Circulating human dendritic cells differentially express high levels of a 55-kD actin-bundling protein

Abstract

This study was initiated to examine the differential expression of an evolutionary conserved human 55-kd actin-bundling (p55) protein that is induced in B lymphocytes by Epstein-Barr virus infection. Our study demonstrates that p55 is specifically expressed at constitutively high levels in human peripheral blood dendritic cells and lymph node (interdigitating) dendritic cells. Blood dendritic cells constitute a minority (< 2%) of all blood leukocytes but are a distinct population of potent antigen-presenting cells. Immunofluorescence microscopy with a monoclonal antibody specific for p55 showed that 87% of peripheral blood dendritic cells stained brightly in the cytoplasm and in the veiled cytoplasmic extensions. In contrast, monocytes, granulocytes, T cells, and B lymphocytes showed no expression of the p55 protein. Western blot analysis confirmed that only the dendritic cell component of peripheral blood expressed high levels of p55. Staining of human lymph node sections demonstrated selective expression of the p55 antigen by dendritic cells in the T-cell-dependent areas but not in the B cell follicles. p55 is likely to be involved in the organization of a specialized microfilament cytoskeleton in the dendritic cells, and the anti-p55 antibody should be useful for further characterization of this important population of antigen-presenting cells in clinical transplantation, HIV-1 pathogenesis, and autoimmune diseases.

Full text

American
journal
of
Patbology,
Vol.
148,
No.
2,
February
1996
Copyright
©)
Americani
Society
for
Investigative
Pathology
Circulating
Human
Dendritic
Cells
Differentially
Express
High
Levels
of
a
55-kd
Actin-Bundling
Protein
George
Mosialos,*
Mark
Birkenbach,t
Seyoum
Ayehunie,*
Fumio
Matsumura,§
Geraldine
S.
Pinkus,11
Elliott
Kieff,*
and
Erik
Langhoffl
From
the
Department
of
Microbiology
and
Molecular
Genetics
and
Medicine,*
Harvard
Medical
School,
the
Department
of
Human
Retrovirology,t
Dana-Farber
Cancer
Institute,
the
Department
of
Pathology,"1
Brigham
and
Women's
Hospital,
and
the
Renal
Unit,9
Massachusetts
General
Hospital,
Boston,
Massachusetts;
The
Mariorie
B.
Kovler
Viral
Oncology
Laboratories,t
The
University
of
Chicago,
Chicago,
Illinois;
and
the
Department
of
Molectular
Biology
and
Biochemistry,5
Rutgers
Universitv,
Piscataway,
New
Jersey
This
study
was
initiated
to
examine
the
differential
expression
of
an
evolutionary
conserved
human
55-kd
actin-bundling
(p55)
protein
that
is
induced
in
B
lymphocytes
by
Epstein-Barr
virus
infection.
Our
study
demonstrates
that
p55
is
specifically
expressed
at
constitutively
high
levels
in
human
peripheral
blood
dendritic
ceUls
and
lymph
node
(interdigitating)
dendritic
ceUls.
Blood
dendritic
cells
constitute
a
minority
(<2%)
of
aUl
blood
leu-
kocytes
but
are
a
distinct
population
of
potent
an-
tigen-presenting
cells.
Immunofluorescence
mi-
croscopy
with
a
monoclonal
antibody
specific
for
p55
showed
that
87%
ofperipheral
blood
dendritic
cels
stained
brightly
in
the
cytoplasm
and
in
the
veiled
cytoplasmic
extensions.
In
contrast,
mono-
cytes,
granulocytes,
T
ceUls,
and
B
lymphocytes
showed
no
expression
of
the
p55
protein.
Western
blot
analysis
confirmed
that
only
the
dendritic
ceUl
component
ofperipheral
blood
expressed
high
lev-
els
of
p55.
Staining
of
human
lymph
node
sections
demonstrated
selective
expression
of
the
p55
anti-
gen
by
dendritic
cells
in
the
T-ceUl-dependent
areas
but
not
in
the
B
cell
foUlicles.
p55
is
likely
to
be
involved
in
the
organization
of
a
specialized
mi-
crofilament
cytoskeleton
in
the
dendritic
cells,
and
the
anti-pSS
antibody
should
be
usefulforfurther
characterization
of
this
important
population
of
antigen-presenting
cells
in
clinical
transplantation,
HIV-1
pathogenesis,
and
autoimmune
diseases.
(Am
J
Pathol
1996
14&8593-600)
Blood
dendritic
cells
(DCs)
constitute
less
than
2%
of
peripheral
blood
leukocytes
and
are
highly
potent
antigen presenting
cells.1
In
lymphoid
organs,
circu-
lating
DCs
localize
to
the
T-cell-dependent
area
as
interdigitating
DCs
(IDCs).
DCs
are
15
to
20
,um
and
are
motile
with
minimal
phagocytic
activity.
DCs
ex-
press
high
levels
of
class
and
11
major
histocompat-
ibility
complex
proteins
and
are
bone-marrow-de-
rived
cells.1
2
A
different
type
of
cell
with
dendritic
morphology
localizes
in
the
B
cell
zones
(germinal
centers)
as
follicular
DCs
(FDCs).
FDCs
are
not
leu-
kocytes
and do
not
act
as
traditional
antigen-pre-
senting
cells
for
T
cells
but
are
involved
in
B
cell
proliferation
and
differentiation.3'4
The
study
of
human
blood
DCs
has
been
severely
restricted
by
lack
of
selective
markers
for
this
highly
specialized
subset
of
antigen-presenting
cells.
