Mechanism of secretion and function of heat shock protein-90 (Hsp90) family genes

PDF

Download

Open document

Flip pages

Copy asset link

Request this asset

Transcript (if available)

Content
1

MECHANISM
OF
SECRETION
AND
FUNCTION
OF
HEAT

SHOCK
PROTEIN-‐90
(HSP90)
FAMILY
GENES

by

Priyamvada
Jayaprakash

A
Dissertation
Presented
to
the

FACULTY
OF
THE
USC
GRADUATE
SCHOOL

UNIVERISTY
OF
SOUTHERN
CALIFORNIA

In
Partial
Fulfillment
of
the

Requirements
for
the
Degree

DOCTOR
OF
PHILOSOPHY

(GENETIC,
MOLECULAR,
AND
CELLULAR
BIOLOGY)

August
2016

Copyright
2013

Priyamvada
Jayaprakash

2

TABLE
OF
CONTENTS

Acknowledgements
3

List
of
Figures
4

Chapter
1:
Introduction
6

The
Hsp90
protein
family
6

Hsp90α
and
Hsp90β
isoforms
perform
similar

as
well
as
distinct
functions

7

The
contrasting
knockout
phenotypes
of

Hsp90α
and
Hsp90β

8

Extracellular
roles
of
Hsp90α
and
Hsp90β
10

Secreted
Hsp90α
accelerates
wound
healing
11

Mechanism
of
action
of
secreted
Hsp90α
14

Secreted
Hsp90α
and
its
downstream
effectors
in
cancer
15

Mechanism
of
Hsp90α
secretion
by
cells
18

Mechanisms
of
exosome
biogenesis
19

Upstream
stimuli
driving
exosome
secretion
20

Proline-‐Rich
Akt
substrate
of
40
kDa
(PRAS40)
22

Chapter
2:
PRAS40
connects
microenvironmental
stress
signaling
to

exosome
secretion

36

Abstract
36

Introduction
36

Results
40

Discussion
54

References
57

Chapter
3:
Hsp90α
and
Hsp90β
Co-‐Operate
a
Stress-‐Response
Mechanism

to
Cope
With
Hypoxia
and
Nutrient
Paucity
during
Wound
Healing

65

Abstract
65

Introduction
65

Results
68

Discussion
79

References
81

Chapter
4:
Breast
Cancer
MDA-‐MB-‐231
Cells
Use
Secreted
Heat
Shock

Protein-‐90alpha
(Hsp90α)
to
Survive
a
Hostile
Hypoxic
Environment

87

Abstract
87

Introduction
87

Results
90

Discussion
103

References
105

Chapter
5:
Conclusions
113

Chapter
6:
Methods
120

3

Acknowledgements

My
PhD
would
not
have
been
possible
without
the
tremendous
support
of
many

people.
First,
I
am
thankful
to
my
mentor,
Dr.
Wei
Li
for
giving
me
the
opportunity
to

work
on
extremely
interesting
projects
and
for
helping
me
to
appreciate
and
enjoy

doing
science.
I
learnt
being
persistent
and
strong
in
the
face
of
failed
experiments

and
to
make
it
beyond
all
challenges.
I
am
grateful
to
Dr.
Mei
Chen
for
making
sure

we
had
the
resources
to
carry
out
experiments.
I
also
thank
my
past
and
present

colleagues,
Ayesha,
Patrick,
Divya,
Fred
and
Kate
for
their
scientific
suggestions
and

helpful
collaborations
as
well
as
the
Stallcup
lab
for
providing
me
with
valuable

reagents
and
scientific
help
every
time
I
needed
it.
I
am
indebted
to
my
past
and

current
committee
members,
Dr.
Michael
Stallcup,
Dr.
Louis
Dubeau
and
Dr.
Amy

Lee
for
always
being
ready
to
help
and
provide
valuable
advice.

My
mom,
dad,
sister
and
Ashwat
have
been
great
support
systems,
without
whom,
I

would
not
be
where
I
am
today.
I
am
also
extremely
thankful
to
my
friends
who

made
sure
of
providing
me
with
a
home
away
from
home
and
made
every
step

toward
my
PhD
easier
than
it
could
ever
be.

Finally,
none
of
this
would
have
been
possible
without
the
grace
of
God
and
I
thank

Him
for
giving
me
the
strength
and
the
drive
to
keep
going
and
achieve
what
I
set

out
to
do.

4

LIST
OF
FIGURES

Figure
1-‐1:
A
model
of
how
eHsp90
promotes
re-‐epithelialization
and

recruits
dermal
cells
into
the
wound
during
wound
healing

13

Figure
1-‐2:
Schematic
representation
of
extracellular
Hsp90α-‐mediated

signaling
pathway

15

Figure
1-‐3:
Structure
of
Hsp90
highlighting
domains
required
for
its

intracellular
versus
extracellular
functions

18

Figure
1-‐4:
Exosome
secretory
pathway

20

Figure
1-‐5:
Domain
structure
of
PRAS40
23

Figure
2-‐1:
Identification
of
differentially
activated
pathways
between
TGFα

and
EGF
stimulated
human
keratinocytes
(HKCs)

42

Figure
2-‐2:
PRAS40
knockdown
inhibits
Hsp90α
secretion
in
response
to

TGFα,
hypoxia
and
oxidative
stress
in
HKCs

45

Figure
2-‐3:
PRAS40
down
regulation
decreases
TGFα-‐driven
exosome

secretion
in
HKCs

47

Figure
2-‐4:
PRAS40
regulates
constitutive
and
hypoxia-‐driven
exosome

secretion
in
MDA-‐MB-‐231
breast
cancer
cells

48

Figure
2-‐5:
TGFα-‐stimulated
exosome
secretion
requires
T246

phosphorylation
of
PRAS40

50

Figure
2-‐6:
Constitutive
exosome
secretion
in
MDA-‐MB-‐231
cells
is

independent
of
T246
phosphorylation
of
PRAS40

51

Figure
2-‐7:
Hypoxia
increases
PRAS40
levels
in
MDA-‐MB-‐231
cells

52

Figure
2-‐8:
PRAS40
does
not
associate
with
exosomes

53

Figure
3-‐1:
Distinct
requirements
for
Hsp90α
and
Hsp90β
for
hypoxia-‐
triggered
cell
migration

70

Figure
3-‐2:
Secreted
Hsp90α,
not
Hsp90β,
mediates
hypoxia-‐triggered
HDF

migration
and
promotes
wound
healing

72

Figure
3-‐3:
Only
Hsp90β
stabilizes
the
LRP-‐1
receptor

75

Figure
3-‐4:
Exogenously
expressed
Hsp90β
rescues
endogenous
Hsp90β-‐ 78

5

down
regulated
HDF
motility

Figure
4-‐1:
Selection
of
MDA-‐MB-‐231
breast
cancer
cell
line
as
the
model
of

study

92

Figure
4-‐2:
Generation
of
Hsp90α
knockout
MDA-‐MB-‐231
cells

94

Figure
4-‐3:
CRISPR-‐cas9
knockout
of
Hsp90α
sensitizes
MDA-‐MB-‐231
cells
to

hypoxia-‐driven
killing

96

Figure
4-‐4:
Rescue
of
Hsp90α-‐knockout
cells
from
hypoxia-‐driven
killing
by

extracellular
Hsp90α,
but
not
Hsp90β,
protein
via
LRP-‐1
receptor
signalling

99

Figure
4-‐5:
Monoclonal
antibody,
1G6-‐D7,
binds
to
secreted
Hsp90α
and

neutralizes
its
function
in
vitro
and
in
vivo

101

Figure
4-‐6.
mAb
1G6-‐D7
neutralizes
secreted
Hsp90α
function
and
renders

MDA-‐MB-‐231
cells
susceptible
to
hypoxia-‐driven
cell
death

103

6

Chapter
1:
Introduction

Hsp90
was
initially
reported
as
an
intracellular
protein
whose
cellular
level

increases
in
response
to
heat
(1).
Since
then,
Hsp90
has
been
found
to
be
present
in

most
cells
and
has
been
characterized
as
an
intracellular
chaperone
protein
that

assists
the
conformational
activation
of
a
long
list
of
client
proteins
under
both

physiological
and
stress
conditions
(2,3).
Hsp90
proteins
constitute
2-‐3%
of
total

cellular
proteins
(19).
The
basic
structure
of
Hsp90
proteins
includes
an
N-‐terminal

ATP-‐binding
domain,
a
charged
linker
region,
a
client
binding
middle
domain
and
a

C-‐terminal
dimerization
domain.
Hsp90
proteins
form
multi-‐protein
complexes
with

proteins
called
“co-‐chaperones”.
The
chaperone-‐co-‐chaperone
complexes
co-‐
operate
with
the
ubiquitin-‐proteasome
system
and
target
misfolded
proteins
for

degradation,
thereby
enabling
protein
homeostasis
(70,71,72).
In
the
chaperone

cycle,
the
co-‐chaperones
Hsp70
and
Hsp40
form
an
“early
complex”
with
the

substrate
to
be
folded.
This
is
followed
by
formation
of
an
“intermediate
complex”

with
Hsp90
and
finally
a
“late
complex”
along
with
other
co-‐chaperones
like
p23
and

PPIase
and
ATP.
ATP
hydrolysis
results
in
the
release
of
the
properly
folded

substrate
(73).
Hsp90
has
around
200
client
proteins,
including
protein
kinases
and

nuclear
receptors
(74,
75).

The
Hsp90
protein
family

The
90
kDa
heat
shock
protein
family
comprises
the
cytosolic
isoforms,
Hsp90α
and

Hsp90β,
the
endoplasmic
reticulum
isoform,
GRP94
and
the
mitochondrial
isoform,

TRAP1. The
cytosolic
Hsp90α
and
Hsp90β
isoforms
are
86%
identical
at
the
amino

7

acid
level
and
are
believed
to
have
arisen
due
to
a
gene
duplication
event
(4).

Despite
the
high
degree
of
homology,
the
two
isoforms
have
certain
biochemical
as

well
as
functional
differences.

Structurally,
they
mainly
differ
in
the
C-‐terminal

dimerization
domain.
Hsp90β
dimers
are
relatively
unstable,
hence
mainly
exists
as

monomers.
Hsp90α,
on
the
other
hand,
readily
forms
homodimers.
Dimerization
is

reported
to
be
important
for
its
function
in
vivo
(5).
Apart
from
the
dimerization

potential,
the
isoforms
differ
in
other
amino
acid
stretches
too,
giving
rise
to

differences
in
client
protein
binding.
For
example,
Hsp90α
binds
more
strongly
to

ERK,
A-‐Raf
and
c-‐Src
(6),
while
Hsp90β
and
not
Hsp90α,
is
reported
to
be
the

chaperone
for
cIAP
(7).

Hsp90α
and
Hsp90β
isoforms
perform
similar
as
well
as
distinct
functions

Studies
in
yeast
reveal
that
an
increased
Hsp90β
to
Hsp90α
ratio
makes
the
yeast

more
susceptible
to
the
Hsp90
inhibitor,
radicicol
(8).
The
isoforms
have
been

reported
to
behave
both
in
synergy
and
independent
of
each
other
in
cancer
cells.

Viability
of
multiple
myeloma
cells
was
decreased
when
Hsp90α
and
Hsp90β
were

knocked
down
individually
by
siRNA
and
the
viability
was
further
decreased
by
the

double
knockdown
(9).
The
role
of
Hsp90β
in
anti-‐apoptosis
was
shown
in
newt

testis
where
presence
of
the
prolactin
receptor
inhibited
the
anti-‐apoptotic
signal
of

Hsp90β
(10).
CpG-‐B
ODN,
a
bacterial
component
known
to
trigger
macrophage
and

dendritic
cell
responses
specifically
increases
association
of
Bcl-‐2
with
Hsp90β
and

not
Hsp90α.
Macrophages
and
dendritic
cells
expressing
siRNA
against
Hsp90β

were
no
longer
responsive
to
CpG-‐B
ODN
mediated
anti-‐apoptosis
(11).
Such
studies

8

placed
Hsp90β
as
an
important
player
in
anti-‐apoptotic
responses.
Bouchier-‐Hayes

et
al.
provided
a
possible
mechanism
for
how
Hsp90α
or
Hsp90β
is
involved
in
anti-‐
apoptosis.
Using
RNAi
approach,
they
showed
that
Hsp90α
is
a
key
negative

regulator
of
heat-‐shock-‐induced
caspase-‐2
activation
(12).

The
contrasting
knockout
phenotypes
of
Hsp90α
and
Hsp90β

The
distinct
roles
of
Hsp90α
and
Hsp90β
were
further
highlighted
from
the

knockout
mouse
phenotypes.
Three
papers
on
Hsp90α
knockout
mice
and
one

paper
on
Hsp90β
knockout
mice
have
been
reported.
Hsp90β
null
mice
failed
to

form
a
placental
labyrinth
and
were
embryonic
lethal
(13).
However,
the
limitations

of
this
study
were
a)
the
authors
did
not
verify
that
Hsp90β
was
truly
nullified
in
the

knockout
mice
b)
did
not
check
if
Hsp90α
mRNA
and
protein
levels
were
unaffected.

Nonetheless,
this
finding
suggests
that
Hsp90β
is
essential
for
life
and
implies
one
of

the
following
two
possibilities:
(1)
the
role
of
Hsp90β
is
distinct
and
it
cannot
be

replaced
by
Hsp90α
or
(2)
Hsp90α
and
Hsp90β
together
make
up
a
threshold
of
the

activity
that
carries
out
the
same
functions.
Therefore,
reduction
in
either
Hsp90α

or
Hsp90β
level
would
show
defects.
However,
the
phenotype
of
Hsp90α
-‐knockout

mice
did
not
support
the
“threshold”
possibility
that
presence
of
both
Hsp90α
and

Hsp90β
are
necessary
to
avoid
lethality.

Three
articles
on
Hsp90α
-‐knockout
mice
have
recently
appeared.
Picard's
group

first
reported
generation
of
the
knockout
mice
carrying
a
gene
trap
insertion
in

intron
10
of
the
Hsp90α
gene.
This
insertion
could
potentially
produce
a
truncated

9

Hsp90α
protein
lacking
the
C-‐terminal
36
amino
acids,
but
for
unknown
reasons
it

did
not
occur
and
the
mice
had
an
Hsp90α
knockout-‐like
environment.
The
lack
of

expression
of
truncated
Hsp90α
appeared
not
due
to
a
compromised
stability
of
the

protein
(such
as
due
to
failed
dimerization),
because
Li's
group
detected
little

difference
in
expression
between
a
mutant
Hsp90α
with
deletion
of
the
36
amino

acids
and
the
wild-‐type
Hsp90α
in
human
keratinocytes
(Cheng
et
al.,
unpublished).

Surprisingly,
except
for
the
lack
of
sperms
in
the
male
mice
due
to
an
apparently

higher
rate
of
apoptosis
of
the
spermatocytes
in
the
testes,
these
Hsp90α
-‐knockout

mice
developed
normally
with
only
a
slightly
increased
level
of
Hsp90β
(14).
The

role
of
Hsp90α
in
spermatogenesis
even
in
adult
mice
was
confirmed
by
an

independent
study
(15).
This
finding
suggests
that,
in
contrast
to
Hsp90β,
Hsp90α
is

not
essential
for
life
and,
perhaps
more
interestingly,
has
a
tissue-‐specific
role
that

cannot
be
replaced
by
Hsp90β.

Concurrently,
Imai
et
al.
generated
conditional
Hsp90α-‐knockout
mice
by
floxing
the

exons
9
and
10
in
Hsp90α
gene.
Again,
the
mice
showed
a
normal
phenotype
(16).
In

addition,
there
were
several
additional
interesting
observations
from
this
study:
(1)

Hsp90β
makes
up
at
least
50%
of
the
total
Hsp90
in
the
cells
(assuming
that
the
pan

anti-‐Hsp90
antibody
used
recognized
Hsp90α
and
Hsp90β
with
similar
affinities),

(2)
Hsp90α
is
responsible
for
cytosolic
translocation
of
extracellular
antigen
across

the
endosomal
membrane
into
the
cytosol.

10

Extracellular
roles
of
Hsp90α
and
Hsp90β

Aside
from
the
well-‐documented
role
of
Hsp90α
and
Hsp90β
as
intracellular

chaperones,
recent
studies
imply
a
role
for
Hsp90α
in
the
extracellular
space
as
well

with
fewer
papers
suggesting
such
a
role
for
Hsp90β.
Hsp90α
is
secreted
by
normal

cells
under
stress
and
constitutively
by
cancer
cells.
The
primary
role
of
secreted

Hsp90α
is
to
promote
cell
motility
that
facilitates
both
wound
healing
and
tumor

metastasis.
Upstream
regulators
of
Hsp90α
secretion
identified
so
far
include

hypoxia-‐inducible
factor-‐1α
(HIF-‐1α)
in
human
keratinocytes,
dermal
fibroblasts

and
triple
negative
MDA-‐MB-‐231
breast
cancer
cells
(17,
18,
19);
cytokines
like

TGFα
in
human
keratinocytes;
PDGF-‐BB,
VEGF,
bFGF
and
SDF-‐1
in
endothelial
cells

(30);
p53
in
non
small
cell
lung
cancer
(20)
and
Hectd1
in
the
cranial
mesenchyme

(21).

On
the
other
hand,
Hsp90β
was
expressed
on
the
cell
surface
as
well
in
the

conditioned
medium
of
osteosarcoma
cells
(76).
Hsp90β
was
also
identified
on
the

cell
surface
of
oligodendrocyte
precursor
cells
and
anti-‐Hsp90β
antibodies
targeting

these
cells
were
hypothesized
to
prevent
remyelination
in
multiple
sclerosis

patients
(78).
However,
neither
of
these
studies
investigated
the
expression
of

Hsp90α
on
the
cell
surface
or
conditioned
media.
Using
wound
healing
as
a
model,

our
group
and
Luo’s
group
have
shown
that
only
secreted
Hsp90α,
but
not
Hsp90β,

drives
skin
cell
migration
in
vitro
and
enables
wound
healing
in
vivo
(19,
30).

11

Secreted
Hsp90α
accelerates
wound
healing

Wound
healing
is
a
complex
process,
involving
3
major
skin
cell
types—human

keratinocytes
(HKCs),
human
dermal
fibroblasts
(HDFs)
and
human
dermal

microvascular
endothelial
cells
(HDMECs).
The
microenvironment
of
wounded

tissues
is
hypoxic
and
lacks
continued
nutrient
supply
due
to
vascular
disruption

and
high
oxygen
consumption
by
cells
at
the
wound
edge
(22).
Acute
hypoxia
in

injured
tissues
is
also
a
critical
environmental
cue
that
triggers
initiation
of
the

wound
healing
processes
(23).
For
instance,
it
has
been
shown
that
hypoxia

promotes
migration
of
human
keratinocytes,
the
cell
responsible
for
wound
closure

via
re-‐epithelialization
(24),
and
migration
of
human
dermal
fibroblasts,
which

deposit
new
extracellular
matrices
to
the
wound
and
support
subsequent
wound

remodeling
(25).
On
the
other
hand,
impaired
responses
to
hypoxia
are
associated

with
impaired
wound
healing,
such
as
the
environment
in
chronic
diabetic
wounds

(26).
In
diabetic
foot
ulcers,
in
particular,
the
stability
of
the
hypoxia-‐inducible

factor-‐1alpha
(HIF-‐1α)
protein
is
compromised
due
to
the
hyperglycemic

environment,
albeit
the
mechanism
remains
unclear
(27,
28,
29).
These
in
vitro
and

in
vivo
studies
suggest
that
acute
hypoxia
is
a
natural
shock
signal
to
the
injured

tissues
and,
more
importantly,
a
call
for
an
immediate
jump-‐start
of
the
wound

healing
process.
We
previously
reported
that
hypoxia
drives
Hsp90α
secretion
in

HKCs
(17)
and
HDFs
(18).
This
was
dependent
on
the
stable
expression
of
HIF-‐1α

since
expression
of
a
constitutively
active
HIF-‐1α
mutant
under
normoxia
was
able

to
drive
Hsp90α
secretion.
Once
secreted
the
authors
demonstrated
that
it
binds
to

a
cell
surface
receptor
LRP-‐1.
The
authors
identify
the
hypoxia>HIF1-‐
α>
secretion

12

of
Hsp90α>LRP-‐1
binding>
cell
migration>
wound
healing
loop
in
HKCs.
Anti-‐LRP-‐1

antibodies,
LRP-‐1
down
regulation
using
siRNA
and
re-‐introduction
of
a
siRNA-‐
resistant
mLRP1
receptor
into
the
down
regulated
cells
all
inhibited
secreted

Hsp90α
driven
HKC
migration
(17).
In
addition
to
hypoxia,
tissue-‐injury
released

cytokines
such
as
TGFα
also
drive
Hsp90α
secretion
in
HKCs.
Secreted
Hsp90α
binds

to
LRP-‐1
receptor
on
HDFs
and
HDMECs
to
facilitate
the
wound
healing
process.

Endothelial
cells
activated
by
ECM
proteins
such
as
gelatin
and
fibronectin
and

growth
factors
such
as
bFGF,
VEGF,
PDGF-‐BB
and
chemokines
such
as
SDF-‐1
secrete

Hsp90α,
but
not
Hsp90β.
The
secreted
Hsp90α
promotes
endothelial
cell
migration

in
vitro
and
angiogenesis
in
mice.
In
vivo,
extracellular
Hsp90α
was
localized
on
the

new
vasculature
in
the
wounded
tissue
(30).

Using
deletion
mutagenesis,
Cheng
et
al.,
identified
F-‐5,
a
115
amino
acid
fragment

of
secreted
Hsp90α
that
recapitulated
the
pro-‐motility
activity
of
the
full-‐length

protein.
F-‐5
included
amino
acids
236-‐350
of
Hsp90α,
encompassing
the
charged

linker
region
and
middle
domains,
implying
the
N-‐terminal
ATPase
domain
as
being

dispensable
for
secreted
Hsp90α’s
pro-‐motility
function.
In
line
with
this
finding,

ATPase
mutants
of
Hsp90α
were
as
effective
as
the
wild
type
protein
in
promoting

cell
migration.
F-‐5
was
superior
to
FDA-‐approved
recombinant
PDGF-‐BB

(Becaplermin)
on
both
acute
and
diabetic
wounds
in
mice.
The
superior
effect
of
F-‐5

was
due
to
3
reasons:
1)
F-‐5
acted
on
all
3
cell
types
involved
in
wound
healing-‐

HKCs,
HDFs
and
HDMECs
due
to
the
expression
of
LRP-‐1
on
all
3
cell
types;
2)
F-‐5

overrode
the
inhibition
of
TGFβ3
in
the
wound
bed,
in
contrast
to
PDGF-‐BB
and
3)

13

F-‐5
promoted
cell
migration
even
under
hyperglycemia
(31).
We
also
proved
the

efficacy
of
F-‐5
in
accelerating
wound
healing
in
normal
and
diabetic
pigs
(32).
F-‐5

accelerated
wound
healing
via
re-‐epithelialization,
the
lateral
migration
of
HKCs

across
the
wound
bed.
The
function
of
secreted
Hsp90α
in
wound
healing
is

depicted
in
Figure
1.

Figure
1-‐1:
A
model
of
how
eHsp90
promotes
re-‐epithelialization
and
recruits
dermal
cells
into
the

wound
during
wound
healing.
(Step
1)
Uninjured
intact
skin
with
little
detectable
TGFβ,
cell

migration
or
stress;
(Step
2)
Injury
triggers
release
of
TGFβ
from
several
sources,
the
immotile
to

motile
transition
of
keratinocytes
and
release
of
conventional
growth
factors.
However,
the
growth

factors
will
not
be
able
to
recruit
the
dermal
cells
at
the
wound
edge
to
the
wound
bed
due
to
the

presence
of
TGFβ;
(Step
3)
While
keratinocytes
are
migrating,
they
secrete
Hsp90α.
Whence
the

secreted
Hsp90α
reaches
the
threshold
concentration
of
0.1
μM,
it
will
drive
inward
migration
of

14

HDFs
and
HDMECs;
(Step
4)
The
HKs
are
about
to
close
the
wound
and
the
moved-‐in
HDFs
will
start

to
remodel
the
wound
and
HDMECS
will
start
to
build
new
blood
vessels.
HK,
human
keratinocyte,

HDF,
human
dermal
fibroblast
and
HDMECs,
human
dermal
microvascular
endothelial
cells.
Adapted

from
Li
et
al.
2013.

Mechanism
of
action
of
secreted
Hsp90α

Secreted
Hsp90α
acts
as
a
ligand
for
cell
surface
LRP-‐1
(Low
Density
Lipoprotein

Receptor-‐Related
Protein-‐1)
receptor
and
mediates
transmembrane
signaling.
LRP-‐
1
has
a
515
kDa
extracellular
region
with
four
extracellular
domains
I-‐IV,
an
85
kDa

transmembrane
domain
and
a
100
amino
acid
long
cytoplasmic
tail.
Extracellular

Hsp90α
binds
to
subdomain
II
in
the
extracellular
domain
of
LRP-‐1
and
transmits

the
pro-‐motility
signal
to
Akt1
and
Akt2
to
drive
HDF
motility.
Topical
application
of

recombinant
Hsp90α
does
not
accelerate
wound
closure
in
Akt1
(-‐/-‐)
and
Akt2
(-‐/-‐)

mice.
Since
both
full-‐length
and
F-‐5
bind
LRP-‐1
equally
well,
this
mechanism

proposes
a
chaperone-‐independent
role
for
extracellular
Hsp90α
(33).
This

mechanism
is
shown
in
Figure
2.

15

Figure
1-‐2:
Schematic
representation
of
extracellular
Hsp90α-‐mediated
signaling
pathway.
The
red

arrows
depict
the
flow
of
extracellular
Hsp90α
signals
from
inside
to
outside
the
cell.
Akt
is
essential

for
extracellular
Hsp90α
signaling
to
promote
cell
migration
and
wound
healing.
Adapted
from
Tsen

et
al.,
2013.

Secreted
Hsp90α
and
its
downstream
effectors
in
cancer

Hsp90α
secretion
has
been
reported
in
a
number
of
cancers
including
breast,
colon,

bladder,
glioblastoma,
ovarian
etc.
The
primary
function
of
secreted
Hsp90α
is
to

promote
tumor
migration,
invasion
and
metastasis.
Different
groups
have
reported

different
mechanisms
of
action
of
secreted
Hsp90α
in
cancer.
In
prostate
cancer

cells,
secreted
Hsp90α
acted
in
an
autocrine
manner
through
the
LRP-‐1
receptor
to

16

activate
ERK.
The
secreted
Hsp90-‐LRP1-‐ERK
pathway
mediated
an
epithelial-‐to-‐
mesenchymal
transition
and
also
increased
the
transcription
of
MMPs
2,3
and
9
to

increase
cancer
invasion
(34).
Secreted
Hsp90α
drove
HCT-‐8
colon
cancer
cell

migration
through
LRP-‐1-‐mediated
integrin
αv
expression
in
an
NFκB-‐dependent

manner
(35).
In
glioblastoma,
LRP-‐1
acted
as
a
co-‐receptor
for
the
receptor
tyrosine

kinase
EphA2
and
mediated
secreted
Hsp90
driven
cell
motility,
which
was

independent
of
Hsp90’s
ATPase
activity
(36).

Many
groups
have
identified
different
extracellular
client
proteins
of
Hsp90.
Matrix

metalloproteinases,
MMPs
2
and
9,
are
the
best
characterized.
Jay’s
group
found
that

similar
to
intracellular
Hsp90α,
the
extracellular
pool
functioned
in
a
complex
with

co-‐chaperones
such
as
Hsp70,
Hsp40,
Hop
and
p23
to
increase
MMP2
activation
as

well
as
breast
cancer
cell
migration
and
invasion
(37).
In
conjunction
with
this,

inhibition
of
cell
surface
Hsp90
with
a
monoclonal
antibody,
mAb
4C5
inhibited
the

activation
of
MMP2
and
MMP9
and
decreased
breast
cancer
metastatic
deposits
in

the
lungs
of
SCID
mice
(38).
Stabilization
of
MMP2
was
mediated
by
the
interaction

between
the
middle
domain
of
Hsp90α
and
the
C-‐terminal
hemopexin
domain
of

MMP2.
Recombinant
Hsp90α
treatment
of
MCF7
cells
increased
MMP2
secretion.

Secreted
Hsp90α
from
endothelial
cells
promoted
endothelial
cell
proliferation
and

tube
formation
in
vitro
and
tumor
formation
and
angiogenesis
in
vivo
in
an
MMP2

dependent
manner.
The
effect
of
Hsp90α
was
isoform-‐specific
and
ATP
independent

(39).
In
addition
to
MMPs,
surface
Hsp90
interacted
with
the
extracellular
domain
of

HER2
and
disruption
of
this
interaction
by
mAb
4C5
inhibited
breast
cancer
motility

17

and
invasion
(40).
Surface
Hsp90
interacted
with
surface
Cdc37,
a
50
kDa

chaperone
on
MDA-‐MB-‐231
and
MDA-‐MB-‐453
breast
cancer
cells,
to
stabilize
HER2

and
EGFR
thereby
promoting
cell
migration
and
invasion
(41).
Other
extracellular

clients
include
tissue
plasminogen
activator
(tPA)
that
mediates
the
conversion
of

plasminogen
to
plasmin.
Inhibition
of
extracellular
Hsp90α
function
using
either

antibody
or
cell-‐impermeable
inhibitors
like
DMAG-‐N-‐oxide
decreases
plasminogen

activation
and
cancer
cell
migration
(42).
Using
mass
spectrometry
of
conditioned

media
from
MDA-‐MB-‐231
breast
cancer
cells,
Lysyl
Oxidase
2-‐like
Protein
(LOXL2)

was
identified
as
an
extracellular
client
protein
of
Hsp90α.
STA-‐12-‐7191,
a
cell-‐
impermeable
inhibitor
of
Hsp90,
blocked
breast
cancer
cell
migration,
which
can
be

rescued
by
exogenous
supplementation
with
recombinant
LOXL2
(43).
In
addition

to
its
effects
on
cancer
cells
themselves,
secreted
Hsp90α
also
activates
an

inflammatory
program
in
prostate
stromal
fibroblasts.
Recombinant
full
length
or
F-‐
5
Hsp90α
treatment
of
prostate
fibroblasts
increases
the
transcription
and
secretion

of
IL-‐6
and
IL-‐8
in
an
ERK1/2,
MMP2/9
and
NFκB-‐dependent
manner
(44).

