol Chem
276(23): 20029-20038.

Hoeffler JP, Meyer TE, Yun Y, Jameson JL and Habener JF. (1988) Cyclic AMP-
responsive DNA-binding protein: structure based on a cloned placental cDNA. Science
242(4884): 1430-1433.

Hoffert JD, Leitch V, Agre P and King LS. (2000) Hypertonic induction of aquaporin-
5 expression through an ERK-dependent pathway. J Biol Chem 275(12): 9070-9077

Hough S, Avioli LV, Bergfeld MA, Fallon MD, Slatopolsky E and Teitelbaum SL.
(1981) Correction of abnormal bone and mineral metabolism in chronic streptozotocin-
induced diabetes mellitus in the rat by insulin therapy. Endocrinology 108(6): 2228-2234

Hsu JC, Bravo R and Taub R. (1992) Interactions among LRF-l, JunB, c-Jun, and c-
Fos define a regulatory program in the G1 phase of liver regeneration. Mol Cell Biol
12(10): 4654-4665.

Hui SL, Epstein S and Johnston CC, Jr. (1985) A prospective study of bone mass in
patients with type I diabetes. J Clin Endacrinal Metab 60(1): 74-80

‘ Hunt JV, Smith CC and Wolff SP. (1990) Autoxidative glycosylation and possible
involvement of peroxides and free radicals in LDL modification by glucose. Diabetes
39(11): 1420-1424

Hussy N, Deleuze C, Desarmenien MG and Moos FC. (2000) Osmotic regulation of

neuronal activity: a new role for taurine and glial cells in a hypothalamic neuroendocrine
structure. Prag Neurabial 62(2): 113-134.

114

Igarashi M, Wakasaki H, Takahara N, Ishii H, Jiang ZY, Yamauchi T, Kuboki K,
Meier M, Rhodes CJ and King GL. (1999) Glucose or diabetes activates p38 mitogen-
activated protein kinase via different pathways. J Clin Invest 103(2): 185-195

Inaba M, Terada M, Koyama H, Yoshida O, Ishimura E, Kawagishi T, Okuno Y,
Nishizawa Y, Otani S and Morii H. (1995) Influence of high glucose on 1,25-
dihydroxyvitamin D3-induced effect on human osteoblast-like MG-63 cells. J Bone
Miner Res 10(7): 1050-1056

Ishida T, Haneda M, Maeda S, Koya D and Kikkawa R. (1999) Stretch-induced
overproduction of fibronectin in mesangial cells is mediated by the activation of mitogen-
activated protein kinase. Diabetes 48(3): 595-602

Itoh T, Yamauchi A, Miyai A, Yokoyama K, Kamada T, Ueda N and Fujiwara Y.
(1994) Mitogen-activated protein kinase and its activator are regulated by hypertonic
stress in Madin-Darby canine kidney cells. J Clin Invest 93(6): 2387-2392.

Jochum W, Passegue E and Wagner EF. (2001) AP—l in mouse development and
tumorigenesis. Oncogene 20(19): 2401-2412.

Katai H, Stephenson JD, Simkevich CP, Thompson JP and Raghow R. (1992) An
AP-l-like motif in the first intron of human Pro alpha 1(I) collagen gene is a critical
determinant of its transcriptional activity. Mol Cell Biochem 118(2): 119-129.

Katayama Y, Akatsu T, Yamamoto M, Kugai N and Nagata N. (1996) Role of
nonenzymatic glycosylation of type I collagen in diabetic osteopenia. J Bone Miner Res
11(7): 931-937

Kawasaki H, Schiltz L, Chin R, Itakura K, Taira K, Nakatani Y and Yokoyama
KK. (2000) ATF-2 has intrinsic histone acetyltransferase activity which is modulated by
phosphorylation. Nature 405(6783): 195-200.

Kemink SA, Hermus AR, Swinkels LM, Lutterman JA and Smals AG. (2000)
Osteopenia in insulin-dependent diabetes mellitus; prevalence and aspects of
pathophysiology. J Endacrinal Invest 23(5): 295-303

Kim SJ, Wagner S, Liu F, O’Reilly MA, Robbins PD and Green MR. (1992)
Retinoblastoma gene product activates expression of the human TGF-beta 2 gene through
transcription factor ATF-2. Nature 358(6384): 331-334.