In
the
course
of
studying
the
tissue-restricted
expression
of
a
newly
characterized
human
actin-bundling
protein,
p55,
the
expression
of
which
is
highly
induced
by
Epstein-Barr
virus
infection,
we
discovered
that
p55
was
present
in
cells
of
dendritic
phenotype
in
the
T-cell-dependent
zones
of
the
lymph
nodes.
The
cDNA
for
p55
has
only
recently
been
cloned
and
has
been
characterized
as
an
actin-bundling
protein
originally
described
and
purified
from
HeLa
cells.5,6
A
monoclonal
antibody
against
p55
has
been
de-
rived
and
was
used
in
the
present
study
to
examine
p55
expression
in
human
leukocyte
subsets.
The
cells
expressing
p55
in
lymph
nodes
had
long
cyto-
plasmic
processes
characteristic
of
interdigitating
DCs.
As
these
cells
are
believed
to
be
derived
from
E.
L.
was
supported
by
National
Institutes
of
Health
grant
A128734.
Accepted
for
publication
October
19,
1995.
Address
reprint
requests
to
Dr.
Erik
Langhoff,
Massachusetts
General
Hospital,
Renal
Unit,
Boston,
MA
02114.
593

594
Mosialos
et
al
A/P
February
1996,
Vol.
148,
No.
2
Table
1.
55-kd
Protein
Expression
in
Subsets
of
Primary
Leukotes
from
a
Healthy
Donor
Cell
fraction
Cell
type
Anti-p55
Anti-DR
Anti-CD3
Anti-CD20
Anti-CD14/anti-CD11b
Anti-CD56
PBMCs
Granulocytes
Monocytes
T
cells
DC-enriched
fraction*
DC-depleted
fraction
(B
cell
enriched)
FACS-sorted
B
cells
FACS-sorted
DCs
0.9
<1
6
0
54
3
<1
96
87
(bright)
32
<1
71
2
80
34
>95
97
95
(bright)
51
<1
<1
95
8
24
9
<1
1
<1
6
18
<1
>95
<1
<1
ND,
not
determined.
*From
5
donors;
mean
=
50.4%,
SD
=
6.1%.
peripheral
blood
DCs,
we
set
out
to
investigate
p55
expression
in
peripheral
blood
DCs.
Materials
and
Methods
Cell
Fractionation
and
Cell
Cultures
Techniques
for
isolation
of
leukocyte
subsets
have
previously
been
described.7
Peripheral
blood
mononuclear
cells
(PBMCs)
were
isolated
by
Ficoll-
Hypaque
(Pharmacia,
Uppsala,
Sweden)
density
separation
from
blood
donor
buffy
coats
or
leuko-
paks
(fraction
1,
Table
1).
Granulocytes
were
iso-
lated
from
the
red
cell
pellet
of
the
Ficoll-Paque
gradients
after
lysis
of
the
red
cells
(fraction
2).
Next,
the
PBMC
fraction
was
separated
on
Percoll
density
gradients.
Monocytes
were
low
density
cells
and
were
depleted
of
contaminating
T
cells
by
rosetting
with
neuraminidase-treated
sheep
erythrocytes
(fraction
3).
T
cells
were
high
density
cells
positively
enriched
by
rosetting
with
sheep
erythrocytes
and
were
depleted
of
accessory
leukocytes
and
B
cells
by
nylon
wool
column
separation
(fraction
4).
This
procedure
yielded
>94%
pure
T
cells
as
determined
by
immunostaining.7
The
sheep-erythrocyte-rosette-
negative
cells
were
depleted
of
contaminating
monocytes
by
Fc
panning
of
monocytes
on
IgG-
coated
dishes.7
Low
density
cells
of
this
population
(enriched
for
DCs)
were
isolated
from
14.5%
metri-
zamide
(Sigma
Chemical
Co.,
St.
Louis,
MO)
gradi-
ents
at
500
x
g
for
12
minutes
at
room
temperature.
This
procedure
yielded
a
population
of
cells
that
was
estimated
to
be
50
to
70%
DCs
in
accordance
with
previous
studies7-9
(fraction
5).
The
high
density
cells
from
the
pellet
of
the
metrizamide
gradients
(depleted
of
DCs)
were
also
recovered.
This
fraction
is
enriched
in
B
cells
(fraction
6).
Purified
B
cells
(fraction
7)
were
obtained
by
fluorescein
isothiocya-
nate
(FITC)
FACS
sorting
for
CD19-positive
cells
of
fraction
6.
Purified
DCs
were
obtained
by
FACS
sort-
ing
by
positive
selection
for
large
cells
(high
side
scatter
and
high
forward
scatter).
More
than
95%
of
these
sorted
cells
had
on
the
hemocytometer
the
distinctive
phenotype
of
large
veiled
cells
and
were
strongly
positive
for
HLA-DR
antigen
(fraction
8).
Leukocyte
Cultures
The
function
of
monocytes
and
DCs
was
examined
by
comparing
their
antigen-presenting
functions
in
the
mixed
leukocyte
reaction
(MLR).
FACS-purified
blood
DCs
or
plastic-adherent
(Percoll-isolated)
monocytes
were
used
as
stimulator
cells
for
the
MLR
cultures
as
previously
described.10
The
stimulator
cells
were
irradiated
(3000
rads,
1
rad
=
0.01
Gy)
from
a
Cs127
source.