Recently,
Isaacs’
group
discovered
an
epigenetic
role
for
secreted
Hsp90.
Secreted

Hsp90
mediates,
via
an
ERK1/2-‐dependent
pathway,
transcription
and
increased

expression
of
EZH2,
a
methyltransferase
of
the
polycomb
repressor
complex.
It
also

increased
EZH2
recruitment
to
E-‐cadherin
promoter,
repressing
it.
EZH2
was

required
for
secreted
Hsp90
driven
EMT
events
in
vitro
and
tumor
formation
in
vivo

(45).

Hsp90α
domains
required
for
its
intracellular
vs
extracellular
functions
are
in
Fig
3.

18

Figure
1-‐3:
Structure
of
Hsp90
highlighting
domains
required
for
its
intracellular
versus

extracellular
functions.
Adapted
from
Li
et
al.,
2013

Mechanism
of
Hsp90α
secretion
by
cells

There
are
two
types
of
protein
secretory
pathways—the
conventional
ER/Golgi

pathway
and
the
unconventional
exosome
trafficking
pathway.
Proteins
secreted
by

the
ER/Golgi
pathway
contain
a
short
amino
acid
sequence
of
5-‐30
amino
acids

called
“signal
peptide”
at
their
N-‐termini.
This
peptide
is
required
for
recognition
by

a
signal
recognition
particle
that
binds
to
the
SRP
receptor
on
the
ER.
This
facilitates

19

translocation
of
the
protein
through
the
ER
and
Golgi
followed
by
secretion
outside

the
cells.
In
contrast,
proteins
such
as
Hsp90α
lack
this
signal
peptide
and
are

therefore
secreted
via
exosomes.
Exosomes
are
30-‐150
nm
non-‐plasma
membrane

derived
vesicles
of
endosomal
origin.
They
are
formed
due
to
inward
budding
of
the

limiting
endosomal
membrane,
forming
intraluminal
vesicles
or
ILVs
within

multivesicular
bodies
or
MVBs.
MVBs
have
one
of
two
fates-‐
1)
they
fuse
with
the

lysosomes
and
target
their
contents
to
degradation;
2)
they
fuse
with
the
plasma

membrane
and
release
the
contents
into
the
extracellular
space
(46,
47).
The
main

known
function
of
exosomes
is
cell-‐cell
communication.
There
are
two
proposed

mechanisms
of
action-‐
one
in
which
exosomes
expose
proteins
on
the
surface
that

can
bind
to
cell-‐surface
receptors
on
target
cells
and
trigger
downstream
signaling.

The
second
mechanism
involves
fusion
of
the
exosomal
membrane
with
the
plasma

membrane
of
target
cells
and
delivery
of
the
exosomal
cargo
into
the
recipient
cell

cytosol.

Mechanisms
of
exosome
biogenesis

Exosome
biogenesis
mechanisms
can
be
broadly
divided
into:
ESCRT-‐dependent

and
ESCRT-‐independent
pathways.
ESCRT
(Endosomal
Sorting
Complex
Required

for
Transport)
is
a
multi-‐protein
complex
composed
of
at
least
30
proteins
divided

into
four
complexes-‐
ESCRT
0,
I,
II
and
III.
ESCRT
0
recognizes
ubiquitinated

transmembrane
proteins
and
sequesters
them
into
the
endosomal
membrane.

Complexes
I
and
II
mediate
vesicle
budding,
while
ESCRT
III
mediates
vesicle

scission.
Knockdown
of
ESCRT
components
such
as
Alix
and
TSG101
decreased

20

exosome
secretion
in
tumor
cells
and
Hela
cells
respectively
(48,
49).
However,

concomitant
inactivation
of
all
four
ESCRT
complexes
did
not
block
exosome

secretion
in
melanocytic
cells
where
sorting
of
the
pre-‐melanosomal
protein,
PMEL

occurred
independent
of
ESCRT
(50),
but
dependent
on
CD63
(51).
Two
lipid

metabolism
enzymes,
neutral
sphingomyelinase
(nSMase)
and
phospholipase
D2,

have
been
implicated
in
the
ESCRT-‐independent
mode
of
exosome
secretion.

nSMase
hydrolyzes
sphingomyelin
to
ceramide
(52),
while
phospholipase
D2

hydrolyzes
phosphatidylcholine
to
phosphatidic
acid
(53).
By
generating
lipids
on

the
surface
of
MVBs,
these
enzymes
facilitate
inward
budding
and
ILV
formation.

Figure
1-‐4:
Exosome
secretory
pathway.
Inward
budding
results
in
formation
of
intraluminal

vesicles
within
multivesicular
endosomes
(MVEs).
MVEs
either
fuse
with
lysosome
for
degradation

or
with
the
plasma
membrane
for
exocytosis
and
release
of
vesicles
as
exosomes.
Adapted
from

Raposo
&
Stoorvogel.,
2013

Upstream
stimuli
driving
exosome
secretion

Various
physiological
and
pathological
stimuli
drive
exosome
secretion,
in
order
to

21

facilitate
intercellular
communication.
Stimuli
modulate
both
the
number
and

contents
of
exosomes.
High
glucose
(30
mmol/L)
treatment
of
glomerular
epithelial

cells
in
the
kidney
increased
the
number
of
exosomes
secreted
and
also
increased

TGFβ1
expression
in
exosomes.
TGFβ1
was
taken
up
by
glomerular
mesangial
cells

that
increased
their
proliferation,
α-‐SMA
production
and
fibrosis.
This
study

highlighted
the
negative
impact
of
exosomes
on
diabetic
nephropathy
(54).

Recombinant
Wnt5A
induced
the
secretion
of
IL6,
VEGF
and
MMP2
via
exosomes
in

malignant
melanoma
cells
in
a
Cdc42
dependent
manner
and
the
secreted
exosomes

induced
endothelial
cell
branching
(55).
Genneback
et
al.,
reported
that
stimulation

of
cardiomyocytes
with
TGFβ2
or
PDGF-‐BB
resulted
in
differences
in
exosomal

cargo
content,
without
affecting
the
number
and
morphology
of
exosomes
secreted.

Although
217
transcripts
were
common
in
control
and
growth
factor
treated

cardiomyocytes
with
roles
in
energy
supply
of
the
cell,
growth
factor
treatment

resulted
in
cargo
content
with
roles
in
cell
proliferation
and
hypertrophy
(56).

Hypoxia
regulates
exosome
secretion
both
quantitatively
and
qualitatively.
Hypoxia

increased
the
number
of
exosomes
secreted
in
a
HIF1α-‐dependent
manner
in
three

different
breast
cancer
cell
lines,
MDA-‐MB-‐231,
MCF-‐7
and
Skbr3
(57).
Exosomes

isolated
from
hypoxic
glioblastoma
cell
lines
increased
endothelial
cell
proliferation

and
migration
and
enhanced
the
paracrine
activation
of
vascular
pericytes
by

endothelial
cells.
These
effects
led
to
accelerated
tumor
growth
of
GBM
xenografts

(58).
Hypoxia
activated
the
transcription
of
Rab22A,
a
protein
of
the
Rab
GTPase

family
to
increase
exosome
release
and
subsequent
tumor
invasion
and
metastasis

22

in
breast
cancer
(59).
In
hepatocellular
carcinoma,
hypoxia
induced
the
expression

of
Rab11-‐family
interacting
protein
4
(Rab11-‐FIP4)
in
a
HIF1α-‐dependent
manner

that
in
turn
phosphorylated
PRAS40
to
mediate
hypoxia-‐driven
invasion
and
tumor

metastasis
in
mice
(60).

Proline-‐Rich
Akt
substrate
of
40
kDa
(PRAS40)

The
proline-‐rich
Akt
substrate
of
40
kDa
(PRAS40)
was
initially
identified
as
one
of

the
direct
substrates
for
the
Akt
family
kinases
and
a
14-‐3-‐3-‐binding
protein
upon

phosphorylation
at
threonine-‐246
by
Akt
in
insulin-‐stimulated
rat
hepatoma
cell
line,

H4IIE
(61).

Deduced
amino
acid
sequences
from
isolated
cDNAs
of
both
rat
and

human
(AKT1S1)
PRAS40
revealed
a
proline-‐rich
molecule
without
any
major

homology
to
other
proteins
in
the
database
or
any
previously
reported
functional

motifs
from
signaling
molecules
(61,
62).
PRAS40
is
a
component
and
substrate
of

mTORC1
and
required
for
inhibiting
mTOR
activity
in
the
absence
of
stimuli.
PRAS40

plays
a
critical
role
in
linking
insulin
signaling
to
the
mTOR
(mammalian
target
of

rapamycin)
pathway,
leading
to
protein
synthesis
and
cell
growth.
In
growth-‐
arrested
cells,
such
as
serum
starvation
or
mitochondrial
metabolic
inhibition,

PRAS40
binds,
via
the
raptor
subunit,
to
mTORC1
and
inhibits
mTORC1
function.

Insulin
stimulation
activates
Akt
kinases
that
phosphorylate
the
mTORC1-‐associated

PRAS40
at
threonine-‐246
in
its
C-‐terminus
and
cause
dissociation
of
PRAS40
from

mTORC1
and
association
with
14-‐3-‐3.
The
activated
mTORC1
now
can

phosphorylate
and
activate
S6K1
and
4E-‐BP1
and
promote
protein
synthesis
and
cell

growth

(63,
64,
65,
66).
PRAS40
binding
to
raptor
also
appeared
to
require
the

23

phosphorylation
of
PRAS40
at
serine-‐183
by
mTORC1’s
kinase
activity
(64).
In

addition,
a
dozen
more
phosphorylation
sites
in
PRAS40,
mostly
by
mTORC1
kinase,

were
reported
and
their
functions
remain
unknown
(67,
68).
Deregulation
of
PRAS40

phosphorylation
has
been
reported
in
cancer
and
insulin-‐resistance
in
diabetes
(67,

69).

Figure
1-‐5:
Domain
structure
of
PRAS40

Hypothesis
and
goals

As
emphasized
in
the
previous
sections,
lots
of
questions
remain
with
respect
to

delineating
the
distinct
functions
of
Hsp90α
and
Hsp90β
as
well
as
identifying
the

relative
importance
of
extracellular
and
intracellular
Hsp90
pools
under
diverse

pathological
conditions.
In
addition,
how
Hsp90α
secretion
is
regulated
is
a
huge

area
of
investigation.
The
goal
of
my
thesis
was
to
address
these
questions.
The

findings
will
shed
light
on
the
probable
reasons
as
to
why
Hsp90
inhibitors

targeting
its
chaperone
activity
failed
in
clinical
trials,
facilitate
design
of
isoform-‐
specific
inhibitors
that
leave
the
chaperone
function
untouched
and
help
identify

regulators
of
Hsp90α
secretion
that
might
act
as
biomarkers
as
well
as
therapeutic

24

targets
for
the
treatment
of
cancers
that
rely
on
secreted
Hsp90α
function.

In
summary,
my
thesis
includes
three
main
projects,
beginning
with
identifying
the

mechanism
of
Hsp90α
secretion
in
response
to
diverse
stress
signals
(chapter
1),

investigating
the
communication
between
Hsp90α
and
Hsp90β
isoforms
in
wound

healing
(chapter
2)
and
uncovering
a
novel
role
of
secreted
Hsp90α
as
a
tumor

survival
factor
under
hypoxia
(chapter
3).

References:

1. Ritossa,
F.,
(1996).
Discovery
of
the
heat
shock
response.
Cell
Stress

Chaperones
1,
97–98.

2. Young,
J.C.,
Moarefi,
I.,
Hartl,
F.U.,
(2001).
Hsp90:
a
specialized
but
essential

protein-‐folding
tool.
J.
Cell.
Biol.
154,
267–273.

3. Whitesell,
L.,
Lindquist,
S.L.,
(2005).
HSP90
and
the
chaperoning
of
cancer.

Nat.
Rev.
Cancer
5,
761–772.

4. Gupta,
RS.,
(1995).
Phylogenetic
analysis
of
the
90
kD
heat
shock
family
of

protein
sequences
and
an
examination
of
the
relationship
among
animals,

plants,
and
fungi
species.
Mol
Biol
Evol.

12
(6),
1063-‐73.

5. Minami,
Y.,
Kimura,
Y.,
Kawasaki,
H.,
Suzuki,
K.,
&
Yahara,
I.
(1994).
The

carboxy-‐terminal
region
of
mammalian
HSP90
is
required
for
its

dimerization
and
function
in
vivo.
Mol.
Cell.
Biol.
14(2),
1459–1464.

25

6. Taherian,
A.,
Krone,
PH.,
&
Ovsenek,
N.
(2008).
A
comparison
of
Hsp90alpha

and
Hsp90beta
interactions
with
cochaperones
and
substrates.
Biochem
Cell

Biol.

86
(1),
37-‐45

7. Didelot,
C.,
Lanneau,
D.,
Brunet
M.,
Bouchot
A,
Cartier,
J.,

Jacquel,
A
.,
et
al.

(2008).
Interaction
of
heat-‐shock
protein
90β
isoform
(HSP90β)
with
cellular

inhibitor
of
apoptosis
1
(c-‐IAP1)
is
required
for
cell
differentiation.
Cell
Death

and
Differentiation
(2008)
15,
859–86

8. Millson,
S.
H.,
Truman,
A.
W.,
Rácz,
A.,
Hu,
B.,
Panaretou,
B.,
Nuttall,
J.,
et
al.

(2007),
Expressed
as
the
sole
Hsp90
of
yeast,
the
α
and
β
isoforms
of
human

Hsp90
differ
with
regard
to
their
capacities
for
activation
of
certain
client

proteins,
whereas
only
Hsp90β
generates
sensitivity
to
the
Hsp90
inhibitor

radicicol.
FEBS
Journal,
274,
4453–4463.

9. Chatterjee,
M.,
Jain,
S.,
Stühmer,
T.,
Andrulis,
M.,
Ungethüm,
U.,
Kuban,
RJ.,
et

al.,
(2007).
STAT3
and
MAPK
signaling
maintain
overexpression
of
heat

shock
proteins
90alpha
and
beta
in
multiple
myeloma
cells,
which
critically

contribute
to
tumor-‐cell
survival.
Blood.

109
(2),
720-‐8

10. Saribek,
B.,
Jin,
Y.,
Saigo,
M.,
Eto,
K.,
Abe,
S.,
(2006).
HSP90beta
is
involved
in

signaling
prolactin
induced
apoptosis
in
newt
testis.
Biochem.
Biophys.
Res.

Commun.
349,
1190–1197.

11. Kuo,
C.C.,
Liang,
C.M.,
Lai,
C.Y.,
Liang,
S.M.,
(2007).
Involvement
of
heat
shock

protein
Hsp90
beta
but
not
Hsp90
alpha
in
antiapoptotic
effect
of
CpG-‐B

oligodeoxynucleotide.
J.
Immunol.
178,
6100–6108.

26

12. Bouchier-‐Hayes,
L.,
Oberst,
A.,
McStay,
G.P.,
Connell,
S.,
Tait,
S.W.,
Dillon,
C.P.,

et
al.,
(2009).
Characterization
of
cytoplasmic
caspase-‐2
activation
by

induced
proximity.
Mol.
Cell.
35,
830–840.

13. Voss,
A.K.,
Thomas,
T.,
Gruss,
P.,
(2000).
Mice
lacking
HSP90beta
fail
to

develop
a
placental
labyrinth.
Development
127,
1–11.

14. Grad,
I.,
Cederroth,
C.R.,
Walicki,
J.,
Grey,
C.,
Barluenga,
S.,
Winssinger,
N.,
et

al.,
(2010).
The
molecular
chaperone
Hsp90α
is
required
meiotic
progression

of
spermatocytes
beyond
pachytene
in
the
mouse.
PLoS
One
5,
e15770.

15. Kajiwara,
C.,
Kondo,
S.,
Uda,
S.,
Dai,
L.,
Ichiyanagi,
T.,
Chiba,
T.,
et
al.,
(2012).

Spermatogenesis
arrest
caused
by
conditional
deletion
of
Hsp90α
in
adult

mice.
Biol.
Open
1,
977–982.

16. Imai,
T.,
Kato,
Y.,
Kajiwara,
C.,
Mizukami,
S.,
Ishige,
I.,
Ichiyanagi,
T.,
et
al.,

(2011).
Heat
shock
protein
90
HSP90
contributes
to
cytosolic
translocation

of
extracellular
antigen
for
cross-‐presentation
by
dendritic
cells.
Proc.
Natl.

Acad.
Sci.
USA
108,
16363–16368.

17. Woodley,
DT.,
Fan,
J.,
Cheng,
CF.,
Li,
Y.,
Chen,
M.,
Bu,
G.
and
Li
W.
(2009)

Participation
of
the
lipoprotein
receptor
LRP1
in
hypoxia-‐HSP90alpha

autocrine
signaling
to
promote
keratinocyte
migration.
J
Cell
Sci
122,
1495-‐
1498.

18. Li,
W.,
Li,
Y.,
Guan,
S.,
Fan,
J.,
Cheng,
C.,
Bright,
AM.,
Chinn,
C.,
Chen,
M.
and

Woodley
DT.
(2007)
Extracellular
heat
shock
protein-‐90alpha:
linking

hypoxia
to
skin
cell
motility
and
wound
healing.
EMBO
J
26,
1221-‐1233.

27

19. Sahu,
D.,
Zhao,
Z.,
Tsen,
F.,
Cheng,
CF.,
Park,
R.,
Situ,
AJ.,
Dai,
J.,
Eginli,

A.,
Shams,
S.,
Chen,
M.
et
al.
(2012)
A
potentially
common
peptide
target
in

secreted
heat
shock
protein-‐90α
for
hypoxia-‐inducible
factor-‐1α-‐positive

tumors.
Mol
Biol
Cell
23,
602-‐613

20. Yu,
X.,
Harris,
S.L.,
Levine,
A.J.,
(2006).
The
regulation
of
exosome
secretion:
a

novel
function
of
the
p.53
protein.
Cancer
Res.
66,
4795–4801.

21. Sarkar,
A.A.,
Zohn,
I.E.,
(2012).
Hectd1
regulates
intracellular
localization
and

secretion
of
Hsp90
to
control
cellular
behavior
of
the
cranial
mesenchyme.
J.

Cell.
Biol.
196,
789–800.

22. Pai,
M.
P.
and
Hunt,
T.
K.
(1972).
Effect
of
varying
oxygen
tensions
on
healing

of
open
wounds.
Surg.
Gynecol.
Obstet.
135,
756-‐758.

23. Tandara,
AA.
and
Mustoe,
TA.
(2004)
Oxygen
in
wound
healing-‐-‐more
than
a

nutrient.
World
J
Surg
28,
294-‐300.

24. O’Tool,
EA.,
Marinkovich,
MP.,
Peavey,
CL.,
Amieva,
MR.,
Furthmayr,
H.,

Mustoe,
TA.
and
Woodley
DT.
(1997)
Hypoxia
increases
human
keratinocyte

motility
on
connective
tissue
.
J.
Clin.
Invest.
100,
2881-‐2891.

25. Mogford,
JE.,
Tawil,
N.,
Chen,
A.,
Gies,
D.,
Xia,
Y.
and
Mustoe
TA.
(2002)
Effect

of
age
and
hypoxia
on
TGFbeta1
receptor
expression
and
signal

transduction
in
human
dermal
fibroblasts:
impact
on
cell
migration.
J
Cell

Physiol
190,
259-‐265.

26. Botusan,
IR.,
Sunkari,
VG.,
Savu,
O.,
Catrina,
AI.,
Grünler,
J.,
Lindberg,
S.,

Pereira,
T.,

Yla-‐Herttuala,
S.,
Poellinger,
L.,
Brismar,
K.
et
al.
(2008)

28

Stabilization
of
HIF-‐1alpha
is
critical
to
improve
wound
healing
in
diabetic

mice.
Proc
Natl
Acad
Sci.
USA.
105,19426-‐19431.

27. Catrina,
SB.,
Okamoto,
K.,
Pereira,
T.,
Brismar,
K.
and
Poellinger,
L.
(2004)

Hyperglycemia
regulates
hypoxia-‐inducible
factor-‐1alpha
protein
stability

and
function.
Diabetes
53,
3226–3232.

28. Fadini,
GP.,
Sartore,
S.,
Schiavon,
M.,
Albiero,
M.,
Baesso,
I.,
Cabrelle,

A.,
Agostini,
C.,
&
Avogaro,
A.
(2006)
Diabetes
impairs
progenitor
cell

mobilisation
after
hindlimb
ischaemia-‐reperfusion
injury
in
rats.

Diabetologia
49,
3075-‐3084.

29. Gao,
Z.,
Sasaoka,
T.,
Fujimori,
T.,
Oya,
T.,
Ishii,
Y.,
Sabit,
H.,
Kawaguchi,

M.,
Kurotaki,
Y.,
Naito,
M.,
Wada
T.
et
al.
(2005)
Deletion
of
the
PDGFR-‐beta

gene
affects
key
fibroblast
functions
important
for
wound
healing.
J
Biol

Chem
280,
9375-‐9389.

30. Song,
X.,
&
Luo,
Y.
(2010).
The
regulatory
mechanism
of
Hsp90alpha

secretion
from
endothelial
cells
and
its
role
in
angiogenesis
during
wound

healing.
Biochem
Biophys
Res
Commun.

398
(1),
111-‐7

31. Cheng,
CF.,
Fan,
J.,
Fedesco,
M.,
Guan,
S.,
Li,
Y.,
Bandyopadhyay,
B.,
et
al.,

(2008).
Transforming
Growth
Factor
α
(TGFα)-‐Stimulated
Secretion
of

HSP90α:
Using
the
Receptor
LRP-‐1/CD91
To
Promote
Human
Skin
Cell

Migration
against
a
TGFβ-‐Rich
Environment
during
Wound
Healing.
Mol
Cell

Biol.
28(10),
3344–3358.

32. O’Brien,
K.,
Bhatia,
A.,
Tsen,
F.,
Chen,
M.,
Wong,
AK.,
Woodley,
DT.,
et
al

(2014).
Identification
of
the
Critical
Therapeutic
Entity
in
Secreted
Hsp90α

29

That
Promotes
Wound
Healing
in
Newly
Re-‐Standardized
Healthy
and

Diabetic
Pig
Models.
PLoS
One
9(12):
e113956.

33. Tsen,
F.,
Bhatia,
A.,
O'Brien,
K.,
Cheng
CF,
Chen
M,
Hay
N.,
et
al.,
(2013).

Extracellular
heat
shock
protein
90
signals
through
subdomain
II
and
the

NPVY
motif
of
LRP-‐1
receptor
to
Akt1
and
Akt2:
a
circuit
essential
for

promoting
skin
cell
migration
in
vitro
and
wound
healing
in
vivo.
Mol
Cell

Biol.
(24),
4947-‐59

34. Hance,
M.
W.,
Dole,
K.,
Gopal,
U.,
Bohonowych,
J.
E.,
Jezierska-‐Drutel,
A.,

Neumann,
C.
A.,
et
al.,
(2012).
Secreted
Hsp90
Is
a
Novel
Regulator
of
the

Epithelial
to
Mesenchymal
Transition
(EMT)
in
Prostate
Cancer.
J.

Biol.Chem.
287(45),
37732–37744.

35. Chen,
J.-‐S.,
Hsu,
Y.-‐M.,
Chen,
C.-‐C.,
Chen,
L.-‐L.,
Lee,
C.-‐C.,
&
Huang,
T.-‐S.
(2010).

Secreted
Heat
Shock
Protein
90α
Induces
Colorectal
Cancer
Cell
Invasion

through
CD91/LRP-‐1
and
NF-‐κB-‐mediated
Integrin
αV
Expression.
J.

Biol.Chem,
285(33),
25458–25466.

36. Gopal,
U.,
Bohonowych,
J.
E.,
Lema-‐Tome,
C.,
Liu,
A.,
Garrett-‐Mayer,
E.,
Wang,

B.,
et
al.,
(2011).
A
Novel
Extracellular
Hsp90
Mediated
Co-‐Receptor
Function

for
LRP1
Regulates
EphA2
Dependent
Glioblastoma
Cell
Invasion.
PLoS

ONE,
6(3),
e17649.

37. Sims,
J.
D.,
McCready,
J.,
&
Jay,
D.
G.
(2011).
Extracellular
Heat
Shock
Protein

(Hsp)70
and
Hsp90α
Assist
in
Matrix
Metalloproteinase-‐2
Activation
and

Breast
Cancer
Cell
Migration
and
Invasion.
PLoS
ONE,
6(4),
e18848.

30

38. Stellas,
D.,
El
Hamidieh,
A.,
&
Patsavoudi,
E.
(2010).
Monoclonal
antibody
4C5

prevents
activation
of
MMP2
and
MMP9
by
disrupting
their
interaction
with

extracellular
HSP90
and
inhibits
formation
of
metastatic
breast
cancer
cell

deposits.
BMC
Cell
Biology,
11,
51.

39. Song,
X.,
Wang,
X.,
Zhuo,
W.,
Shi,
H.,
Feng,
D.,
Sun,
Y.,
et
al.
(2010).
The

Regulatory
Mechanism
of
Extracellular
Hsp90α
on
Matrix
Metalloproteinase-‐
2
Processing
and
Tumor
Angiogenesis.
J.
Biol.Chem,
285
(51),
40039–40049.

40. Sidera,
K.,
Gaitanou,
M.,
Stellas,
D.,
Matsas,
R.,
&
Patsavoudi,
E.
(2008).
A

Critical
Role
for
HSP90
in
Cancer
Cell
Invasion
Involves
Interaction
with
the

Extracellular
Domain
of
HER-‐2.
J.
Biol.Chem,
283,2031-‐2041.

41. El
Hamidieh,
A.,
Grammatikakis,
N.,
&
Patsavoudi,
E.
(2012).
Cell
Surface

Cdc37
Participates
in
Extracellular
HSP90
Mediated
Cancer
Cell
Invasion.

PLoS
ONE,
7(8),
e42722.

42. McCready,
J.,
Sims,
J.
D.,
Chan,
D.,
&
Jay,
D.
G.
(2010).
Secretion
of
extracellular

hsp90α
via
exosomes
increases
cancer
cell
motility:
a
role
for
plasminogen

activation.
BMC
Cancer,
10,
294.

43. McCready,
J.,
Wong,
D.
S.,
Burlison,
J.
A.,
Ying,
W.,
&
Jay,
D.
G.
(2014).
An

Impermeant
Ganetespib
Analog
Inhibits
Extracellular
Hsp90-‐Mediated

Cancer
Cell
Migration
that
Involves
Lysyl
Oxidase
2-‐like

Protein.
Cancers,
6(2),
1031–1046.

44. Bohonowych,
J.,
Hance,
M.,
Nolan,
K.,
Defee,
M.,
Parsons,
C.,
&
Isaacs,
J.
(2014).

Extracellular
Hsp90
mediates
an
NF-‐κB
dependent
inflammatory
stromal

31

program:
Implications
for
the
prostate
tumor
microenvironment.
The

Prostate,
74(4),
395–407.

45. Nolan,
K.
D.,
Franco,
O.
E.,
Hance,
M.
W.,
Hayward,
S.
W.,
&
Isaacs,
J.
S.
(2015).

Tumor-‐secreted
Hsp90
Subverts
Polycomb
Function
to
Drive
Prostate
Tumor

Growth
and
Invasion.
J.
Biol.Chem.,
290(13),
8271–8282.

46. Jaiswal,
JK.,
Andrews,
NW.
&
Simon
SM.
(2002).
Membrane
proximal

lysosomes
are
the
major
vesicles
responsible
for
calcium-‐dependent

exocytosis
in
nonsecretory
cells.
J.
Cell
Biol.
159,
625–35

47. Raposo,
G.,
Nijman,
HW.,
Stoorvogel,
W.,
Liejendekker,
R.,
Harding,
CV.,
Melief,

CJ.,
et
al.
1996.
B
lymphocytes
secrete
antigen-‐presenting
vesicles.
J.
Exp.
Med.

183,
1161–72

48. Baietti,
MF.,
Zhang,
Z.,
Mortier,
E.,
Melchior,
A.,
Degeest,
G.,
Geeraerts
A.,
et
al.

(2012).
Syndecan-‐syntenin-‐ALIX
regulates
the
biogenesis
of
exosomes.
Nat.

Cell
Biol.
14,
677–85

49. Colombo,
M.,
Moita,
C.,
van
Niel,
G.,
Kowal
J,
Vigneron
J,

Benaroch
P
et
al.

(2013).
Analysis
of
ESCRT
functions
in
exosome
biogenesis,
composition
and

secretion
highlights
the
heterogeneity
of
extracellular
vesicles.
J.
Cell
Sci.
126,

5553–65

50. Theos,
A.
C.,
Truschel,
S.
T.,
Tenza,
D.,
Hurbain,
I.,
Harper,
D.
C.,
Berson,
J.
F.,
et

al.,
(2006).
A
novel
pathway
for
sorting
to
intralumenal
vesicles
of

multivesicular
endosomes
involved
in
organelle
morphogenesis.
Dev.

Cell,
10(3),
343–354.

32

51. Van
Niel,
G.,
Charrin,
S.,
Simoes,
S.,
Romao,
M.,
Rochin,
L.,
Saftig,
P.,
et
al.

(2011).
The
tetraspanin
CD63
regulates
ESCRT-‐independent
and
dependent

endosomal
sorting
during
melanogenesis.
Dev.
Cell,
21(4),
708–721.

52. Trajkovic,
K.,
Hsu,
C.,
Chiantia,
S.,
Rajendran,
L.,
Wenzel,
D.,
Wieland,
F.,
et
al.

(2008).
Ceramide
triggers
budding
of
exosome
vesicles
into
multivesicular

endosomes.
Science
319,
1244–47

53. Ghossoub,
R.,
Lembo,
F.,
Rubio,
A.,
Gaillard,
CB.,
Bouchet,
J.,
Vitale,
N.,
et
al.

(2014).
Syntenin-‐ALIX
exosome
biogenesis
and
budding
into
multivesicular

bodies
are
controlled
by
ARF6
and
PLD2.
Nat.
Commun.
5,
3477

54. Wu,
X.,
Gao,
Y.,
Cui,
F.,
&
Zhang,
N.
(2016).
Exosomes
from
high
glucose-‐
treated
glomerular
endothelial
cells
activate
mesangial
cells
to
promote
renal

fibrosis.
Biol.
Open.
0,
1-‐8.

55. Ekström,
E.
J.,
Bergenfelz,
C.,
von
Bülow,
V.,
Serifler,
F.,
Carlemalm,
E.,
Jönsson,

G.,
et
al.,
(2014).
WNT5A
induces
release
of
exosomes
containing
pro-‐
angiogenic
and
immunosuppressive
factors
from
malignant
melanoma

cells.
Molecular
Cancer,
13,
88.