Knepper MA, and Rector, F.C., Jr. (1996) in The Kidney (Brenner, B.M., ed) 5th Ed.
pp. 532-570 W.B. Saunders Co., Philadelphia

115

 

Koe RC, Clohisy J C, Tyson DR, Pulumati MR, Cook TF and Partridge NC. (1997)
Parathyroid hormone versus phorbol ester stimulation of activator protein-1 gene family
members in rat osteosarcoma cells. Calcif Tissue Int 61(1): 52-58

Kolch W. (2000) Meaningful relationships: the regulation of the Ras/Raf/MEK/ERK
pathway by protein interactions. Biochem J 351(Pt 2): 289-305

Kouzarides T and Ziff E. (1988) The role of the leucine zipper in the fos-jun
interaction. Nature 336(6200): 646-651

Krakauer J C, McKenna MJ, Buderer NF, Rao DS, Whitehouse FW and Parfitt
AM. (1995) Bone loss and bone turnover in diabetes. Diabetes 44(7): 775-782

Kreisberg JI, Radnik RA, Ayo SH, Garoni J and Saikumar P. (1994) High glucose
elevates c-fos and c-jun transcripts and proteins in mesangial cell cultures. Kidney Int
46(1): 105-112.

Kwon HM, Yamauchi A, Uchida S, Preston AS, Garcia-Perez A, Burg MB and
Handler JS. (1992) Cloning of the cDNa for a Na+lmyo-inositol cotransporter, a
hypertonicity stress protein. J Biol Chem 267(9): 6297-6301.

Larkins RG and Dunlop ME. (1992) The link between hyperglycaernia and diabetic
nephropathy [published erratum appears in Diabetologia 1992 Nov;35(11):1100].
Diabetologia 35(6): 499-504

Lee K, Deeds JD, Chiba S, Un-No M, Bond AT and Segre GV. (1994) Parathyroid
hormone induces sequential c-fos expression in bone cells in vivo: in situ localization of
its receptor and c-fos messenger ribonucleic acids. Endocrinology 134(1): 441-450

Lee KA, Hai TY, SivaRaman L, Thimmappaya B, Hurst HC, Jones NC and Green
MR. (1987) A cellular protein, activating transcription factor, activates transcription of
multiple ElA-inducible adenovirus early promoters. Prac Natl Acad Sci U S A 84(23):
8355-8359.

Levin ME, Boisseau VC and Avioli LV. (1976) Effects of diabetes mellitus on bone
mass in juvenile and adult-onset diabetes. N Engl J Med 294(5): 241-245

Lian JB, Stein GS, Bortell R, Owen TA (1991) Phenotype suppression: a postulated
molecular mechanism for mediating the relationship of proliferation and differentiation
by Fos/Jun interactions at AP-l sites in steroid responsive promoter elements of tissue-
specific genes. J Cell Biochem 45(1):9-14

Liu F, Thompson MA, Wagner S, Greenberg ME and Green MR. (1993) Activating

transcription factor-1 can mediate Ca(2+)- and CAMP-inducible transcriptional
activation. J Biol Chem 268(9): 6714-6720.

116

Locatto ME, Abranzon H, Caferra D, Fernandez MC, Alloatti R and Puche RC.
(1993) Growth and development of bone mass in untreated alloxan diabetic rats. Effects
of collagen glycosylation and parathyroid activity on bone turnover. Bone Miner 23(2):
129-144

Lytle C and Forbush B, 3rd. (1996) Regulatory phosphorylation of the secretory Na—K-
Cl cotransporter: modulation by cytoplasmic Cl. Am J Physiol 270(2 Pt 1): C437-448.

Machwate M, Jullienne A, Moukhtar M, Lomri A and Marie PJ. (1995) c-fos
protooncogene is involved in the mitogenic effect of transforming growth factor-beta in
osteoblastic cells. Mol Endacrinal 9(2): 187-198

Maekawa T, Sakura H, Kanei-Ishii C, Sudo T, Yoshimura T, Fujisawa J, Yoshida
M and Ishii S. (1989) Leucine zipper structure of the protein CRE-BPl binding to the
cyclic AMP response element in brain. Embo J 8(7): 2023-2028.