By
limiting
dilution,
graded
doses
of
stimulator
cells
(25,000
to
1,500
DCs
or
monocytes)
were
added
to
cultures
of
50,000
allo-
geneic
T
cells
in
round-bottom
96-well
plates
(Lim-
bro,
Flow
Laboratories,
Inc,
McLean,
VA)
with
growth
medium
of
RPMI
1640
and
10%
fetal
calf
serum.
The
MLR
cultures
were
cultured
for
5
days
before
addi-
tion
of
[3H]thymidine
(1
gCi/well;
DuPont-NEN,
Wilm-
ington,
DE)
for
20
hours.7'10
Phytohemagglutinin
(20
,Lg/ml;
Difco
Laborato-
ries,
Detroit,
Ml)
was
used
for
stimulation
of
T
cell
cultures
(fraction
3)
and
pokeweed
mitogen
(1:10
and
1:100;
Gibco,
Grand
Island,
NY)
was
used
for
stimulation
of
FACS-sorted
B
cells
(fraction
7)
for
5
days
before
harvest
of
the
cells
for
immunohisto-
chemistry.
1
1
Western
Blot
Analysis
The
presence
of
the
55-kd
actin-bundling
protein
was
analyzed
by
Western
blot
analysis.
Whole
cell
lysates
(4
x
105
primary
cells)
were
generated
by
boiling
the
cells
in
sodium
dodecyl
sulfate
(SDS)
1.
2.
3.
4.
5.
6.
7.
8.
23/ND
6/91
81/ND
6/ND
6/ND
12/ND
3/ND
<1/ND
2
<1
<1
<1
ND
<1

Leukocyte
Expression
of
Actin-Bundling
Protein
595
AJP
February
1996,
Vol.
148,
No.
2
loading
buffer
for
10
minutes
followed
by
centrifuga-
tion
to
remove
insoluble
material.
Whole
cell
isolates
(1
x
105
cell
equivalents)
were
analyzed
on
an
8.5%
SDS
polyacrylamide
gel
followed
by
electrophoretic
transfer
onto
nitrocellulose
membrane
and
immuno-
blotted
using
the
anti-p55
monoclonal
antibody
di-
luted
1:500
in
phosphate-buffered
saline
(PBS)
con-
taining
0.05%
Tween-20
and
3%
nonfat
dry
milk.
An
alkaline-phosphatase-conjugated
goat
anti-mouse
antibody
(Promega,
Madison,
WI)
was
used
at
1:5000
dilution
as
secondary
antibody
for
the
chro-
mogenic
detection
of
protein
bands
reactive
with
the
anti-p55
monoclonal
antibody.
Purified
55-kd
actin-
bundling
protein
from
HeLa
cells
was
run
in
parallel
control
lanes.5
Monoclonal
Antibodies
and
Immunohistochemistry
A
monoclonal
antibody
to
the
55-kd
protein
was
raised
as
previously
described.12
Female
BALB/57
mice
were
immunized
intraperitoneally
with
20
,ug
of
purified
55-kd
protein
in
complete
Freund's
adjuvant
and
after
1
month
with
20
,ug
of
55-kd
protein
in
incomplete
Freund's
adjuvant.
The
mice
were
re-
boosted
12
times
before
harvest
of
the
spleens
for
fusion
with
the
mouse
myeloma
cell
line
NS-1
and
enzyme-linked
immunosorbent
assays
were
used
to
screen
for
positive
clones.12
One
clone,
K-2
(IgG1),
of
three
clones
reactive
with
the
55-kd
protein,
was
used
for
the
present
studies.
Culture
supernatants
of
the
K-2
hybridoma
were
used
directly
as
an
antibody
source
for
the
staining
reactions.
Ascites
fluid
was
also
obtained
by
intraperitoneal
injection
of
the
K-2
hybridoma
into
BALB/c
mice
(Hybridoma
Core
Fa-
cility,
Dana-Farber
Cancer
Institute).
Both
hybridoma
supernatant
(undiluted)
and
ascites
(diluted
1:500
to
1:1000)
were
used
in
the
studies
and
gave
consis-
tent
results.
Antibodies
against
leukocyte
differentiation
anti-
gens
were
used
according
to
the
manufacturer's
instructions
unless
otherwise
specified:
anti-HLA-DR
(9.3F10,
IgG2,,
American
Type
Culture
Collection
(ATCC),
Rockville,
MD,
or
Anti-HLA-DR,
lgG21,
Bec-
ton
Dickinson
(BD),
San
Jose,
CA),
anti-CD3
(leu
4,
IgG1,
BD),
anti-CD19/CD20
(leu
12,
IgG1,
BD/Leu
16,
IgG1,
BD),
anti-CD14
IgG1),
anti-CD14
(3C-
10,lgG2a,
ATCC,
or
leu-M3,
IgG1,
BD),
anti-CD11b
(OKM1,
IgG1,
ATCC)
or
anti-CD56
(leu
19,
IgG1,
BD),
anti-CD1a
(OKT6,
IgG1,
ATCC).
The
antibody
specific
to
the
Epstein-Barr
virus
nuclear
protein
2
(EBNA
2)
was
PE2
(IgG1,
Young
et
al13).
For
immunohistochemical
tissue
localization
of
p55
or
CD1a,
acetone-fixed
cryostat
sections
of
the
reactive
lymph
node
were
incubated
with
monoclo-
nal
antibodies
for
1
hour.
