B. Toews. Departments of Internal Medicine. (L.P.N., J.C.W., G.B.T.) and Pathology. (M.F.L.),. The University ...... Elias,. J.A.,. Kamoun,. M., Daniele,. R.P., and Rossan,. M.D.. Alveolar macrophages, blood ... 1986. 16. Kaltreider,. H.B., Caldwell,.
Journal
of Leukocyte
Mononuclear Cells From Human Lung Support Antigen-Induced T Lymphocyte Laurent
P. Nicod,
Departments
Mary
of Internal
Medicine
We
previously
have
F. Lipscomb, (L.P.N.,
J.C.W.,
Jonathan
demonstrated
that
there
(M.F.L.),
and
Galen
The University
of Texas
of loosely
adherent
is a subpopulation
45:336-344
(1989)
Parenchyma Proliferation
C. Weissler,
G.B.T.) and Pathology Center at Dallas, Dallas
Biology
B. Toews Health
Science
pulmonary mononuclear cells that can be isolated from minced and enzyme-digested lung tissue with a potent capacity to stimulate allogeneic T lymphocyte proliferation. We now demonstrate that these cells are also capable of stimulating an autologous mixed leukocyte reaction (AMLR) and presenting antigen to autologous T lymphocytes. These loosely adherent mononuclear cells (LAM) were more effective than either alveolar macrophages or monocytes as antigen-presenting cells. Depletion of phagocytic or Fc receptor-positive cells from the LAM population enhanced the stimulation of an reaction AMLR while preserving antigen-induced T lymphocyte proliferation. These results indicate that there are nonphagocytic, Fc receptor-negative accessory cells in human lung parenchyma capable of activating resting T cells in an AMLR and supporting antigenspecific T lymphocyte proliferation. The Identity of these cells is uncertain, but the data strongly suggest that the cell is not a classical monocyte-derived macrophage. These antigen-presenting cells may be critical in the initiation of immune responses within the lung. Key words:
antigen
presentation,
human
INTRODUCTION The expression is essential
of an immune
to enhance
a host’s
response
within
the lung
capacity
to clear
certain
infectious agents [1,10]. Immune responses develop in hilar lymph nodes [17] in response to antigens delivered to the lung via the tracheobronchial tree. Whether immunity can be initiated in the parenchyma of the lung is still controversial [15,25]. However, the involvement of the pulmonary interstitium in diseases such as hypersensitivity
pneumonitis
[20]
or sarcoidosis
[14]
to identify macrophages The
majority
antigens
[11 ,38].
Previous
studies
an APC in the lung examined (AM) obtained by bronchoalveolar of previous
studies
demonstrated
pulmonary
Interstitium
human lung tissue which were potent stimulators of an allogeneic mixed leukocyte reaction (MLR) [29]. Accessory cell function was significantly enriched in a loosely adherent population of PMC, suggesting that loosely adherent mononuclear cells (LAM) in human lung parenehyma might present soluble antigen. However, some authors have suggested that the ability of an accessory cell to stimulate a MLR may not correlate with presentation of soluble antigen [28,311. Indeed, in a prior study from our laboratory PMC from mechanically digested
suggests
that under certain circumstances cellular and/or humoral immune responses occur in the interstitium of human lung. The generation of an immune response requires the activation of T lymphocytes by an antigen-presenting cell (APC) which must express class II major histocompatibility
lung,
Galen
attempting
alveolar lavage. that
Received
sity
August
B. Toews
3,
is now
of Michigan.
Laurent
P. Nicod
1988;
accepted
at Department
Ann
Arbor,
is now
Cantonal
de Geneva,
1211
collected from most normal volunteers are poor stimulators of antigen-induced T lymphocyte proliferation when compared to peripheral blood monocytes (Mo) [6,16,
ionathan
C.
is the
37,40,41].
We have mononuclear
isolated a population of pulmonary (PMC) from enzymatically digested
© 1989 Alan R. Liss, Inc.
of Internal
1988. Medicine,
Univer-
Geneva,
of Internal
Medicine,
Hopital
Switzerland.
recipient
of
Clinical
Investigator
Award
01797.
Reprint
recently cells
Weissler
21,
48109.
MI.
at Department
AM
HL
November
Medicine, Medical
requests:
ionathan
Pulmonary Center,
Dallas,
C. Weissler, Division, TX
M.D.,
University 75235-9034.
