| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Technical Advances |
-Secreting T-Cell Populations in Rejecting Murine Cardiac Allografts
From the Department of Pathology,*
Immunology Division,
and Department of Medicine,§
Brigham and
Women's Hospital, Boston, Massachusetts, the Department of
Surgery,
Osaka University Medical School,
Osaka, and Department of Cardiothoracic
Surgery,
Tokyo Medical and Dental University,
Tokyo, Japan
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, and interleukin (IL)-4 in acutely rejecting murine
cardiac allografts. We also measured the cytokine production via
ELISA, RPA, and semiquantitative competitive RT-PCR.
The number of CD4+ cells producing IFN- increased as
rejection proceeded, in agreement with previous data; we
detected no IL-4 production at any time, although relatively
low numbers of IL-10-producing cells were identified. In
addition, a high percentage of CD8+ cells,
which outnumber CD4+ cells at day 6 after
transplant, also produce IFN-, suggesting that
cytotoxic lymphocytes contribute significantly to the local cytokine
milieu. This new application of intracellular cytokine staining
provides a powerful methodology for studying transplantation
immunology. The method may also be easily adapted to the study of other
immune-mediated processes.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Helper T cells can differentiate into two different functional subtypes
when exposed to antigen.5
These subtypes express and
secrete different panels of cytokines and therefore are believed to
play different roles in immunologically mediated processes. T-helper-1
(Th1) cells are effectors of cell-mediated immunity; they secrete
primarily interleukin (IL)-2, which stimulates the growth and
activation of T lymphocytes, and interferon (IFN)-, which, among
other functions, activates macrophages.5
Th1 cytokines
likely stimulate the development of acute rejection, and their
transcript levels have been correlated with acute graft rejection
severity.6-8
Conversely, T-helper-2 (Th2) cells express
IL-4, IL-5, and IL-10 and appear to antagonize the effects of Th1 cells
in vivo.5
For example, Th2 cells may modulate
acute rejection and the development of allograft
tolerance.2
Th2 cytokine transcripts correlate with milder
rejection episodes.6,9
Similarly, cytotoxic T lymphocytes
may differentiate into T cytotoxic (Tc)1 and Tc2 populations with
cytokine profiles analogous to those seen in T helper
subpopulations.4
The degree to which strict Th1/2 and Tc1/2
differentiation takes place in humans, and the precise roles that these
cells play in rejection processes, still remains unclear.
Experimental methods used to examine the presence and contribution of discrete T cell subpopulations in acute graft rejection include reverse transcriptase polymerase chain reaction (RT-PCR), RNase protection assay (RPA), in situ hybridization, immunohistochemistry, radioimmunoassay (RIA), and ELISA. Each method has important limitations. For example, levels of cytokine transcript detected by RT-PCR may not always correlate with cytokine protein levels; many important cytokine transcripts contain AU sequences in the 3' untranslated region that confer mRNA instability.10 In addition, RT-PCR is at best semiquantitative and is usually performed on whole-tissue homogenates, giving no information about the source of the detected transcript. Likewise, RPA cannot distinguish the source of particular cytokines. Although in situ hybridization can be used to identify secreting cell types, it also detects RNA levels, not protein. It is technically difficult, and levels of cytokine transcript present in vivo may be insufficient for detection. Immunohistochemistry also allows identification of cytokine-secreting cell type and detects cytokine protein directly; however, it gives variable results and lacks a high degree of sensitivity, and the process of generating statistically meaningful data are tedious. Ex vivo methods such as RIA, ELISA, and bioassay detect the net secretions of heterogeneous, captive, and therefore artificial, culture populations that have often spent days outside the allograft environment; these methods also provide no information about secreting cell type. The need clearly exists for techniques that can quantify cytokine protein expression, generate statistically analyzable data, and allow simultaneous identification of the source of cytokine protein.