Slides
were
washed
with
Tris-buffered
saline
and
then
incubated
sequentially
with
rabbit
anti-mouse
immunoglobulin
antibodies
(Dako
Corp.,
Carpinteria,
CA;
1:40
dilution,
using
0.1
mol/L
Tris/HCL
buffer,
pH
7.6,
supplemented
with
4%
human
AB
serum
as
diluent)
and
calf
alkaline
phosphatase
anti-calf
alkaline
phosphatase
immune
complexes
(APAAP;
Dako
Corp;
1:50
dilution).
Anti-
body
localization
was
effected
using
an
alkaline
phosphatase
reaction
with
naphthol
AS-MX
phos-
phate
(Sigma)
as
substrate
and
Fast
Red
TR
salt
(Sigma)
as
chromogen,
as
described
by
Cordell
et
al.14
For
negative
controls,
isotype-specific
mouse
immunoglobulin
or
Tris
buffer,
respectively,
were
substituted
for
the
primary
antibody
in
sequential
sections.
Cytocentrifuge
preparations
of
peripheral
blood
DCs
were
processed
as
positive
controls.
Cy-
tofluorography
of
cell
isolates
was
performed
as
pre-
viously
described.7
Cytospin
centrifuge
preparations
were
made
of
all
isolated
leukocyte
subsets
(20,000
cells
per
slide;
Shandon,
Southern
Instruments,
Pittsburgh,
PA).
Procedures
for
cytospin
immunofluorescence
have
previously
been
described.815
Briefly,
cytocentri-
fuge
preparations
were
fixed
for
30
minutes
in
ice-
cold
methanol.
After
drying,
the
slides
were
stored
at
-70°C
until
use.
Immunofluorescence
localization
of
the
55-kd
protein
and
leukocyte
differentiation
anti-
gens
was
done
on
cytocentrifuge
slides
presoaked
in
PBS
with
1%
human
serum
and
1%
fetal
calf
serum.
The
slides
were
then
sequentially
incubated
with
titered
primary
antibody
reagents
followed
by
FITC-conjugated
secondary
goat
anti-mouse
anti-
body
(1:200;
Sigma).
For
negative
controls,
isotype-
specific
mouse
immunoglobulin
or
PBS
with
1%
fetal
calf
serum,
respectively,
were
substituted
for
the
primary
antibody
in
sequential
cytospin
prepara-
tions.
For
each
antibody
used,
the
frequency
(percent-
age)
of
reactive
cells
was
counted
independently
by
two
investigators
using
results
of
cell
counts
from
at
least
five
fields
or
more
than
250
cells.
Results
The
reactivity
of
the
anti-p55
antibody
was
examined
by
immunofluorescence
staining
of
cultures
of
PB-
MCs
and
purified
subsets
of
granulocytes,
mono-
cytes,
T
cells,
B
cells,
enriched
DCs,
and
FACS-
purified
DCs.
Table
1
presents
results
of
analysis
of
sequential
steps
of
leukocyte
purification
from
a
healthy
donor.
A
panel
of
monoclonal
antibodies

596
Mosialos
et
al
AJP
Febrary
1996,
Vol.
148,
No.
2
*.4
..,.-s
e.
....
..^....X
.::..........
:4
...,.
.:xeX
......
a
c
Figure
1.
Jnmuno/fiuorescence
FITC
staining
of
bulk
PBMCs
ancd
FACS-ptinfied
blood
DCs.
a:
Phase
contrast
coittrol
(X
200)
of
PBMCs.
The
arrow
shous
a
cell
that
expresses
the
p55
antigen
by
immunofluorescence
(b).
b:
p55
FITCstaining
of
the
PBMCfield
shoun
in
a.
c:
Phase
contrast
conttrol
(X
630)
of
FACS-purified
blood
DCs.
d:
p55
FITC
stainzitng
of
FACS-purified
DC,.
specific
for
leukocyte
differentiation
antigens
was
used
in
parallel
with
the
anti-p55
antibody.
Among
unseparated
PBMCs,
0.9%
of
the
cells
were
found
to
be
brightly
positive
by
immunofluores-
cence
microscopy
for
p55
(Table
1,
fraction
1;
Figure
1,
a
and
b).
In
a
second
experiment,
<2%
of
PBMCs
were
positive
for
p55.
When
isolated
leukocyte
sub-
sets
were
stained
for
p55,
most
of
the
p55-positive
cells
were
in
the
metrizamide-separated
fraction
of
enriched
DCs
(fraction
5).
In
this
fraction,
54%
of
the
cells
were
positive
for
p55,
but
this
preparation
also
contains
some
cell
contaminants
from
other
leuko-
cyte
subsets.
Similar
yields
of
DCs
have
been
ob-
tained
in
repeated
experiments
(mean
=
50.4,
SD
=
6.1,
n
=
5).
To
eliminate
cell
contaminants
from
other
leukocyte
subsets,
DCs
were
further
purified
by
FACS
sorting
for
large
cells;
87%
of
this
population
of
purified
DCs
stained
strongly
for
p55
(Table
1,
frac-
tion
8,
Figure
1,
c
and
d).
When
both
bright
and
medium
intensity
cells
were
counted,
96%
of
the
cells
were
positive
for
p55.
As
expected,
95%
of
these
highly
purified
DCs
were
also
strongly
positive
for
HLA-DR
(not
shown).
A
second
experiment
of
FACS
purification
of
DCs
yielded
95%
cells
positive
(bright
and
medium)
for
p55
and
less
than
1
%
reac-
tive
with
a
cocktail
of
anti-CD14,
anti-CD3,
and
anti-
CD57.