Department of
Texas
of Internal Southwestern
Antigen lung tissue
stimulated
an MLR
but did not present
antigen
Presentation
[29] with Type!
in Lung
collagenase
Parenchyma
(150
units/mI)
337 and Type
hA
[411.
elastase
The purpose of the current study was to determine whether PMC from minced and enzyme-digested lung could present soluble antigen to autologous peripheral blood T lymphocytes. In contrast to previous studies using mechanical digestion alone, PMC were more effective in presenting antigen than AM. Further enrichment of APC was obtained in the loosely adherent population of cells (LAM), and APC function was maintained following depletion of either Fe receptorpositive (FcR+) cells or phagocytic cells, suggesting that the APC was not a classical macrophage. These results demonstrate that there are mononuclear cells in human lung parenchyma capable of inducing antigenspecific lymphoproliferation in response to foreign antigen.
Louis, MO) for 90 mm. The enzyme-digested lung was then tapped gently through a stainless steel screen to dissociate the tissue into single cells. A Ficoll-Hypaque gradient (Ficoli from Sigma Chemical Co., St. Louis,
MATERIALS Subjects
normal
Grossly surgical
AND METHODS
specimens
pulmonary resected
tissue from
was
obtained
patients
with
from primary
lung carcinoma. 90 ml of blood bronchoalveolar
To isolate lymphocytes and monocytes, was drawn. Four patients also underwent lavage at the time of thoracatomy after
their
consent
informed
smokers. pulmonary
was
obtained.
All
No patient had clinical evidence infection at the time of surgery.
Preparation
of Cells
From
patients
of
were
active
lung
specimen
was
units/ml,
Hypaque
both
from
from
Sigma
Chemical
Winthrop-Breon,
New
Co,
York,
St.
NY)
was used to separate the dissociated cells. The pulmonary mononuclear cells at the interface were washed three times in HBSS. Viability was assessed by trypan blue and was >90%. Differential counts were performed on Wright-stained smears: 70-75% were macrophagelike; 20-25% were lymphocytes. To evaluate the contamination
of
PMC
by
blood
mononuclear
cells,
the
number of blood mononuclear cells/g of hemoglobin in peripheral blood and the hemoglobin content of the lung digests
was
determined.
The
number
of blood
mononu-
clear cells in the lung cell population was determined by multiplying the lung digest hemoglobin by the number of mononuclear cells/g of hemoglobin in peripheral blood. Maximal contamination of PMC by peripheral blood cells averaged 3% (range 1-8%). PMC were fractionated into low-density, loosely adherent fractions of mononuclear cells (LAM) and firmly adherent mononuclear cells (FAM) according to the method of VanVoorhis et al. [39]. Briefly, 150-350 X 106 cells were incubated in complete medium containing 10% heat-inactivated fetal calf serum (FCS, Gibco LabChagrin
Falls,
OR),
on
100-mm
tissue
culture
dishes (Corning Glass Works, Corning, NY) in aliquots of 15-20 x l0 cells/dish at 37#{176}C.The nonadherent
Lung
initially
MO;
oratories,
Bronchoalveolar lavage was performed in a lobe adjacent to the lobe to be resected using a total of 200 ml of 0.9% normal saline as previously described [37]. Cells from bronchoalveolar lavage (BAL) were washed twice with Hanks’ balanced salt solution (HBSS; Microbiological Associates, Walker, MD). The cells were resuspended in complete medium (RPM! 1640 with 25 mM HEPES (Inland Laboratory, Austin, TX), 2 mM glutamine, 100 i.g/ml penicillin, 100 pg/ml streptomycm (all from Gibco Laboratories, Grand Island, NY), and 24 g/ml gentamiein (Eikins-Sinn, Cherry Hill, NJ)) plus 20% heat, inactivated pooled human serum (HS). Viability was assessed by trypan blue exclusion and was greater than 95%. BAL consisted of more than 90% large cells with complex cytoplasm consistent with AM. It has been previously shown that isolating AM from BAL does not change their inability to present antigen or stimulate an MLR [37]. The
(10
lightly
pressed
and
rinsed with HBSS to remove residual blood. PMC were obtained from whole lung by mincing the 30 g of tissue with scissors in a petri dish filled with complete medium. The lung fragments were digested as reported previously
cells
were
removed
Nonadherent
cells
after
1 h with
represented
half
three of the
rinses
of HBSS.