Intracellular cytokine staining (ICCS) with flow cytometric analysis was recently described as a technique for examining cytokine expression in cloned T cell lines.11-13 Briefly, intracellular cytokine staining involves incubating cells with an anti-cytokine antibody in the presence of a mild detergent that permeabilizes the cell surface and allows the antibody to pass through cytoplasmic and organelle membranes and bind to intracellular cytokine protein. Cells can then be resealed and stained with monoclonal antibodies against cell-surface markers, such as CD4 or CD8, so that cytokine expression can be correlated with cell type using multicolor flow cytometry.
When applied to the study of acute graft rejection, this method offers many advantages. First, cell-surface staining may be used to identify a specific subpopulation (eg, CD4+ or CD8+) of graft-infiltrating lymphocytes from within a heterogeneous mixture, without the need for physical isolation of the cells. Using intracellular staining techniques and multicolor flow cytometry, one may define the particular cytokines produced by a specific cell type and the number of cells producing a given cytokine. Flow cytometry permits analysis of a large number of cells and the generation of statistically significant data.
To evaluate the applicability of this technique to the study of solid
tissue processes, we developed a method for isolating and stimulating
cells from acutely rejecting allografts and optimized intracellular
staining conditions for the isolated lymphocyte populations. We defined
the CD4+ and CD8+ lymphocyte populations within
rejecting allografts by surface staining and demonstrated an increase
in both populations as acute rejection proceeded. Finally, we used ICCS
to determine the relative numbers of CD4+ and
CD8+ cells within rejecting allografts producing IFN- or
IL-4 at each time point.
Our results demonstrate increasing Th1-type and Tc1-type infiltrates as
acute rejection proceeds, without any detectable IL-4. Moreover, we
find that, at the time of graft failure, CD8+ T lymphocytes
predominate as a source of IFN-.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
C57/BL6 (B6, H-2b) and BALB/c wild-type mice were obtained from
Taconic Farms (Germantown, NY) or the Jackson Laboratories (Bar Harbor,
ME). All mice were maintained in the pathogen-free Harvard Medical
School facility on acidified water; all experiments conformed with
approved animal care protocols. The Riboquant multiprobe RNase
protection assay (RPA) kit, in vitro transcription kit and
mCK-1 template set, biotinylated anti-IFN- (rat IgG1
clone XMG1.2), biotinylated anti-IL-4 (rat IgG1, clone
BVD6–24G2), biotinylated anti-IL-10 (rat IgG2b, clone
JES5–16E3), biotinylated control antibody (anti-IgE rat
IgG1 clone R35–92 or IgG1 isotype-matched
control, clone R3–34), fluorescein isothiocyanate (FITC)- conjugated
anti-CD4 (clone RM4–5), Cy-chrome (CyC)- conjugated anti-CD4
(clone RM4–5), phycoerythrin (PE)-conjugated anti-CD8 (clone 53-6.7),
Cy-C-conjugated anti-CD8 (clone 53-6.7), PE-conjugated streptavidin,
unlabeled anti-IFN- (rat IgG1, clone R4–6A2), unlabeled
anti-IL-4 (rat IgG1, clone 11B11), and unlabeled
anti-CD16/CD32 (clone 2.4G2, Fc block) were obtained from PharMingen
(San Diego, CA). Horseradish peroxidase (HRP)-conjugated ultra-avidin
was from Leinco Technologies, Ballwin, MO. Ionomycin, phorbol myristate
acetate (PMA), brefeldin A (BFA), collagenase (C1030, type I), saponin,
bovine serum albumin (BSA),
2,2'-azino-bis(3-ethylbenz-thiazoline-6-sulfonic acid), and
paraformaldehyde were obtained from Sigma Chemical Co. (St. Louis, MO).