56. Gennebäck,
N.,
Hellman,
U.,
Malm,
L.,
Larsson,
G.,
Ronquist,
G.,
Waldenström,

A.,
et
al.,
(2013).
Growth
factor
stimulation
of
cardiomyocytes
induces

changes
in
the
transcriptional
contents
of
secreted
exosomes.
J.
Extracell.

Vesicles,
2,
10.3402/jev.v2i0.20167.

57. King,
H.
W.,
Michael,
M.
Z.,
&
Gleadle,
J.
M.
(2012).
Hypoxic
enhancement
of

exosome
release
by
breast
cancer
cells.
BMC
Cancer,
12,
421.

33

58. Kucharzewska,
P.,
Christianson,
H.
C.,
Welch,
J.
E.,
Svensson,
K.
J.,
Fredlund,
E.,

Ringnér,
M.,
et
al.,
(2013).
Exosomes
reflect
the
hypoxic
status
of
glioma
cells

and
mediate
hypoxia-‐dependent
activation
of
vascular
cells
during
tumor

development.
Proc
Natl
Acad
Sci.
USA.
110(18),
7312–7317.

59. Wang,
T.,
Gilkes,
D.
M.,
Takano,
N.,
Xiang,
L.,
Luo,
W.,
Bishop,
C.
J.,
et
al.,

(2014).
Hypoxia-‐inducible
factors
and
RAB22A
mediate
formation
of

microvesicles
that
stimulate
breast
cancer
invasion
and
metastasis.
Proc
Natl

Acad
Sci.
USA,
111(31),
E3234–E3242.

60. Hu,
F.,
Deng,
X.,
Yang,
X,
Jin,
H.,
Gu,
D.,
Lv,
X.,

et
al.,
(2015).

Hypoxia
upregulates
Rab11-‐family
interacting
protein
4
through
HIF-‐1α
to

promote
the
metastasis
of
hepatocellular
carcinoma.
Oncogene.

34(49),

6007-‐17

61. Kovacina,
KS.,
Park,
GY.,
Bae,
SS.,
Guzzetta,
AW.,
Schaefer,
E.,
Birnbaum
MJ,
et

al.
(2003).
Identification
of
a
proline-‐rich
Akt
substrate
as
a
14-‐3-‐3
binding

partner.
J
Biol
Chem
278,
10189–10194.

62. Beausoleil,
SA.,
Jedrychowski,
M.,
Schwartz,
D.,
Elias,
JE.,
Villen,
J.,
Li,
J.,
et
al.,

(2004).
Large-‐scale
characterization
of
HeLa
cell
nuclear
phosphoproteins.

Proc
Natl
Acad
Sci
USA
101,
12130–12135.

63. Sancak,
Y.,
Thoreen,
CC.,
Peterson,
TR.,
Lindquist,
RA.,
Kang,
SA.,
Spooner,
E.,

et
al.,
(2007).
PRAS40
is
an
insulin-‐regulated
inhibitor
of
the
mTORC1

protein
kinase.
Mol
Cell
25:
903–915

64. Oshiro,
N.,
Takahashi,
R.,
Yoshino,
K.,
Tanimura,
K.,
Nakashima,
A.,
Eguchi,
S.,

et
al.,
(2007).
The
proline-‐rich
Akt
substrate
of
40
kDa
(PRAS40)
is
a

34

physiological
substrate
of
mammalian
target
of
rapamycin
complex
1.
J
Biol

Chem
282,
20329–20339

65. Vander
Haar,
E.,
Lee,
SI.,
Bandhakavi,
S.,
Griffin,
TJ.
&
Kim,
DH.
(2007).
Insulin

signalling
to
mTOR
mediated
by
the
Akt/PKB
substrate
PRAS40.
Nat
Cell
Biol

9,
316–323.

66. Thedieck,
K.,
Polak,
P.,
Kim,
ML.,
Molle,
KD.,
Cohen,
A.,
Jeno,
P.,
et
al.,
(2007).

PRAS40
and
PRR5-‐like
protein
are
new
mTOR
interactors
that
regulate

apoptosis.
PLoS
One
2:
e1217.

67. Wiza,
C.,
Nascimento,
EBM.
&
Ouwens,
DM.
(2012).
Role
of
PRAS40
in
Akt
and

mTOR
signaling
in
health
and
disease.
Am
J
Physiol
Endocrinol
Metab
302:

E1453–E1460.

68. Wang,
H.,
Zhang,
Q.,
Wen,
Q.,
Zheng,
Y.,
Philip,
L.,
Jiang,
H.,
et
al.,
(2012).

Proline-‐rich
Akt
substrate
of
40kDa
(PRAS40):
a
novel
downstream
target
of

PI3k/Akt
signaling
pathway.
Cell
Signal
24,
17–24.

69. Malla,
R.,
Wang,
Y.,
Chan,
WK.,
Tiwari,
AK.,
&
Faridi,
JS.
(2015).
Genetic

ablation
of
PRAS40
improves
glucose
homeostasis
via
linking
the
AKT
and

mTOR
pathways.
Biochem
Pharmacol.
96
(1),
65-‐75

70. Pratt,
WB.,
Morishima,
Y.,
Murphy,
M.
&
Harrell,
M.
(2006).
Chaperoning
of

glucocorticoid
receptors.
Handb.
Exp.
Pharmacol.
172,
111–138.

71. Pratt,
WB.,
Morishima,
Y.,
Peng,
HM.,
&
Osawa,
Y.
(2010).
Proposal
for
a
role

of
the
Hsp90/Hsp70-‐based
chaperone
machinery
in
making
triage
decisions

when
proteins
undergo
oxidative
and
toxic
damage.
Exp.
Biol.
Med.
235,
278–
289.

35

72. Gamerdinger,
M.,
Hajieva,
P.,
Kaya,
AM.,
Wolfrum,
U.,
Hartl,
FU.,
&
Behl,
C.

(2009).
Protein
quality
control
during
aging
involves
recruitment
of
the

macroautophagy
pathway
by
BAG3.
EMBO
J.
28,
889–901.

73. Li,
J.,
Richter,
K.,
&
Buchner,
J.
(2011).
Mixed
Hsp90-‐cochaperone
complexes

are
important
for
the
progression
of
the
reaction
cycle,
Nat.
Struct.
Mol.
Biol.

18,
61–66

74. Smith,
D.F.,
Stensgard,
BA.,
Welch,
WJ.,

&
Toft,
D.O.
(1992).
Assembly
of

progesterone
receptor
with
heat
shock
proteins
and
receptor
activation
are

ATP
mediated
events,
J.
Biol.
Chem.
267,
1350–1356.

75. Ziemiecki,
A.,
Catelli,
MG.,
Joab,
I.,
&
Moncharmont,
B.
(1986).
Association
of

the
heat
shock
protein
hsp90
with
steroid
hormone
receptors
and
tyrosine

kinase
oncogene
products,
Biochem.
Biophys.
Res.
Commun.
138,
1298–1307

76. Suzuki,
S.,
&
Kulkarni,
AB.
(2010).
Extracellular
Heat
Shock
Protein
HSP90β

Secreted
by
MG63
Osteosarcoma
Cells
Inhibits
Activation
of
Latent
TGF-‐β1.

Biochem.
Biophys.
Res.
Commun.
398(3),
525–531.

77. Sidera,
K.,
Samiotaki,
M.,
Yfanti,
E.,
Panayotou,
G.,
&
Patsavoudi,
E.
(2004).

Involvement
of
cell
surface
HSP90
in
cell
migration
reveals
a
novel
role
in
the

developing
nervous
system.
J
Biol
Chem.

279
(44),
45379-‐88

78. Cid,
C.,
Alvarez-‐Cermeño,
JC.,
Camafeita,
E.,
Salinas,
M.,
&
Alcázar,
A.
(2004).

Antibodes
reactive
to
heat
shock
protein
90

induce
oligodendrocyte
precursor
cell
death
in
culture.
Implications
for

demyelination
in
multiple
sclerosis.
FASEB
J.

18
(2),
409-‐11

36

Chapter
2:
PRAS40
connects
microenvironmental
stress
signaling
to

exosome
secretion

Abstract

Exosomes
are
30-‐150
nm
sized
membrane
vesicles
of
endosomal
origin
that
are

formed
within
multivesicular
bodies
inside
the
cell
and
are
released
into
the

extracellular
space
following
fusion
with
the
plasma
membrane.
The
main
function

of
exosomes
is
to
facilitate
intercellular
communication.
Various
environmental

stimuli
such
as
growth
factors,
hypoxia,
radiation,
oxidative
stress
etc.
increase

exosome
secretion.
However,
how
these
diverse
signals
converge
on
driving

exosome
secretion
is
not
known.
Here,
we
report
identification
of
proline-‐rich
Akt

substrate
of
40kDa
(PRAS40)
as
a
central
regulator
of
exosome
secretion
in

response
to
growth
factor,
hypoxia
and
H2O2
stimulation
in
both
normal
and
cancer

cells.
Down
regulation
of
PRAS40
blocked
exosome
secretion
in
response
to
diverse

stresses.
PRAS40
does
not
physically
associate
with
exosomes
and
mediates

exosome
secretion
by
both
Threonine
246
(T246)
phosphorylation
dependent
and

PRAS40
total
protein
level
dependent
pathways.
Thus,
PRAS40
controls
both
stress-‐
induced
and
constitutive
secretion
of
exosomes
in
normal
and
tumor
cells

respectively.

Introduction

Based
on
differences
in
the
content,
size
and
membrane
composition,
cell-‐secreted

extracellular
vesicles
(EVs)
are
classified
into
three
subgroups,
namely,
apoptotic

bodies,
microvesicles/ectosomes
and
exosomes
(1-‐4).
Exosomes
or
intraluminal

37

vesicles
(ILVs)
are
membrane
vesicles
of
endosomal
origin,
30-‐150
nm
in
diameter

and
formed
within
multivesicular
bodies
(MVB).
MVB-‐derived
exosomes
can
fuse

with
the
plasma
membrane
to
release
their
cargo
molecules
into
the
extracellular

space.
This
release
process
was
reported
to
include
1)
sorting
into
smaller
vesicles;

2)
fusing
with
the
cell
membrane;
and
3)
releasing
the
vesicles
to
the
extracellular

space.
Exosomes
can
be
enriched
from
cell
culture
supernatants
of
virtually
any
cell

type
through
a
series
of
steps
of
high-‐speed
and
ultra-‐speed
centrifugations
and

express
the
tetraspanin
protein
family
including
CD9,
CD63
and
CD81,
as
well
as

carry
cargo
proteins
including
heat
shock
protein
90
and
70,
DNA,
mRNA
and

miRNA
(5,6).

The
main
known
function
of
exosomes
is
cell-‐cell
communication
(7,
8).

There
are

two
proposed
mechanisms
of
action-‐
one
in
which
exosomes
expose
proteins
on
the

surface
that
can
bind
to
cell-‐surface
receptors
on
target
cells
and
trigger

downstream
signaling.
The
second
mechanism
involves
fusion
of
the
exosomal

membrane
with
the
plasma
membrane
of
target
cells
and
delivery
of
the
exosomal

cargo
into
the
recipient
cell
cytosol.
Exosomes
play
an
important
role
in
intercellular

communication
during
host
immune
responses,
tissue
repair
and
cancer

progression.
During
innate
immune
response,
for
instance,
infected
macrophages

secrete
exosomes
containing
pro-‐inflammatory
cytokines
and
chemokines
that
in

turn
recruit
other
immune
cells
to
the
site
of
infection
(9).
Natural
killer
(NK)
cell-‐
derived
exosomes
contain
cytotoxic
proteins
such
as
Fas
ligand
and
perforin
that

facilitate
tumor
cell
lysis.
Activation
of
T
cells
induces
exosome
release
that
triggers

38

proliferation
of
resting
T
cells
and
alters
their
cytokine
profile
to
a
more
cytotoxic

CD8+
phenotype

(10).

Mesenchymal
stem
cell
(MSC)-‐secreted
exosomes
play
an
important
role
in
wound

healing
by
enhancing
migration
and
proliferation
of
normal
and
diabetic
dermal

fibroblasts
and
augmenting
angiogenesis
by
endothelial
cells,
thereby
facilitating
the

wound
healing
process
(11).
MSC
exosomes
have
also
been
reported
to
alleviate

acute
lung
injury
caused
by
E.coli
LPS.
Co-‐injection
of
exosomes
along
with
LPS

decreased
pulmonary
edema,
influx
of
neutrophils
and
levels
of
pro-‐inflammatory

cytokines
such
as
TNFα
(12).
Exosomes
released
by
human
umbilical
cord
stem
cells

decrease
kidney
injury
mediated
by
cisplatin
treatment
in
rats.
Intrarenal
delivery

of
exosomes
following
cisplatin
treatment
decreased
renal
cell
apoptosis,
increased

their
proliferation
and
decreased
oxidative
stress
as
measured
by
increased
GSH

and
decreased
8-‐OHdG
levels
(13).
Other
cell
types
in
the
wound
bed
such
as

keratinocytes
(14)
and
circulating
monocytes
home
to
the
wound
bed

and
secrete

exosomes
to
activate
fibroblasts
(22).
Keratinocyte
exosomes
express
stratifin,
a

protein
of
the
14-‐3-‐3
family.
These
exosomes
are
taken
up
by
fibroblasts
and
induce

MMP-‐1
up
regulation
through
the
p38-‐MAPK
pathway
(14,
15).
Bone-‐marrow

derived
progenitor
cells
secrete
exosomes
that
promoted
proliferation
of

keratinocytes
and
fibroblasts,
migration
of
keratinocytes
and
tube
formation
by

endothelial
cells
in
vitro.
In
addition,
exosome
treatment
enabled
wound
healing
in

diabetic
mice
by
triggering
angiogenesis,
keratinocyte
proliferation
and
ECM

deposition
(16).

39

The
pathological
role
of
exosomes
in
tumor
progression
and
metastasis
has
been

reported
in
a
number
of
cancers.
Due
to
their
small
size
and
surrounding
membrane

coat,
exosomes
are
stable
in
the
extracellular
environment
and
can
migrate
to

distant
sites
to
set
up
metastatic
niches.
In
pancreatic
ductal
adenocarcinoma,

“educating”
mice
with
cancer
exosomes
prior
to
tumor
cell
injection
increased
liver

metastatic
burden.
The
authors
report
that
Kupffer
cells
in
the
liver
take
up
these

exosomes
and
activate
hepatic
stellate
cells
to
secrete
TGFβ.
This
increases

fibronectin
production
and
recruitment
of
BM-‐derived
macrophages
to
liver,

creating
a
pro-‐inflammatory
&
pre-‐metastatic
niche
(22).
Hypoxia
is
a
characteristic

feature
of
solid
tumors
with
50%
of
solid
tumors
reportedly
being
hypoxic.
Hypoxia

has
been
shown
to
increase
the
secretion
as
well
as
alter
the
proteome
of
exosomes

in
a
variety
of
studies
(17,
18,
19).
High
exosome
levels
in
blood
plasma
are

correlated
with
decreased
patient
survival

(20).
Exosomes
secreted
in
the
tumor

microenvironment
facilitate
tumor
migration
and
invasion
in
a
variety
of
cancers

such
as
breast
(17,
19,
21),
pancreas
(22)
and
glioma
(18,
23).

Proline-‐rich
Akt
substrate
of
40
kDa
(PRAS40)
is
an
inhibitor
and
substrate
of
the

mTORC1
complex.
In
unstimulated
cells,
PRAS40
binds
to
the
raptor
protein
of
the

mTORC1
complex
and
inhibits
mTOR
kinase
activity
and
downstream
signaling.

When
cells
are
subjected
to
growth
factors
such
as
insulin,
Akt
is
phosphorylated
on

Ser473
by
mTORC2
and
on
T308
by
PDK1,
which
in
turn
phosphorylates
PRAS40
on

Threonine-‐246
(T246).
mTOR
itself
phosphorylates
PRAS40
on
Serine-‐183
(S183).

These
phosphorylations
sequester
PRAS40
away
from
mTORC1,
thereby
relieving

40

the
inhibitory
constraint
on
mTORC1.
This
activates
the
kinase
activity
of
mTOR
to

phosphorylate
its
downstream
substrates
such
as
4E-‐BP1
and
S6K1
that
regulate

processes
such
as
cell
growth
and
protein
synthesis.
The
sequestered
PRAS40
binds

to
14-‐3-‐3
sigma
protein.
Thus,
the
known
functions
of
PRAS40
so
far
is
its
regulation

of
mTORC1
complex,
influencing
insulin
signaling,
cell
cycle
and
pathological

conditions
such
as
cancer
and
insulin
resistance.

Although
there
has
been
considerable
advance
in
understanding
the
growing

functions
of
exosomes
in
intercellular
communication,
mediating
physiological
and

pathological
responses,
few
studies
have
focused
on
identifying
upstream
regulators

of
exosome
secretion.
In
this
study,
we
identify
proline-‐rich
Akt
substrate
of
40
kDa

(PRAS40)
as
a
central
regulator
of
exosome
secretion
in
response
to
diverse
signals

such
as
cytokine,
hypoxia
and
oxidative
stress.
PRAS40
down
regulation
blocked

exosome
secretion
in
different
cell
types.
PRAS40-‐mediated
exosome
secretion
was

both
through
its
T246
phosphorylation
dependent
and
total
protein
level
dependent

pathways,
depending
on
the
stress
stimuli
involved.
Considering
the
importance
of

exosomes
in
tumor
progression,
PRAS40
might
act
as
a
potential
biomarker
to

identify
the
aggressiveness
of
tumors
and
act
as
an
indication
of
the
effectiveness
of

treatment.

Results

Identification
of
PRAS40
as
a
regulator
of
TGFα-‐stimulated
Hsp90α
secretion

To
identify
regulator(s)
of
exosome
secretion,
we
initially
used
secretion
of
Hsp90α

41

as
readout
since
it
is
a
widely
reported
exosomal
cargo
protein.
To
establish
a
system

to
identify
such
regulator(s),
we
took
advantage
of
an
observation
made
previously

in
our
lab
that
in
human
keratinocytes
(HKCs),
only
TGFα,
but
not
EGF,
drives

Hsp90α
secretion
(Figure
1A).
This
was
interesting
since
both
TGFα
and
EGF
bind
to

the
same
cell
surface
EGFR
and
activate
many
common
downstream
pathways
to

drive
cell
proliferation,
migration,
adhesion
and
angiogenesis.
We
hypothesized
that

following
binding
to
EGFR,
only
TGFα
and
not
EGF
activated
a
pathway(s)
that

contributed
to
Hsp90α
secretion.
In
order
to
investigate
if
this
secretion
required

new
protein
synthesis
or
TGFα
drove
the
secretion
of
the
pre-‐existing
Hsp90α
pool,

we
took
advantage
of
cycloheximide
(CHX)
that
interferes
with
protein
biosynthesis

by
interfering
with
translational
elongation.
We
found
that
CHX
treatment
did
not

block
TGFα-‐stimulated
Hsp90α
secretion,
implying
that
TGFα
stimulated
the

secretion
of
the
pre-‐existing
Hsp90α
pool
(Fig
1B).
Since
our
goal
was
to
identify

differentially
activated
pathways
between
TGFα
and
EGF,
we
took
advantage
of
the

human
phospho-‐kinase
array
from
R&D
Systems.
This
array
simultaneously

compares
the
activation
of
47
different
kinases
and
two
related
proteins.
The

experimental
design
is
illustrated
in
Fig
1D.
Lysates
from
untreated,
TGFα
treated

and
EGF
treated
HKCs
were
added
to
the
array
and
procedures
were
carried
out
as

per
manufacturer’s
instructions.
At
the
2
min
time
point,
we
detected
four
signaling

molecules,
including
PRAS40,
EGFR,
ERK1/2
and
Akt
(T-‐308),
that
exhibited

significantly
stronger
phosphorylation
in
TGFα-‐stimulated
cells
than
EGF-‐stimulated

cells.
We
eliminated
the
involvement
of
EGFR
and
ERK1/2,
because
the
difference

was
due
to
different
kinetics
of
their
phosphorylation
in
response
to
the
two
growth

42

factors.
We
decided
to
focus
on
PRAS40,
because
it
has
been
reported
that

phosphorylation
at
threonine-‐308
of
Akt
phosphorylates
PRAS40
at
threonine246

and
activates
PRAS40
(24).
In
addition,
TGFα,
but
not
EGF,
stimulates
theonine-‐308

phosphorylation
of
Akt
in
human
keratinocytes
(see
later
section).

Figure
2-‐1.
Identification
of
differentially
activated
pathways
between
TGFα
and
EGF

stimulated
human
keratinocytes
(HKCs)

(A) Western
blotting
showing
Hsp90α
secretion
only
in
response
to
TGFα
but
not
EGF

(B) Western
blotting
showing
that
cycloheximide
(10μg/ml)
did
not
inhibit
TGFα-‐stimulated

Hsp90α
secretion

(C) Model
depicting
the
rationale
behind
choosing
HKCs.
TGFα
activates
certain
pathway
(s)
not

activated
by
EGF
and
this
might
contribute
to
Hsp90α
secretion

(D) Phospho
kinase
array
for
comparing
pathways
amongst
unstimulated,
TGFα
or
EGF

43

stimulated
HKCs

We
first
verified
the
results
of
the
array
through
immunoblotting
(Figure
2).
We

treated
HKCs
with
TGFα
or
EGF
for
different
times
and
investigated
the
activation
of

PRAS40,
RSK
and
Akt
at
Threonine
308
(T308)
and
Serine
473
(S473).

Consistent

with
the
array
results,
PRAS40
was
phosphorylated
at
T246
only
by
TGFα
but
not

EGF
(panel
c).
This
correlated
with
Akt
phosphorylation
at
T308
only
by
TGFα

(panel
b).
On
the
other
hand,
Akt
was
phosphorylated
at
S473
by
both
TGFα
and

EGF
(panel
a).
Similarly,
another
protein
on
the
array,
RSK
was
also
phosphorylated

by
both
the
growth
factors
(panel
d).
Also,
TGFα-‐stimulated
PRAS40

phosphorylation
was
dependent
on
EGFR
signaling
since
knockdown
of
EGFR
in

HKCs
blocked
TGFα-‐stimulated
PRAS40
phosphorylation
(Figure
2B).
Thus,
we

propose
the
model
that
in
HKCs,
TGFα
but
not
EGF
phosphorylates
Akt
on
T308,

which
in
turn
phosphorylates
PRAS40
on
T246
(Figure
2C).

The
next
question
was
if
PRAS40
was
required
for
TGFα-‐stimulated
Hsp90α

secretion.
To
this
end,
we
down
regulated
PRAS40
in
HKCs
(Figure
2D).
We

subjected
serum-‐starved
control
and
PRAS40
knockdown
cells
to
TGFα
or
EGF

stimulation
and
investigated
Hsp90α
secretion
by
analyzing
conditioned
media.
We

found
that
while
TGFα-‐stimulated
Hsp90α
secretion
in
control
HKCs
(lane
5),

PRAS40
knockdown
significantly
decreased
the
secretion
(lane
5),
implying
PRAS40

was
required
for
TGFα-‐stimulated
Hsp90α
secretion
in
HKCs.
Prior
reports
have

shown
that
various
environmental
stresses
such
as
hypoxia
and
oxidative
stresses

44

also
drive
Hsp90α
secretion.
To
examine
if
PRAS40
also
regulated
Hsp90α
secretion

in
response
to
other
stress
signals,
we
focused
on
hypoxia
and
H2O2
.
We
subjected

control
and
PRAS40
knockdown
HKCs
to
either
normoxia
or
hypoxia
(1%
O2)
for
16

hours
and
collected
conditioned
media.
We
found
that
hypoxia-‐driven
Hsp90α

secretion
was
significantly
decreased
on
PRAS40
knockdown
(Figure
2F
lanes
3
vs

4).
In
addition,
while
treatment
with
H2O2
drove
Hsp90α
secretion
in
control
cells,
it

was
significantly
reduced
on
PRAS40
knockdown
(Figure
2G
lanes
3
vs
4).
A

schematic
representation
of
the
findings
is
shown
in
Figure
2H.

45

Figure
2-‐2.
PRAS40
knockdown
inhibits
Hsp90α
secretion
in
response
to
TGFα,
hypoxia
and

oxidative
stress
in
HKCs

(A) Time
course
treatment
of
HKCs
with
TGFα
or
EGF
to
investigate
phosphorylation
of

downstream
substrates

(B) EGFR
knockdown
in
HKCs
(top)
and
conditioned
media
from
control
&
EGFR
knockdown

HKCs
stimulated
with
TGFα
or
EGF
(bottom)
to
investigate
PRAS40
phosphorylation

(C) Model
depicting
our
current
understanding.
After
binding
to
EGFR,
only
TGFα
activates
Akt

on
T308,
resulting
in
activation
of
PRAS40
at
T246.

46

(D) PRAS40
knockdown
in
HKCs.
Knockdown
decreases
Hsp90α
secretion
in
response
to
TGFα

(E),
hypoxia
(F)
and
H2O2
(G).

PRAS40
regulates
exosome
secretion

We
next
asked
how
PRAS40
regulated
Hsp90α
secretion.
From
studies
of
others
and

our
lab,
we
know
that
secretion
of
Hsp90α
is
mediated
by
the
unconventional

exosome
trafficking
pathway,
due
to
the
absence
of
a
signal
peptide
in
Hsp90α,
that

is
required
for
secretion
by
the
ER/Golgi
pathway.
Using
inhibitors
specific
to
the

ER-‐Golgi
pathway
(Brefeldin
A
or
BFA)
and
the
exosome
trafficking
pathway

(Dimethyl
ameloride
or
DMA),
several
groups
including
ours
have
reported
that

only
DMA
blocks
Hsp90α
secretion
(25,
26).

To
investigate
if
PRAS40
regulates

exosome
secretion,
we
treated
control
and
PRAS40
knockdown
HKCs
with
TGFα
co-‐
treated
with
either
BFA
or
DMA.
Neither
BFA
nor
DMA
inhibited
PRAS40

phosphorylation
in
response
to
TGFα
(Figure
3B
panel
a).

We
also
isolated

exosomes
from
serum-‐free
conditioned
media
using
ultracentrifugation.
We

quantitated
the
number
of
exosomes
secreted
using
Nanosight
tracking
analysis

(NTA).
We
found
that
TGFα
treatment
increased
the
number
of
exosomes
secreted

in
control,
but
not
in
PRAS40
knockdown
HKCs
(Figure
3A).
This
increase
was

blocked
by
co-‐treatment
with
DMA
but
not
BFA.
We
also
investigated
a
panel
of

exosome
markers
(CD63,
CD81,
CD9
and
flotillin)
in
TGFα-‐stimulated
control
and

PRAS40
knockdown
HKCs.
Consistent
with
the
NTA
analysis,
we
found
that
TGFα

stimulated
the
secretion
of
Hsp90α
(panel
d),
CD63
(panel
e),
flotillin
(panel
f),

CD81
(panel
g)
and
CD9
(panel
h)
in
control
HKCs,
but
the
secretion
of
all
the

47

markers
were
significantly
reduced
when
PRAS40
was
down
regulated.
To

investigate
if
the
regulation
by
PRAS40
was
specific
to
the
exosome
trafficking

pathway,
we
looked
at
the
secretion
of
MMP9,
which
is
known
to
be
secreted
by
the

classical
ER-‐Golgi
protein
secretory
pathway
(27).
We
found
that
PRAS40

knockdown
failed
to
inhibit
MMP9
secretion.
Thus
PRAS40
regulates
exosome

secretion
in
response
to
TGFα
in
HKCs.
While
PRAS40
knockdown
decreased

secretion
of
exosomes,
it
resulted
in
intracellular
accumulation
of
exosomes.
As

shown
in
Figure
3C,
intracellular
CD63
levels
were
higher
in
PRAS40
knockdown

cells
compared
to
control
cells
(panel
a).
However,
levels
of
other
exosome
markers

such
as
CD9
(panel
b)
and
flotillin
(panel
c)
were
unaffected.

Figure
2-‐3.
PRAS40
down
regulation
decreases
TGFα-‐driven
exosome
secretion
in
HKCs

(A) Nanosight
Tracking
Analysis
(NTA)
to
investigate
number
of
exosomes
secreted
in
control

HKCs
treated
with
TGFα,
co-‐treated
with
BFA
or
DMA
(bars
1-‐4).
NTA
of
exosomes
from

control
&
PRAS40
down
regulated
HKCs
in
response
to
TGFα
stimulation
(bars
5-‐8).

48

(B) PRAS40
phosphorylation
in
lysates
of
TGFα-‐treated
control
and
PRAS40
knockdown
HKCs.

Levels
of
exosome
markers
from
the
conditioned
media
of
the
same
cells.

(C) Intracellular
levels
of
exosome
markers
in
TGFα-‐treated
control
and
PRAS40
knockdown

HKCs

PRAS40
regulates
constitutive
exosome
secretion
in
MDA-‐MB-‐231
breast

cancer
cells

We
previously
reported
that
MDA-‐MB-‐231,
a
triple
negative
breast
cancer
cell
line

secretes
Hsp90α
constitutively
due
to
constitutive
HIF1α
expression
(2).
To

examine
the
role
of
PRAS40
in
this
process,
we
down
regulated
PRAS40
in
MDA-‐MB-‐
231
cells
(Figure
4A).
We
subjected
control
and
PRAS40
knockdown
cells
to
either

normoxia
or
hypoxia
and
isolated
exosomes
from
the
different
conditions.
PRAS40

knockdown
significantly
decreased
both
constitutive
and
hypoxia-‐driven
exosome

secretion
in
MDA-‐MB-‐231
cells
(Figure
4B).
Thus,
PRAS40
is
a
central
regulator
of

exosome
secretion
in
response
to
diverse
stresses
in
normal
and
cancer
cells.