Manduteanu I, Voinea M, Serban G and Simioneecu M. (1999) High glucose induces
enhanced monocyte adhesion to valvular endothelial cells via a mechanism involving
ICAM-1, VCAM-l and CD18. Endathelium 6(4): 315-324

McCabe LR, Banerjee C, Kundu R, Harrison RJ, Dobner PR, Stein JL, Lian JB
and Stein GS. (1996) Developmental expression and activities of specific fos and jun
proteins are functionally related to osteoblast maturation: role of Fra-2 and Jun D during
differentiation. Endocrinology 137(10): 4398-4408

McCabe LR, Kockx M, Lian J, Stein J and Stein G. (1995) Selective expression of
fos- and jun-related genes during osteoblast proliferation and differentiation. Exp Cell
Res 218(1): 255-262

McCabe LR, Last TJ, Lynch M, Lian J, Stein J and Stein G. (1994) Expression of
cell growth and bone phenotypic genes during the cell cycle of normal diploid osteoblasts
and osteosarcoma cells. J Cell Biochem 56(2): 274-282.

McCarthy AD, Etcheverry SB, Bruzzone L and Cortizo AM. (1997) Effects of
advanced glycation end-products on the proliferation and differentiation of osteoblast-
like cells. Mol Cell Biochem 170(1-2): 43-51

McCauley LK, Koh-Paige AJ, Chen H, Chen C, Ontiveros C, Irwin R and McCabe
LR. (2001) Parathyroid hormone stimulates fra-2 expression in osteoblastic cells in vitro
and in vivo. Endocrinology 142(5): 1975-1981.

McCracken M, Lemons JE, Rahemtulla F, Prince CW and Feldman D. (2000) Bone

response to titanium alloy implants placed in diabetic rats. Int J Oral Maxillafac Implants
15(3): 345-354

117

Merriman HL, La Tour D, Linkhart TA, Mohan S, Baylink DJ and Strong DD.
(1990) Insulin-like growth factor-1 and insulin-like growth factor-H induce c-fos in
mouse osteoblastic cells. Calcif Tissue Int 46(4): 258-262

Meyer HE, Tverdal A and Falch JA. (1993) Risk factors for hip fracture in middle-
aged Norwegian women and men. Am J Epidemiol 137(11): 1203-1211

Morooka H, Bonventre JV, Pombo CM, Kyriakis JM and Force T. (1995) Ischemia
and reperfusion enhance ATF-2 and c-Jun binding to cAMP response elements and to an
AP-l binding site from the c-jun promoter. J Biol Chem 270(50): 30084-30092.

Morrisey K, Steadman R, Williams JD, Phillips A0. (1999) Renal proximal tubular
cell fibronectin accumulation in response to glucose is polyol pathway dependent.Kidney
Int 55(1): 160-7

Morrison L and Bogan I. (1927) Bone development in diabetic children: a roentgen
study. Am J Med Sci 174:313-319

Musti AM, Treier M and Bohmann D. (1997) Reduced ubiquitin-dependent
degradation of c-Jun after phosphorylation by MAP kinases. Science 275(5298): 400-402.

Nadkami V, Gabbay KH, Bohren KM and Sheikh-Hamad D. (1999) Osmotic
response element enhancer activity. Regulation through p38 kinase and mitogen-activated
extracellular signal-regulated kinase kinase. J Biol Chem 274(29): 20185-20190.

O’Donnell ME. (1993) Role of Na-K-Cl cotransport in vascular endothelial cell volume
regulation. Am J Physiol 264(5 Pt 1): C1316-1326.

O’Donnell ME, Martinez A and Sun D. (1995) Endothelial Na-K-Cl cotransport
regulation by tonicity and hormones: phosphorylation of cotransport protein. Am J
Physiol 269(6 Pt 1): C1513-1523.

Ohta S, Hiraki Y, Shigeno C, Suzuki F, Kasai R, Ikeda T, Kohno H, Lee K, Kikuchi
H, Konishi J and et al. (1992) Bone morphogenetic proteins (BMP-2 and BMP-3)
induce the late phase expression of the proto-oncogene c-fos in murine osteoblastic
MC3T3-E1 cells. FEBS Lett 314(3): 356-360

Olive M, Krylov D, Echlin DR, Gardner K, Taparowsky E and Vinson C. (1997) A
dominant negative to activation protein-1 (API) that abolishes DNA binding and inhibits
oncogenesis. J Biol Chem 272(30): 18586-18594.