In
both
experiments,
the
expression
of
p55
as
measured
by
immunofluorescence
was
comparable
to
that
of
HLA-DR
expression
by
DCs
and
p55
by
Epstein-Barr-virus-infected
cell
lines.
Cytofluorogra-
phy
was
performed
in
two
experiments
and
showed
no
surface
expression
of
p55
by
isolated
DCs
or
T
cells.
p55
localized
diffusely
to
the
cytoplasm
and
was
also
evident
along
dendritic
projections.
In
the
p55-
positive
cells,
the
p55
staining
coincided
with
actin
staining
(not
shown).
p55
staining
was
specific
as
isotype-matched
control
antibodies
(anti-CD3,
anti-
CD20,
and
anti-CD56;
Table
1)
failed
to
stain
DCs.
Monocytes,
granulocytes,
T
cells,
or
B
cells
(frac-
tions
2,
3,
4,
6,
and
7)
were
not
positive
for
p55.
In
two
sets
of
experiments,
phytohemagglutinin-stimu-
lated
cultures
of
T
cells
and
pokeweed-mitogen-
stimulated
cultures
of
FACS-sorted
B
cells
yielded
the
same
low
frequency
(<1%)
of
p55-positive
cells
after
5
days
of
stimulation
as
that
of
unstimulated
cultures
(<1%).
Figure
2
shows
Western
blot
analysis
of
p55
ex-
pression
in
isolated
leukocyte
subsets.
The
Western
blot
analysis
confirmed
the
results
of
the
immunoflu-
orescence
staining.
Thus,
by
Western
blot
analysis,
p55
was
found
to
be
abundant
in
the
DC
fraction
but

Leukocyte
Expression
of
Actin-Bundling
Protein
597
A/P
Februiary
1996,
Vol.
148,
No.
2
1
2 3
4
5
-97
kD
-
66
kD
-45
kD
Figure
2.
Western
blot
analysis
ofpiS
expression
in
v'arious
primnary
lenktoc
te
sub.sets.
The
following
samples
uere
anialyzed:
lane
1,
whole
ccell
extracftromn
1
x
105
DCs;
lane
2,
wbole
cell
extractfromn
I
X
10J
mnoniocytes;
lane
3.
whole
cell
extractifomn
1
x
10
enrliched
B
cells.
lane
4,
nhole
cell
extract
from
I
x
10"
bulk
T
cells;
lane
5,
purified
55-kd
protein
from
HeLa
cells.
The
arrow
indicates
the
position
()/'the
55-kd
actini-btonidlinig
protein.
was
not
detectable
in
other
examined
leukocyte
sub-
sets.
The
functional
capacity,
as
antigen-presenting
cells,
of
the
FACS-sorted
95%
HLA-DR-positive/87%
p55-positive
DCs
was
examined
(fraction
8).
For
these
studies,
the
mixed
lymphocyte
stimulatory
ac-
tivity
(MLR)
of
DCs
was
compared
with
that
of
iso-
lated
monocytes
(Table
1,
fraction
3).
Figure
3
shows
the
results
of
the
MLR
studies
using
limiting
dilutions
of
DCs
or
monocytes
as
stimulator
cells
and
alloge-
neic
T
cells
as
responder
cells.
By
comparison,
the
half-maximal
stimulatory
capacity
of
DCs
was
20-fold
higher
than
that
of
monocytes
(Figure
3).
In
cryostat
sections
of
a
lymph
node,
the
reactivity
for
p55
was
localized
to
mononuclear
cells
in
inter-
follicular
areas
(T
cell
zones)
of
the
node
(Figure
4a).
The
immunoreactive
cells
revealed
a
dendritic
ap-
pearance,
consistent
with
IDCs,
and
exhibited
strong,
cytoplasmic
staining
for
the
p55
(Figure
4b).
The
p55-positive
cells
in
the
T-cell-dependent
zone
constituted
<3%
of
the
cell
population
in
tissue
sec-
tions.
Lymphoid
cells
in
the
T
and
B
cell
zones
were
uniformly
nonreactive.
Similar
results
have
been
ob-
tained
from
lymph
node
specimens
from
four
differ-
ent
sources.
Follicular
DCs
of
B
cell
zones
were
nonreactive,
although
focally,
weak
staining
was
noted
for
occasional
cells.
Studies
of
sequential
sec-
30
Q
C.)
0-
co
C)
CD
20
E
10
-ce
I-
2
1.5
x
103
3
x
103
6
x
103
12.5
x
103
25
x
103
Stimulator
Cells/Well
Figure
3.
Mixced
leukocyte
reaction.
The
MLR
stinmlatory'
capacity
qf
FACS-enriched
DCs
(-)
and
enicbhed
monocytes
(0)
uere
co-culltulred
wvitb
50,
000
allogeneic
T
cells.
As
stimullator
cells
1,
500
to
25,
000
cells
were
used
in
the
experiment.
Day
5
PHithymidine
incorporation
is
sbo'wn
on
the
ordinate.
Number
of
added
(y-irradiated)
stimnlator
DCs
or
inoniocytes
is
shon'n
on?
the
abscissa.
Day
S
13Hithymnidine
incotporation
of
50,000
Tcell.s
alone
was
120
cpmt?
(niot
shown'i).
tions
using
monoclonal
antibody
(OKT6
(CDla)),
re-
vealed
only
small
numbers
of
reactive
cells
in
the
same
distribution
as
the
p55-positive
cells
(Figure
4c).