Figure
2-‐4.
PRAS40
regulates
constitutive
and
hypoxia-‐driven
exosome
secretion
in
MDA-‐MB-‐
231
breast
cancer
cells

(A) Western
blot
verification
of
PRAS40
down
regulation
in
MDA-‐MB-‐231
cells

49

(B) Effect
of
PRAS40
down
regulation
on
constitutive
&
hypoxia-‐induced
exosome
secretion
in

MDA-‐MB-‐231
cells

TGFα-‐stimulated
exosome
secretion
in
HKCs
requires
T246
phosphorylation

of
PRAS40

TGFα
stimulates
the
phosphorylation
of
PRAS40
at
T246
(Figure
2A).
The
schematic

showing
the
structure
of
PRAS40
is
shown
in
Fig
5A.
PRAS40
consists
of
two

proline-‐rich
regions
followed
by
TOS
and
RAIP
motifs,
implicated
in
mTORC1

binding
(41).
To
investigate
the
requirement
of
T246
phosphorylation
of
PRAS40
in

TGFα-‐stimulated
exosome
secretion,
we
generated
phosphorylation-‐deficient

(T246A)
and
phosphorylation-‐mimic
(T246E)
mutants.
We
overexpressed
WT,

T246A
or
T246E
cDNA
into
PRAS40
down
regulated
HKCs
(Figure
5B
lanes
3,4,5
vs

lane
2).
We
isolated
exosomes
from
control,
PRAS40
down
regulated
HKCs
and

PRAS40
down
regulated
HKCs
in
which
WT,
T246A
or
T246E
constructs
were

overexpressed,
left
untreated
or
treated
with
TGFα.
We
investigated
the
secretion
of

Hsp90α
and
an
exosome
marker,
CD9.
We
found
that
PRAS40
knockdown
decreased

TGFα-‐stimulated
exosome
secretion,
which
was
rescued
by
WT
cDNA

overexpression
(lane
6
vs
lane
4)
The
T246A
phosphorylation
deficient
mutant

failed
to
rescue
the
secretion
(lane
8
vs
lane
4).
Interestingly,
the
T246E
mutant

rescued
secretion
only
in
response
to
TGFα
and
not
constitutively
even
in
untreated

cells
(lane
10
vs
lane
4).
This
implies
that
PRAS40
phosphorylation
at
T246
was

necessary,
but
not
sufficient
for
TGFα-‐stimulated
exosome
secretion
and
TGFα
has

additional
downstream
effectors
to
drive
exosome
secretion.

50

Figure
2-‐5.
TGFα-‐stimulated
exosome
secretion
requires
T246
phosphorylation
of
PRAS40.

(A) Structure
of
PRAS40
highlighting
the
location
of
Threonine
246
at
the
C-‐terminus
and
design

of
different
mutants

(B) Overexpression
of
PRAS40
WT
and
mutants
in
PRAS40
down
regulated
HKCs

(C) Exosome
secretion
in
response
to
TGFα
by
the
cells
shown
in
(B).

Constitutive
exosome
secretion
in
MDA-‐MB-‐231
cells
is
independent
of
T246

phosphorylation
of
PRAS40

In
contrast
to
induced
secretion
in
normal
cells,
MDA-‐MB-‐231
breast
cancer
cells

constitutively
secrete
exosomes
due
to
constitutive
HIF-‐1α
expression
(25).
In
these

cells,
PRAS40
is
constitutively
phosphorylated
at
T246
(28).
To
investigate
the

requirement
of
T246
phosphorylation
in
exosome
secretion,
we
overexpressed
WT,

T246A
or
T246E
cDNA
into
PRAS40
down
regulated
MDA-‐MB-‐231
(Figure
6
panels

a
&
b).
We
investigated
the
constitutive
secretion
of
Hsp90α
and
exosome
markers,

51

CD9
and
CD63
in
control,
PRAS40
down
regulated
MDA-‐MB-‐231
and
PRAS40
down

regulated
MDA-‐MB-‐231
in
which
WT,
T246A
or
T246E
constructs
were

overexpressed.
Interestingly,
in
contrast
to
our
data
on
HKCs,
we
found
that
all
the

three
constructs
were
able
to
rescue
exosome
secretion
in
PRAS40
down
regulated

MDA-‐MB-‐231
cells
(Figure
6
panels
c,d,e).
There
are
two
possible
explanations
for

this
finding.
The
first
possibility
is
that
normal
and
cancer
cells
utilize
PRAS40

differently
to
regulate
exosome
secretion.
The
second
possibility
is
that
different

stresses
regulate
PRAS40
differently
to
mediate
exosome
secretion.
Thus,
TGFα-‐
stimulated
exosome
secretion
requires
T246
phosphorylation,
while
hypoxia-‐
induced
secretion
does
not.
The
second
hypothesis
can
be
verified
in
HKCs,
which

secrete
exosomes
in
response
to
hypoxia.

Figure
2-‐6.
Constitutive
exosome
secretion
in
MDA-‐MB-‐231
cells
is
independent
of
T246

phosphorylation
of
PRAS40

(a) and
(b).
Overexpression
of
PRAS40
WT
and
mutant
cDNA
in
PRAS40
down
regulated
MDA-‐
MB-‐231
cells.

52

(c),
(d)
and
(e).
Secreted
Hsp90α,
CD9
and
CD63
levels
in
exosomes
from
cells
in
(a).

Hypoxia
increases
PRAS40
levels
in
MDA-‐MB-‐231
cells

Since
hypoxia-‐driven
exosome
secretion
in
MDA-‐MB-‐231
cells
was
independent
of

the
T246
phosphorylation
of
PRAS40,
we
hypothesized
that
the
rescue
of
secretion

in
PRAS40
down
regulated
MDA-‐MB-‐231
cells
was
due
to
increased
total
protein

levels,
rather
than
the
T246
phosphorylation
status.
To
this
end,
we
subjected
the

cells
to
1%
O2
for
increasing
periods
of
time
and
investigated
PRAS40
levels.
We

found
that
PRAS40
levels
were
increased
in
response
to
hypoxia
at
3h
and
6h
and

started
decreasing
at
8h
(Figure
7
panel
a).
We
observed
an
increase
in
HIF-‐1α

levels
at
around
1h
and
3h,
followed
by
a
decrease
with
increasing
time
of
hypoxic

exposure
at
6h
and
8h
(panel
b).
Thus,
hypoxia
up
regulates
PRAS40,
which
in
turn,

might
regulate
exosome
secretion.

Figure
2-‐7.
Hypoxia
increases
PRAS40
levels
in
MDA-‐MB-‐231
cells

(a) PRAS40
levels
in
MDA-‐MB-‐231
cells
subjected
to
hypoxia
for
different
periods
of
time

(b) HIF1-‐α
levels
in
the
cells
in
(a)

53

PRAS40
does
not
interact
with
exosomes
inside
the
cells

In
order
to
investigate
how
PRAS40
regulates
exosome
secretion
and
to
examine

any
potential
interactions
between
PRAS40
and
exosomes
before
their
fusion
with

the
plasma
membrane,
we
performed
co-‐immunoprecipitation
(co-‐IP)
assays.
We

used
anti-‐PRAS40
antibody
to
pull
down
the
protein
from
MDA-‐MB-‐231
cells
and

performed
western
blotting
for
markers
such
as
CD63
and
CD9.
We
found
no

association
between
PRAS40
and
both
the
markers
tested
(Figure
8).
This
is

consistent
with
the
fact
that
PRAS40
is
not
secreted
by
exosomes
(Figure
3).
Hence,

the
regulation
of
exosome
secretion
by
PRAS40
is
indirect
and
involves
additional

players
between
PRAS40
and
the
exosome
pathway.

Figure
2-‐8.
PRAS40
does
not
associate
with
exosomes

(a) CD63
association
after
immunoprecipitation
with
increasing
amounts
of
PRAS40
antibody

(b) CD9
association
after
immunoprecipitation
with
increasing
amounts
of
PRAS40
antibody

(c) PRAS40
levels
following
immunoprecipitation
with
increasing
amounts
of
PRAS40
antibody

54

Discussion

We
have
identified
PRAS40
as
a
new
regulator
of
exosome
secretion
in
response
to

diverse
stress
signals
in
different
cell
types.
We
initially
took
advantage
of
the

differential
responses
of
HKCs
to
TGFα
and
EGF
treatment,
where
despite
being
to

the
same
EGFR,
only
TGFα
drove
Hsp90α
secretion.
Comparison
of
pathways

between
TGFα
and
EGF
treated
HKCs
helped
us
identify
PRAS40.
Since
cells
secrete

Hsp90α
in
response
to
other
stresses
such
as
hypoxia
and
oxidative
stress,
we
also

investigated
and
verified
the
importance
of
PRAS40
in
exosome
secretion
under

these
stresses
in
different
cell
types.
PRAS40
works
in
both
T246
phosphorylation

dependent
and
protein
level
dependent
pathways
to
mediate
exosome
secretion.
We

found
that
TGFα-‐induced
exosome
secretion
depended
on
T246
phosphorylation
of

PRAS40
in
HKCs.
On
the
other
hand,
constitutive
and
hypoxia-‐driven
exosome

secretion
in
MDA-‐MB-‐231
cells
was
independent
of
phosphorylation
at
this
site
and

instead
depended
on
PRAS40
expression
levels.
PRAS40
did
not
associate
with

exosomes
inside
cells,
ruling
out
the
possibility
of
a
direct
interaction
between

PRAS40
and
exosomes
being
involved
in
regulating
secretion.

PRAS40
is
well
studied
as
an
inhibitor
and
substrate
of
mTORC1
complex.
It

possesses
two
short
motifs
implicated
in
mTORC1
binding,
i.e.,
a
TOS
motif
(amino

acids
129–133
of
the
human
256-‐amino
acid
PRAS40
protein)
and
a
Lys-‐Ser-‐Leu-‐
Pro
sequence
(amino
acids
182–185)
showing
resemblance
to
the
RAIP
motif.
In

addition
to
PRAS40,
mTORC1
complex
includes
the
regulatory-‐associated
protein
of

mTOR
(raptor),
mLST8
and
deptor.
Raptor
acts
as
a
scaffold
protein,
facilitating

55

assembly
of
the
mTORC1
complex.
PRAS40
competes
with
mTOR
substrates,
4E-‐
BP1
and
S6K1,
for
substrate
binding
sites
on
raptor.
Dissociation
of
PRAS40
from

raptor
by
growth
factor
induced
PRAS40
phosphorylation
enables
the
binding
of
4E-‐
BP1
and
S6K1
to
the
substrate
binding
sites
on
raptor,
activating
mTORC1
resulting

in
increased
cell
growth
and
protein
synthesis
(29).
PRAS40
is
mainly
regulated

through
extensive
phosphorylation
at
T246
and
a
panel
of
mTORC1

phosphorylation
sites
such
as
S221,
S183,
S202
S203,
S211
and
S212.
Akt
is
the

principal
kinase
phosphorylating
PRAS40
at
T246
since
PI3K
inhibitors
like

wortmannin,
but
not
mTOR
inhibitors
like
rapamycin
blocked
insulin-‐stimulated

T246
phosphorylation
in
skeletal
and
cardiac
muscles
in
vivo
and
in
vitro
(35).

PDGF-‐stimulated
T246
phosphorylation
was
blocked
in
MEFs
lacking
Akt1
and

Akt2.
Inducible
activation
of
Akt
was
sufficient
to
promote
T246
phosphorylation
in

NIH3T3
fibroblasts
(34).
However,
Akt
independent
pathways
can
also
drive
T246

phosphorylation.
Protein
kinase
A
(PKA)
has
been
shown
to
phosphorylate
T246
in

thyroid
cells
in
response
to
thyroid
hormone
treatment
and
intracellular
cAMP

levels
(36).
The
proto-‐oncogene,
PIM1,
phosphorylates
T246
in
non-‐small
cell
lung

cancer
cells
and
promotes
radioresistance
(37).

In
the
past
decade,
exosomes
have
been
gradually
recognized
as
important
vehicles,

which
are
able
to
carry
various
kinds
of
cargos
such
as
proteins,
lipids
and
nucleic

acids,
for
intercellular
communication,
thereby
influencing
various
physiological

and
pathological
functions
of
both
donor
and
recipient
cells.
Various
environmental

stimuli
regulate
not
only
secretion
of
exosomes,
but
also
modify
their
cargoes.
In

56

adipose
mesenchymal
stem
cells,
treatment
with
growth
factors
such
as
PDGF-‐BB

increased
the
number
of
exosomes
secreted
and
modulated
their
cargo
to
a
more

pro-‐angiogenic
state
(30).
Similarly,
TGFβ2
and
PDGF-‐BB
treatment
of

cardiomyocytes
resulted
in
secretion
of
exosomes
with
higher
expression
of

transcripts
involved
in
proliferation
and
hypertrophy
(31).
Stresses
such
as
hypoxia

increase
exosome
secretion
from
endothelial
cells
(32)
and
various
tumor
cells
(17,

18,
19,
33),
facilitating
angiogenesis
and
tumor
metastasis
respectively.
Considering

the
growing
number
of
roles
of
exosomes
under
normal
and
pathological
conditions,

there
is
a
great
need
to
identify
regulators
of
exosome
secretion.
So
far,
Rab
family

GTPases
have
been
widely
studied
for
their
role
in
regulating
exosome
secretion.

Rab27a
and
b
regulate
exosome
secretion
in
Hela
cells
by
controlling
different
steps

of
the
exosome
trafficking
pathway.
Rab27a
knockdown
resulted
in
intracellular

accumulation
of
exosomal
vesicles
and
increased
their
size,
while
Rab27b

knockdown
resulted
in
a
perinuclear
distribution
of
exosomes.
These
resulted
in

reduced
docking
of
MVBs
at
the
plasma
membrane,
a
pre-‐requisite
for
exosome

secretion
(38).
Rab27a
was
overexpressed
in
malignant
melanoma
cells
and
Rab27a

knockdown
decreased
the
number
of
exosomes
secreted,
though
it
did
not
affect
the

protein
content,
implying
Rab27a
regulated
exosome
secretion
quantitatively,
but

not
qualitatively
(39).
In
ER
positive
breast
cancer
cell
lines,
only
Rab27b
but
not

Rab27a
regulated
Hsp90α
secretion
and
promoted
cell
proliferation,
invasion
and

tumorigenesis
in
mice
(40).

Rab22A
mediates
formation
of
microvesicles
that

increase
breast
cancer
metastasis
(17).
Thus,
the
importance
of
an
individual
Rab

GTPase
in
exosome
secretion
varies
depending
on
the
cell
type.
Interestingly,
Bobrie

57

et
al.,
reported
that
Rab27a
modulated
the
tumor
microenvironment
of
mouse

mammary
carcinoma
by
both
exosome-‐dependent
and
independent
pathways.

Using
two
different
mouse
mammary
carcinoma
cell
lines,
4T1
and
TS/A,
the

authors
show
that
only
Rab27a,
but
not
Rab27b
regulated
exosome
secretion,

primary
tumor
formation
and
lung
colonization
by
4T1
cells
in
mice.
Rab27a
also

regulated
the
secretion
of
MMP9,
a
protein
secreted
through
the
conventional

ER/Golgi
pathway.
These
functions
facilitated
the
influx
of
pro-‐tumoral
neutrophils

and
thereby
tumor
progression
(42).
In
contrast,
we
found
that
PRAS40
specifically

regulates
exosome
secretion
in
different
cell
types
and
under
diverse
environmental

stresses.
Targeting
PRAS40
in
cancer
will
therefore
be
a
valid
approach
to
block

exosome
secretion
and
overcome
exosome-‐mediated
tumor
invasion,
angiogenesis

and
metastasis.

References:

1. Gould
SJ,
Raposo
G.
(2013).
As
we
wait:
coping
with
an
imperfect

nomenclature
for
extracellular
vesicles.
J
Extracell
Vesicles.
2,
20389

2. Colombo
M,
Raposo
G,
Thery
C.
(2014)
Biogenesis,
secretion,
and

intercellular
interactions
of
exosomes
and
other
extracellular
vesicles.
Ann

Rev
Cell
Dev
Biol.
30,
255-‐89.

3. Kalra
H,
Adda
CG,
Liem
M,
Ang
CS,
Mechler
A,
Simpson
RJ,
et
al.
(2013).

Comparative
proteomics
evaluation
of
plasma
exosome
isolation
techniques

58

and
assessment
of
the
stability
of
exosomes
in
normal
human
blood
plasma.

Proteomics.
13,
3354-‐64.

4. Tauro
BJ,
Greening
DW,
Mathias
RA,
Ji
H,
Mathivanan
S,
Scott
AM,
et
al.

(2013).
Comparison
of
ultracentrifugation,
density
gradient
separation,
and

immunoaffinity
capture
methods
for
isolating
human
colon
cancer
cell
line

LIM1863-‐derivedexosomes.
Methods.
56,
293-‐304.

5. Théry,
C.,
Regnault,
A.,
Garin,
J.,
Wolfers,
J.,
Zitvogel,
L.,
Ricciardi-‐Castagnoli,

P.,
et
al.
(1999).
Molecular
Characterization
of
Dendritic
Cell-‐Derived

Exosomes:
Selective
Accumulation
of
the
Heat
Shock
Protein
Hsc73.The
J.

Cell.
Biol.
147(3),
599–610.

6. Hegmans,
J.
P.
J.
J.,
Bard,
M.
P.
L.,
Hemmes,
A.,
Luider,
T.
M.,
Kleijmeer,
M.
J.,

Prins,
J.-‐B.,
et
al.
(2004).
Proteomic
Analysis
of
Exosomes
Secreted
by
Human

Mesothelioma
Cells.
The
American
Journal
of
Pathology,
164(5),
1807–1815.

7. Raposo,
G.,
Nijman,
HW.,
Stoorvogel,
W.,
Liejendekker,
R.,
Harding,
CV.,
et
al.

(1996).
B
lymphocytes
secrete
antigen-‐presenting
vesicles.
J.
Exp.
Med.
183,

1161–72

8. Zitvogel,
L.,
Regnault,
A.,
Lozier,
A.,
Wolfers,
J.,
Flament,
C.,
et
al.
1998.

Eradication
of
established
murine
tumors
using
a
novel
cell-‐free
vaccine:

dendritic
cell-‐derived
exosomes.
Nat.
Med.
4,
594–600

9. Singh,
P.
P.,
Smith,
V.
L.,
Karakousis,
P.
C.,
&
Schorey,
J.
S.
(2012).
Exosomes

isolated
from
mycobacteria-‐infected
mice
or
cultured
macrophages
can

recruit
and
activate
immune
cells
in
vitro
and
in
vivo.
J.
Immunol.
189(2),

777–785.

59

10. Wahlgren,
J.,
Karlson,
T.
D.
L.,
Glader,
P.,
Telemo,
E.,
&
Valadi,
H.
(2012).

Activated
Human
T
Cells
Secrete
Exosomes
That
Participate
in
IL-‐2
Mediated

Immune
Response
Signaling.
PLoS
ONE,
7(11),
e49723.

11. Shabbir,
A.,
Cox,
A.,
Rodriguez-‐Menocal,
L,
Salgado
M,
&
Van
Badiavas
E.

(2015).
Mesenchymal
Stem
Cell
Exosomes
Induce
Proliferation
and

Migration
of
Normal
and
Chronic
Wound
Fibroblasts,
and
Enhance

Angiogenesis
In
Vitro.
Stem
Cells
Dev.

24(14),
1635-‐47

12. Zhu,
Y.,
Feng,
X.,
Abbott,
J.,
Fang,
X.,
Hao,
Q.,
Monsel,
A.,
et
al.
(2014).
Human

Mesenchymal
Stem
Cell
Microvesicles
for
Treatment
of
E.coli
Endotoxin-‐
Induced
Acute
Lung
Injury
in
Mice.
Stem
Cells
32(1),
116–125.

13. Zhou,
Y.,
Xu,
H.,
Xu,
W.,
Wang,
B.,
Wu,
H.,
Tao,
Y.,
et
al.
(2013).
Exosomes

released
by
human
umbilical
cord
mesenchymal
stem
cells
protect
against

cisplatin-‐induced
renal
oxidative
stress
and
apoptosis
in
vivo
and
in

vitro.
Stem
Cell
Research
&
Therapy,
4(2),
34.

14. Chavez-‐Muñoz,
C.,
Morse,
J.,
Kilani,
R.
&
Ghahary,
A.
(2008).
Primary
human

keratinocytes
externalize
stratifin
protein
via
exosomes.
J.
Cell.
Biochem.,
104:

2165–2173.

15. Lam
E,
Kilani
RT,
Li
Y,
Tredget
EE,
Ghahary
A.
(2005).
Stratifin-‐induced

matrix
metalloproteinase-‐1
in
fibroblast
is
mediated
by
c-‐fos
and
p38

mitogen-‐activated
protein
kinase
activation.
J
Invest
Dermatol.

125(2),
230-‐
8.

60

16. Geiger
A,
Walker
A,
Nissen
E.
(2015).
Human
fibrocyte-‐derived
exosomes

accelerate
wound
healing
in
genetically
diabetic
mice.
Biochem
Biophys
Res

Commun.

467(2),
303-‐9

17. Wang,
T.,
Gilkes,
D.
M.,
Takano,
N.,
Xiang,
L.,
Luo,
W.,
Bishop,
C.
J.,
et
al.
(2014).

Hypoxia-‐inducible
factors
and
RAB22A
mediate
formation
of
microvesicles

that
stimulate
breast
cancer
invasion
and
metastasis.
Proc.
Natl.
Acad.
Sci.

USA,
111(31),
E3234–E3242.

18. Kucharzewska,
P.,
Christianson,
H.
C.,
Welch,
J.
E.,
Svensson,
K.
J.,
Fredlund,
E.,

Ringnér,
M.,
et
al.
(2013).
Exosomes
reflect
the
hypoxic
status
of
glioma
cells

and
mediate
hypoxia-‐dependent
activation
of
vascular
cells
during
tumor

development.
Proc.
Natl.
Acad.
Sci.
USA,
110(18),
7312–7317.

19. King,
HW.,
Michael,
MZ.
&
Jonathan
M
Gleadle
(2012).
Hypoxic

enhancement
of
exosome
release
by
breast
cancer
cells.
BMC
Cancer
12,

421

20. Logozzi,
M.,
De
Milito,
A.,
Lugini,
L.,
Borghi,
M.,
Calabrò,
L.,
Spada,
M.,
…
Fais,
S.

(2009).
High
Levels
of
Exosomes
Expressing
CD63
and
Caveolin-‐1
in
Plasma

of
Melanoma
Patients.
PLoS
ONE,
4(4),
e5219.

21. Harris,
D.
A.,
Patel,
S.
H.,
Gucek,
M.,
Hendrix,
A.,
Westbroek,
W.,
&
Taraska,
J.

W.
(2015).
Exosomes
Released
from
Breast
Cancer
Carcinomas
Stimulate
Cell

Movement.
PLoS
ONE,
10(3),
e0117495.

22. Costa-‐Silva
B,
Aiello
NM,
Ocean
AJ,
Singh
S,
Zhang
H,
Thakur
BK.,
et
al.
(2015).

Pancreatic
cancer
exosomes
initiate
pre-‐metastatic
niche
formation
in
the

liver.
Nat
Cell
Biol.

17(6),
816-‐26

61

23. Arscott,
W.
T.,
Tandle,
A.
T.,
Zhao,
S.,
Shabason,
J.
E.,
Gordon,
I.
K.,
Schlaff,
C.
D.,

et
al.
(2013).
Ionizing
Radiation
and
Glioblastoma
Exosomes:
Implications
in

Tumor
Biology
and
Cell
Migration.
Translational
Oncology,
6(6),
638–648.

24. Vincent,
E.
E.,
Elder,
D.
J.
E.,
Thomas,
E.
C.,
Phillips,
L.,
Morgan,
C.,
Pawade,
J.,
et

al.,
(2011).
Akt
phosphorylation
on
Thr308
but
not
on
Ser473
correlates
with

Akt
protein
kinase
activity
in
human
non-‐small
cell
lung
cancer.
Br.
J.

Cancer,
104(11),
1755–1761

25. Sahu,
D.,
Zhao,
Z.,
Tsen,
F.,
Cheng,
C.-‐F.,
Park,
R.,
Situ,
A.
J.,
et
al.,
(2012).
A

potentially
common
peptide
target
in
secreted
heat
shock
protein-‐90α
for

hypoxia-‐inducible
factor-‐1α–positive
tumors.
Mol.
Biol.
Cell.

23(4),
602–613.

26. McCready,
J.,
Sims,
J.
D.,
Chan,
D.,
&
Jay,
D.
G.
(2010).
Secretion
of
extracellular

hsp90α
via
exosomes
increases
cancer
cell
motility:
a
role
for
plasminogen

activation.
BMC
Cancer,
10,
294.
27. Duellman,
T.,
Burnett,
J.,
Shin,
A.,
&
Yang,
J.
(2015).
LMAN1
(ERGIC-‐53)
is
a

potential
carrier
protein
for
matrix
metalloproteinase-‐9
glycoprotein

secretion.
Biochem.
Biophys.
Res.
Commun.
464(3),
685-‐91
28. Zou,
M.,
Bhatia,
A.,
Dong,
H.,
Jayaprakash,
P.,
Sahu,
D.,
Tsen,
F.,
et
al.,

Conserved
Dual
Lysines
Determine
Non-‐Chaperone
Function
of
Secreted

HSP90α
in
Tumor
Progression.
Oncogene
(in
submission).
29. Wang
L,
Harris
TE,
Roth
RA,
Lawrence
JC
Jr.
(2007).
PRAS40
regulates

mTORC1
kinase
activity
by
functioning
as
a
direct
inhibitor
of
substrate

binding.
J
Biol
Chem.

282
(27),
20036-‐44

62

30. Lopatina,
T.,
Bruno,
S.,
Tetta,
C.,
Kalinina,
N.,
Porta,
M.,
&
Camussi,
G.
(2014).

Platelet-‐derived
growth
factor
regulates
the
secretion
of
extracellular

vesicles
by
adipose
mesenchymal
stem
cells
and
enhances
their
angiogenic

potential.Cell
Commun.
Signal,
12,
26.

31. Gennebäck,
N.,
Hellman,
U.,
Malm,
L.,
Larsson,
G.,
Ronquist,
G.,
Waldenström,

A.,
&
Mörner,
S.
(2013).
Growth
factor
stimulation
of
cardiomyocytes
induces

changes
in
the
transcriptional
contents
of
secreted
exosomes.
J.
Extracell.

Vesicles,
2,
10.3402

32. De
Jong,
O.
G.,
van
Balkom,
B.
W.
M.,
Gremmels,
H.,
&
Verhaar,
M.
C.
(2016).

Exosomes
from
hypoxic
endothelial
cells
have
increased
collagen

crosslinking
activity
through
up-‐regulation
of
lysyl
oxidase-‐like
2.
J.
Cell
Mol.r

Med.
20(2),
342–350.

33. Hu,
F.,
Deng,
X.,
Yang,
X.,
Jin,
H.,
Gu,
D.,
Lv,
X.,
et
al.
(2015).
Hypoxia

upregulates
Rab11-‐family
interacting
protein
4
through
HIF-‐1α
to
promote

the
metastasis
of
hepatocellular
carcinoma.
Oncogene.

34
(49),
6007-‐17.
34. Kovacina,
KS.,
Park,
GY.,
Bae,
SS.,
Guzzetta,
AW.,
Schaefer,
E.,
Birnbaum,
MJ.,
&

Roth,
RA.
(2003).
Identification
of
a
proline-‐rich
Akt
substrate
as
a
14-‐3-‐3

binding
partner.
J
Biol
Chem
278,
10189–10194.
35. Nascimento,
EB.,
Fodor,
M.,
van
der
Zon,
GC.,
Jazet,
IM.,
Meinders,
AE.,
Voshol,

PJ.,
et
al.
(2006).
Insulin-‐mediated
phosphorylation
of
the
proline-‐rich
Akt

substrate
PRAS40
is
impaired
in
insulin
target
tissues
of
high-‐fat
diet-‐fed

rats.
Diabetes.
55(12),
3221-‐8.

63

36. Blancquaert,
S.,
Wang,
L.,
Paternot,
S.,
Coulonval,
K.,
Dumont,
JE.,
Harris
TE,
et

al.
(2010).
cAMP-‐dependent
activation
of
mammalian
target
of
rapamycin

(mTOR)
in
thyroid
cells.
Implication
in
mitogenesis
and
activation
of
CDK4.

Mol
Endocrinol
24,
1453–1468.
37. Kim,
W.,
Youn,
H.,
Seong,
KM.,
Yang,
HJ.,
Yun,
YJ.,
Kwon,
T.,
et
al.
(2011).
PIM1-‐
activated
PRAS40
regulates
radioresistance
in
non-‐small
cell
lung
cancer

cells
through
interplay
with
FOXO3a,
14-‐3-‐3
and
protein
phosphatases.

Radiat
Res.

176(5),
539-‐52
38. Ostrowski,
M.,
Carmo,
NB.,
Krumeich,
S.,
Fanget,
I.,
Raposo,
G.,
Savina,
A.,
et
al.

(2010).
Rab27a
and
Rab27b
control
different
steps
of
the
exosome
secretion

pathway.
Nat
Cell
Biol.

12(1),
19-‐30
39. Peinado
H.,
Alečković
M.,
Lavotshkin
S.,
Matei
I.,
Costa-‐Silva
B.,
Moreno-‐
Bueno
G.,
et
al.,
(2012).Melanoma
exosomes
educate
bone
marrow

progenitor
cells
toward
a
pro-‐metastatic
phenotype
through
MET.
Nature

Medicine,
18(6),
883–891.

40. Hendrix
A,
Maynard
D,
Pauwels
P,
Braems
G,
Denys
H,
Van
den
Broecke
R
et

al.,
(2010).
Effect
of
the
Secretory
Small
GTPase
Rab27B
on
Breast
Cancer

Growth,
Invasion,
and
Metastasis.
J
Natl
Cancer
Inst.
102(12):
866-‐80

41. Wiza,
C.,
Nascimento,
EB.,
&
Ouwens
DM.
(2012).
Role
of
PRAS40
in
Akt
and

mTOR
signaling
in
health
and
disease.
Am
J
Physiol
Endocrinol

Metab.

302(12),
E1453-‐60

42. Bobrie,
A.,
Krumeich,
S.,
Reyal,
F.,
Recchi,
C.,
Moita,
LF.,
Seabra,
MC.,
et
al.,

(2012).
Rab27a
supports
exosome-‐dependent
and
-‐independent
mechanisms

64

that
modify
the
tumor
microenvironment
and
can
promote
tumor

progression.
Cancer
Res.
72(19),
4920-‐30

65

Chapter
3:
Hsp90α
and
Hsp90β
Co-‐Operate
a
Stress-‐Response
Mechanism
to

Cope
With
Hypoxia
and
Nutrient
Paucity
during
Wound
Healing

Abstract

The
lack
of
blood
supply
makes
the
wound
microenvironment
predominantly

hypoxic,
deprived
of
oxygen
and
nutrients.
Heat
shock
proteins
of
Hsp90
family
are

secreted
under
hypoxia
and
enable
wound
healing
and
tissue
repair.
The
cytosolic

Hsp90
isoforms,
Hsp90α
and
Hsp90β
are
86%
identical
at
amino
acid
levels
and

thought
to
be
redundant
in
their
function.
Here,
we
report
a
unique
communication

between
Hsp90α
and
Hsp90β
in
driving
dermal
fibroblast
migration.
Hsp90β
acts
as

an
intracellular
chaperone
protein
for
LRP-‐1
receptor,
stabilizing
it
at
the
cell

surface.
Hsp90α,
on
the
other
hand,
gets
secreted
outside
the
cell
to
bind
LRP-‐1

extracellularly.
This
novel
communication
facilitates
cell
migration
and
wound

healing
and
has
the
potential
to
be
applicable
to
other
non-‐cutaneous
wounds
as

well.