IL-2 from the supernatant of X63 cell cultures was the generous gift of
Dr. Andrew Lichtman, Brigham and Women's Hospital, Boston, MA. RPMI
1640 media was provided by BioWhittaker (Walkersville, MD),
nonessential amino acids, L-glutamine, HEPES buffer,
minimal essential medium sodium pyruvate, penicillin/streptomycin,
2-mercaptoethanol, heat-inactivated fetal calf serum, Superscript
reverse transcriptase, Taq polymerase, 50 mmol/L MgCl
solution, TRIzol, and 10X PCR buffer were from Gibco BRL/Life
Technologies (Grand Island, NY). Ficoll lymphocyte separation medium
was from Organon Teknika (Durham, NC). IFN- and IL-4 primers for PCR
were as published by Murray et al,14
and dNTPs were from
Pharmacia (Piscataway, NJ). The competitive fragment for quantitation
of cytokine mRNA was the generous gift of Dr. Jim Lederer, Brigham and
Women's Hospital.
Heterotopic Murine Heart Transplant
Vascularized cardiac allografts were generated using the abdominal heterotopic murine cardiac transplantation model originally described by Corry et al15 and detailed by Nagano et al.16 Total surgical time was roughly 60 minutes, and heart ischemic time was approximately 20 minutes. At the time of graft harvest, the host was sacrificed by exsanguination under methoxyflurane anesthesia, and the graft was recovered.
BALB/c (H2d) donor hearts were transplanted into nonimmunosuppressed C57BL/6 (H2b) recipients; using this total allogeneic mismatch strain combination, acute graft rejection is evident histologically by 4 to 6 days, and graft failure (cessation of palpable contractions) due to rejection occurs by 7 ± 1 days.17 Animals were sacrificed 3 hours after surgery to establish the baseline of cells that immediately infiltrated the graft or were already resident. Additional animals were sacrificed on days 4, 6, and 8 after transplantation. Transplanted hearts were harvested at day 7 for RNA extraction for both competitive RT-PCR and RPAs.
Extraction of Lymphocytes from Spleens or Cardiac Allografts
In most experiments, recovered cardiac tissue was minced with a sterile razor blade and placed in 10 ml of borate-buffered saline with 2% BSA and 2 mg/ml collagenase. This mixture was rocked at 37°C for 2 to 3 hours and then strained through a 70-µm nylon cell strainer (Becton Dickinson, Franklin Lakes, NJ). Dead lymphocytes and red blood cells were removed by centrifugation through Ficoll for 10 minutes at 2000 rpm; resulting interface lymphocytes were washed in RPMI and resuspended in C/10 media (RPMI 1640 supplemented with 1% nonessential amino acids, 1% L-glutamine, 1% HEPES buffer, 1% minimal essential medium sodium pyruvate, 1% penicillin/streptomycin, 0.1% 2-mercaptoethanol, and 10% heat-inactivated fetal calf serum). Mechanical dissociation of heart tissue through a cytoscreen, without collagenase treatment, was also used to extract heart cells from some samples.
Spleens from transplant recipient animals were passed through a cytoscreen into 7 ml of C/10, and cells and residue were pelleted at 1200 rpm for 10 minutes and resuspended in 5 ml of ammonium chloride buffer (5 mmol/L Tris, 0.83% NH4Cl, pH 7.2) at 37°C for 5 minutes to lyse red blood cells. Splenocytes were washed twice with RPMI medium and resuspended in C/10 for subsequent stimulation.
Intracellular Cytokine Staining (Figure 1)
Extracted cells were stimulated with 25 µmol/L ionomycin and 10 ng/ml PMA for 4 hours at 37°C under a 5% CO2 humidified atmosphere, and 10 µg/ml of the fungal antimetabolite BFA was added for the duration of the culture to block cytokine secretion and thereby improve cytokine detection.