In
preliminary
studies,
tissue
thymic
DCs
have
also
been
shown
to
be
positive
for
p55
with
predom-
inant
localization
to
the
medulla.
Few
cells
in
kidney
sections
have
staining
for
p55
at
a
frequency
com-
parable
to
cells
expressing
HLA-DR.
Work
is
cur-
rently
in
progress
to
further
characterize
those
p55-
positive
cells.
Discussion
The
present
results
demonstrate
that
among
periph-
eral
blood
cells,
DCs
express
high
levels
of
p55
whereas
monocytes,
T
lymphocytes,
B
lymphocytes,
and
granulocytes
do
not
express
p55.
In
PBMC
cul-
tures
(Table
1),
0.9%
of
the
cells
were
p55
positive,
which
is
in
agreement
with
the
expected
frequency
of
DCs
reported
in
other
studies.1'2
Thus,
the
present
results
suggest
that
p55
expression
is
a
useful
marker
for
identifying
and
potentially
for
quantitating
the
number
of
DCs
in
peripheral
blood
or
in
leuko-
cyte
subsets.
Although
virtually
all
FACS-isolated
DCs
expressed
at
least
moderate
levels
of
p55
(95%
of
the
FACS-sorted
cells
versus
97%
expressing
high
levels
of
HLA-DR),
87%
of
the
FACS-sorted
cells
displayed
high
levels
of
p55.
DCs
that
express
only
a
moderate
level
of
p55
may
constitute
a
distinct
fraction
of
peripheral
blood
DCs
at
a
different
stage
4
-
=00.

598
Mosialos
et
al
AJP
February
1996,
Vol.
148,
No.
2
Figure
4.
Lymph
node,
cryostat
section.
a:
At
lonl!
miagnification
(X200),
frequent
mononuclear
cells
immunoreactiveforpSi
are
observed
in
the
interfollicular
area
of
the
node.
Lymphoid
cells
are
nonreactive.
b:
Higher
miagtnificationi
demonstrates
strong
cytoplasmic
reactivity
for
p55
in
the
population
of
DCs
in
the
interfollictular
regioni
of
the
node.
Note
the
comnplex
elonigated
cellprocesses
of
the
immunoreactive
cells.
c:
In
a
sequential
section
of
the
reactive
niode,
only
small
numbers
of
mononuiclear
cells,
morphbologically
consistent
with
DCs,
are
reactivefor
OKT6(
CDla).
These
cells
are
also
localized
to
inteifollicular
areas.
Immunoalkaline
phosphatase
technique;
methyl
greeni
counterstain.
M1,
niantle
zone:
G,
germninial
center.
of
differentiation
or
of
the
cell
cycle.
Mosialos
et
al5
have
previously
shown
that
p55
is
undetectable
by
Western
blot
analysis
in
two
actively
proliferating
Epstein-Barr-virus-negative
B
lymphoma
cell
lines
(BJAB
and
BL41)
and
is
also
undetectable
by
North-
ern
blot
analyis
in
three
Epstein-Barr-virus-negative
B
lymphoma
cell
lines
(loukes,
BL30,
and
BL41)
and
two
T
cell
lines
(Jurkat
and
MOLT4).
These
results
further
support
the
selective
expression
of
p55,
al-
though
we
cannot
exclude
the
possibility
that
a
rare
B
cell
or
T
cell
population
(less
than
1%
of
the
total
B
and
T
cell
population)
may
express
p55.
In
phytohe-
magglutinin-activated
PBMC
cultures
containing
DCs,
a
transient
up-regulation
of
p55
RNA
(14
to
48
hours)
has
been
reported.16
The
results
also
show
that
the
expression
of
p55
in
lymph
node
sections
is
restricted
to
cells
resembling
IDCs
and
is
not
found
expressed
in
the
follicles.
This
is
in
agreement
with
the
current
theory
that
IDCs
are
derived
from
a
cell
lineage
different
from
FDCs.34
IDCs
in
the
lymph
node
are
thought
to
be
derived
from
peripheral
blood
DCs
migrating
to
the
lymphoid
tissues.
17,18
A
small
number
of
CDla-positive
DCs
was
found
in
lymph
node
sections
in
the
T
cell
zones.
CDla
is
thought
to
be
an
early
maturation
marker
present
on
bone-marrow-derived
DCs.
The
low
frequency
of
CDla-positive
DCs
could
be
explained
by
a
small
number
of
immature
DCs
in
the
lymphoid
tissues
or
a
loss
of
the
CDla
antigen
upon
maturation
in
periph-
eral
blood
and
in
lymph
nodes.19
20
The
extent
of
the
reactivity
with
cell
subsets
in
other
tissues
is
pres-
ently
under
investigation.
Both
in
vitro
studies
and
tissue
studies
have
not
demonstrated
reactivity
with
activated
B
cells
in
culture
or
staining
with
B
cell
follicles
that
contain
activated
B
cells.
It
is
possible
that
the
p55
antigen
is
also
expressed
in
immature
bone-marrow-derived
DCs
and
Langerhans
cells.
Initial
studies
(not
shown)
have
indicated
that
Lang-
erhans
cells
express
the
p55
antigen,
but
additional
studies
are
in
progress
to
asses
the
extent
of
p55
expression
in
immature
DCs.