Introduction

The
wound
environment
is
hypoxic
due
to
the
loss
of
blood
supply
and
excessive

consumption
of
oxygen
by
cells
at
the
wound
edge
(1).
Hypoxia
is
an
important

driver
of
skin
cell
migration,
promoting
migration
of
all
the
three
skin
cell
types,

including
human
keratinocytes
(2,
3),
human
dermal
fibroblasts
(4,
5)
and
human

dermal
microvascular
endothelial
cells
(6).
Impaired
hypoxia
results
in
impaired

wound
healing
as
in
the
case
with
diabetic
foot
ulcers
(7).
Hyperglycemia

destabilizes
HIF1-‐α
or
impairs
the
function
of
HIF1-‐α
(8,9,10)
resulting
in
impaired

66

wound
healing
responses.
In
vitro,
knockdown
of
HIF1-‐α,
blocked
migration
of

keratinocytes
and
human
dermal
fibroblasts
whereas
the
overexpression
of
both

wild
type
and
a
constitutively
active
mutant
of
HIF1-‐α
rescued
the
defect
(3,4).

Hypoxia
facilitates
the
stabilization
of
hypoxia-‐inducible
factor-‐1
α
(HIF1-‐α),
which

is
an
upstream
regulator
of
Hsp90α
secretion
(3,4,11).
The
secreted
Hsp90α,
in
turn,

promotes
wound
healing
and
tissue
repair.
Supplementation
with
recombinant

Hsp90α
drives
fibroblast
migration
under
hyperglycemia,
supporting
the
finding

that
HIF1-‐α
driven
Hsp90α
secretion
is
required
for
wound
healing
under
both

normal
and
diabetic
conditions
(6).
Recombinant
Hsp90α
was
found
to
accelerate

wound
closure
in
both
rodent
and
porcine
models
(12).
Using
site-‐directed

mutagenesis,
Cheng
et
al.,
identified
that
the
full
pro-‐motility
activity
of
secreted

Hsp90α
could
be
recapitulated
by
F-‐5,
a
115
amino
acid
region
encompassing
the

charged
linker
and
middle
domains,
independent
of
its
N-‐terminal
ATPase
domain

(6).
Topical
application
with
F-‐5,
enhanced
wound
re-‐epithelialization
compared
to

control
and
improved
the
quality
of
healing
in
normal
and
diabetic
pigs
and
mice

(11,12).

Different
mechanisms
of
action
of
secreted
Hsp90α
have
been
reported.
One

mechanism
is
that
secreted
Hsp90α
acts
by
binding
to
the
cell
surface
LRP-‐1
(Low

Density
Lipoprotein
Receptor
Related
Protein-‐1)
receptor,
driving
either
wound

healing
(13)
or
tumor
progression
(11,13).
LRP-‐1
is
a
member
of
the
LDL
receptor

family
that
includes
seven
members.
It
is
synthesized
as
a
600kDa
precursor
that
is

67

proteolytically
cleaved
into
a
515kDa
extracellular
ligand
binding
α
subunit,
an

85kDa
transmembrane
β
subunit
and
a
100
amino
acid
long
cytoplasmic
tail
(30).

The
extracellular
ligand
binding
domain
is
divided
into
four
domains,
I,
II,
III
and
IV.

Most
ligands
bind
to
domains
II
and
IV
(34).
LRP-‐1
is
a
scavenger
receptor
that
is

involved
in
the
uptake
of
around
30
ligands
including
extracellular
matrix

macromolecules,
active
proteinases
and
proteinase/inhibitor
complexes
(31).
It
is

also
reported
to
bind
heat
shock
proteins
such
as
Hsp90,
Hsp70,
gp96
and

calreticulin
(33).
LRP-‐1
knockout
in
mice
results
in
embryonic
lethality
(32).
Tsen
et

al.
reported
that
secreted
Hsp90α
binds
to
the
extracellular
domain
II
of
LRP-‐1,

activates
downstream
Akt
kinases
and
promotes
wound
healing
(13).
However,

other
reports
propose
that
extracellular
Hsp90α
acts
as
a
chaperone
stabilizing
its

extracellular
client
proteins
such
as
MMP2
and
MMP9,
activating
them
and
driving

tumor
cell
motility
and
invasion
(14,
16).
Cell
surface
HER2
is
another
reported

client
of
extracellular
Hsp90α
(18,
29).

There
are
two
cytosolic
Hsp90
isoforms,
Hsp90α
and
Hsp90β,
which
are
86%

identical
at
the
amino
acid
level.
Despite
the
high
degree
of
homology,
only
secreted

Hsp90α,
but
not
Hsp90β,
possesses
pro-‐motility
activity.
In
order
to
investigate
the

non-‐redundancy
of
the
two
isoforms
in
wound
healing,
we
used
human
dermal

fibroblasts
(HDFs)
as
a
model.
HDFs
offer
a
good
model
owing
to
three

characteristics:
1)
they
secrete
Hsp90α
under
hypoxia,
2)
have
the
ability
to
migrate

in
response
to
recombinant
Hsp90α
stimulation
and
3)
express
high
levels
of
LRP-‐1.

We
report
that
Hsp90α
and
Hsp90β
have
distinct
functions
in
the
wound
healing

68

process
and
cannot
compensate
for
each
other.
Hsp90β
acts
inside
the
cell
as
a

chaperone
for
LRP-‐1,
in
order
to
stabilize
it
at
the
cell
surface.

Hsp90α,
on
the
other

hand,
is
secreted
outside
where
it
acts
as
a
ligand
for
LRP-‐1,
mediating

transmembrane
signaling
and
driving
cell
motility.

Results

Secreted
Hsp90α
and
intracellular
Hsp90β
are
required
for
HDF
migration

In
order
to
understand
the
specific
roles
of
Hsp90α
and
Hsp90β
and
investigate
the

existence
of
a
compensatory
mechanism,
we
generated
lentiviral
knockdown
of

either
Hsp90α
or
Hsp90β
in
HDFs.
Fig
1a
shows
the
verification
of
antibody

specificity
using
recombinant
proteins,
showing
that
the
Hsp90α
and
Hsp90β

antibodies
do
not
cross
react.
We
then
used
these
antibodies
to
verify
specific

knockdown
of
the
Hsp90
proteins.
Fig
1b
depicts
that
the
knockdown
of
Hsp90α

and
Hsp90β
were
both
efficient
(>95%
down
regulation)
and
specific
(not
affecting

the
other
isoform).
We
subjected
these
cells
to
the
colloidal
gold
migration
assay.

This
assay
measures
the
migration
of
individual
cells
denoted
by
tracks
created
on
a

gold
surface.
HDFs
infected
with
control,
Hsp90α
or
Hsp90β
shRNA
were
stimulated

with
either
PDGF-‐BB
(physiological
condition)
or
hypoxia
(stress
condition)
and
left

to
migrate
overnight.
As
shown
in
Fig
1c,
all
three
groups
of
cells
had
basal
motility

under
serum-‐free
conditions.
PDGF-‐BB
induced
a
robust
migration
of
all
the
three

cell
types,
consistent
with
its
role
as
a
potent
mitogen
and
motogen
for
HDFs,

binding
and
signaling
through
the
cell
surface
PDGF
receptor.
Interestingly,
the

migration
of
Hsp90α
knockdown
cells
was
even
slightly
higher
than
control
cells.

69

However,
when
the
same
cells
were
subjected
to
hypoxia
(1%
O2),
only
the
control

cells
responded
with
increased
migration.
In
contrast,
both
Hsp90α
and
Hsp90β

knockdown
inhibited
hypoxia-‐driven
migration.
Since,
we
have
previously
reported

that
hypoxia
drives
the
secretion
of
Hsp90
proteins,
we
asked
the
question
if

extracellular
supplementation
with
recombinant
Hsp90α
and
Hsp90β
proteins

rescued
the
migration
defect
under
hypoxia.
We
found
that
recombinant
Hsp90α

but
not
Hsp90β
rescued
the
migration
defect
of
Hsp90α
knockdown
HDFs.
In

contrast,
neither
Hsp90α
nor
Hsp90β
protein
was
able
to
rescue
the
migration

defect
of
Hsp90β
knockdown
HDFs.
These
findings
indicate
that
extracellular

Hsp90α
and
intracellular
Hsp90β
are
required
for
hypoxia-‐driven
HDF
motility.
The

quantitation
of
migration
is
shown
in
Fig
1d.

70

Figure
3-‐1.
Distinct
requirements
for
Hsp90α
and
Hsp90β
for
hypoxia-‐triggered
cell

migration.
(A)
The
specificity
of
anti-‐Hsp90α
and
anti-‐Hsp90β
antibodies
was
confirmed
using

purified
recombinant
Hsp90α
and
Hsp90β
proteins
in
Western
blot
analysis.
(B)
Lentiviral

infection-‐mediated
down
regulation
of
endogenous
Hsp90α
and
Hsp90β
in
HDFs,
shown
by

Western
blots.(C)
The
above
cells
were
serum-‐starved
for
16
hours
and
subjected
to
colloidal
gold

migration
assay.
Motility
was
visualized
as
“migration
tracks”
indicated
by
dotted
circles.
Human

recombinant
(hr)
Hsp90α
and
Hsp90β
proteins
were
used
to
rescue
Hsp90
down
regulation-‐caused

cell
migration
defects
in
response
to
hypoxia.
(D)
Quantitation
of
the
cell
migration
(in
C)
as

Migration
Index
(MI,
%).
n
=
4,
p
<
0.05

Secreted
Hsp90α,
not
Hsp90β
mediates
hypoxia-‐driven
HDF
migration

In
order
to
verify
if
secreted
Hsp90α
is
necessary
for
hypoxia-‐driven
HDF
motility,

we
used
a
neutralizing
antibody
approach.
As
shown
in
Fig
2a,
we
subjected
wild

type
HDFs
to
the
colloidal
gold
migration
assay
in
response
to
either
PDGF-‐BB
or

hypoxia.
We
co-‐treated
the
cells
with
antibodies
specifically
targeting
either

71

secreted
Hsp90α
or
Hsp90β.
We
found
that
hypoxia
promoted
the
migration
of
wild

type
HDFs
compared
to
serum-‐free
control.
This
migration
was
dose
dependently

blocked
by
anti-‐Hsp90α,
but
not
anti-‐Hsp90β
antibody.
Interestingly,
neither
of
the

antibodies
blocked
PDGF-‐BB
driven
migration,
implying
that
physiological
and

stress
conditions
use
different
mechanisms
to
drive
HDF
motility.

We
also
found
that
secreted
Hsp90α
has
a
chemotactic
ability,
comparable
to
PDGF-‐
BB
stimulation
(Fig
2b).
Treatment
with
recombinant
Hsp90α
enabled
HDFs
to

migrate
toward
10%
FBS
through
a
collagen
coated
matrix
in
a
transwell
assay.
In

order
to
compare
the
relative
importance
of
Hsp90α
and
Hsp90β
in
vivo,
we
used
a

porcine
wound
healing
model
since
pig
skin
is
physiologically
closer
to
human
skin

than
rodents’
(12).
We
created
1.5
x
1.5
cm
wounds
and
topically
treated
them
with

either
carboxymethyl
cellulose
(CMC)
vehicle,
recombinant
Hsp90α
or
Hsp90β.
We

found
that
at
day
7
post
wounding,
control
CMC
treated
wounds
were
about
50%

closed.
Treatment
with
recombinant
Hsp90α
accelerated
wound
closure
to

approximately
75%.
Interestingly,
recombinant
Hsp90β
also
caused
a
moderate

acceleration
of
wound
closure,
though
significantly
lesser
than
recombinant
Hsp90α

treated
wounds.
The
major
difference
between
the
two
proteins
was
evident
in
the

quality
of
wound
healing,
as
determined
by
histological
analyses.
Recombinant

Hsp90α
treated
wounds
showed
a
similar
epidermal
thickness
as
unwounded
skin,

while
CMC
treated
wounds
showed
a
significant
reduction
in
thickness.

Recombinant
Hsp90β
treated
wounds
however,
displayed
the
poorest
quality
of

healing
amongst
the
three
groups,
with
an
extremely
thin
epidermis
(Fig
1d).
Thus,

72

recombinant
Hsp90α
is
a
superior
wound
healing
agent
than
recombinant
Hsp90β.

The
moderate
acceleration
of
wound
closure
by
Hsp90β
treatment
might
be
due
to

its
effects
on
wound
contraction
or
its
lower
promotility
effect
on
HDFs
(6).

Figure
3-‐2.
Secreted
Hsp90α,
not
Hsp90β,
mediates
hypoxia-‐triggered
HDF
migration
and

promotes
wound
healing.

(A)
HDF
migration
under
indicated
conditions.
Anti-‐Hsp90α
antibody
inhibited
hypoxia-‐triggered

HDF
migration
(bars
5
and
6).

(B)
Comparison
of
PDGF-‐BB-‐
and
Hsp90α-‐
induced
chemotaxis
relative
to
untreated
cells
using
the

transwell
assay.
The
percentage
of
cells
that
penetrated
that
membrane
is
calculated.

73

(C)
hrHsp90α
and
hrHsp90β
proteins
were
compared
for
their
effects
on
promoting
pig
wound

healing.
Topical
application
of
Hsp90
proteins
(100μg/ml)
or
control
vehicle
(CMC)
was
carried
out

once
on
day
0.
n
=
3,
p
<
0.05

(D)
H&E
staining
of
fully
closed
wounds
on
day
21.
n
=
20-‐24
(sections)
per
treatment.

Hsp90β
chaperones
LRP-‐1
and
stabilizes
it
on
the
cell
surface

The
findings
so
far
indicate
that
extracellular
Hsp90α
and
intracellular
Hsp90β
are

indispensable
for
hypoxia-‐driven
HDF
motility.
Our
next
question
was
to
identify

the
mechanism
by
Hsp90β
played
a
role
in
this
process.
We
focused
on
LRP-‐1,
the

cell
surface
receptor
required
for
extracellular
Hsp90α’s
transmembrane
signaling

and
pro-‐motility
activity
(19,
3,
13).
We
hypothesized
that
Hsp90α
and
Hsp90β
bind

differentially
to
LRP-‐1
receptor.
To
investigate
which
isoform
binds
to
the

cytoplasmic
tail
of
LRP-‐1,
we
took
advantage
of
RAP
(receptor-‐associated
protein),
a

universal
antagonist
of
the
extracellular
ligand-‐binding
domain
of
LRP-‐1
(20).
We

incubated
HDFs
with
increasing
amounts
of
RAP,
in
order
to
prevent
binding
of

Hsp90
proteins
to
LRP-‐1’s
extracellular
domain
following
cell
lysis.
We
then

performed
immunoprecipitation
with
LRP-‐1
antibody
and
looked
at
the
association

of
Hsp90α
and
Hsp90β.
As
shown
in
Fig
3a,
we
found
that
RAP
treatment

specifically
blocked
the
association
of
Hsp90α
to
LRP-‐1.
In
contrast,
association
of

Hsp90β
was
unaffected.
The
slight
increase
in
Hsp90β
association
with
increasing

amounts
of
RAP
might
be
due
to
lesser
internalization
of
LRP-‐1
following
RAP

binding.
Based
on
this
finding,
we
concluded
that
Hsp90α
binds
to
the
extracellular

ligand-‐binding
domain
of
LRP-‐1
while
Hsp90β
binds
to
its
intracellular
cytoplasmic

tail.
We
hypothesized
that
Hsp90β
binds
to
the
cytoplasmic
tail
of
LRP-‐1
and
acts
as

74

a
chaperone
to
stabilize
it
on
the
cell
surface.
Therefore,
we
investigated
LRP-‐1

levels
in
control,
Hsp90α
knockdown
and
Hsp90β
knockdown
HDFs.
We
found
that

only
the
knockdown
of
Hsp90β
significantly
decreased
LRP-‐1
levels,
verifying
that

Hsp90β
acts
as
a
chaperone
for
LRP-‐1.
In
contrast,
the
levels
of
PDGFRβ
and
EGFR

were
either
unaffected
or
slightly
reduced
in
the
Hsp90β
knockdown
HDFs

respectively.
We
also
verified
that
Hsp90β
regulated
LRP-‐1
post-‐translationally

since
the
mRNA
levels
were
unaffected
in
Hsp90β
knockdown
HDFs
(Fig
3c).

Due
to
the
high
degree
of
homology,
Hsp90α
and
Hsp90β
are
thought
to
compensate

for
each
other’s
function.
To
investigate
if
the
migration
defect
in
Hsp90β

knockdown
HDFs
could
be
rescued
by
physical
replacement
of
the
missing
Hsp90β

with
Hsp90α
to
increase
the
total
amount
of
Hsp90
proteins
in
the
cells,
we
used

lentiviral
infection
to
overexpress
Hsp90α
in
Hsp90β
knockdown
HDFs.
As
shown
in

Fig
3d,
there
was
a
compensatory
up
regulation
of
Hsp90α
in
Hsp90β
knockdown

HDFs
(panel
b
lane
2
vs
lane
1)
and
has
also
been
reported
by
others
(21).
We
also

achieved
a
successful
overexpression
of
Hsp90α
in
Hsp90β
knockdown
HDFs
(panel

b
lane
3
vs
lane
1).
Interestingly,
we
found
that
even
the
overexpression
of
Hsp90α

failed
to
rescue
LRP-‐1
levels
(panel
c
lane
2
vs
lane
3).
To
verify
the
specific

chaperone
function
for
Hsp90β,
not
Hsp90α,
we
also
rescued
Hsp90β’s
expression

in
Hsp90β
knockdown
HDFs
through
lentiviral
overexpression.
The
overexpression

is
shown
in
Fig
3e
(panel
a
lane
2
vs
lane
3).
We
found
that
Hsp90β
overexpression

rescued
LRP-‐1
levels
(panel
c
lane
2
vs
lane
3).
Based
on
these
two
gene
rescue

experiments,
we
concluded
that
Hsp90β
is
the
specific
chaperone
for
LRP-‐1
and

75

even
increasing
the
total
amount
of
Hsp90
proteins
in
the
cell
through
Hsp90α

overexpression
cannot
overcome
the
absence
of
Hsp90β
to
regulate
LRP-‐1
levels.

Figure
3-‐3.
Only
Hsp90β
stabilizes
the
LRP-‐1
receptor.

(A)
HDFs
in
150
mm
dishes
were
serum-‐starved
for
16
hours
and
incubated
with
increasing

amounts
of
RAP
to
block
LRP-‐1’s
extracellular
domain.
Cell
lysates
were
immunoprecipitated
with

anti-‐LRP-‐1
antibody.
Anti-‐LRP-‐1
immunoprecipitates
were
divided
into
three
portions:
45%
for

blotting
with
anti-‐Hsp90α,
45%
for
blotting
with
anti-‐Hsp90β
and
10%
for
blotting
with
anti-‐LRP-‐1

antibodies.

(B)
Hsp90α-‐
or
Hsp90β-‐down
regulated
HDFs
were
examined
for
their
effects
on
expression
of
LRP-‐
1
(panel
a),
PDGFRβ
(panel
b)
and
EGFR
(panel
c),
in
comparison
to
parental
HDFs
(lanes
1)
by

Western
blot
analyses.

(C)
RT-‐PCR
analysis
of
LRP-‐1
mRNA
in
the
cells.

76

(D)
Hsp90β-‐down
regulated
HDFs
were
re-‐infected
with
a
lentiviral
vector
carrying
a
wild
type

Hsp90α
gene.
The
cellular
levels
of
Hsp90β
(panel
a),
Hsp90α
(panel
b)
and
LRP-‐1
(panel
c)
were

examined
by
Western
blot
analysis
with
corresponding
antibodies.

(E)
HDFs
with
Hsp90β
downregulation
were
re-‐infected
with
a
lentiviral
vector
carrying
a
wild
type

Hsp90β
gene.
The
cellular
levels
of
Hsp90β
(panel
a),
Hsp90α
(panel
b)
and
LRP-‐1
(panel
c)
were

examined
by
Western
blot
analysis
with
corresponding
antibodies.

Exogenous
expression
of
Hsp90β
or
LRP-‐1
rescues
migration
defect
of
Hsp90β

knockdown
HDFs
in
response
to
extracellular
Hsp90α
and
hypoxia

To
directly
investigate
if
Hsp90β
regulated
LRP-‐1
levels
contributed
to
hypoxia-‐
driven
migration,
we
subjected
control,
Hsp90β
knockdown,
Hsp90β
knockdown

cells
with
exogenous
Hsp90α
overexpression
and
Hsp90β
knockdown
cells
with

exogenous
Hsp90β
overexpression
to
a
migration
assay
in
response
to
hypoxia
and

recombinant
Hsp90α
stimulation.
Consistent
with
the
rescue
of
LRP-‐1
levels,
only

overexpression
of
Hsp90β
rescued
the
migration
defect
(Fig
4a).

In
order
to
verify
that
LRP-‐1
stabilization
downstream
Hsp90β
is
required
for
HDF

migration,
we
tested
if
overexpression
of
LRP-‐1
was
sufficient
to
rescue
the

migration
defect
of
Hsp90β
knockdown
cells.
The
13-‐kb
cDNA
for
the
human
LRP-‐1

encodes
a
515-‐kDa
extracellular
α
subunit
and
a
transmembrane
85-‐kDa
β
subunit

that
has
a
100-‐amino-‐acid-‐long
cytoplasmic
tail.
It
is
too
large
to
be
accommodated

and
expressed
by
any
existing
mammalian
cDNA
expression
systems
(32,
35).
LRP-‐
1’s
extracellular
domain
is
composed
of
four
independent
ligand-‐binding

subdomains
(I
to
IV).
We
therefore
made
use
of
four
hemagglutinin
(HA)-‐tagged

77

mini-‐LRP-‐1
receptors
(mLRP1-‐I
to
mLRP1-‐IV),
in
which
each
of
the
four

extracellular
subdomains
was
fused
to
the
p85
subunit
gene
and
examined
which

subdomain(s)
was
sufficient
to
mediate
eHsp90α
signaling.
We
identified
mLRP1-‐II

to
be
sufficient
to
bind
and
signal
secreted
Hsp90α’s
pro-‐motility
activity
(22,
23,

13).
We
therefore
overexpressed
mLRP1-‐II
in
Hsp90β
knockdown
HDFs
and
found

that
it
was
sufficient
to
rescue
their
motility
defect
in
response
to
recombinant

Hsp90α.
The
overexpression
is
shown
in
Fig
4b
and
the
migration
quantitation
is

shown
in
Fig
4c.
As
previously
reported
by
us,
PDGF-‐BB
signaling
was
unaffected
by

mLRP1-‐II
overexpression,
implying
PDGF-‐BB
and
secreted
Hsp90α
utilize
different

pathways
to
drive
cell
motility
(13).
Fig
4d
is
a
depiction
of
a
working
model,

indicating
the
differential
roles
of
Hsp90α
and
Hsp90β.

78

Figure
3-‐4.
Exogenously
expressed
Hsp90β
rescues
endogenous
Hsp90β-‐down
regulated

HDF
motility.

(A)
HDFs
with
control
sh-‐LacZ
(bars
1-‐3),
Hsp90β-‐down
regulation
(bars
4-‐6),
Hsp90β-‐down

regulation
and
Hsp90β
re-‐expression
(bars
7-‐9)
or
Hsp90β-‐down
regulation
and
Hsp90α
re-‐
expression
(bars
10-‐12)
were
subjected
to
the
migration
assay
in
response
to
the
stimulations

indicated.
Quantitation
of
the
data
as
migration
index
(MI,
%)
is
shown.
n
=
3,
p
<
0.05

(B)
Overexpression
of
LRP-‐1-‐II
in
Hsp90β-‐down
regulated
cells.

(C)
LRP-‐1-‐II
rescues
the
motility
defect
of
Hsp90β-‐down
regulated
cells
in
response
to
extracellular

Hsp90α
stimulation.

79

(D)
Schematically,
when
tissue
is
damaged,
acute
hypoxia
triggers
cells
in
the
wound
edge
to
secrete

Hsp90α.
Hsp90β
stabilizes
LRP-‐1
and
the
secreted
Hsp90α
binds
and
signals
through
the
LRP-‐1

receptor
to
promote
cell
migration
and
wound
healing.

Discussion

This
study
provides
a
novel
working
relationship
between
extracellular
Hsp90α
and

intracellular
Hsp90β
in
driving
HDF
migration
under
stress
conditions
such
as

hypoxia
and
nutrient
paucity.
Hsp90α
is
secreted
outside
the
cells
under
the
control

of
HIF-‐1α.
The
secreted
Hsp90α
binds
to
the
extracellular
domain
of
LRP-‐1
and

through
activation
of
downstream
kinases
such
as
Akt,
activates
cell
migration
and

wound
healing.
This
process
requires
the
presence
of
intracellular
Hsp90β,
which

binds
to
the
cytoplasmic
tail
of
LRP-‐1
and
stabilizes
it
on
the
cell
surface.
Thus,

Hsp90α
and
Hsp90β
have
distinct
roles
in
wound
healing.

For
decades,
scientists
have
not
distinguished
the
functions
of
the
two
isoforms,

which
were
thought
to
have
redundant
functions.
However,
the
contrasting

knockout
phenotypes
of
Hsp90α
and
Hsp90β
and
the
emerging
roles
of
secreted

Hsp90α
in
tissue
repair
highlight
the
need
to
study
the
two
isoforms
as
independent

entities.
Hsp90β
knockout
mice
were
embryonic
lethal
due
to
their
failure
to
form
a

placental
labyrinth
(24).
Three
papers
report
the
generation
of
Hsp90α
knockout

mice.
Picard’s
group
generated
a
gene
trap
insertion
in
intron
10
of
the
Hsp90α

gene.
However,
instead
of
the
expected
C-‐terminally
truncated
protein,
the
authors

reported
an
Hsp90α
knockout
mouse
(25).
These
mice
developed
normally,
but

80

showed
defects
in
spermatogenesis.
This
observation
was
confirmed
by
a
second

group,
using
the
Cre/LoxP
system
to
remove
exons
9
and
10,
creating
an
Hsp90α

knockout
mouse
(26).
Additionally,
Hsp90α’s
role
in
spermatogenesis
was
also

verified
in
adult
mice
(27).
These
reports
bring
to
light
the
following:
1)
Hsp90β
is

essential
for
life
while
Hsp90α
is
not;
2)
the
tissue-‐specific
functions
of
Hsp90α

cannot
be
compensated
for
by
Hsp90β.
Hsp90α
was
reported
to
have
specific
roles

as
a
chaperone
for
AR
and
promoting
piRNA
biogenesis
(28).
Meanwhile,
the

extracellular
pro-‐motility
function
of
Hsp90α
in
driving
wound
healing
and
tumor

progression
is
becoming
increasingly
popular.
We
reported
that
recombinant

Hsp90α
accelerated
wound
closure
of
acute
and
diabetic
wounds
in
both
mice
(6)

and
pigs
(12).
Interestingly,
extracellular
Hsp90β
has
no
pro-‐motility
activity.
Based

on
these
studies,
we
suggest
that
the
long
recognized
chaperone
function
of
Hsp90

proteins
is
mainly
attributed
to
the
Hsp90β
isoform.
Hsp90α,
on
the
other
hand,
is

more
important
in
the
extracellular
space
where
it
acts,
independent
of
its
ATPase

function,
in
driving
cell
motility
(19).

Subsequent
studies
in
our
lab
identified
that
the
pro-‐motility
activity
of
Hsp90α

involved
the
presence
of
two
lysines
K270
and
K277
in
the
F-‐5
region.
The
lack
of

pro-‐motility
activity
in
Hsp90β
was
due
to
the
replacement
of
these
residues
by

glycine-‐262
and
threonine-‐269
respectively.
Substituting
these
residues
with

lysines
converted
Hsp90β
into
an
Hsp90α-‐like
pro-‐motility
factor.
Thus,
Hsp90α

functions
as
an
extracellular
pro-‐motility
factor
independent
of
its
N-‐terminal

81

ATPase
activity,
but
dependent
on
these
two
lysine
residues
(Jayaprakash
et
al.,
in

submission).

References

1.
Pai,
M.
P.
and
Hunt,
T.
K.
(1972).
Effect
of
varying
oxygen
tensions
on
healing
of

open
wounds.
Surg.
Gynecol.
Obstet.
135,
756-‐758.

2.
O’Toole,
EA.,
Marinkovich,
MP.,
Peavey,
CL.,
Amieva,
MR.,
Furthmayr,
H.,
Mustoe,

TA.
and
Woodley
DT.
(1997)
Hypoxia
increases
human
keratinocyte
motility
on

connective
tissue
.
J.
Clin.
Invest.
100,
2881-‐2891.

3.
Woodley,
DT.,
Fan,
J.,
Cheng,
CF.,
Li,
Y.,
Chen,
M.,
Bu,
G.
and
Li
W.
(2009)

Participation
of
the
lipoprotein
receptor
LRP1
in
hypoxia-‐HSP90alpha
autocrine

signaling
to
promote
keratinocyte
migration.
J
Cell
Sci
122,
1495-‐1498.

4.
Li,
W.,
Li,
Y.,
Guan,
S.,
Fan,
J.,
Cheng,
C.,
Bright,
AM.,
Chinn,
C.,
Chen,
M.
and

Woodley
DT.
(2007)
Extracellular
heat
shock
protein-‐90alpha:
linking
hypoxia
to

skin
cell
motility
and
wound
healing.
EMBO
J
26,
1221-‐1233.

5.
Mogford,
JE.,
Tawil,
N.,
Chen,
A.,
Gies,
D.,
Xia,
Y.
and
Mustoe
TA.
(2002)
Effect
of

age
and
hypoxia
on
TGFbeta1
receptor
expression
and
signal
transduction
in

human
dermal
fibroblasts:
impact
on
cell
migration.
J
Cell
Physiol
190,
259-‐265.

6.
Cheng,
CF.,

Sahu,
D.,
Tsen,
F.,
Zhao,
Z.,
Fan,
J.,
Kim,
R.,
Wang,
X.,
O’Brien,
K.,
Li,
Y.,

Kuang
Y.
et
al.
(2011)
A
fragment
of
secreted
Hsp90α
carries
properties
that
enable

it
to
accelerate
effectively
both
acute
and
diabetic
wound
healing
in
mice.
J
Clin

Invest.
121,
4348-‐4361

82

7.
Botusan,
IR.,
Sunkari,
VG.,
Savu,
O.,
Catrina,
AI.,
Grünler,
J.,
Lindberg,
S.,
Pereira,
T.,

Yla-‐Herttuala,
S.,
Poellinger,
L.,
Brismar,
K.
et
al.
(2008)
Stabilization
of
HIF-‐1alpha

is
critical
to
improve
wound
healing
in
diabetic
mice.
Proc
Natl
Acad
Sci.
USA.