Owen TA, Aronow M, Shalhoub V, Barone LM, Wilming L, Tassinari MS,
Kennedy MB, Pockwinse S, Lian JB and Stein GS. (1990) Progressive development of
the rat osteoblast phenotype in vitro: reciprocal relationships in expression of genes
associated with osteoblast proliferation and differentiation during formation of the bone
extracellular matrix. J Cell Physiol 143(3): 420-430

118

Park CW, Kim JH, Lee JW, Kim YS, Ahn HJ, Shin YS, Kim SY, Choi EJ, Chang
YS and Bang BK. (2000) High glucose-induced intercellular adhesion molecule-1
(ICAM-1) expression through an osmotic effect in rat mesangial cells is PKC-NF-kappa
B-dependent. Diabetologia 43(12): 1544-1553.

Parker JC. (1993) In defense of cell volume? Am J Physiol 265(5 Pt 1): C1191-1200

Pearman AT, Chou WY, Bergman KD, Pulumati MR and Partridge NC. (1996)
Parathyroid hormone induces c-fos promoter activity in osteoblastic cells through

phosphorylated CAMP response element (CRE)-binding protein binding to the major
CRE J Biol Chem 271(41): 25715-25721.

Quarles LD, Yohay DA, Lever LW, Caton R and Wenstrup RJ. (1992) Distinct
proliferative and differentiated stages of murine MC3T3-E1 cells in culture: an in vitro
model of osteoblast development. J Bone Miner Res 7(6): 683-692

Raingeaud J, Whitmarsh AJ, Barrett T, Derijard B and Davis RJ. (1996) MKK3-
and MKK6-regulated gene expression is mediated by the p38 mitogen-activated protein
kinase signal transduction pathway. Mol Cell Biol 16(3): 1247-1255.

Raskin P, Stevenson MR, Barilla DE and Pak CY. (1978) The hypercalciuria of
diabetes mellitus: its amelioration with insulin. Clin Endocrinol (Oxf) 9(4): 329-335

Rehfuss RP, Walton KM, Loriaux MM and Goodman RH. (1991) The CAMP-
regulated enhancer-binding protein ATF-1 activates transcription in response to CAMP-
dependent protein kinase A J Biol Chem 266(28): 18431-18434.

Reirnold AM, Grusby MJ, Kosaras B, Fries JW, Mori R, Maniwa S, Clauss IM,
Collins T, Sidman RL, Glimcher MJ and Glimcher LH. (1996) Chondrodysplasia and
neurological abnormalities in ATF-2-deficient mice. Nature 379(6562): 262-265.

Rizoli SB, Kapus A, Fan J, Li YH, Marshall JC and Rotstein OD. (1998)
Immunomodulatory effects of hypertonic resuscitation on the development of lung
inflammation following hemorrhagic shock. J Immunol 161(11): 6288-6296.

Rizoli SB, Rotstein OD and Kapus A. (1999) Cell volume-dependent regulation of L-
selectin shedding in neutrophils. A role for p38 mitogen-activated protein kinase. J Biol
Chem 274(31): 22072-22080.

Rizoli SB, Rotstein OD, Parodo J, Phillips MJ and Kapus A. (2000) Hypertonic
inhibition of exocytosis in neutrophils: central role for osmotic actin skeleton remodeling.
Am J Physiol Cell Physiol 279(3): C619-633.

Roger F, Martin PY, Rousselot M, Favre H and Feraille E. (1999) Cell shrinkage
triggers the activation of mitogen-activated protein kinases by hypertonicity in the rat

119

kidney medullary thick ascending limb of the Henle’s loop. Requirement of p38 kinase
for the regulatory volume increase response. J Biol Chem 274(48): 34103-34110

Rosette C and Karin M. (1996) Ultraviolet light and osmotic stress: activation of the
IN K cascade through multiple growth factor and cytokine receptors. Science 274(5290):
1 194-1 197.

Ruther U, Garber C, Komitowski D, Muller R and Wagner EF. (1987) Deregulated
c-fos expression interferes with normal bone development in transgenic mice. Nature
325(6103): 412-416.

Rydziel S, Durant D and Canalis E. (2000) Platelet-derived growth factor induces
collagenase 3 transcription in osteoblasts through the activator protein 1 complex. J Cell
Physiol 184(3): 326-333.

Ryseck RP and Bravo R. (1991) C-JUN, JUN B, and JUN D differ in their binding
affinities to AP-l and CRE consensus sequences: effect of FOS proteins. Oncogene 6(4):
533-542

Sabatakos G, Sims NA, Chen J, Aoki K, Kelz MB, Amling M, Bouali Y,
Mukhopadhyay K, Ford K, Nestler EJ and Baron R. (2000) Overexpression of
DeltaFosB transcription factor(s) increases bone formation and inhibits adipogenesis. Nat

Med 6(9): 985—990.