|
Flow Cytometry
Flow cytometry was performed on a FACScan flow cytometer (Becton Dickinson, Mountain View, CA), equipped with a 15-mW argon laser and filter settings for FITC (530 nm), CyC (650 nm), and PE (585 nm), using CellQuest software (Macintosh). Lymphocytes were distinguished from other cardiac-infiltrating cells on the basis of light-scatter characteristics. Scatter regions for lymphocytes were established before each collection using stimulated splenocytes. Collection of cytokine staining data from allografts was restricted to this gated region (to exclude nonlymphocyte cells). Data were collected on 2,000 to 10,000 cells within the lymphocyte scatter region.
mRNA Isolation, RPAs, and Semiquantitative RT-PCR
TRIzol was used to perform mRNA isolation according to the manufacturer's protocol. The RNA pellet was dissolved in 50 µl of diethylpyrocarbonate (DEPC)-treated water and stored at -20°C. For RPAs, the RNA was quantitated by optical densitometry, and approximately 20 µg of RNA from fresh tissue or 5 µg of RNA from extracted and stimulated lymphocytes was assayed using the Riboquant protocol (PharMingen). Specifically, a panel of radiolabeled probes of staggered sizes, complementary to cytokine mRNAs of interest, are hybridized to sample mRNA. Any remaining single-strand probe or RNA is enzymatically digested; resulting mRNA protected probes are electrophoresed on a sequencing gel where each cytokine migrates a specific distance based on the size of its probe. The gel is then exposed to radiographic film or a phospor plate for detection and quantitation. Semiquantitative competitive RT-PCR was performed as described by Platzer et al.18 Briefly, cDNA was created from the mRNA samples using specific cytokine primers according to the Superscript reverse transcriptase protocol. Serial dilutions of known concentration of a competitive fragment (designed to produce products slightly larger than the native cytokine mRNA were added to equal quantities of cDNA, and PCR was performed with specific primers to cytokines of interest. When the intensity of the authentic cytokine band is equal to the intensity of the competitive fragment band, the concentrations of the two are approximately equal.
Generation of Th2 and Th1 Control Cells in Vitro
To validate both our RPA and ICCS techniques, we generated
Th2-like (IL-4-secreting) and Th1-like (IFN--secreting) control
cells in vitro. Th2 cells were generated by incubating
splenocytes from C57BL/6 mice with equal numbers of irradiated BALB/c
splenocytes and exogenously added IL-4 (2000 U/ml) and anti-IFN- (40
µg/ml) for 5 days. C57BL/6 splenocytes tend to differentiate into Th1
cells without additional cytokine stimulation, so our IFN--secreting
cells were generated by incubating splenocytes from C57BL/6 mice with
equal numbers of irradiated BALB/c splenocytes for 5 days without
additional cytokines. After washing and resting overnight in medium
(with 200 U/ml IL-4 for the Th2 cells), the cells were treated like the
graft-infiltrating cells, and either they were re-stimulated with
PMA/ionomycin for 4 hours in the presence of BFA for subsequent ICCS or
their RNA was extracted for RPA
Quantitation of Mononuclear Cells in Allografts
Leukocyte infiltrates in rejecting allografts were quantitated by immunohistochemical staining against leukocyte common antigen (LCA) or CD45, followed by counting positively stained cells. Sections of allografts were frozen in OCT compound (Ames, Division of Miles Laboratories, Elkhart, IN) and stored at -80°C. Frozen, 4- to 5-µm-thick sections of heart were fixed in acetone for 10 minutes and incubated with rat anti-CD45 by standard techniques. The sections were then incubated with rabbit biotinylated anti-rat IgG, developed by the avidin-biotin complex method, and counterstained with hematoxylin. Positively stained cells were counted and averaged from 10 high-power fields (x100) per allograft; two allografts were analyzed at each time point (days 0, 4, 6, and 8 after transplant).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Initial reports of the technique of ICCS used stable in vitro lymphocyte cell lines as a source of cytokine-producing cells.19,20 These cells represent a homogeneous population of selected, transformed lymphocytes that produce high concentrations of cytokine protein. Primary cells are more vulnerable to cell death and may produce cytokines with lesser frequency and at lower concentrations than cultured cells. Therefore, it was necessary to re-optimize isolation and stimulation conditions for maximal lymphocyte survival and cytokine detection in cells extracted directly from cardiac allografts.