105,19426-‐19431.

8.
Catrina,
SB.,
Okamoto,
K.,
Pereira,
T.,
Brismar,
K.
and
Poellinger,
L.
(2004)

Hyperglycemia
regulates
hypoxia-‐inducible
factor-‐1alpha
protein
stability
and

function.
Diabetes
53,
3226–3232.

9.
Fadini,
GP.,
Sartore,
S.,
Schiavon,
M.,
Albiero,
M.,
Baesso,
I.,
Cabrelle,
A.,
Agostini,

C.,
&
Avogaro,
A.
(2006)
Diabetes
impairs
progenitor
cell
mobilisation
after

hindlimb
ischaemia-‐reperfusion
injury
in
rats.
Diabetologia
49,
3075-‐3084.

10.
Gao,
Z.,
Sasaoka,
T.,
Fujimori,
T.,
Oya,
T.,
Ishii,
Y.,
Sabit,
H.,
Kawaguchi,

M.,
Kurotaki,
Y.,
Naito,
M.,
Wada
T.
et
al.
(2005)
Deletion
of
the
PDGFR-‐beta
gene

affects
key
fibroblast
functions
important
for
wound
healing.
J
Biol
Chem
280,

9375-‐9389.

11.
Sahu,
D.,
Zhao,
Z.,
Tsen,
F.,
Cheng,
CF.,
Park,
R.,
Situ,
AJ.,
Dai,
J.,
Eginli,
A.,
Shams,

S.,
Chen,
M.
et
al.
(2012)
A
potentially
common
peptide
target
in
secreted
heat

shock
protein-‐90α
for
hypoxia-‐inducible
factor-‐1α-‐positive
tumors.
Mol
Biol
Cell

23,
602-‐613

12.
O’Brien,
K.,
Bhatia,
A.,
Tsen,
F.,
Chen,
M.,
Wong,
AK.,
Woodley,
DT.
and
Li
W.

(2014).
Identification
of
the
Critical
Therapeutic
Entity
in
Secreted
Hsp90α
That

Promotes
Wound
Healing
in
Newly
Re-‐Standardized
Healthy
and
Diabetic
Pig

Models.
PLoS
One
9(12):e113956.

83

13.
Tsen,
F.,
Bhatia,
A.,
O'Brien,
K.,
Cheng
CF,
Chen
M,
Hay
N.,
et
al.,
(2013).

Extracellular
heat
shock
protein
90
signals
through
subdomain
II
and
the
NPVY

motif
of
LRP-‐1
receptor
to
Akt1
and
Akt2:
a
circuit
essential
for
promoting
skin
cell

migration
in
vitro
and
wound
healing
in
vivo.
Mol
Cell
Biol.
(24),
4947-‐59

14.
Sims,
J.
D.,
McCready,
J.,
&
Jay,
D.
G.
(2011).
Extracellular
Heat
Shock
Protein

(Hsp)70
and
Hsp90α
Assist
in
Matrix
Metalloproteinase-‐2
Activation
and
Breast

Cancer
Cell
Migration
and
Invasion.
PLoS
ONE,
6(4),
e18848

15.
McCready,
J.,
Sims,
J.
D.,
Chan,
D.,
&
Jay,
D.
G.
(2010).
Secretion
of
extracellular

hsp90α
via
exosomes
increases
cancer
cell
motility:
a
role
for
plasminogen

activation.
BMC
Cancer,
10,
294.

16.
Song,
X.,
Wang,
X.,
Zhuo,
W.,
Shi,
H.,
Feng,
D.,
Sun,
Y.,
et
al.,
(2010).
The
Regulatory

Mechanism
of
Extracellular
Hsp90α
on
Matrix
Metalloproteinase-‐2
Processing
and

Tumor
Angiogenesis.
J
Biol
Chem.
285(51),
40039–40049.

17.
Eustace
BK,
Sakurai
T,
Stewart
JK,
Yimlamai
D,
Unger
C,
Zehetmeier
C.,
et
al.,

(2004).
Functional
proteomic
screens
reveal
an
essential
extracellular
role
for

hsp90
alpha
in
cancer
cell
invasiveness.
Nat
Cell
Biol.
6(6),
507-‐14

18.
El
Hamidieh,
A.,
Grammatikakis,
N.,
&
Patsavoudi,
E.
(2012).
Cell
Surface
Cdc37

Participates
in
Extracellular
HSP90
Mediated
Cancer
Cell
Invasion.
PLoS
ONE,
7(8),

e42722.

19.
Cheng,
C.-‐F.,
Fan,
J.,
Fedesco,
M.,
Guan,
S.,
Li,
Y.,
Bandyopadhyay,
B.,
et
al.,
(2008).

Transforming
Growth
Factor
α
(TGFα)-‐Stimulated
Secretion
of
HSP90α:
Using
the

Receptor
LRP-‐1/CD91
To
Promote
Human
Skin
Cell
Migration
against
a
TGFβ-‐Rich

Environment
during
Wound
Healing.
Mol
Cell
Biol.
28(10),
3344–3358.

84

20.
Bu,
G.
and
Marzolo,
M.
P.
(2000).
Role
of
rap
in
the
biogenesis
of
lipoprotein

receptors.
Trends
Cardiovasc.
Med.
10,
148-‐155.

21.
Chatterjee,
M.,
Jain,
S.,
Stu¨hmer,
T.,
Andrulis,
M.,
Ungethu¨m,
U.,
Kuban,
R.
J.,
et
al.

(2007).
STAT3
and
MAPK
signaling
maintain
overexpression
of
heat
shock
proteins

90a
and
b
in
multiple
myeloma
cells,
which
critically
contribute
to
tumor-‐cell

survival.
Blood
109,
720-‐728.

22.
Obermoeller-‐McCormick
LM,
Li
Y,
Osaka
H,
FitzGerald
DJ,
Schwartz
AL,
Bu
G.

(2001).
Dissection
of
receptor
folding
and
ligand-‐binding
property
with
functional

minireceptors
of
LDL
receptor-‐related
protein.
J.
Cell
Sci.
114,
899–908.

23.
Song
H,
Bu
G.
(2009).
MicroRNA-‐205
inhibits
tumor
cell
migration
through

down-‐regulating
the
expression
of
the
LDL
receptor-‐related
protein
1.
Biochem.

Biophys.
Res.
Commun.
388,
400–405.

24.
Voss,
A.
K.,
Thomas,
T.
and
Gruss,
P.
(2000).
Mice
lacking
HSP90beta
fail
to

develop
a
placental
labyrinth.
Development
127,
1-‐11.

25.
Grad,
I.,
Cederroth,
C.
R.,
Walicki,
J.,
Grey,
C.,
Barluenga,
S.,
Winssinger,
N
et
al.,

(2010).
The
Molecular
Chaperone
Hsp90α
Is
Required
for
Meiotic
Progression
of

Spermatocytes
beyond
Pachytene
in
the
Mouse.
PLoS
ONE,5(12),
e15770.

26.
Imai,
T.,
Kato,
Y.,
Kajiwara,
C.,
Mizukami,
S.,
Ishige,
I.,
Ichiyanagi,
T.,
et
al.,
(2011).

Heat
shock
protein
90
(HSP90)
contributes
to
cytosolic
translocation
of

extracellular
antigen
for
cross-‐presentation
by
dendritic
cells.
Proc.
Natl.
Acad.
Sci.

USA,
108
(39),
16363–16368.

85

27.
Kajiwara,
C.,
Kondo,
S.,
Uda,
S.,
Dai,
L.,
Ichiyanagi,
T.,
Chiba,
T.,
et
al.,
(2012).

Spermatogenesis
arrest
caused
by
conditional
deletion
of
Hsp90α
in
adult
mice.
Biol.

Open,
1(10),
977–982.

28.
Ichiyanagi,
T.,
Ichiyanagi,
K.,
Ogawa,
A.,
Kuramochi-‐Miyagawa,
S.,
Nakano,
T.,

Chuma,
S.,
et
al.,
(2014).
HSP90α
plays
an
important
role
in
piRNA
biogenesis
and

retrotransposon
repression
in
mouse.
Nucleic
Acids
Res.
42(19),
11903–11911.

29.
Sidera
K,
Gaitanou
M,
Stellas
D,
Matsas
R,
Patsavoudi
E.
(2008).
A
critical
role
for

HSP90
in
cancer
cell
invasion
involves
interaction
with
the
extracellular
domain
of

HER-‐2.
J
Biol
Chem.
283(4),
2031-‐41.

30.
Strickland,
DK.,
Ashcom,
JD.,
Williams,
S.,
Burgess,
WH.,
Migliorini,
M.,
&

Argraves
WS.
(1990).
Sequence
identity
between
the
alpha
2-‐macroglobulin

receptor
and
low
density
lipoprotein
receptor-‐related
protein
suggests
that

this
molecule
is
a
multifunctional
receptor.
J
Biol
Chem.
265,
17401-‐4.

31.
Emonard,
H.,
Bellon,
G.,
de
Diesbach,
P.,
Mettlen,
M.,
Hornebeck,
W.,
&
Courtoy
PJ.

(2005).
Regulation
of
matrix
metalloproteinase
(MMP)
activity
by
the
low-‐density

lipoprotein
receptor-‐related
protein
(LRP).
A
new
function
for
an
“old
friend”.

Biochimie

87,
369-‐76

32.
Herz,
J.,
&
Strickland,
DK.
(2001).
LRP:
a
multifunctional
scavenger
and
signaling

receptor.
J
Clin
Invest.
108,
779-‐84.

33.
Basu,
S.,
R.
J.
Binder,
T.
Ramalingam,
&
P.
K.
Srivastava.
(2001).
CD91
is

a
common
receptor
for
heat
shock
proteins
gp96,
hsp90,
hsp70,
and

calreticulin.
Immunity
14,
303–313.

34.
Obermoeller-‐McCormick,
LM.,
Yonghe,
Li.,
Osaka,
H.,
FitzGerald,
DJ.,
Schwartz,

86

AL.,
et
al.,
(2001).
Dissection
of
receptor
folding
and
ligand-‐binding
property
with

functional
minireceptors
of
LDL
receptor-‐related
protein.
J
Cell
Sci.

114,
899-‐908.

35.
Lillis,
AP.,
Mikhailenko,
I.,
and
Strickland,
DK.
(2005).
Beyond
endocytosis:
LRP

function
in
cell
migration,
proliferation
and
vascular
permeability.
J.
Thromb.

Haemost.
3,
1884–1893.

87

Chapter
4:
Breast
Cancer
MDA-‐MB-‐231
Cells
Use
Secreted
Heat
Shock
Protein-‐
90
alpha
(Hsp90α)
to
Survive
a
Hostile
Hypoxic
Environment

Abstract

Rapidly
growing
tumors
consume
large
amounts
of
nutrients
and
oxygen
as
well
as

quickly
outgrow
the
surrounding
blood
supply,
making
them
intrinsically
hypoxic.

However,
despite
being
a
harsh
microenvironment,
hypoxia
correlates
with

increased
tumor
growth
and
progression.
This
implies
that
tumor
cells
adapt
self-‐
sustenance
mechanisms
to
survive
the
hypoxic
tumor
microenvironment.
The
lack

of
blood
supply
to
the
growing
tumor
indicates
that
the
factor(s)
playing
a
role
in

this
“adaptation”
is
secreted
either
by
the
tumor
cells
themselves
or
the
surrounding

stromal
cells.
Constitutive
hypoxia
drives
the
secretion
of
heat
shock
protein-‐90α

(Hsp90α)
in
triple
negative
MDA-‐MB-‐231
cells.
Here,
we
report
that
tumor
cells

secrete
Hsp90α
to
protect
themselves
from
hypoxia-‐driven
cell
death.
Hsp90α

knockout
sensitized
the
cells
to
hypoxia-‐driven
cell
death,
which
could
be
rescued

by
supplementation
with
recombinant
Hsp90α,
but
not
Hsp90β,
protein.

Neutralization
of
secreted
Hsp90α
using
a
monoclonal
antibody
resulted
in

increased
cell
death
of
parental
MDA-‐MB-‐231
cells.
Finally,
secreted
Hsp90α-‐LRP-‐1

autocrine
signaling
was
required
for
protecting
cells
from
hypoxia-‐driven
cell
death.

Thus,
secreted
Hsp90α
is
a
potential
therapeutic
target
to
inhibit
the
growth
of
HIF-‐
1α
overexpressing
tumors.

Introduction

50%
of
tumors
are
hypoxic
and
constitutively
express
HIF1α
(1,
2,
3).
While
normal

88

cells
express
HIF1α
only
under
hypoxic
conditions,
some
tumors
constitutively

express
HIF1α
either
due
to
constant
intratumoral
hypoxia,
over
activated

oncogenes
or
mutated
tumor
suppressor
genes
(2,3).
HIF1α
activates
the

transcription
of
around
800
target
genes,
encompassing
those
contributing
to
tumor

proliferation
and
stem
cell
maintenance
(eg.
TGFα,
IGF2
and
TERT),
(4,5,6)

angiogenesis
(eg.
VEGF
and
SDF-‐1)
as
well
as
tumor
invasion
and
metastasis
(MMPs

and
LOX
family
proteins
that
remodel
the
ECM)
(7,8).
HIF
expression
has
been

correlated
to
resistance
to
chemotherapy
(9)
and
radiotherapy
(10).
HIF-‐1
mediates

the
switch
from
oxidative
phosphorylation
to
glycolysis,
reducing
cellular
ROS
levels

(11),
resulting
in
resistance
to
chemotherapy
(12).

In
addition,
target
genes
of
HIF

include
genes
of
the
ABC
family
of
transporters
that
mediate
efflux
of
cancer
drugs

(13,
14).
HIF1
decreases
expression
of
pro-‐apoptotic
proteins
such
as
Bid
and
Bax
as

well
as
caspases
3,8
and
10
and
increases
expression
of
anti-‐apoptotic
proteins
such

as
Bcl2
and
Birc5
(15-‐19).
There
are
a
number
of
HIF1
inhibitors,
that
target
both

HIF1α
stability
at
the
mRNA
and
protein
levels
(20,
21)
as
well
as
its
transcriptional

activity
by
blocking
its
DNA
binding
(22,
23)
or
transactivation
properties
(24-‐26).

However,
most
of
these
drugs
have
failed
in
clinical
trials,
even
when
used
at
the

maximum
tolerated
dose.
In
addition,
they
work
only
for
some
but
not
all
cancers.

The
failure
of
these
drugs
has
been
attributed
to
1)
not
high
enough
concentration

to
inhibit
HIF1
function
2)
other
compensatory
pathways
in
response
to
drug

treatment
3)
the
targeted
pathway
does
not
contribute
to
HIF’s
action.
Thus,
there
is

a
growing
need
for
identifying
new
targets
that
offer
a
more
viable
alternative
to

HIF
inhibition.

89

Heat
shock
proteins
of
Hsp90
family
have
been
widely
studied
as
intracellular

chaperone
proteins
that
help
a
wide
array
of
client
proteins
fold
into
their
mature,

functional
states.
They
have
been
found
to
be
either
qualitatively
overactive
or

quantitatively
overexpressed
in
a
wide
range
of
cancers
(27-‐30).
Since
Hsp90

stabilizes
a
variety
of
growth
factors
and
oncogenes,
targeting
Hsp90
offered
a
good

therapeutic
approach
to
overcome
drug
resistance
in
response
to
a
single
drug
in

cancer
cells.
Ansamycin
family
drugs
such
as
geldanamycin
and
its
derivatives,
17-‐
AAG
and
the
water-‐soluble
17-‐DMAG
target
the
N-‐terminal
ATPase
of
Hsp90

proteins,
inhibiting
its
chaperone
function.
However,
these
drugs
present
severe

hepatotoxicity
and
none
of
them
have
cleared
clinical
trials.
Over
the
last
decade,
a

new
role
and
cellular
localization
for
Hsp90
proteins,
especially
Hsp90α
have

emerged-‐
the
secreted
pool
of
Hsp90α.
We
reported
that
normal
cells
secrete

Hsp90α
under
stress
conditions
such
as
hypoxia
(31,
32)
and
serum-‐derived
growth

factors
such
as
TGFα
(33).
Hsp90α
is
secreted
constitutively
by
a
number
of
cancer

cells
including
HT-‐1080
fibrosarcoma
cells
and
MDA-‐MB-‐231
breast
cancer
cells

(34-‐36);
MCF-‐7
breast
cancer
cells
(36),
HCT-‐8
colorectal
cancer
cells
(37),
T24

bladder
cancer
cells,
B16
melanoma
cells
and
PC3
prostate
cancer
cells
(38),
SKBR3,

MDA-‐MB-‐453,
MDA-‐MB
435
and
MDA-‐MB-‐468
breast
cancers,
CaoV-‐3
ovarian

cancer
and
HepG2
hepatoma
(34,
36,
39),
A172
glioblastoma
and
SUM159
breast

cancer
(34).
The
main
role
of
secreted
Hsp90α,
reported
so
far,
is
to
drive
tumor
cell

invasion
and
metastasis
(40).

90

In
this
study,
we
report
a
new
role
for
secreted
Hsp90α
as
a
survival
factor
for
MDA-‐
MB-‐231
breast
cancer
cells
under
hypoxia.
Constitutive
HIF1α
expression
drives

constitutive
secretion
of
Hsp90α
in
these
cells.
The
secreted
Hsp90α
acts
via
the

LRP-‐1
receptor
to
protect
the
cells
from
hypoxia-‐driven
cell
death.

Results

Selection
of
a
suitable
cell
line
as
the
model
of
study

Breast
cancers
are
highly
heterogenous
and
are
broadly
classified
into
ER+
PR+,

ER+PR+HER2+
(triple
positive)
and
ER-‐PR-‐HER2-‐
(triple
negative).
In
order
to

identify
the
best
cell
model
for
our
study,
we
investigated
the
intracellular
levels
of

Hsp90α
and
Hsp90β
as
well
as
that
of
LRP-‐1,
the
cell
surface
receptor
for
secreted

Hsp90α.
This
highlighted
the
heterogeneity
of
breast
cancers
with
all
cell
lines

expressing
comparable
levels
of
Hsp90α
and
Hsp90β
except
MDA-‐MB-‐468,
which

expressed
significantly
lesser
levels
of
Hsp90β
(Fig
1a).
LRP-‐1
levels
varied

significantly
amongst
the
cell
lines
with
three
of
the
eight
cell
lines
screened

expressing
no
LRP-‐1,
MDA-‐MB-‐231
and
T47D
expressing
intermediate
levels
and

HS-‐578T
expressing
the
highest
levels
of
LRP-‐1
(Fig.
1d).
We
also
found
differing

levels
of
secretion
with
HS-‐578T
secreting
undetectable
levels
of
both
Hsp90α
and

Hsp90β,
Skbr3
secreting
only
Hsp90α
and
MDA-‐MB-‐231
secreting
both
Hsp90α
and

Hsp90β
(Fig
1b).
We
have
previously
reported
that
constitutive
HIF1α
expression
in

MDA-‐MB-‐231
cells
drives
the
secretion
of
Hsp90α
that
binds
to
cell
surface
LRP-‐1

receptor
and
this
autocrine
signaling
pathway
is
crucial
for
tumor
cell
invasion
in

vitro
and
tumor
metastasis
in
vivo
(39).
In
line
with
this
finding,
we
found
that
cell

91

lines
lacking
this
pathway
failed
to
invade
in
a
Matrigel
invasion
assay
(Fig
1c).
For

example,
though
HS-‐578T
cells
expressed
the
highest
LRP-‐1
levels,
they
did
not

secrete
both
Hsp90α
and
Hsp90β
and
were
therefore
unable
to
invade
through
the

Matrigel
barrier.
MDA-‐MB-‐468
cells
secreted
Hsp90α
and
Hsp90β,
but
lacked
LRP-‐1

and
failed
to
invade
in
vitro.
Previously,
we
also
reported
that
MDA-‐MB-‐468
cells

fail
to
form
tumors
in
vivo
(39).
The
only
exception
was
T47D
cells
which
expressed

comparable
levels
of
secreted
Hsp90α
and
Hsp90β
and
LRP-‐1
as
MDA-‐MB-‐231
but

failed
to
invade,
highlighting
that
secreted
Hsp90α
might
not
be
used
by
all
cancers

to
invade.
An
alternative
possibility
is
that
for
certain
cancers,
secreted
Hsp90α
is

necessary,
but
not
sufficient
for
invasion.
Based
on
these
findings,
we
decided
to
use

MDA-‐MB-‐231
as
the
cell
model
for
our
study.
We
found
that
exposure
of
these
cells

to
hypoxia
(1%
O2)
for
16
hours
resulted
in
an
increase
of
both
intracellular
and

secreted
Hsp90α
and
Hsp90β.

92

Figure
4-‐1.
Selection
of
MDA-‐MB-‐231
breast
cancer
cell
line
as
the
model
of
study

Western
blots
for
Hsp90α
and
Hsp90β
in
total
lysates
(TL)
(A)
or
conditioned
media
(100x)
(CM)
(B)

and
the
invasiveness
(C)
of
the
indicated
breast
cancer
and
control
cell
lines.
Western
blots
for
LRP-‐1

receptor
among
cell
lines
(D)
and
the
total
(E)
and
the
secreted
(F)
Hsp90α
and
Hsp90β
under

hypoxia
in
MDA-‐MB-‐231.
Intracellular
(G)
and
secreted
(H)
ratios
between
Hsp90α
and
Hsp90β
in

MDA-‐MB-‐231
cells.

93

Generation
of
CRISPR-‐Cas9
knockout
of
Hsp90α
in
MDA-‐MB-‐231

In
order
to
investigate
the
role
of
Hsp90α
in
cell
survival
under
hypoxia,
we

generated
CRISPR-‐Cas9
knockout
of
Hsp90α
in
MDA-‐MB-‐231
cells.
We
verified
the

knockout
at
three
levels—DNA,
mRNA
and
sequencing
levels.
We
found
that

compared
to
parental
MDA-‐MB-‐231
cells,
Hsp90α
mRNA
was
slightly
down

regulated
in
the
knockout
cells
(Figure
2A).
This
decrease
in
mRNA
levels
in

CRISPR/Cas-‐9
knockout
cells
is
consistent
with
other
reports
using
this
technology.

We
found
complete
depletion
of
Hsp90α
protein
by
the
CRISPR/Cas9
system

(Figure
2B).
Interestingly,
there
was
an
up
regulation
of
Hsp90β
in
Hsp90α

knockout
cells
and
has
also
been
reported
by
others
(41).
Consistently,
secretion
of

Hsp90α
was
abolished
(Figure
2C
panel
a
lane
5
vs
lane
4)
while
that
of
Hsp90β
was

unaffected
(Figure
2C
panel
b
lane
5
vs
lane
4).
Sequencing
of
the
Cas9
target
site
in

Hsp90α
gene
revealed
the
presence
of
a
premature
stop
codon,
contributing
to
the

lack
of
protein
in
the
knockout
cells
(Figure
2D
and
2E).
The
target
site
is
located
in

exon
1
of
Hsp90α
gene,
explaining
the
lack
of
detection
of
even
a
truncated
protein.

We
found
that
the
knockout
did
not
affect
the
cellular
morphology
(Fig
2F)
or
the

proliferation
profiles
of
the
cells
(Fig
2G).

94

Figure
4-‐2.
Generation
of
Hsp90α
knockout
MDA-‐MB-‐231
cells

(A) RT-‐PCR
from
parental
and
Hsp90α
KO
MDA-‐MB-‐231
cells.
cDNA
was
synthesized
by
the

SuperScript®
III
RT
kit(Invitrogen)
from
total
RNA
isolated

from
the
two
cell
lines
.
PCR

was
set
up
for
Hsp90α
using
the
primers
5’-‐ATGCCCCCGTGTTCG-‐3’
(sense)
and
5’-‐
CTGAAAGGCGAACGTCTC-‐3’
(antisense)
and
GAPDH
using
primers
GAPDH
5′-‐
CCATCACCATC-‐TTCCAGGAG-‐3′
(sense)
and
5′-‐CCTGCTTCACCACCTTCTTG-‐3′
(antisense).

(B) Western
blotting
for
Hsp90α
and
Hsp90β
in
lysates
of
parental
and
Hsp90α
KO
MDA-‐MB-‐
231
cells

(C) Western
blotting
for
Hsp90α
and
Hsp90β
in
conditioned
media
from
parental
and
Hsp90α

KO
MDA-‐MB-‐231
cells

(D) &
(E)
Sequencing
of
CRISPR-‐Cas9
target
site
from
parental
and
Hsp90α
KO
MDA-‐MB-‐231

cells.
2μl
of
the
RT-‐PCR
from
MDA-‐MB-‐231
parental
and
Hsp90α
KO
cells
was
ligated
into

95

the
TOPO
vector
using
the
TOPO
TA
cloning
kit
following
manufacturer’s
protocol

(Invitrogen).
Positive
colonies
were
subjected
to
sequencing
using
the
M13
reverse
primer.

(F) Morphology
images
of
MDA-‐MB-‐231
parental
and
Hsp90α
KO
cells

(G)
Proliferation
profiles
of
MDA-‐MB-‐231
parental
and
Hsp90α
KO
cells

Hsp90α
knockout
sensitizes
MDA-‐MB-‐231
cells
to
hypoxia-‐driven
cell
death

We
initially
investigated
the
viability
of
parental
MDA-‐MB-‐231
cells
under
varying

concentrations
of
oxygen,
ranging
from
20%
(normoxia)
down
to
varying
levels
of

hypoxia
up
to
0%
(Fig
3A
and
B).
Results
of
calcein-‐AM
and
ethidium
homodimer-‐1

staining
to
detect
live
and
dead
cells
respectively
revealed
that
cell
viability
was

comparable
under
normoxia
and
hypoxia
upto
1%
O2.
Decreasing
the
oxygen

content
further
to
0.5%
and
0%
significantly
reduced
the
cell
viability.

We
then
used
knockout
cells
to
investigate
their
response
to
hypoxia-‐driven
cell

death.
We
found
that
Hsp90α
knockout
cells
survived
normally
under
both

normoxia
and
2%
oxygen
(hypoxia)
conditions.
However,
there
was
a
drastic
drop

in
cell
viability
when
the
oxygen
content
was
lowered
to
1%
and
lower.
As
shown
in

Fig
3C
and
D,
1%
oxygen
resulted
in
>50%
reduction
in
cell
viability,
with
0.5%
and

0%
resulting
in
further
decline.
We
also
found
1%
oxygen
exposure
for
48
hours
led

to
a
statistically
significant
difference
in
cell
viability
between
MDA-‐MB-‐231

parental
and
Hsp90α
knockout
cells
(p<
0.05,
n=3).

96

Figure
4-‐3.
CRISPR-‐cas9
knockout
of
Hsp90α
sensitizes
MDA-‐MB-‐231
cells
to
hypoxia-‐driven

killing.

(A)
Cell
viability
by
fluorescence
microscopy
(panels
a
to
e)
and
flow
cytometry
(panels
a’
to
e’).
(B)

Quantitation
of
viability
data.
(C)
Viability
of
Hsp90α-‐knockout
cells
under
normoxia
or
various

degrees
of
hypoxia
(D)
Quantitation
of
viability
data.
n
=
3,
*
p
<
0.05.

Secreted
Hsp90α,
not
Hsp90β,
protects
MDA-‐MB-‐231
cells
from
hypoxia-‐
mediated
cell
death

Hsp90α
knockout
depletes
both
intracellular
and
extracellular
pools
of
the
protein.

As
mentioned
above
(Fig
1F),
hypoxia
drives
the
secretion
of
both
Hsp90α
and

Hsp90β.
In
order
to
distinguish
which
pool
contributes
to
protecting
cells
from

hypoxia-‐driven
cell
death,
we
supplemented
Hsp90α
knockout
cells
with
either

97

recombinant
Hsp90α
or
recombinant
Hsp90β
and
subjected
them
to
varying

concentrations
of
oxygen.
As
shown
in
Figure
4A,
normoxia
did
not
affect
cell

viability
of
Hsp90α
knockout
cells
(panels
a
and
a’).
Treatment
with
recombinant

Hsp90α
or
Hsp90β
did
not
alter
the
viability
(panels
b,
b’
and
c,
c’).
However

hypoxia
caused
a
dramatic
decrease
in
viability
of
Hsp90α
knockout
cells
(panels
d

and
d’).
Interestingly,
treatment
with
recombinant
Hsp90α
(panels
e
and
e’
vs
d
and

d’),
but
not
recombinant
Hsp90β
(panels
f
and
f’
vs
d
and
d’),
rescued
cell
viability.

The
quantitation
is
shown
in
Figure
4B
and
the
proteins
used
for
the
rescue

experiments
are
shown
via
Coomassie
blue
staining
in
Figure
4C.
This
implies
that

tumor
cells
use
secreted
Hsp90α
and
not
Hsp90β
to
protect
themselves
from

hypoxia-‐driven
cell
death.
The
secreted
and
not
the
intracellular
chaperone
function

of
Hsp90α
is
therefore
the
driver
of
this
process.

We
next
investigated
the
mechanism
by
which
secreted
Hsp90α
mediates
tumor
cell

survival
under
hypoxia.
We
focused
on
the
cell
surface
receptor
LRP-‐1
(low
density

lipoprotein
receptor-‐related
protein-‐1).
We
have
previously
shown
that
in
both

normal
skin
cells
and
in
MDA-‐MB-‐231
cells,
secreted
Hsp90α
binds
to
LRP-‐1
and

drives
tumor
cell
invasion
in
vitro
and
in
vivo.
Inhibition
of
LRP-‐1
either
through

lentiviral
knockdown
or
via
a
universal
antagonist
of
its
ligand-‐binding
domain
led

to
decreased
cell
motility
in
vitro
and
a
reduction
in
both
wound
healing
and
tumor

metastasis
(31,
33,
39).
We
hypothesized
that
secreted
Hsp90α
used

transmembrane
signaling
through
LRP-‐1
to
mediate
its
pro-‐survival
function
as

well.
To
this
end,
we
generated
stable
knockdown
of
LRP-‐1
in
MDA-‐MB-‐231
cells

98

(Fig
4D).
We
subjected
these
cells
to
either
normoxia
or
hypoxia
(1%
O2)
with
or

without
exogenous
supplementation
with
recombinant
Hsp90α
protein.
Hypoxia
led

to
a
dramatic
reduction
in
cell
viability
of
LRP-‐1
down
regulated
cells
(Fig

4E
panels

b
and
b’
vs
a
and
a’)
similar
to
the
phenotype
observed
in
Hsp90α
knockout
cells.