Sakamoto MK, Suzuki K, Takiya S, Yoshimura Y, Imai T, Matsumoto A and
Nakamura S. (1998) Developmental profiles of phosphorylated and unphosphorylated
CREBs in murine calvarial MC3T3-E1 cells. J Biochem (Tokyo) 123(3): 399-407

Sasaki T, Kaneko H, Ramamurthy NS and Golub LM. (1991) Tetracycline
administration restores osteoblast structure and function during experimental diabetes.
Anat Rec 231(1): 25-34

Schaffler A, Arndt H, Scholmerich J and Palitzsch KD. (2000) Amelioration of
hyperglycenric and hyperosmotic induced vascular dysfunction by in vivo inhibition of
protein kinase C and p38 MAP kinase pathway in the rat mesenteric microcirculation.
Eur J Clin Invest 30(7): 586-593

Schneider LE and Schedl HP. (1972) Diabetes and intestinal calcium absorption in the
rat. Am J Physiol 223(6): 1319-1323

Sheikh-Hamad D, Di Mari J, Suki WN, Safirstein R, Watts BA, 3rd and Rouse D.
(1998) p38 kinase activity is essential for osmotic induction of mRNAs for HSP70 and
transporter for organic solute betaine in Madin-Darby canine kidney cells. J Biol Chem
273(3): 1832-1837.

120

Sheng M, McFadden G and Greenberg ME. (1990) Membrane depolarization and
calcium induce C-fos transcription via phosphorylation of transcription factor CREB.
Neuron 4(4): 571-582.

Shimizu M, Nomura Y, Suzuki H, Ichikawa E, Takeuchi A, Suzuki M, Nakamura T,
Nakajima T and Oda K. (1998) Activation of the rat cyclin A promoter by ATF2 and
Jun family members and its suppression by ATF4. Exp Cell Res 239(1): 93-103.

Shires R, Avioli LV, Bergfeld MA, Fallon MD, Slatopolsky E and Teitelbaum SL.
(1980) Effects of semistarvation on skeletal homeostasis. Endocrinology 107(5): 1530-
1535

Smeal T, Binetruy B, Mercola D, Grover-Bardwick A, Heidecker G, Rapp UR and
Karin M. (1992) Oncoprotein-mediated signalling cascade stimulates c-Jun activity by
phosphorylation of serines 63 and 73. Mol Cell Biol 12(8): 3507-3513

Spencer EM, Khalil M and Tobiassen O. (1980) Experimental diabetes in the rat
causes an insulin-reversible decrease in renal 25-hydroxyvitamin D3-1 alpha-hydroxylase
activity. Endocrinology 107( 1): 300-305

Stricker E, and Verbalis JG. (1999) Fluid intake and homeostasis. In: Fundamental
Neuroscience, edited by Zigmond MJ, Bloom FE, Landis SC, Roberts JL, and Squire
LR.. San Diego: Academic. pp. 1111-1126

Sudo H, Kodama HA, Amagai Y, Yamamoto S and Kasai S. (1983) In vitro
differentiation and calcification in a new Clonal osteogenic cell line derived from
newborn mouse calvaria. J Cell Biol 96(1): 191-198.

Sugiura T, Yamauchi A, Kitamura H, Matusoka Y, Horio M, Imai E and Hori M.
(1998) Effects of hypertonic stress on transforming growth factor-beta activity in normal
rat kidney cells. Kidney Int 53(6): 1654-1660.

Sun AM, Saltzberg SN, Kikeri D and Hebert SC. (1990) Mechanisms of cell volume
regulation by the mouse medullary thick ascending limb of Henle. Kidney Int 38(6):
1019-1029.

Sunters A, McCluskey J and Grigoriadis AE. (1998) Control of cell cycle gene
expression in bone development and during c-Fos-induced osteosarcoma formation. Dev
Genet 22(4): 386-397

Takekawa M, Posas F and Saito H. (1997) A human homolog of the yeast Ssk2/Ssk22
MAP kinase kinase kinases, MTKl, mediates stress-induced activation of the p38 and
JNK pathways. Embo J 16(16): 4973-4982.