Several
studies
of
tis-
sue
sections
from
thymus
(two
samples),
lymph
node
(three
samples),
and
tonsil
(one
sample)
have
shown
that
cells
of
dendritic
morphology
and
reac-
tive
with
the
p55
antibody
consistently
localize
in
the
T-cell-dependent
areas
of
the
lymph
node
and
tonsil
as
well
as
in
the
medulla
of
the
thymus.
In
kidney
sections
(four
samples),
occasional
cells
of
dendritic
morphology
in
the
interstitium
stain
for
the
p55
anti-
gen.
Proliferating
cells
in
histiocytosis
X
are
thought
to
be
derived
from
cells
of
the
dendritic
lineage.21'22
Preliminary
data
(Birkenbach
et
al,
personal
commu-
nication)
have
shown
that
cells
from
histiocytosis
X
lesions
express
the
p55
antigen.
p55
is
one
of
the
few
markers
for
human
DCs.
Recent
work
has
described
expression
of
the
HB-15
antigen
by
DCs
of
the
blood
and
thymus.23
However,
the
HB-15
antigen
is
also
expressed
by
FDCs.
Fur-
thermore,
the
HB-15
antigen
is
expressed
by
acti-
vated
T
cells.23
The
S-100
antigen
has
a
wide
tissue
distribution
and
is
expressed
in
many
cell
types
of
different
somatic
origin
including
DCs.24
I&

Leukocyte
Expression
of
Actin-Bundling
Protein
599
AJP
February
1996,
Vol.
148,
No.
2
The
expression
of
p55
by
DCs
adds
to
the
com-
plexity
of
the
differentiation-specific
expression
of
this
particular
actin-bundling
protein.
Actin
bundling
proteins
regulate
rearrangements
of
the
cytoskeleton
or
interaction
between
the
cytoskeleton
and
mem-
branes
in
response
to
extra-
or
intracellular
sig-
nals.25
Actin-bundling
proteins
direct
the
three-
dimensional
polymerization
of
actin
filaments
in
mi-
grating
cells.
Related
to
cell
motility
is
also the
ability
to
process
and
present
antigen.25
Thus,
the
migra-
tory
nature
and
unique
antigen-presenting
proper-
ties
of
DCs
suggest
a
critical
role
of
p55
in
this
specialized
subset
of
leukocytes.
p55
is
highly
con-
served
from
sea
urchin
through
Drosophila
to
mam-
malian
species.5
In
Drosophila,
mutations
in
p55
re-
sult
in
a
singed
hair
phenotype
and
sterility.
In
humans,
p55
expression
is
found
also
in
the
den-
drites
of
specific
neurons
in
the
central
nervous
sys-
tem.
In
rats,
a
high
level
of
expression
has
also
been
demonstrated
in
neural
cells
in
vitro
and
in
neural
growth
cones.5
Migration
of
axons
during
the
devel-
opment
of
the
fetal
nervous
system
is
functionally
related
to
the
process
of
cell
migration.25
In
vivo
engagement
of
DCs
in
disease
is
sug-
gested
from
studies
in
rodents,
but
studies
in
hu-
mans
have
been
limited
by
the
lack
of
a
strongly
reactive
and
specific
marker
for
these
specialized
cells.
Of
immediate
clinical
interest
is
the
potential
use
of
the
p55
protein
as
a
marker
for
the
possible
engagement
of
dendritic
leukocytes
in
organ
trans-
plantation,
HIV-1
pathogenesis,
and
autoimmunity.
Acknowledgments
We
thank
A.
M.
Bruzzese
for
technical
assistance.
References
1.
Steinman
RM:
The
dendritic
cell
system
and
its
role
in
immunogenicity.
Annu
Rev
Immunol
1991,
9:271-296
2.
Steinman
RM,
Cohn
ZA:
Identification
of
a
novel
cell
type
in
peripheral
lymphoid
organs
of
mice.
J
Exp
Med
1974,
139:380-397
3.
Schreiver
F,
Nadler
LM:
The
central
role
of
follicular
dendritic
cells
in
lymphoid
tissues.
Adv
Immunol
1992,
51:243-283
4.
Klaus
GGB,
Humphrey
JH,
Kunki
A,
Dongworth
DW:
The
follicular
dendritic
cell:
its
role
in
antigen
presen-
tation
in
the
generation
of
immunological
memory.
Im-
munol
Rev
1980,
53:243-283
5.
Mosialos
G,
Yamashiro
S,
Baughman
R,
Matsudaira
R,
Matsumura
F,
Vara
L,
Kief
E,
Birkenbach
M:
Epstein-
Barr
virus
infection
induces
expression
in
B
lympho-
cytes
of
a
novel
gene
encoding
an
evolutionarily
con-
served
55-kd
actin-bundling
protein.
J
Virol
1994,
68:
7320-7324
6.
Yamashiro-Matsumura
S,
Matsumura
F:
Purification
and
characterization
of
an
F-actin-bundling
55-kilodal-
ton
protein
from
HeLa
cells.
J
Biol
Chem
1985,
260:
5087-5097
7.
Langhoff
E,
Steinman
RM:
Clonal
expansion
of
human
T
lymphocytes
isolated
by
dendritic
cells.
J
Exp
Med
1989,
169:315-320
8.
Langhoff
E,
Terwilliger
EF,
Bos
HJ,
Kalland
KH,
Poznansky
MC,
Bacon
OML,
Haseltine
WA:
Replication
of
human
immunodeficiency
virus
type
1
in
primary
dendritic
cells.