Lymphocytes were obtained by digesting minced heart tissue with collagenase and then pushing the fragments through a cell strainer and removing dead lymphocytes and red blood cells by centrifugation through Ficoll. Estimates based on the number of mononuclear cells in standard tissue histology sections and the number of sections per mouse heart indicate there are 5 x 106 to 40 x 106 graft-infiltrating mononuclear cells per acutely rejecting heart. Lymphocyte extraction by mechanical grinding, without collagenase treatment, yielded approximately 25,000 to 50,000 cells within the lymphocyte scatter region, or approximately 0.5% to 5% of the anticipated total. Collagenase digestion of the heart tissue increased the yield of mononuclear cells to greater than 106 per acutely rejecting heart (data not shown).
Some existing methods for studying rejection use a separate enrichment
step, such as cell sorting, to isolate lymphocytes from rejecting
tissue. This additional step is likely to reduce cell yield and might
alter cytokine expression. We therefore took advantage of the
distinctive pattern of light scatter by lymphocytes in flow cytometry
and selected a lymphocyte-enriched population from a heterogeneous
group of cells on the basis of forward-scatter and side-scatter
characteristics alone (Figure 2)
. Ficoll
fractionation of collagenase-digested explants, as routinely used,
resulted in decreased numbers of nonlymphocytic cells (eg, myocytes;
not shown). Cells obtained after Ficoll centrifugation or after
dissociation were directly cultured.
|
shows control stains with
stimulated primary lymphocytes to demonstrate our ability to stain for
both intracellular IL-4 (Figure 3A)
and IFN- (Figure 3B)
. The
staining intensity of the isotype-matched control antibody (solid dark
line in both panels) was used to set the threshold for the
anti-cytokine-stained samples. The threshold for positive cytokine
signal was set where fewer than 5% of cells showed autofluorescence or
background staining with control antibody; background signal was
subtracted from cytokine-positive signal.
|
. Cell survival was judged by
number, viability, and surface staining characteristics as detected by
flow cytometry. In initial experiments, optimal cell survival was
obtained with shorter culture times (ie, 4 to 6 hours versus
12 to 48 hours). PMA/ionomycin stimulation has been reported to reduce
CD4 expression;21
however, we found, in our protocol, that
the PMA/ionomycin treatment did not affect the percentage of cells
staining for CD4 or CD8, although it occasionally and variably
decreased the signal intensity (Figure 4C)
. The loss of signal
intensity was most pronounced with CD4 staining of highly stimulated
lymphocytes (eg, from acutely rejecting hearts shortly before they
stopped beating), with less effect seen on Thy1 or CD8 staining.
|
. IL-2, IL-10, and IFN- transcripts
are detected in both unstimulated (lane 2) and stimulated (lane 3)
extracted cells. No new cytokine message was induced by the stimulation
protocol. The loss of IL-6 and IL-15 signals in the RNA from the
stimulated cells is presumably due to decreased recovery of macrophages
after the 4-hour culture time required for PMA/ionomycin stimulation.
The ratio of IFN- to GAPDH housekeeping signal in the stimulated
cells was 10 times greater than that seen in nonstimulated lymphocytes
recovered directly from homogenized tissue. Notably, there is no IL-4
signal either before or after stimulation of lymphocytes from acutely
rejecting hearts although IL-4 is detected in the control RNA (lane 1).
|
Cytokine Expression in Acute Rejection
To demonstrate the applicability of ICCS to the study of transplant immunology, we used the technique to define the cytokine production of infiltrating lymphocytes during the course of acute cardiac allograft rejection. Total allogeneic-mismatched heterotopic allografts in nonimmunosuppressed recipients proceed to graft failure within approximately 8 days.17 Grafts were harvested from multiple animals at intervals after transplantation (days 0, 4, 6, and 8); samples were analyzed by flow cytometry for the percentage of CD4+ and CD8+ cells infiltrating the graft, the CD4/CD8 ratio, the total cytokine production by the cells, and the percentage contributed by each population. In addition, the recovered cells were stained with additional surface markers to determine the other cell types infiltrating the grafts. The total number of infiltrating mononuclear cells in these allografts was assessed by enumerating LCA-positive cells in high-power fields.