Interestingly,
in
contrast
to
Hsp90α
knockout
cells
whose
viability
was
rescued
by

exogenous
supplementation
with
Hsp90α
protein,
the
viability
of
LRP-‐1
down

regulated
cells
was
not.
This
implies
that
LRP-‐1
was
indispensable
for
MDA-‐MB-‐231

cell
survival
under
hypoxia
and
secreted
Hsp90α
uses
signaling
through
LRP-‐1

receptor
to
protect
tumor
cells
under
hypoxia.

99

Figure
4-‐4.
Rescue
of
Hsp90α-‐knockout
cells
from
hypoxia-‐driven
killing
by
extracellular

Hsp90α,
but
not
Hsp90β,
protein
via
LRP-‐1
receptor
signalling

(A) Cell
viability
data
showing
representative
images
and
FACS
plots
of
Hsp90α
knockout
cells
under

normoxia
and
hypoxia
with
extracellular
supplementation
of
Hsp90α
and
Hsp90β.

(B) Quantitation
of
viability
data
from
three
independent
experiments.

100

(C) Coomassie
blue
stained
SDS-‐PAGE
gel
showing
the
purified
proteins
used
in
the
rescue

experiments.

(D) Down-‐regulation
of
LRP-‐1
shown
by
Western
blot.

(E) Cell
viability
data
showing
representative
images
and
FACS
plots
using
LRP-‐1
down
regulated

cells.
The
cell
viability
percentage
is
indicated
on
the
figure.

(F) Quantitation.
n=3,
*
p
<
0.05.

Design
of
a
neutralizing
antibody
against
secreted
Hsp90α

In
order
to
further
verify
the
requirement
of
secreted
Hsp90α
in
protecting
tumor

cells
from
hypoxia-‐driven
cell
death,
we
used
an
antibody
neutralization
approach.

We
generated
a
monoclonal
antibody,
1G6-‐D7,
targeting
the
F-‐5
region
of
secreted

Hsp90α.
F-‐5
is
a
115
amino
acid
peptide
that
recapitulates
the
full
pro-‐motility

activity
of
the
full
length
Hsp90α
protein
(42).
Fig
5A
shows
the
location
of
F-‐5
in

Hsp90α
against
which
1G6-‐D7
was
designed.
1G6-‐D7
was
able
to
pull
down
His-‐
tagged
recombinant
F-‐5
fragment
(Figure
5B).
In
addition,
using
co-‐
immunoprecipitation,
we
proved
that
1G6-‐D7
was
capable
of
binding
to

endogenously
secreted
Hsp90α
in
the
conditioned
media
of
MDA-‐MB-‐231
cells.
As

shown
in
Figure
5C,
1G6-‐D7
bound
strongly
to
secreted
Hsp90α
and
weakly
to

secreted
Hsp90β.
Functionally,
1G6-‐D7
was
effective
in
decreasing
Hsp90α-‐driven

MDA-‐MB-‐231
cell
invasion
(Figure
5D
panels
b,
c
vs
a,
d)
and
migration
(Figure
5E

panel
d
vs
panel
a)
in
vitro
and
tumor
formation
in
mice
(Figure
5F).
The
inhibitory

effect
of
1G6-‐D7
on
invasion
and
migration
was
rescued
by
F-‐5
addition
(Figure
5D

panel
e
vs
panels
f
&
g;
Figure
5E
panel
d
vs
panel
e).
Neutralization
of
secreted

101

Hsp90β
by
an
anti-‐Hsp90β
antibody
did
not
inhibit
tumor
cell
migration
(Figure
5E

panel
c
vs
panel
a).

Figure
4-‐5.
Monoclonal
antibody,
1G6-‐D7,
binds
to
secreted
Hsp90α
and
neutralizes
its

function
in
vitro
and
in
vivo.

(A) Location
of
the
antigen
F-‐5
in
Hsp90α
against
which
1G6-‐D7
is
targeted

(B) Immunoprecipitation
showing
1G6-‐D7
binding
to
recombinant
F-‐5
peptide

(C) Co-‐
immunoprecipitation
using
1G6-‐D7
antibody
showing
the
ability
of
the
antibody
to
bind

endogenous
secreted
Hsp90α

(D) Representative
images
from
an
invasion
assay
showing
that
1G6-‐D7
inhibits
basal
invasion

of
MDA-‐MB-‐231
cells,
which
is
rescued
by
addition
of
increasing
amounts
of
F-‐5.

(E) Representative
images
from
a
migration
assay
showing
the
effect
of
1G6-‐D7
vs
control

mouse
IgG
and
an
Hsp90β
neutralizing
antibody.
The
black
tracks
outlined
by
white
circles

are
an
indication
of
cell
motility.

102

(F
&
G)
Tumor
growth
as
measured
over
4
weeks
from
the
time
MDA-‐MB-‐231
tumor
cells
were

injected
into
nude
mice.
1G6-‐D7
(red)
is
compared
to
normal
mouse
IgG
(black)
as
a
control.

Neutralization
of
secreted
Hsp90α
by
1G6-‐D7
sensitizes
parental
MDA-‐MB-‐
231
cells
to
hypoxia-‐driven
cell
death

We
hypothesized
that
if
secreted
Hsp90α
was
indispensable
for
tumor
cell
survival

under
hypoxia,
then
treatment
with
1G6-‐D7
will
make
parental
MDA-‐MB-‐231
cells

more
susceptible
to
hypoxia-‐induced
cell
death.
We
treated
parental
cells
with

either
control
mouse
IgG
or
1G6-‐D7
and
subjected
them
to
either
normoxia
or

hypoxia
(1%
O2)
for
48h.

As
shown
in
Figure
6A,
while
control
IgG
treatment
under

hypoxia
did
not
affect
cell
viability
(panels
c
and
c’),
treatment
with
1G6-‐D7
led
to
a

dramatic
reduction
in
cell
viability
under
hypoxia
(panels
d
and
d’),
again
proving

that
the
secreted
pool
of
Hsp90α
contributes
to
tumor
cell
survival
in
the
hypoxic

tumor
microenvironment.
Also,
addition
of
F-‐5
overcame
this
loss
of
cell
viability

implying
that
the
inhibition
mediated
by
1G6-‐D7
was
due
to
its
neutralization
of

secreted
Hsp90α
function
(panels
e
and
e’
vs
d
and
d’).
The
quantitation
is
shown
in

Figure
6B.
A
working
model
explaining
our
findings
is
depicted
in
Figure
6C.
Under

hypoxia
in
the
tumor
microenvironment,
HIF1α
levels
are
stabilized,
which
drives

Hsp90α
secretion.
The
secreted
Hsp90α
binds
to
the
cell
surface
LRP-‐1
receptor
and

possibly
through
activating
Akt,
mediates
tumor
cell
survival,
invasion
and

metastasis.

103

Figure
4-‐6.
mAb
1G6-‐D7
neutralizes
secreted
Hsp90α
function
and
renders
MDA-‐MB-‐231
cells

susceptible
to
hypoxia-‐driven
cell
death

(A) Cell
viability
data
showing
representative
images
from
fluorescence
microscopy
and
FACS

plots
of
parental
MDA-‐MB-‐231
cells
under
normoxia
or
hypoxia
with
1G6-‐D7
treatment.

(B) Quantitation
of
cell
viability

(C) Model
showing
the
requirement
of
HIF1α-‐secreted
Hsp90α-‐LRP-‐1
autocrine
loop
in
cellular

adaptation
to
hypoxia

Discussion

Approximately
50%
of
solid
tumors
are
hypoxic
and
constitutively
express
HIF-‐1α,

regardless
of
the
oxygen
content
(43).
HIF-‐1α
is
an
upstream
regulator
of
Hsp90α

secretion
in
both
normal
(31,
32,
33)
and
tumor
cells
(39).
We
have
previously

shown
that
constitutive
secretion
of
Hsp90α
contributes
to
MDA-‐MB-‐231
cell

migration
and
invasion
in
vitro
and
tumor
formation
in
mice
(39).
Here,
we
report
a

new
function
of
secreted
Hsp90α
in
protecting
tumor
cells
from
hypoxia-‐driven
cell

death.
Constitutive
HIF-‐1α
expression
in
MDA-‐MB-‐231
cells
drives
constitutive

104

secretion
of
both
Hsp90α
and
Hsp90β.
Exogenous
supplementation
of
Hsp90α

knockout
cells
with
recombinant
Hsp90α
protein,
but
not
Hsp90β
protein,
rescued

cell
viability,
while
neutralizing
endogenously
secreted
Hsp90α
made
parental

MDA-‐MB-‐231
cells
more
susceptible
to
hypoxia-‐driven
killing.
Similar
to
the

requirement
of
LRP-‐1
receptor
in
mediating
secreted
Hsp90α’s
pro-‐motility

function,
it
was
also
important
for
its
pro-‐survival
function.
Thus,
MDA-‐MB-‐231
cells

use
secreted
Hsp90α
to
thrive
in
the
harsh
hypoxic
tumor
microenvironment
and

subsequently
invade
and
metastasize.

Hypoxia
in
tumors
is
correlated
with
increased
invasiveness
(44)
and
HIF-‐1α
levels

in
tumors
are
correlated
with
increased
mortality
in
cervical
(45),
lung
(46),
breast

(47,48)
and
ovarian
cancers
(49).
HIF-‐1α
is
a
master
transcription
factor
that

triggers
expression
of
a
number
of
genes
involved
in
driving
tumor
angiogenesis

(eg.
VEGF),
tumor
metabolism
(eg.
GLUT1)
and
therefore,
tumor
metastasis.
HIFs

are
crucial
for
tumor
progression
in
both
autochthonous
and
orthotopic
models
of

breast
cancer.
HIFs
also
drive
secretion
of
various
proteins
via
the
unconventional

exosome
trafficking
pathway
that
mediate
tumor
progression.
However,
targeting

HIF
directly
has
been
tough
since
it
is
an
intracellular
protein
(43).
A
better
strategy

will
be
to
target
downstream
effectors
of
HIF-‐mediated
tumor
progression.
Secreted

Hsp90α
is
one
such
molecule.
Secreted
Hsp90α
is
an
important
driver
of
metastasis

in
different
types
of
cancer
such
as
breast
(35,36,39),
colon
(37),
prostate
(38)
and

ovarian
(36),
to
name
a
few.
In
this
study,
we
identify
that
in
addition
to

contributing
to
tumor
metastasis,
secreted
Hsp90α
also
plays
a
role
in
tumor
cell

105

survival
under
hypoxia.
Hence
targeting
secreted
Hsp90α
with
a
neutralizing

antibody
such
as
1G6-‐D7
will
help
contain
both
primary
tumor
formation
(Figure
5)

and
inhibit
metastasis
(Zou
et
al.,
in
submission).

References:

1. Semenza,
G.L.,
(2007).
Evaluation
of
HIF-‐1
inhibitors
as
anticancer
agents.

Drug
Discov.
Today
12,
853–859.

2. Semenza,
G.L.,
(2012a).
Hypoxia-‐inducible
factors
in
physiology
and

medicine.
Cell
148,
399–408.

3. Semenza,
G.L.,
(2012b).
Molecular
mechanisms
mediating
metastasis
of

hypoxic
breast
cancer
cells.
Trends
Mol.
Med.
18,
534–543.

4. Gunaratnam
L.,
Morley
M.,
Franovic
A.,
de
Paulsen
N.,
Mekhail
K.,
Parolin
DA.,

et
al.
(2003).
Hypoxia
inducible
factor
activates
the
transforming
growth

factor-‐α/epidermal
growth
factor
receptor
growth
stimulatory
pathway
in

VHL–/–
renal
cell
carcinoma
cells.
J.
Biol.
Chem.
278,
44966–44974.

5. Zhang
L.,
Zhou
W.,
Velculescu
VE.,
Kern
SE.,
Hruban
RH.,
Hamilton
SR.,
et
al.

Gene
expression
profiles
in
normal
and
cancer
cells.
Science.
276:1268–1272.

6. Feldser
D,
Agani
F,
Iyer
NV,
Pak
B,
Ferreira
G,
Semenza
GL.
(1999).
Reciprocal

positive
regulation
of
hypoxia-‐inducible
factor
1α
and
insulin-‐like
growth

factor
2.
Cancer
Res.
59,
3915–3918.

7. Erler
JT,
et
al.
(2006).
Lysyl
oxidase
is
essential
for
hypoxia-‐induced

metastasis.
Nature.
440,
1222–1226.

106

8. Erler
JT.,
Bennewith
KL.,
Nicolau
M.,
Dornhöfer
N.,
Kong
C.,
Le
QT.,
et
al.

(2011).
Hypoxia-‐inducible
factor
1
is
a
master
regulator
of
breast
cancer

metastatic
niche
formation.
Proc.
Natl.
Acad.
Sci.
USA.
108,
16369–16374.

9. Rohwer
N.,
Cramer
T.
(2011).
Hypoxia-‐mediated
drug
resistance:
novel

insights
on
the
functional
interaction
of
HIFs
and
cell
death
pathways.
Drug

Resist
Updat.
14,
191–201.

10. Moeller
BJ.,
Richardson
RA.,
Dewhirst
MW.
(2007).
Hypoxia
and

radiotherapy:
opportunities
for
improved
outcomes
in
cancer
treatment.

Cancer
Metastasis
Rev.
26,
241–248

11. Semenza
GL.
(2009).
Regulation
of
metabolism
by
hypoxia-‐inducible
factor
1.

Semin
Cancer
Biol.
19(1),
12-‐6

12. Santamaría
G.,
Martínez-‐Diez
M.,
Fabregat
I.,
Cuezva
JM.
(2006).
Efficient

execution
of
cell
death
in
non-‐glycolytic
cells
requires
the
generation
of
ROS

controlled
by
the
activity
of
mitochondrial
H+-‐ATP
synthase.
Carcinogenesis.

27,
925–935.

13. Comerford
KM.,
Wallace
TJ.,
Karhausen
J.,
Louis
NA.,
Montalto
MC.,
Colgan
SP.

(2002).
Hypoxia-‐inducible
factor-‐1-‐dependent
regulation
of
the
multidrug

resistance
(MDR1)
gene.
Cancer
Res.
62,
3387–3394.

14. Krishnamurthy
P.,
Ross
DD.,
Nakanishi
T.,
Bailey-‐Dell
K.,
Zhou
S.,
Mercer

KE.,
et
al.
(2004).
The
stem
cell
marker
Bcrp/ABCG2
enhances
hypoxic
cell

survival
through
interactions
with
heme.
J.
Biol.
Chem.
279,
24218–24225.

15. Erler
JT,
et
al.
(2004)
Hypoxia-‐mediated
down-‐regulation
of
Bid
and
Bax
in

tumors
occurs
via
hypoxia
inducible
factor
1-‐dependent
and
-‐independent

107

mechanisms
and
contributes
to
drug
resistance.
Mol.
Cell.
Biol.
24:2875–
2889.

16.
Peng
XH,
et
al.
(2006)
Cross-‐talk
between
epidermal
growth
factor
receptor

and
hypoxia-‐inducible
factor-‐1α
signal
pathways
increases
resistance
to

apoptosis
by
up-‐regulating
survivin
gene
expression.
J.
Biol.
Chem.
281,

25903–25914.

17. Brown
LM,
Cowen
RL,
Debray
C,
Eustace
A,
Erler
JT,
Sheppard
FC.
(2006).

Reversing
hypoxic
cell
chemoresistance
in
vitro
using
genetic
and
small

molecule
approaches
targeting
hypoxia-‐inducible
factor
1.
Mol.
Pharmacol.

69,
411–418.

18. Liu,
L.,
Ning,
X.,
Sun,
L.,
Zhang,
H.,
Shi,
Y.,
Guo,
C.,
et
al.
(2008).
Hypoxia-‐
inducible
factor-‐1α
contributes
to
hypoxia-‐induced
chemoresistance
in

gastric
cancer.
Cancer
Science,
99,
121–128

19. Flamant,
L.,
Notte,
A.,
Ninane,
N.,
Raes,
M.,
Michiels,
C.
(2010).
Anti-‐apoptotic

role
of
HIF-‐1
and
AP-‐1
in
paclitaxel
exposed
breast
cancer
cells
under

hypoxia.
Molecular
Cancer,
9,
191.

20. Terzuoli,
E.,
Puppo,
M.,
Rapisarda,
A.,
Uranchimeg,
B.,
Cao,
L.,
Burger,
A.
M.,
…

(2010).
Aminoflavone,
a
ligand
of
the
Aryl
Hydrocarbon
Receptor
(AhR),

inhibits
HIF-‐1α
expression
in
an
AhR-‐independent
fashion.
Cancer

Research,
70(17),
6837–6848.

21. Shackelford,
D.
B.,
Vasquez,
D.
S.,
Corbeil,
J.,
Wu,
S.,
Leblanc,
M.,
Wu,
C.-‐L.,
et
al.

(2009).
mTOR
and
HIF-‐1α-‐mediated
tumor
metabolism
in
an
LKB1
mouse

model
of
Peutz-‐Jeghers
syndrome.
PNAS,
106(27),
11137–11142.

108

22. Lee,
K.,
Qian,
D.
Z.,
Rey,
S.,
Wei,
H.,
Liu,
J.
O.,
&
Semenza,
G.
L.
(2009).

Anthracycline
chemotherapy
inhibits
HIF-‐1
transcriptional
activity
and

tumor-‐induced
mobilization
of
circulating
angiogenic
cells.
PNAS,
106(7),

2353–2358.

23. Wang,
Y.,
Liu,
Y.,
Malek,
S.
N.,
Zheng,
P.,
&
Liu,
Y.
(2011).
Targeting
HIF1α

eliminates
cancer
stem
cells
in
hematological
malignancies.
Cell
Stem
Cell,

8(4),
399–411.

24. Molineaux
SM.
Molecular
pathways:
targeting
proteasomal
protein

degradation
in
cancer.
(2012).
Clin.
Cancer
Res.
18,
15–20.

25. Kaluz,
S.,
Kaluzová,
M.,
&
Stanbridge,
E.
J.
(2006).
Proteasomal
Inhibition

Attenuates
Transcriptional
Activity
of
Hypoxia-‐Inducible
Factor
1
(HIF-‐1)
via

Specific
Effect
on
the
HIF-‐1α
C-‐Terminal
Activation
Domain.
Molecular
and

Cellular
Biology,
26(15),
5895–5907.

26. Befani
CD,
Vlachostergios
PJ,
Hatzidaki
E,
Patrikidou
A,
Bonanou
S,
Simos

G,
et
al.,
(2015).
Bortezomib
enhances
the
radiosensitivity
of
hypoxic
cervical

cancer
cells
by
inhibiting
HIF-‐1α
expression.
International
Journal
of
Clinical

and
Experimental
Pathology,
8(8),
9032–9041.

27. Workman,
P.,
Burrows,
F.,
Neckers,
L.
&
Rosen,
N.
(2007).
Drugging
the

cancer
chaperone
HSP90:
combinatorial
therapeutic
exploitation
of
oncogene

addiction
and
tumor
stress.
Ann
N
Y
Acad
Sci
1113,
202–216

28. Trepel,
J.,
Mollapour,
M.,
Giaccone,
G.
&
Neckers,
L.
(2010).
Targeting
the

dynamic
HSP90
complex
in
cancer.
Nat
Rev
Cancer
10,
537–549

109

29. Kamal
A.,
Thao
L.,
Sensintaffar
J.,
Zhang
L.,
Boehm
MF.,
Fritz
LC.,
et
al.
(2003).

A
high-‐affinity
conformation
of
Hsp90
confers
tumour
selectivity
on
Hsp90

inhibitors.
Nature
425,
407–410

30. Li,
W.,
Tsen,
F.,
Sahu,
D.,
Bhatia,
A.,
Chen,
M.,
Multhoff,
G.,
&
Woodley,
D.
T.

(2013).
Extracellular
Hsp90
(eHsp90)
as
the
Actual
Target
in
Clinical
Trials:

Intentionally
or
Unintentionally.
IRCMB.
303,
203–235.

31. Woodley,
DT.,
Fan,
J.,
Cheng,
C.-‐F.,
Li,
Y.,
Chen,
M.,
Bu,
G.,
et
al.,
(2009).

Participation
of
the
lipoprotein
receptor
LRP1
in
hypoxia-‐HSP90α
autocrine

signaling
to
promote
keratinocyte
migration.
Journal
of
Cell
Sci.
122
(10),

1495–1498.

32. Li,
W.,
Li,
Y.,
Guan,
S.,
Fan,
J.,
Cheng,
C.-‐F.,
Bright,
A.
M.,
et
al.,
(2007).

Extracellular
heat
shock
protein-‐90α:
linking
hypoxia
to
skin
cell
motility
and

wound
healing.

EMBO
J,
26(5),
1221–1233.

33. Cheng,
C.-‐F.,
Fan,
J.,
Fedesco,
M.,
Guan,
S.,
Li,
Y.,
Bandyopadhyay,
B.,
…
Li,
W.

(2008).
Transforming
Growth
Factor
α
(TGFα)-‐Stimulated
Secretion
of

HSP90α:
Using
the
Receptor
LRP-‐1/CD91
To
Promote
Human
Skin
Cell

Migration
against
a
TGFβ-‐Rich
Environment
during
Wound
Healing.
Mol.
Cell.

Biol.
28(10),
3344–3358.

34. McCready,
J.,
Sims,
J.
D.,
Chan,
D.,
&
Jay,
D.
G.
(2010).
Secretion
of
extracellular

hsp90α
via
exosomes
increases
cancer
cell
motility:
a
role
for
plasminogen

activation.
BMC
Cancer,
10,
294.

110

35. Eustace
BK,
Sakurai
T,
Stewart
JK,
Yimlamai
D,
Unger
C,
Zehetmeier
C.,
et
al.,

(2004).
Functional
proteomic
screens
reveal
an
essential
extracellular
role

for
hsp90
alpha
in
cancer
cell
invasiveness.
Nat
Cell
Biol.
6(6),
507-‐14

36. Wang,
X.,
Song,
X.,
Zhuo,
W.,
Fu,
Y.,
Shi,
H.,
Liang,
Y.,
et
al.
(2009).
The

regulatory
mechanism
of
Hsp90α
secretion
and
its
function
in
tumor

malignancy.
PNAS,
106(50),
21288–21293.

37. Chen,
J.-‐S.,
Hsu,
Y.-‐M.,
Chen,
C.-‐C.,
Chen,
L.-‐L.,
Lee,
C.-‐C.,
&
Huang,
T.-‐S.
(2010).

Secreted
Heat
Shock
Protein
90α
Induces
Colorectal
Cancer
Cell
Invasion

through
CD91/LRP-‐1
and
NF-‐κB-‐mediated
Integrin
αV
Expression.
J.
Biol.

Chem.,
285(33),
25458–25466.

38. Tsutsumi,
S.,
Scroggins,
B.,
Koga,
F.,
Lee,
M.J.,
Trepel,
J.,
Felts,
S.,
et
al.,
(2008).

A
small
molecule
cell-‐impermeant
Hsp90
antagonist
inhibits
tumor
cell

motility
and
invasion.
Oncogene
27,
2478–2487.

39. Sahu,
D.,
Zhao,
Z.,
Tsen,
F.,
Cheng,
C.-‐F.,
Park,
R.,
Situ,
A.
J.,
et
al.,
(2012).
A

potentially
common
peptide
target
in
secreted
heat
shock
protein-‐90α
for

hypoxia-‐inducible
factor-‐1α–positive
tumors.
MBoC.

23(4),
602–613.

40. Li,
W.,
Sahu,
D.,
&
Tsen,
F.
(2012).
Secreted
Heat
Shock
Protein-‐90
(Hsp90)
in

Wound
Healing
and
Cancer.
BBA,
1823(3),
730–741.

41. Chatterjee,
M.,
Jain,
S.,
Stu¨hmer,
T.,
Andrulis,
M.,
Ungethu¨m,
U.,
Kuban,
R.
J.,

et
al.
(2007).
STAT3
and
MAPK
signaling
maintain
overexpression
of
heat

shock
proteins
90a
and
b
in
multiple
myeloma
cells,
which
critically

contribute
to
tumor-‐cell
survival.
Blood
109,
720-‐728.

111

42. Cheng,
C.-‐F.,
Sahu,
D.,
Tsen,
F.,
Zhao,
Z.,
Fan,
J.,
Kim,
R.,
et
al.,
(2011).
A

fragment
of
secreted
Hsp90α
carries
properties
that
enable
it
to
accelerate

effectively
both
acute
and
diabetic
wound
healing
in
mice.

J.
Clin.
Invest.

121(11),
4348–4361.

43. Semenza
GL
(2003).
Targeting
HIF-‐1
for
cancer
therapy.
Nat.
Rev.
Cancer.

3(10),
721-‐32.

44. Gilkes,
DM
&
Semenza,
GL
(2013).
Role
of
hypoxia-‐inducible
factors
in
breast

cancer
metastasis.
Future
Oncol
9(11),
1623–1636

45. Birner,
P.,
Schindl,
M.,
Obermair,
A.,
Plank,
C.,
Breitenecker,
G.
&
Oberhuber
G

(2000).
Overexpression
of
hypoxia-‐inducible
factor
1α
is
a
marker
for
an

unfavorable
prognosis
in
early-‐stage
invasive
cervical
cancer.
Cancer
Res.
60,

4693–4696

46. Volm,
M.
&
Koomagi,
R.
(2000).Hypoxia-‐inducible
factor
(HIF-‐1)
and
its

relationship
to
apoptosis
and
proliferation
in
lung
cancer.
Anticancer
Res.
20,

1527–1533

47. Schindl
M,
Schoppmann
SF,
Samonigg
H,
Hausmaninger
H,
Kwasny
W,
Gnant

M,
et
al.
(2002).
Overexpression
of
hypoxia-‐inducible
factor
1α
is
associated

with
an
unfavorable
prognosis
in
lymph
node-‐positive
breast
cancer.
Clin.

Cancer
Res.
8,
1831–1837

48. Bos
R,
van
der
Groep
P,
Greijer
AE,
Shvarts
A,
Meijer
S,
Pinedo
HM
et
al.

(2003).
Levels
of
hypoxia-‐inducible
factor-‐1α
independently
predict

prognosis
in
patients
with
lymph
node
negative
breast
carcinoma.
Cancer
97,

1573–1581

112

49. Birner,
P.,
Schindl,
M.,
Obermair,
A.,
Breitenecker,
G.
&
Oberhuber
G
(2001).

Expression
of
hypoxia-‐inducible
factor
1α
in
epithelial
ovarian
tumors:
its

impact
on
prognosis
and
on
response
to
chemotherapy.
Clin.
Cancer
Res.
7,

1661–1668

113

Chapter
5:
Conclusions

It
is
now
becoming
clear
that
the
cytosolic
Hsp90
isoforms,
Hsp90α
and
Hsp90β

possess
distinct
functions.
The
contrasting
knockout
mice
phenotypes
and
the

specific
function
of
secreted
Hsp90α,
but
not
Hsp90β,
in
wound
healing
and
tumor

progression
highlight
this
further.
For
decades,
researchers
have
not
studied
the

isoforms
in
isolation
and
inhibitors
do
not
distinguish
between
them.
Using
human

dermal
fibroblasts,
my
work
shows
that
Hsp90α’s
primary
role
is
to
act
as
an

extracellular
ligand
for
the
LRP-‐1
receptor
and
promote
cell
migration
under

hypoxia.
Hsp90α
does
not
bind
LRP-‐1
inside
the
cell
and
is
not
required
to

chaperone
LRP-‐1.
Hsp90β,
on
the
other
hand,
functions
as
an
intracellular

chaperone
for
LRP-‐1
and
stabilizes
it
on
the
cell
surface.
Though
secreted
under

hypoxia,
Hsp90β
alone
does
not
have
pro-‐motility
activity.
Thus,
extracellular

Hsp90α
and
intracellular
Hsp90β
have
exclusive
functions,
which
cannot
be

compensated
for
by
the
other
isoform.

The
discovery
of
isoform-‐specific
functions

of
Hsp90α
and
Hsp90β
has
important
implications
in
cancer
therapy.
Hsp90

proteins
have
been
found
to
be
either
qualitatively
over
activated
or
quantitatively

overexpressed
in
cancer
cells.
In
addition,
targeting
Hsp90
was
proposed
to

simultaneously
target
and
shut
down
multiple
oncogenic
signaling
pathways
and

overcome
the
problem
of
drug
resistance.
Thus,
they
provide
a
valid
therapeutic

target
(1,2,3,4).
Current
Hsp90
inhibitors
target
the
N-‐terminal
ATPase,
which

inhibits
both
Hsp90α
and
Hsp90β
isoforms.
Despite
numerous
clinical
trials,
few
of

the
inhibitors
have
been
approved
for
human
use
(2,8,9).
The
failure
of
the

inhibitors
have
been
attributed
to
their
poor
stability
and
toxicity
(10).
My
work,
in

114

conjunction
with
others,
shows
the
importance
of
Hsp90β’s,
but
not
Hsp90α’s

chaperone
function
in
maintaining
homeostasis
in
normal
cells.
At
the
same
time,

secreted
Hsp90α
is
not
required
for
normal
cells
under
physiological
conditions.

Thus,
we
propose
that
specific
targeting
of
secreted
Hsp90α
using
neutralizing

antibodies
or
membrane-‐impermeable
inhibitors
instead
of
its
chaperone
function

in
cancer,
as
being
currently
pursued
will
provide
safer
and
more
selective

therapeutic
options.

We
identified
a
novel
role
for
secreted
Hsp90α
as
a
survival
factor
for
tumors,

protecting
them
from
the
hypoxic
microenvironment-‐mediated
cell
death.
We
used

CRISPR/Cas9
to
generate
Hsp90α
knockout
MDA-‐MB-‐231
cells.
We
found
an

increased
susceptibility
of
these
cells
to
hypoxia
driven
cell
death
that
could
be

rescued
by
exogenous
Hsp90α,
but
not
Hsp90β,
protein.
Additionally,
we
developed

a
monoclonal
antibody,
1G6-‐D7,
targeted
against
the
F-‐5
region
of
Hsp90α.
1G6-‐D7

sensitized
parental
MDA-‐MB-‐231
cells
to
hypoxia-‐triggered
cell
death.
In

conjunction
with
this,
1G6-‐D7
inhibited
the
ability
of
MDA-‐MB-‐231
cells
to
form

tumors
in
vivo.
This
finding
further
emphasizes
the
importance
of
targeting
secreted

Hsp90α,
not
intracellular
Hsp90
chaperone,
to
block
tumor
formation
and

metastasis

Considering
the
vast
number
of
stresses
driving
Hsp90α
secretion,
it
is
important
to

understand
the
central
regulator
(s)
that
facilitates
the
communication
between

these
upstream
stimuli
and
downstream
secretion.
We
identified
PRAS40
(Proline-‐

115

Rich
Akt
Substrate
of
40kDa)
as
being
at
least
one
of
these
molecules.
PRAS40

regulated
exosome-‐mediated
Hsp90α
secretion
in
response
to
TGFα,
H2O2
and

hypoxia
in
normal
and
cancer
cells.
PRAS40
worked
in
both
T246-‐phosphorylation

dependent
and
total
protein
level
dependent
pathways,
depending
on
the
nature
of

the
upstream
stimuli.
Exosomes
play
an
important
role
in
intercellular

communication
and
are
gaining
attention
especially
in
the
field
of
tumor

progression
due
to
their
ability
to
act
locally
and
at
distant
sites,
establishing
pre-‐
metastatic
niches.
They
have
roles
in
tumor
immune
evasion,
metastasis,
drug

resistance
etc.
Few
upstream
regulators
of
exosome
secretion
have
been
reported
in

the
literature,
mainly
focusing
on
the
Rab
GTPase
family.
However,
diverse
Rabs

regulate
exosome
secretion
in
distinct
cell
types,
making
them
a
difficult
drug
target

to
block
exosome
secretion.
My
work
has
helped
identify
PRAS40
as
being
a
central

regulator
of
exosome
secretion,
independent
of
cell
type
and
upstream
stimuli.