Proc
Natl
Acad
Sci
USA
1990,
88:7998-
8002
9.
Langhoff
E,
Kalland
KH,
Haseltine
WA:
Early
molecular
replication
of
human
immunodeficiency
virus
type
1
in
cultured-blood-derived
T
helper
dendritic
cells.
J
Clin
Invest
1993,
91
:2721-2726
10.
Langhoff
E,
Jakobsen
BK,
Platz
P,
Ryder
LP,
Sveigaard
A:
The
impact
of
low
donor-specific
MLR
versus
HLA-DR
compatibility
on
kidney
graft
survival.
Trans-
plantation
1985,
39:18-21
11.
Langhoff
E,
Ladefoged
J,
Dickmeiss
E:
The
immuno-
suppressive
potency
of
various
steroids
on
peripheral
lymphocytes,
T
cells,
NK
and
K
cells.
Int
J
Immuno-
pharmacol
1985,
7:483-489
12.
Yamashiro-Matsumura
S,
Matsumura
F:
Intracellular
lo-
calization
of
the
55-kd
actin-bundling
protein
in
cultured
cells:
spatial
relationships
with
actin,
a-actinin,
tropomy-
osin,
and
fimbrin.
J
Cell
Biol
1986,
103:631-640
13.
Young
L,
Alfieri
C,
Hennessy
K,
Evans
H,
O'Hara
C,
Anderson
KC,
Ritz
J,
Shapiro
RS,
Rickinson
A,
Kieff
E,
Cohen
JI:
Expression
of
Epstein-Barr
virus
transforma-
tion-associated
genes
in
tissues
of
patients
with
EBV
lymphoproliferative
disease.
N
EngI
J
Med
1989,
321:
1080-1085
14.
Cordell
JL,
Falini
B,
Erber
WN,
Ghosh
AK,
Abudulaziz
Z,
MacDonald
S,
Pulford
KAF,
Stein
H,
Mason
DY:
lmmunoenzymatic
labeling
of
monoclonal
antibodies
using
immune
complexes
of
alkaline
phosphatase
and
monoclonal
anti-alkaline
phosphatase
(APAAP
com-
plexes).
J
Histochem
Cytochem
1984,
32:219-229
15.
Kalland
KH,
Szilvay
AM,
Langhoff
E,
Haukeness
G:
Subcellular
distribution
of
human
immunodeficiency
vi-
rus
type
1
Rev
and
colocalization
of
Rev
with
RNA
splicing
factors
in
a
speckled
pattern
on
the
nucleo-
plasm.
J
Virol
1994,
68:1475-1485
16.
Duh
F-H,
Latif
F,
Weng
Y,
Geil
L,
Modi
W,
Stackhouse
T,
Matsumura
F,
Duan
DR,
Linehan
WM,
Lerman
Ml,
Gnarra
JR:
cDNA
cloning
and
expression
of
the
human
homolog
of
the
sea
urchin
fascin
and
Drosophila
singed
genes
which
encode
an
actin-bundling
protein.
DNA
Cell
Biol
1994,
13:821-827
17.
Macatonia
SE,
Edwards
AJ,
Knight
SC:
Dendritic
cells
and
the
initiation
of
contact
sensitivity
to
fluorescein
isothiocyanate.
Immunology
1986,
59:509-514
18.
Larsen
C,
Steinman
R,
Witmer-Pack
M,
Hankins
D,
Morris
P,
Austyn
J:
Skin
migration
and
maturation
of

600
Mosialos
et
al
AJP
February
1996,
Vol.
148,
No.
2
Langerhans
cells
in
skin
transplants
and
explants.
J
Exp
Med
1990,
172:1483-1493
19.
Schuler
G,
Steinman
RM:
Murine
epidermal
Langer-
hans
cells
mature
into
potent
immunostimulatory
den-
dritic
cells
in
vitro.
J
Exp
Med
1985,
161
:526-546
20.
Romani
N,
Lenz,
A,
Glassel
H,
Stbssel
H,
Stanzl
U,
Majdic
0,
Fritsch
P,
Schuler
G:
Cultured
human
Langerhans
cells
resemble
lymphoid
dendritic
cells
in
phenotype
and
function.
J
Invest
Dermatol
1989,
93:600-609
21.
Beckstead
JH,
Wood
GS,
Turner
RR:
Histiocytosis
X
and
Langerhans
cells.
Hum
Pathol
1984,
15:826-833
22.
Willman
CH,
Busque
L,
Griffith
BB,
Blaise
MS,
Favara
E,
McGlain
KL,
Duncan
MH,
Gilliland
DG.
Langerhans'-
cell
histiocytosis
(histiocytosis
X):
a
clonal
proliferative
disease.
N
Engl
J
Med
1994,
331:154-157
23.
Zhou
L-J,
Scharting
R,
Smith
HM,
Tedder
TF:
A
novel
cell-surface
molecule
expressed
by
human
interdigitat-
ing
reticulum
cells,
Langerhans
cells
and
activated
lymphocytes
is
a
new
member
of
this
immunoglobulin
superfamily.
J
Immunol
1992,
149:735-742
24.
Vanstapel
M-J,
Gatter
KC,
de
Wolfe-Peeters
C,
Manson
DY,
Desmet
VD:
New
sites of
human
S-100
immunore-
activity
detected
with
monoclonal
antibodies.
Am
J
Clin
Pathol
1986,
85:160
25.
Stossel
TS:
On
the
crawling
of
animal
cells.
Science
1993,
260:1086-1094