Figure 6
shows typical staining results
from an acutely rejecting allograft harvested 7 days after transplant.
We did not routinely measure the cytokine production of other cell
types using ICCS because surface stains showed that the major cell
population other than T lymphocytes was macrophages (Figure 6A)
. In
general, there were only 1% to 3% natural killer cells (potential
IFN- and IL-4 secretors) and less than 1% B cells infiltrating
these grafts. As shown in Figure 6B
, there are approximately 38%
CD4+ cells, of which 58% are IFN- positive (with
background subtracted), and none are IL-4 positive. Approximately 26%
of the gated cells stain positively for CD8, and 68% of these
CD8+ cells are IFN- positive; none of the
CD8+ cells stain for IL-4. In all, approximately 49% of
the extracted cells are IFN- positive. Figure 7
demonstrates that the ICCS technique is
also useful for identifying cells producing other cytokines, such as
IL-10. Thus, we saw fairly faint IL-10 bands using the RPA technique
(Figure 5A)
compared with the IFN- signal; accordingly, we see a
comparatively reduced number of IL-10-positive cells relative to
IFN--positive cells by ICCS (10% versus 49%).
|
|
. We found a strong IFN- signal, with no detectable IL-4,
consistent with the ICCS (Figure 6)
and RPA data (Figure 5A)
. Figure 5B
, lane 1, shows that there is between 1 and 0.3 ng/µl IFN- RNA
in day 7 allograft tissue samples, corresponding to approximately 15 ng
of IFN- transcript in 70 mg of wet heart tissue, or approximately
0.6 ng per 5000 T cells; lane 2 shows no IL-4 transcript in the RNA
from the same heart. RNA from an IL-4-producing T cell clone (I3L6), as
a positive control, showed that IL-4 could be detected at approximately
3 ng/µl corresponding to 1.5 ng from 5000 I3L6 cells (lane 3).
As shown in Figure 2B
, before the onset of acute rejection, few
lymphocytes were present within the allograft. Throughout the
time-course study of acute rejection, the number of leukocytes
infiltrating the graft steadily increased (Figure 8A)
. Assuming that an average mouse heart
weighs 100 µg and that this is predominantly water, an average volume
per heart is 100 mm3. Each high-power field has a diameter
of 0.2 mm and a thickness of 5 µm, giving a volume of 6.3 x
10-4 mm3. Therefore, an average of 40 to 250
cells/high-power field corresponds to 6.4 x 106
to
40 x 106
cells/heart. In addition, as the total
numbers of infiltrating leukocytes increased, and the percentages of
CD4+ and CD8+ cells remained approximately
equal, the absolute number of both CD4+ and
CD8+ lymphocytes increased as rejection proceeded. Between
days 0 and 4, lymphocytes began to infiltrate the graft, with
CD4+ cells slightly outnumbering CD8+ cells.
During the interval between 4 and 6 days, we observed a sharp increase
in the CD8+ infiltrate and a concomitant inversion in the
CD4/CD8 ratio (Figure 8B)
. By day 8, when histological sections
demonstrate fulminant cellular rejection,16,17
most
graft-infiltrating cells assessed using the size-gated population from
Figure 2
were T lymphocytes, with CD8+ cells outnumbering
the CD4+ cells.
|
and IL-4 at each time point. Very few
IFN--producing cells were present in the cardiac allografts before
the onset of rejection between days 0 and 4 (not shown). Between days 4
and 8, when histological evidence of rejection normally appears,
IFN- production by both CD4+ and CD8+ cells
increased concomitant with the increasing number of these cells present
in the grafts (Figure 8C)
. By day 6, when CD8+ cells
outnumber CD4+ cells within the rejecting allograft,
IFN- was largely a product of CD8+ cells (Figure 8D)
.