Takeshita N, Ishida H, Yamamoto T, Koh G, Kurose T, Tsuji K, Okamoto Y, Ikeda
H and Seino Y. (1993) Circulating levels and bone contents of bone gamma-

121

carboxyglutamic acid-containing protein in rat models of non-insulin-dependent diabetes
mellitus. Acta Endocrinol (Copenh) 128(1): 69-73

Taylor AM, Sharma AK, Avasthy N, Duguid IG, Blanchard DS, Thomas PK and
Dandona P. (1987) Inhibition of somatomedin-like activity by serum from
streptozotocin-diabetic rats: prevention by insulin treatment and correlation with skeletal
growth. Endocrinology 121(4): 1360-1365

Terada M, Inaba M, Yano Y, Hasuma T, Nishizawa Y, Morii H and Otani S. (1998)
Growth-inhibitory effect of a high glucose concentration on osteoblast-like cells. Bane
22(1): 17-23

Terada Y, Tomita K, Homma MK, Nonoguchi H, Yang T, Yamada T, Yuasa Y,
Krebs EG, Sasaki S and Marumo F. (1994) Sequential activation of Raf-1 kinase,
mitogen-activated protein (MAP) kinase kinase, MAP kinase, and S6 kinase by
hyperosmolality in renal cells. J Biol Chem 269(49): 31296-31301.

Thomas DM, Maher F, Rogers SD and Best JD. (1996 a) Expression and regulation by
insulin of GLUT 3 in UMR 106-01, a Clonal rat osteosarcoma cell line. Biochem Biophys
Res Commun 218(3): 789-793

Thomas DM, Rogers SD, Ng KW and Best JD. (1996 b) Dexamethasone modulates
insulin receptor expression and subcellular distribution of the glucose transporter GLUT
1 in UMR 106-01, a clonal osteogenic sarcoma cell line. J Mol Endacrinal 17(1): 7-17

Thomas MJ, Umayahara Y, Shu H, Centrella M, Rotwein P and McCarthy TL.
(1996) Identification of the CAMP response element that controls transcriptional
activation of the insulin-like growth factor-I gene by prostaglandin E2 in osteoblasts. J

Biol Chem 271(36): 21835-21841.

Tsai EY, Jain J, Pesavento PA, Rao A and Goldfeld AE. (1996) Tumor necrosis factor
alpha gene regulation in activated T cells involves ATF-2/Jun and NFATp. Mol Cell Biol
16(2): 459-467.

Uchida S, Kwon HM, Yamauchi A, Preston AS, Marumo F and Handler JS. (1992)
Molecular Cloning of the cDNA for an MDCK cell Na(+)- and Cl(-)—dependent taurine
transporter that is regulated by hypertonicity. Proc Natl Acad Sci U S A 89(17): 8230-
8234.

van Dam H, Duyndam M, Rottier R, Bosch A, de Vries-Smits L, Herrlich P,
Zantema A, Angel P and van der Eb AJ. (1993) Heterodimer formation of CJun and
ATF-2 is responsible for induction of C-jun by the 243 amino acid adenovirus ElA
protein. Embo J 12(2): 479-487.

122

van Dam H, Wilhelm D, Herr I, Steffen A, Herrlich P and Angel P. (1995) ATF-2 is
preferentially activated by stress-activated protein kinases to mediate C-jun induction in
response to genotoxic agents. Embo J 14(8): 1798-1811.

Varghese S, Rydziel S and Canalis E. (2000) Basic fibroblast growth factor stimulates
collagenase-3 promoter activity in osteoblasts through an activator protein-l-binding site.
Endocrinology 141(6): 2185-2191.

Verhaeghe J, Suiker AM, Visser WJ, Van Herck E, Van Bree R and Bouillon R.
(1992) The effects of systemic insulin, insulin-like growth factor-1 and growth hormone
on bone growth and turnover in spontaneously diabetic BB rats. J Endacrinal 134(3):
485-492

Verhaeghe J, van Herck E, Visser WJ, Suiker AM, Thomasset M, Einhom TA,
Faierman E and Bouillon R. (1990) Bone and mineral metabolism in BB rats with long-
term diabetes. Decreased bone turnover and osteoporosis. Diabetes 39(4): 477-482

Volonte D, Galbiati F, Pestell RG and Lisanti MP. (2001) Cellular stress induces the
tyrosine phosphorylation of caveolin-l (Tyr(l4)) via activation of p38 mitogen-activated
protein kinase and C-Src kinase. Evidence for caveolae, the actin cytoskeleton, and focal
adhesions as mechanical sensors of osmotic stress. J Biol Chem 276(11): 8094-8103.