PMC;
30-40%
were lymphocytes, 10-20% were unidentifiable cells, and the remaining cells were mostly macrophages as determined by Wright-stained cytocentrifuge smear preparations. The adherent PMC were incubated for an additional 14-16 h at 37#{176}C in complete medium with 10% FCS. About half of the cells failed to remain adherent after three rinses with HBSS. These cells were recuitured
on
plastic
culture
dishes
for
1 h.
About
one-third of the cells readhered and the remainder were removed by washing with HBSS. These loosely adherent, twice released cells were resuspended in complete medium and floated on bovine plasma albumin of density
of
1 .080
g/ml
(Pentese;
Miles
Lab,
Naperville,
IL). After centrifugation at lO,000g for 15 mm at 4#{176}C the low-density fraction cells at the interface were harvested. These cells are referred to as loosely adherent mononuclear cells (LAM). The population of cells that remained adherent after overnight incubation constituted the firmly adherent mononuclear (FAM) fraction. After an incubation of 30 mm at 4#{176}C in cold phosphate buffered saline, FAM were
338
Nicod
removed
from
et al the dishes
by gentle
scraping
with a rubber
policeman.
Preparation
of FcR-
Sheep
red
blood
subagilutinating
(Difco
cells
doses
Lab,
MI)
FeR + cells, with 5-10
at 900 rpm x room temperature,
The cell gradient. were
in the
were
opsonized
antibody
with
against
as previously
SRBC
described
[35].
washed, opsonized SRBC (1%) x 106 LAM and gently pelleted
10 mm. After the sediment
mixture Rosetting
LAM
(SRBC)
of rabbit
Detroit,
To remove were mixed
cells
or FcR +
30 mm of was gently
was separated (FeR +) and sediment
and
on
incubation resuspended.
at
a Ficoll-Hypaque
nonrosetting
at the
(FcR-)
interface,
respec-
of Nonphagocytic
LAM
Silica particles (0. 1 pm) were a gift from Dr. K. Rabock, Stein-Kohlenberg-Bauverian, 43 Essen-Krey, West Germany. Two milliliters of LAM at a concentration of 5 x l06/ml were mixed with 250 ig/ml of silica in 17
x
100
Plastics
polypropylene
Co.,
culture
Oxnard,
CA).
tubes
The
mixture
rpm x 10 mm, then at 37#{176}C. A Ficoll-Hypaque gradient pelleted the
at 900
cell/silica
negative
mixture.
(Sil-)
silica-laden the gradient.
cells
phagocytic
The
Peripheral donors were washed
3 x
Blood
in HBSS.
Monocytes
reincubating
plastic culture lymphocytes,
3 of
rosetted
ester with
treated
(AET)
X
in HBSS.
accessory as previously
and
column
silica
were
were and
with described
SRBC
to remove rabbit
were
lung and
isolated
by
nonadherent by washing
recovered by acid (EDTA),
removing
the
were
cells
further
5.0 mM L-leucine [22]. T cells were
at 4#{176}C followed
bromideby
sedimentation
After lysis of SRBC with 0. 15 were passed through a nylon residual
treated
B lymphocytes
with
complement
as
anti-HLA-DR previously
and
described
Antigen-Induced T Lymphocyte Proliferation Assay Cultures
were plates
Purified
FAM,
T cells
numbers LAM,
(l0)
of
were
cultured
accessory
FeR+,
alone
cells.
FeR-,
and
or
AM,
Mo,
cells
were
Sil-
irradiated (4,500 rads) in a cesium source. Cultures were incubated for 6 days in 5% CO2 without or with soluble antigen in a concentration previously determined to give optimal
T
cell
proliferation.
Antigens
utilized
were
soluble mumps antigen (M.A. Bioproducts, Walkersyule, MD) at a final dilution of 1:2 or tetanus toxoid (Mass. Biologic Laboratories, Boston, MA) at a final dilution of 1:50. Eighteen hours before termination of cultures,
0.5
cultures.
Cells
1iCi
and
counter.
Data
(average
CPM
wells
without
eration
and
3H-thymidine
were counted were
as total
Monoclonal The
in
in wells
a
for
added
z
scintillation
counts
per
minute
CPM
antigen-induced in wells
leukocyte
to
cell
liquid
antigen-average T cell
without
(AMLR).
antibodies
used
in
prolif-
antigen
reaction
for the
Antibodies
mouse
monoclonal
(anti HLA-DR, from cell line Lampson, et al. [19]); MO2 generated (antihuman
was
an automated
beta as
with
CPM
mixed
with
expressed
antigen)
autoiogous
(3HTdR)
collected
against AM
by Dr.