Since
PRAS40
is
an
upstream
regulator
of
exosome
secretion,
PRAS40
levels
or

T246
phosphorylation
can
act
as
a
biomarker
for
more
aggressive
cancers.
In

addition,
PRAS40
is
an
ideal
target
for
therapeutic
intervention
to
block
exosome

secretion
and
subsequent
tumor
metastasis
and
immune
evasion.

Future
directions

Our
findings
identify
PRAS40
as
being
an
important
regulator
of
exosome
secretion,

independent
of
cell
type
or
stress.
However,
co-‐immunoprecipitation
experiments

and
absence
of
PRAS40
in
exosomes
collected
by
ultracentrifugation
rule
out
a

direct
interaction
between
PRAS40
and
exosomes.
Hence,
additional
molecule
(s)

116

mediate
the
communication
between
PRAS40
and
the
exosome
trafficking
pathway.

In
order
to
identify
potential
interacting
proteins,
we
are
generating
a
PRAS40
GST

fusion
protein.
Using
GST
pull
down
assays
followed
by
mass
spectrometry,
we

propose
to
identify
PRAS40
interacting
proteins
under
unstimulated,
growth
factor-‐,

hypoxia-‐
and
oxidative-‐stress
-‐treated
conditions.
We
hypothesize
that
different

stimuli
will
modify
PRAS40
interactions
differently,
leading
to
exosome
secretion.
It

will
also
be
interesting
to
compare
PRAS40
interacting
partners
amongst
different

cell
types.
Once
potential
protein
(s)
have
been
identified,
we
need
to
verify
their

importance
in
facilitating
PRAS40-‐regulated
exosome
secretion
through
lentiviral

knockdown
and
rescue
experiments.
We
are
also
generating
GST
fusion
proteins
of

PRAS40
T246A
and
T246E
mutants
to
investigate
how
phosphorylation
at
this

residue
alters
its
interactions.

The
ultimate
evidence
that
PRAS40
is
a
central
regulator
of
exosome
secretion
will

come
from
mouse
knockout
studies.
Two
reports
on
PRAS40
knockout
mice
exist

(5,6).
Exosomes
can
be
collected
from
the
plasma
of
wild
type
and
PRAS40
knockout

mice
to
study
if
PRAS40
regulates
exosome
secretion
in
vivo.
Since
exosomes
have

been
reported
to
drive
wound
healing
(7)
as
well
as
tumor
metastasis,
it
will
be

interesting
to
investigate
if
PRAS40
knockout
mice
have
defects
in
wound
healing

and/or
tumor
metastasis
and
if
these
defects
could
be
rescued
by
either
topical

application
or
intratumoral
injection
of
exosomes
respectively.

117

Further
studies
are
warranted
in
delineating
the
isoform-‐specific
functions
of

Hsp90α
and
Hsp90β.
First,
the
unique
communication
between
intracellular
Hsp90β

and
extracellular
Hsp90α
needs
to
be
tested
in
other
cell
types,
including
cancer

cells,
some
of
which
constitutively
secrete
both
Hsp90α
and
Hsp90β.
Second,
the

domains
required
for
Hsp90β-‐LRP-‐1
interaction
need
to
be
identified.
Although

preliminary
data
show
that
Hsp90β
binds
to
the
intracellular
cytoplasmic
tail
of

LRP-‐1,
it
is
unknown
which
domain
(s)
of
Hsp90β
are
required
for
this
interaction.

Also,
hypoxia
drives
Hsp90β
secretion.
Though
extracellular
Hsp90β,
on
its
own,

lacks
pro-‐motility
activity,
it
needs
to
be
tested
if
extracellular
Hsp90β
is
able
to

assist
extracellular
Hsp90α
in
its
pro-‐motility
action.
This
involves
testing
if
the
two

isoforms
dimerize
and
if
this
dimerization
is
required
for
secreted
Hsp90α-‐driven

migration.

These
studies
will
help
elucidate
the
communication
between
intracellular
and

secreted
Hsp90
proteins
and
their
regulation
by
different
environmental
stimuli.

They
will
facilitate
the
identification
of
targets
that
modulate
both
wound
healing

and
tumor
microenvironments.

References:

1. Workman,
P.,
Burrows,
F.,
Neckers,
L.
&
Rosen,
N.
(2007).
Drugging
the

cancer
chaperone
HSP90:
combinatorial
therapeutic
exploitation
of
oncogene

addiction
and
tumor
stress.
Ann
N
Y
Acad
Sci
1113,
202–216

118

2. Trepel,
J.,
Mollapour,
M.,
Giaccone,
G.
&
Neckers,
L.
(2010).
Targeting
the

dynamic
HSP90
complex
in
cancer.
Nat
Rev
Cancer
10,
537–549

3. Kamal
A.,
Thao
L.,
Sensintaffar
J.,
Zhang
L.,
Boehm
MF.,
Fritz
LC.,
et
al.
(2003).

A
high-‐affinity
conformation
of
Hsp90
confers
tumour
selectivity
on
Hsp90

inhibitors.
Nature
425,
407–410

4. Li,
W.,
Tsen,
F.,
Sahu,
D.,
Bhatia,
A.,
Chen,
M.,
Multhoff,
G.,
&
Woodley,
D.
T.

(2013).
Extracellular
Hsp90
(eHsp90)
as
the
Actual
Target
in
Clinical
Trials:

Intentionally
or
Unintentionally.
IRCMB.
303,
203–235.

5. Malla,
R.,
Wang,
Y.,
Chan,
WK.,
Tiwari,
AK.
&
Faridi,
JS.
(2015).
Genetic

ablation
of
PRAS40
improves
glucose
homeostasis
via
linking
the
AKT
and

mTOR
pathways.
Biochemical
Pharmacology
96,
65–75

6. Xiong,
X.,
Xie,
R.,
Zhang,
H.,
Gu,
L.,
Xie,
W.,
Cheng,
M.,
et
al.,
(2014).
PRAS40

plays
a
pivotal
role
in
protecting
against
stroke
by
linking
the
Akt
and
mTOR

pathways.
Neurobiol
Dis.
66,
43-‐52

7. Shabbir,
A.,
Cox,
A.,
Rodriguez-‐Menocal,
L,
Salgado
M,
&
Van
Badiavas
E.

(2015).
Mesenchymal
Stem
Cell
Exosomes
Induce
Proliferation
and

Migration
of
Normal
and
Chronic
Wound
Fibroblasts,
and
Enhance

Angiogenesis
In
Vitro.
Stem
Cells
Dev.

24(14),
1635-‐47

8. Solit,
DB.,
and
Chiosis,
G.
(2008)
Development
and
application
of
Hsp90

inhibitors.
Drug
Discov
Today
13,
38-‐43.

9. Sidera,
K.,
&
Patsavoudi,
E.
(2014)
HSP90
inhibitors:
current
development

and
potential
in
cancer
therapy.
Recent
Pat
Anticancer
Drug
Discov.
9,
1-‐20.

10. Samuni,
Y.,
Ishii,
H.,
Hyodo,
F.,
Samuni,
U.,
Krishna,
M.
C.,
Goldstein,
S.,
et
al.,

119

(2010).
Reactive
oxygen
species
mediate
hepatotoxicity
induced
by
the

Hsp90
inhibiting
anti-‐cancer
geldanamycin
and
its
analogs.
Free
Radical

Biology
&
Medicine,
48(11),
1559–1563.

120

Chapter
6:
Methods

Antibodies
and
reagents

rhPDGF-‐BB
was
purchased
from
R&D
Systems.
Antibodies
against
PDGFR-‐β
(3169)

and
EGFR
(4267)
were
from
Cell
Signaling
Technologies
(Dancers,
MA).
Anti-‐
LRP1/CD91
antibody
(37-‐7600)
was
purchased
from
Life
Technologies
(Grand

Island,
NY).
We
purchased
mouse
monoclonal
antibodies
against
Hsp90α
(CA1023)

from
Calbiochem
(Billerica,
MA)
and
Hsp90β
(SMC
107)
from
Stressmarq

Biosciences
(Victoria,
BC,
Canada).
Anti-‐GAPDH
antibody
(GTX28245)
antibody
was

from
Genetex
(Irvine,
CA).
Anti-‐PRAS40
(MAB6408),
anti-‐p-‐PRAS40
T246

(MAB6890)
and
anti-‐p-‐RSK
S380
antibodies
(MAB79671)
were
from
R&D
systems

(Minneapolis,
MN).
Anti-‐CD63
antibody
(EXOAB
CD63-‐A1)
and
anti-‐CD9
antibodies

(13403)
were
from
System
Biosciences
(Mountain
View,
CA)
and
Cell
Signaling

Technology
(Danvers,
MA)
respectively.
Anti-‐flottilin-‐1
(3253)
and
anti-‐CD81

antibodies
(EXOAB
CD81A-‐1)
were
from
Cell
Signaling
Technology
Inc.
(Danvers,

MA)
and
System
Biosciences
(Mountain
View,
CA)
respectively.
Anti-‐phospho-‐Akt

S473
(4060)
and
anti-‐phospho-‐Akt
T308
(4087)
were
from
Cell
Signaling

Technology
Inc.
(Danvers,
MA). Anti-‐cyclin-‐D1
(GTX61845)
was
from
Genetex

(Irvine,
CA).
Anti-‐EGFR
antibody
(4267)
was
from
Cell
Signaling
Technology
Inc.

(Danvers,
MA).
Brefeldin
A
(BFA)
and
dimethyl
ameloride
(DMA)
were
purchased

from
Sigma
Aldrich
(St.
Louis,
MO).
TGFα
and
EGF
were
purchased
from
Fitzgerald

Industries
International
(Acton,
MA).
The
PRAS40
WT
cDNA
was
purchased
from

Addgene
(Plasmid
#14950).
The
recombinant
RAP
protein
was
as
previously

121

described
(Cheng
et
al.,
2008).
XL-‐10
Gold
Ultra
competent
cells
(XL-‐10
Gold)
were

from
Stratagene.

Cell
lines

Primary
human
dermal
fibroblasts
were
purchased
from
Clonetics
(San
Diego,
CA)

and
were
cultured
in
DMEM
supplemented
with
10%
fetal
bovine
serum.
The
3
rd

or

4
th

passages
were
used
in
cell
migration
assays.
Human
keratinocytes
(HKCs)
were

cultured
in
EpiLife
medium
with
added
human
keratinocyte
growth
supplements.

The
2
nd

or
3
rd

passages
of
cells
were
used.
MDA-‐MB-‐231
cells
were
cultured
in

DMEM
supplemented
with
10%
FBS.
Mouse
hepatocytes
were
a
kind
gift
from
Dr.

Bangyan
Stiles
(University
of
Southern
California,
Los
Angeles).
Mouse
lung

epithelial
cell
line
(MLE
15)
was
obtained
from
Dr.
Zea
Borok’s
lab
(University
of

Southern
California,
Los
Angeles).
Eight
human
breast
cancer
cell
and
a
control

(untransformed)
mammary
epithelial
cell
lines
were
gifts
from
Dr.
Michael
Press

(University
of
Southern
California,
Los
Angeles).
All
the
cells
were
cultured
in
DMEM

medium
supplemented
with
10%
fetal
bovine
serum
(FBS),
as
well
as
ATCC-‐
suggested
media
for
some
of
the
cell
lines,
such
as
McCoy’s
5A
for
Skbr3.

Cell
migration
assay

The
colloidal
gold
migration
assay
was
modified
and
described
previously
in
details

by
us
(Li
et
al.,
2004).
Initially,
1%
BSA
is
used
to
coat
cover
slips
and
the
excess
BSA

washed
off
by
100%
ethanol.
The
cover
slips
are
dried
and
placed
in
a
8-‐well
plate,

one
cover
slip
per
well.
Gold
salt
solution
(9%
of
gold
salt
combined
with
52%
of

H2O
and
30%
of
the
Na2CO3
solution)
was
heated
in
an
Erlenmeyer
flask
with

constant
swirling
until
boiling,
after
which
an
equal
volume
of
freshly
prepared

122

formaldehyde
(0.1%)
as
the
gold
salt
was
quickly
added
to
the
gold
salt
mixture

with
swirling.
Once
the
mixture
turns
light
purplish-‐brown,
it
is
immediately

pipetted
into
each
well
of
the
8-‐well
plate.
The
plates
were
covered
and
left

undisturbed
over
night
to
let
the
gold
salt
particles
settle.
Following
the
removal
of

the
gold
salt
solution,
the
cover
slips
were
gently
rinsed
1.5
ml
Hank's
buffered
salt

solution
(HBSS)
and
then
incubated
with
1.5
ml
of
HBSS
containing
5
mM
Ca
2+

and

25
μg/mL
rat
tail
type
I
collagen
at
37°C
for
2
hours.
The
collagen
solution
was

removed
and
plates
were
rinsed
twice
with
PBS
prior
to
plating
the
cells.
Media

containing
the
appropriate
conditions
such
as
growth
factor,
recombinant
protein

or
antibody
were
prepared
in
individual
tubes.
Cells
were
trypsinized,
resuspended

in
serum-‐free
media
and
counted.
3500
cells
were
added
to
each
condition
and
left

to
migrate
overnight
at
37
o
C.
The
media
was
removed
and
cells
were
fixed
in
0.1%

formaldehyde.
Migration
was
examined
under
dark
field
optics
and
photographed.

Twenty
randomly
selected
and
non-‐overlapping
fields
were
analyzed
with
an

attached
camera.
Migration
index
is
calculated
as
the
average
area
taken
up
by
the

tracks.
Data
from
independent
experiments
(n
≥
3)
were
averaged
and
calculated

(mean
±
SD,
p
<
0.05).

Cell
viability
assays

We
used
the
LIVE/DEAD
Viability/Cytotoxicity
Kit
for
mammalian
cells
(MP
03224,

Molecular
Probes)
and
followed
its
(two)
protocols:
the
Fluorescence
Microscopy

Protocol
and
the
Flow
Cytometry
Protocol.
Samples
were
analysed
in
triplicate
for

each
condition.

123

Chemotaxis
assay

The
transwell
motility
assay
was
carried
out
according
to
the
manufacturer’s

instruction
(Cat
no.
3422,
Corning
Life
Sciences,
Tewksbury,
MA).
1
×
10
5

serum-‐
starved
HDFs
in
100μL
of
serum-‐free
medium
were
seeded
into
the
upper
chamber

of
the
insert
and
650μL
medium
with
chemoattractant
was
added
to
lower
chamber

in
a
24-‐well
tissue
culture
plate.
Medium
with10%
fetal
bovine
serum
or
PDGF-‐BB

(15
ng/ml)
or
30
μg/ml
recombinant
Hsp90α
were
tested
for
induced
chemotaxis
of

the
cells.
Following
24h,
the
migrated
fibroblasts
were
stained
with
crystal
violet,

visualized
under
microscope
and
quantitated
in
averaged
percentage
(%)
of
cells

over
the
total
number
of
seeded
cells
that
penetrated
the
lower
chamber.

Creating
excision
wounds
in
pigs

To
create
full-‐thickness
excision
wounds
in
pigs,
the
animal
was
shaved
along
its

torso
on
both
sides.
The
pig
was
placed
on
its
right
side
(left
side
up).
The
surgical

skin
site
on
the
torso
was
scrubbed
with
betadine
scrub
and
solution
3
times
using

the
sterile
prep
kit
provided
in
the
Operating
Room.
The
outline
of
the
wounds
was

created
with
a
pre-‐cut
paper
template
and
a
permanent
marker.
Wounds
were
1.5cm

x
1.5cm
with
2.5cm
of
unwounded
skin
between
two
wounds.
The
maximum
number

of
excision
wounds
on
each
side
was
12,
making
a
possible
maximum
of
24
wounds

for
each
experimental
investigation.
Using
number
15
scalpel
blades,
the
wounds

were
cut
to
a
full
thickness
depth;
the
epidermis,
dermis
and
underlying
fat
are

removed
to
expose
the
fascia
layer
below.
The
depths
of
the
wounds
were
measured

at
approximately
5
mm.

124

Exosome
Purification,
characterization
and
analyses

Cells
were
cultured
in
serum
free
media
overnight
and
then
treated
with
different

stimuli
at
indicated
time
points.
Supernatant
were
then
collected
and
spun
at
300xg

for
10min
to
discard
cell
contamination.
Dead
cells
were
then
removed
by

centrifugation
at
2,000xg
for
10min
and
cell
debris
were
removed
by
spinning
at

10,000xg
for
30
min.
Finally,
exosomes
were
collected
by
ultracentrifugation

(Beckman
Coulter
Optima
L-‐100
XP,
Beckman
Coulter,
CA)
at
100,000xg
for
70min

and
pelleted
exosomes
were
washed
in
10ml
of
PBS
and
harvested
again
at
10,000xg

for
70
min
to
get
rid
of
micro-‐vesicles
and
other
contaminating
particles.
Exosome

size
distribution
and
concentration
were
analyzed
using
the
Malvern
NanoSight

(Nanosight,
Malvern
instruments)
aided
by
the
Nanoparticle
Tracking
Analysis

(NTA)
software.

Human
Phospho-‐Kinase
Antibody
Array

Human
Keratinocytes
were
grown
to
80%
confluence
in
15cm
tissue
culture
dishes

and
serum-‐starved
overnight.
Cells
were
then
treated
with
various
stimuli
for

indicated
time
points
and
cells
were
lysed
and
subject
to
Proteome
Profiler
Human

Phospho-‐Kinase
Array
Kit
(Catalog#
ARY003B,
R&D
Systems,
Minneapolis,
MN)

strictly
complying
with
the
manufacturer’s
instructions.

Hypoxia
treatment
and
preparation
of
serum-‐free
conditioned
media

The
OxyCycler
C42
from
BioSpherix
(Redfield,
NY)
was
used
as
the
oxygen
content

controller.
This
equipment
allows
creation
of
any
oxygen
profile
with
full-‐range

oxygen
(0.1–99.9%)
and
CO2
control
(0.1–20.0%).
All
media
used
for
hypoxia

experiments
were
preincubated
in
hypoxia
chambers
with
the
designated
oxygen

125

content
for
16
h
prior
to
their
use
to
replace
normoxic
culture
media
(Li
et
al.,
2007).

Preparation
of
serum-‐free
conditioned
media
was
carried
out
as
previously

described
(Cheng
et
al.,
2008).

Immunoprecipitation

Cells
were
lysed
using
lysis
buffer
and
lysates
prepared.
An
equal
amount
of
protein

lysates
from
each
condition
were
incubated
with
the
appropriate
antibody
overnight

at
4°C.
The
next
day,
protein-‐G
sepharose
beads
(20μl
beads/condition)
were

washed
with
lysis
buffer.
The
samples
were
then
incubated
with
the
beads
for
2

hours
at
4
o
C.
They
were
washed
3x
times
with
lysis
buffer
to
eliminate
any
non-‐
specific
binding.
The
associated
proteins
were
investigated
using
western
blot.

Invasion
assay

We
followed
the
procedures
as
described
by
the
manufacturer’s
instructions
(BD

Biosciences,
Bedford,
MA).
The
Corning
Biocoat
Matrigel
Invasion
Chamber

(Cat#354480)
was
used
as
detailed
previously

21
.
The
%
invasion
was
calculated
as

per
the
formula:

Number
of
cells
invaded

%
Invasion
=
-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐-‐
×
100

Total
number
of
cells
seeded

Lentiviral
systems

Utilization
of
lentiviral
systems
for
down-‐regulation
and
overexpression
of
genes
are

as
previously
described
(Li
et
al.,
2007;
Cheng
et
al.,
2008;
Sahu
et
al.,
2012).
The

pRRLsinh-‐CMV
system
was
used
to
overexpress
exogenous
genes.
The
pHR-‐CMV-‐
puro
RNAi
delivery
system
was
used
to
deliver
shRNA.
The
shRNA
sequences
of
the

126

various
genes
are
as
follows:
PRAS40:
GCTGAGTTCTAAGCTCTAA
(sense),
EGFR:

AGAATGTGGAATACCTAAGG
(sense),
Hsp90α:
GGAAAGAGCTGCATATTAA
(sense)

and
Hsp90β:
GCATCTATCGCATGATCAA
(sense).
The
FG-‐12
system
was
used
to

deliver
shRNA
against
Raptor:
AAAATTTCGAAACAGACTAGCC
(sense)
and
mTOR:

GGCCGCATTGTCTCTATCAT
(sense).

Nanoparticle
Tracking
analysis
(NTA)

Exosome
size
distribution
and
concentration
were
analyzed
using
the
Malvern

NanoSight
(Nanosight,
Malvern
instruments)
aided
by
the
Nanoparticle
Tracking

Analysis
(NTA)
software.
Briefly,
conditioned
media
from
cells
treated
under

different
conditions
was
passed
into
a
sample
chamber
and
subjected
to
a
laser

beam.
Particles
in
the
media
scatter
light
that
is
visualized
by
a
microscope
and

captured
by
a
camera.
The
NTA
software
calculates
the
hydrodynamic
diameter
of

the
particles
using
Stokes-‐Einstein
equation,
providing
a
real-‐time
measurement
of

the
concentration,
particle
size
and
aggregation
of
particles
in
the
media.

Preparation
of
carboxymethylcellulose
gel
with
recombinant
human
Hsp90α

proteins

To
prepare
the
wound
healing
formula,
CMC
powder
(Sigma
Aldrich;
viscosity,
50
to

200cps;
purity,
99.5%,
sodium
salt)
was
dissolved
in
double-‐distilled
water
in
a

tissue
culture
hood
at
a
20%
(wt/vol)
concentration.
The
mixture
was
incubated
for

4
hours
at
37°C
and
then
placed
on
a
shaker
for
24
hours
at
4°C.
After
being

equilibrated
back
to
room
temperature,
the
CMC
solution
was
mixed
at
a
1:1

(vol/vol)
ratio
with
the
indicated
concentrations
of
FPLC
purified
and
filtered

Hsp90α
proteins
and
topically
added
to
wounds.

127

Recombinant
Hsp90α
and
Hsp90β
proteins
production
and
purification

Protein
synthesis
was
induced
by
the
addition
of
IPTG
(isopropyl-‐β-‐D-‐
thiogalactopyranoside)
to
the
bacterium
culture
(optical
density
0.5-‐1)
for
a
final

concentration
of
1
mM
and
incubated
for
5
h
at
25°C.
His-‐tagged
proteins
were
first

purified
by
a
nickel-‐nitrilotriacetic
acid
(Ni-‐NTA)
column
with
a
His-‐Bind
purification

kit
(EMD
Biosciences,
Inc.)
according
to
the
manufacturer's
procedure.
Briefly,

bacteria
were
resuspended
in
binding
buffer
with
protease
inhibitors
and
lysozyme.

They
were
sonicated
for
1.30
minutes
in
bursts
of
10
seconds
with
20
seconds
in

between
bursts.
The
purified
proteins
were
concentrated
in
an
Amicon
filter
tube

(Millipore,
Billerica,
MA)
to
4
ml
and
loaded
onto
a
Superdex
200
HiLoad
gel

filtration
column
(GE
Healthcare,
Piscataway,
NJ)
and
separated
by
fast
protein
liquid

chromatography
(FPLC).
Proteins
were
eluted
by
use
of
Dulbecco's
phosphate-‐
buffered
saline
buffer
with
a
flow
speed
of
1.0
ml/min.
The
fractions
with
Hsp90α

were
further
concentrated
with
Amicon
filter
tubes
to
achieve
a
final
concentration

of
1
mg/ml.
Proteins
were
stored
in
10%
glycerol-‐DPBS
at
−80°C.

Reverse
Transcriptase
PCR
(RT-‐PCR)

Total
RNA
was
extracted
from
control,
Hsp90α-‐
or
Hsp90β-‐
knockdown
human

dermal
fibroblasts
using
Trizol
(Invitrogen).
cDNA
was
extracted
using
the

Superscript
III
First
Strand
synthesis
system,
according
to
the
manufacturer’s

instructions
(Invitrogen).
The
primers
for
LRP1
5’
CTCCCACCGCTATGTGATCC
3’

(forward)
and
5’
ACTCATCTTGTGCTCGGCAA
(reverse)
and
for
GAPDH
(5′-‐
CCATCACCATCTTCCAGGAG-‐3′
(forward)
and
5′-‐CCTGCTTCACCACCTTCTTG-‐3′

128

(reverse)
were
designed
using
the
Primer-‐Blast
software.
cDNAs
were
subject
to
PCR

with
an
initial
denaturation
at
94
o
C
for
2
minutes
followed
by
denaturation
at
94
o
C

for
30
seconds,
annealing
at
55

o
C
for
30
seconds
and
extension
at
68

o
C
for
1
minute.

The
products
were
visualized
on
a
1.5%
agarose
gel
using
Ethidium
Bromide

staining.

Site-‐directed
mutagenesis
on
T246
site
of
PRAS40

The
Quikchange
II
XL
site-‐directed
mutagenesis
kit
(200521-‐5)
from
Agilent

Technologies
was
used
to
mutate
the
T246
site
of
PRAS40.
The
primer
sequence

GGAAGTCGCTGGCGTTAAGCCGCGGC
(sense)
was
used
to
generate
the
T246
to

alanine
(T246A)
mutation
and
the
primer
sequence

GCTTCTGGAAGTCGCTTTCGTTAAGCCGCGGCCGTGG
(sense)
was
used
to
generate
the

T246
to
glutamic
acid
(T246E)
mutation.

Statistical
analyses

Data
are
based
on
three
or
four
independent
experiments.
The
data
are
presented

as
mean
±
s.d.
Matrigel
Invasion
Assay
quantification
was
achieved
by
measuring

five
randomly
selected
fields
per
experimental
condition.
Colloidal
gold
salt

migration
assay
quantification
was
achieved
by
measuring
the
individual
tracks
of

20
randomly
selected
individual
cells
per
experimental
condition,
where
each

condition
in
an
experiment
was
repeated
at
least
three
times.
Flow
cytometry
assay

quantification
was
based
on
triplicate
samples
in
each
experiment
from
three

indepedent
experiments
as
percentage
(%).

The
data
are
presented
as
mean
±
s.d.

Statistical
differences
were
evaluated
using
the
two-‐tailed
Student
t-‐test
for

129

comparisons
of
two
groups,
or
analysis
of
variance
for
comparisons
of
more
than

two
groups.
p
<
0.05
was
considered
significant.

Western
blot

Equal
amounts
of
lysates
(normalized
using
Bradford
assay)
were
loaded
onto
a

SDS-‐PAGE
gel.
Proteins
were
transferred
to
a
nitrocellulose
membrane
using
the

wet
transfer
apparatus
(Bio-‐Rad)
at
90V
for
2
hours.
Membranes
were
stained
with

ponceau
red
to
confirm
transfer
and
equal
loading
of
proteins
after
which,
they

were
washed
2x
times
with
TBS-‐T
and
1x
with
TBS.
Membranes
were
blocked
for
1

hour
at
room
temperature
using
5%
BSA
in
TBS.
Primary
antibodies
were
incubated

overnight
and
then
washed
3x
times
with
TBS-‐T
and
1x
with
TBS.
The
membranes

were
incubated
with
HRP-‐
linked
secondary
antibodies
(Santa
Cruz
biotechnology)

for
1
hour
at
room
temperature
followed
by
TBS-‐T
and
TBS
washes.
Membranes

were
developed
using
the
ECL
kit
(GE
Healthcare
Lifesciences)
in
a
dark
room.

Asset Metadata

Creator Jayaprakash, Priyamvada (author)

Core Title Mechanism of secretion and function of heat shock protein-90 (Hsp90) family genes

Contributor Electronically uploaded by the author (provenance)

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Genetic, Molecular and Cellular Biology

Publication Date 07/29/2016

Defense Date 04/28/2016

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag breast cancer,cell migration,exosomes,Hsp90,OAI-PMH Harvest,wound healing

Format application/pdf (imt)

Language English

Advisor Stallcup, Michael (committee chair), Dubeau, Louis (committee member)

Creator Email jayaprak@usc.edu,priyamvada.jayaprakash@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-287386

Unique identifier UC11280393

Identifier etd-Jayaprakas-4682.pdf (filename),usctheses-c40-287386 (legacy record id)

Legacy Identifier etd-Jayaprakas-4682.pdf

Dmrecord 287386

Document Type Dissertation

Format application/pdf (imt)

Rights Jayaprakash, Priyamvada

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Abstract (if available)

Abstract The cytosolic heat shock proteins of 90kDa family, Hsp90α and Hsp90β, were widely studied for their intracellular chaperone function. Over the past decade, an extracellular promotility function of majorly the Hsp90α isoform has emerged, where it drives cell migration resulting in wound healing or cancer metastasis. We report that despite being 86% homologous at the amino acid level, the Hsp90α and Hsp90β isoforms have distinct functions in driving dermal fibroblast migration. Hsp90β acts inside the cell as a chaperone stabilizing the cell surface LRP-1 receptor while Hsp90α acts outside the cell as a promotility factor binding and signaling through LRP-1. In addition, we also show the distinct functions of Hsp90α and Hsp90β in breast cancer formation and metastasis. Only secreted Hsp90α but not Hsp90β protects tumor cells from hypoxia-driven cell death. Neutralizing its function blocks tumor cell survival in vitro and tumor formation in vivo. In addition, we have identified an important upstream regulator of Hsp90α secretion in response to diverse environmental stresses. This protein named PRAS40 controls exosome-mediated Hsp90α secretion and its lentiviral knockdown inhibits not only exosome secretion, but also exosome-mediated cell migration.