Considering both CD4+ and CD8+ lymphocytes,
there is a marked production of IFN- at the peak of acute graft
rejection.
We could not detect the production of any IL-4, with either the ICCS
technique, semiquantitative RT-PCR, or RPA at any point in these
acutely rejecting hearts. However, we did detect IL-4 in all three
assays with control populations: RT-PCR with I3L6 cells, which
constitutively secrete IL-4 (Figure 5B)
, and RPA and ICCS with Th2
cells generated in vitro from murine splenocytes (Figures 5A and 4A)
. This suggests that the lack of IL-4 signal in acutely
rejecting grafts is due to very low or nonexistent levels of IL-4.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We report here the use of intracellular cytokine staining (ICCS) and flow cytometric analysis of primary ex vivo lymphocytes extracted from solid tissue. This is an extension of a technique initially reported for cytokine analysis in stable, long-term lymphocyte-derived cell lines, which has also been used to measure cytokine production by peripheral blood mononuclear cells, splenocytes, and peritoneal cells.23-25 However, we found that measuring the cytokine production directly from the transplanted heart tissue was necessary, as the cytokine production by splenocytes isolated from graft recipients was very low and generally not representative of the profiles from the allograft hearts. Nadeau et al26 also report a lack of cytokine response in splenocytes from kidney transplant recipients. The application of ICCS to the study of solid-tissue processes required optimization of conditions for cell recovery, culture, and stimulation.
The nature of an optimal lymphocyte recovery protocol depends on the type of tissue under study. For work in cardiac tissue, the addition of a collagenase digestion step resulted in a dramatically higher yield of lymphocytes, which is especially useful when lymphocytes are scarce, ie, during very early graft rejection or during chronic rejection. On the other hand, cells from lymphoid organs are easily recovered using a cytoscreen only, without collagenase treatment; a similar mechanical procedure also allows recovery of adequate numbers of lymphocytes (with representative cytokine profiles; not shown), ie, when lymphocytes are abundant in hearts undergoing fulminant cardiac allograft rejection.
A brief episode of nonspecific stimulation immediately before cell
staining was used to boost cytokine production and thus improve
detection. Our results agree with previous studies that have shown that
nonspecific mitogen stimulation optimizes the secretory potential of T
lymphocytes but does not alter the profile of cytokine secretion from
physiologically stimulated cells.20,22
Therefore, cells
already differentiated to a Th1 phenotype continue to produce IFN-,
whereas Th2 cells will still produce IL-4. However, the kinetics of
cytokine production after mitogen stimulation are not
uniform,20
and it is theoretically possible that the choice
of a particular culture time favors detection of a rapidly accumulated
cytokine but may be too short to allow detection of a cytokine that is
sluggishly expressed. Simultaneous flow cytometric detection of both
IL-4 and IFN- has been shown to be optimal at 6 hours in Th0
(undifferentiated) lymphocytes in vitro, as IFN-
expression peaks at 6 hours and then plateaus, whereas IL-4 expression
peaks at 4 to 8 hours and then diminishes.20
Our choice of
a 4-hour stimulation duration was based on known cytokine expression
kinetics in Th0 cells19,20
and reflected our own
optimization experiments, which showed maximal lymphocyte viability and
cytokine detection at the 4-hour time point.