Wang D, Christensen K, Chawla K, Xiao G, Krebsbach PH and Franceschi RT.
(1999) Isolation and Characterization of MC3T3-E1 preosteoblast subclones with distinct
in vitro and in vivo differentiation/mineralization potential. J Bone Miner Res 14(6): 893-

903.

Watts BA, 3rd, Di Mari JF, Davis RJ and Good DW. (1998) Hypertonicity activates
MAP kinases and inhibits HCO-3 absorption via distinct pathways in thick ascending
limb. Am J Physiol 275(4 Pt 2): F478-486.

Waud CE, Marks SC, Jr., Lew R and Baran DT. (1994) Bone mineral density in the
femur and lumbar vertebrae decreases after twelve weeks of diabetes in spontaneously
diabetic-prone BB/Worcester rats. Calcif Tissue Int 54(3): 237-240

Widmann C, Gibson S, Jarpe MB and Johnson GL. (1999) Mitogen-activated protein
kinase: conservation of a three-kinase module from yeast to human. Physiol Rev 79(1):
143-180

Wilhelm D, van Dam H, Herr I, Baumann B, Herrlich P and Angel P. (1995) Both
ATF-2 and C-Jun are phosphorylated by stress-activated protein kinases in response to
UV irradiation. Immunabiology 193(2-4): 143-148.

Winchester SK, Selvamurugan N, D’Alonzo RC and Partridge NC. (2000)
Developmental regulation of collagenase-3 mRNA in normal, differentiating osteoblasts

123

through the activator protein-l and the runt domain binding sites. J Biol Chem 275(30):
23310-23318.

Wolf BA, Williamson JR, Easom RA, Chang K, Sherman WR and Turk J. (1991)
Diacylglycerol accumulation and rnicrovascular abnormalities induced by elevated
glucose levels. J Clin Invest 87(1): 31-38

Wolff SP and Dean RT. (1987) Glucose autoxidation and protein modification. The
potential role of ’autoxidative glycosylation’ in diabetes. Biochem J 245(1): 243-250

Yamauchi A, Uchida S, Kwon HM, Preston AS, Robey RB, Garcia-Perez A, Burg
MB and Handler JS. (1992) Cloning of a Na(+)- and Cl(-)-dependent betaine
transporter that is regulated by hypertonicity. J Biol Chem 267(1): 649-652.

Yang D, Tournier C, Wysk M, Lu HT, Xu J, Davis RJ and Flavell RA. (1997)
Targeted disruption of the MKK4 gene causes embryonic death, inhibition of C-Jun NH2-
terminal kinase activation, and defects in AP-l transcriptional activity. Prac Natl Acad
Sci U S A 94(7): 3004-3009

Yoshida O, Inaba M, Terada M, Shioi A, Nishizawa Y, Otani S and Morii H. (1995)
Impaired response of human osteosarcoma (MG-63) cells to human parathyroid hormone

induced by sustained exposure to high glucose. Miner Electrolyte Metab 21(1-3): 201-
204

Zayzafoon M, Gao H, Burnett J and McCabe LR. (1998) Jun/Fos expression is
modulated by extracellular glucose in osteoblasts. Bane 23.°S206 23:S206

Zayzafoon M, Stell C, Irwin R and McCabe LR. (2000) Extracellular glucose
influences osteoblast differentiation and c-jun expression. J Cell Biochem 79(2): 301-310

Zhang Z and Cohen DM. (1996) NaCl but not urea activates p38 and jun kinase in
mIMCD3 murine inner medullary cells. Am J Physiol 271(6 Pt 2): F 1234-1238.

Zhu T and Lobie PE. (2000) Janus kinase 2-dependent activation of p38 mitogen-
activated protein kinase by growth hormone. Resultant transcriptional activation of ATF-
2 and CHOP, cytoskeletal reorganization and rnitogenesis. J Biol Chem 275(3): 2103-
2114

Ziyadeh FN, Fumo P, Rodenberger CH, Kuncio GS and Neilson EG. (1995) Role of
protein kinase C and cyclic AMP/protein kinase A in high glucose-stimulated
transcriptional activation of collagen alpha 1 (IV) in glomerular mesangial cells. J
Diabetes Complications 9(4): 255-261.

124