R. Todd
licular
dendritie
the cell originally
line macrophage described and
III et al. [2,18]);
cells
Barbara,
from
CA]);
and
were
L243
243 originally described by (anti-Mo antibody initially p9 [2]); PAM1 kindly provided
DAKO-DRC1 (antifolline R4/23 [Dakopatts,
cell rabbit
anti-Factor
VIII
antise-
rum (to detect endothelial cells as previously described [27] from Dakopatts). The endothelial cells used as a positive control for the anti-Factor VIII antiserum were kindly
provided
by Dr.
(0.25-0.5
X
106)
relevant
monoclonal
monoclonal
antibody
pra,
TX).
Dallas,
After
cell were washed fluorescein-labeled
(Cooper 30 mm
0.2%
tive cells with
Ziff
(Dallas,
suspended
in
antibody
or
(kindly
provided in
sodium
a fluorescent
by Dr.
RPMI rabbit
The
Don
CaA hgG
was added for 3 X in RPM!
percentage
of posi-
at least
equipped
the
at 0#{176}C, the
1640 media. antimouse
by examining
microscope
of
control
for 30 mm
azide.
was determined
Cells
iii
anti-arsonate
incubation
twice F(ab’)2
TX). 50
Biomedical, West Chester, PA) at 0#{176}C. Cells were then washed
containing
cence
Morris
were
with
200 cells epifluores-
optics.
Mo.
(L243)
[29].
microtiter
PMC,
Santa
Monocytes
2-aminoethylisothiouronium
T cells baby
or
interface, whereas to the bottom of
Lymphocytes
cells
through Ficoll-Hypaque. M NH4C1, the lymphocytes Lastly,
to separate
dishes. The were removed
at 37#{176}C for 45 mm,
washing
wool
was used
(Mo)
various
gently
for 30 mm
nonphagoeytic
3x with HBSS. Adherent cells adding 10 mM ethylenediaminetetraacetie
depleted methyl
was
blood mononuclear cells (PBM) from obtained on a Ficoll-Hypaque gradient
adhesion on cells, mostly
by
Falcon
incubated
remained at the cells sedimented
Preparation of Peripheral and Lymphocytes
(2005,
HS.
with
harvester
tively.
Preparation
20%
performed in triplicate in flat-bottomed in 200 pi of complete medium with
RESULTS Comparison of BAL and PMC Antigen-Induced T Lymphocyte In previous
studies
nonenzymatieaily antigen-induced
tively
than
PMC
in Supporting Proliferation
obtained
digested lung lymphoproliferation
BAL
or AM
isolated
from
from did
minced not
any
BAL
but
stimulate more effec-
[41].
In order
Antigen
Presentation
TABLE 1. Comparison of BAL and PMC in Supporting Antigen-Induced T Lymphocyte Proliferationa
T cells
>-
T cells
PMC/BAL
PMC
(ratio)
Experiment
only
1b
414
3.358
11,125
2
50
2,534
5,211
2.0
3b
0
4,849
4.9
+
BAL
+
o
6
BAL
I I.,
983
814
1,303
ll,976
9.2
5
156
1.368
5,316
3.8
autologous
T cells
(l0)
were
cultured
with
BAL
4
o Ea 0.
0
2oo
or PMC -
(2.5 x tTetanus cMumps
339
#{149} LAM
I.-
3.3
4C
ap.unfied
Parenchyma
8-
CPM T cells
in Lung
l0) both with and without soluble was used as the soluble antigen. was
used
as the
soluble
antigen.
antigen.
to improve the yield of APC from the interstitium, lungs were minced and digested with collagenase and elastase. PMC obtained after enzymatic digestion supported antigen-induced T lymphocyte proliferation two to five times more effectively than BAL (Table 1). Enzymatic treatment of BAL with coliagenase or elastase neither diminished nor enhanced their ability to stimulate an MLR (data not shown). Thus the differences between the PMC and BAL could not be explained by enzymeinduced changes on these cells. Comparison Supporting Proliferation
of PMC, FAM, Antigen-Induced
I
Number
of Stimulator
50 Cells
(a 1O)
Fig. 1. Dose-reponse curve with BAL (which consist predominantly of AM) and LAM in supporting antigen-Induced T lymphocyte proliferation. Autologous T cells (105/well) were cultured in triplicate for 6 days with various numbers of APC.
n6