We also found that in vitro stimulation was necessary to detect cytokine. Although PMA and ionomycin stimulation has been shown to decrease CD4 expression,21 this was not a major problem in our hands. We also found that CD8 expression was less affected than CD4 by the PMA and ionomycin treatment and that Thy1, which is expressed by all T cells, was unaffected. Therefore, it may also be possible to obtain a reasonably accurate assessment of CD4 cell numbers by subtracting the CD8 signal from the Thy1 signal. Other protocols for ICCS have used anti-CD3 antibody to stimulate cells (eg, the PharMingen protocol); we observed comparable results using either PMA/ionomycin or anti-CD3 stimulation. We chose not to use the anti-CD3 method because CD3 down-regulation on T cell activation has also been reported.27
The choice of an appropriate anti-cytokine antibody is critical to the success of the intracellular staining technique. As a fixation step is necessary to immobilize the cytokines within the cell before permeabilization, the anti-cytokine antibody must be able to recognize epitopes on paraformaldehyde-fixed cells. Although chromophore-labeled antibodies are now commercially available for use with ICCS, we and others28,29 have also had greater success with biotinylated primary antibodies followed by chromophore-labeled avidin.
Cells stained with irrelevant fluorescent antibody show higher signal intensity, as detected by flow cytometry, than unstained cells, perhaps due to nonspecific interactions between cells and labeled antibody. Such an effect necessitates that a threshold for positive cytokine signal intensity be established using isotype-matched, irrelevant control antibody. We have routinely set our threshold for specificity at the signal intensity where fewer than 5% of cells stained with isotype-matched irrelevant antibody would be considered positive.19,20
IFN- - and IL-4-Secreting Cells in Rejecting Allografts
Because IFN- is a potent macrophage activator and mediator of
cellular immunity, Th1-differentiated cells have been thought to drive
the rejection process.30
High levels of Th1 cytokines
correlate with moderate to severe rejection,30
whereas high
levels of Th2 cytokines correlate with milder rejection
episodes.6,9
In addition, Th2 cells have been identified as
modulators of cellular immunity, and investigators have proposed that
Th2 cells are involved in antagonizing rejection and in the development
of tolerance.2
Although some investigators have been able
to detect significant Th2 cytokine mRNA in graft-infiltrating cells
from acutely rejecting allografts,2
our results agree with
other investigators who report very little to no detectable IL-4 in
acute grafts.7,8
We did, however, reproducibly detect
relatively low-level IL-10 production, which has been correlated with
acute rejection in human renal allografts.1,32
The
differences in detection in IL-4 between this report and earlier
investigators may reflect differences in assay sensitivity or
particular cytokine responses in selected mouse strains.
As Th1 and Th2 cells have been proposed to play such critical roles in
acute allograft rejection, interest in identifying IFN- and IL-4
within allografts has grown. Intracellular cytokine staining should be
a valuable addition to current methods for studying the roles of
IFN-, IL-4, and other cytokines in acute allograft rejection because
of the ability to detect the source of the cytokine proteins.
With increasing time, allografts show increasing numbers of
infiltrating mononuclear cells. Using the intracellular cytokine
staining method, we found that the number of graft-infiltrating
CD4+ cells producing IFN- increased with time as
rejection proceeded, in agreement with previous data.2
We
also identified CD8+ cells as a major cellular source of
IFN-. These cytotoxic T lymphocytes undoubtedly influence the local
cytokine environment within the allograft; however, it is unclear
whether IFN- production by CD8+ cells is a cause or an
effect of Th1-type cell proliferation. By immunohistochemical staining,
the majority of the mononuclear cells in rejecting allografts are
macrophages.16,17
The predominance of T lymphocytes over
macrophages in our analyses results from the initial size gating of the
infiltrating cells, which will tend to increase the relative
percentages of lymphocytes.
In any event, emerging Th1 dominance is clearly not accomplished by CD4+ cells alone. Our observations suggest that CD8+ cells should be viewed as important contributors to the cytokine milieu in rejecting allografts and that disruption of CD8+ cell cytokine should also be considered when developing immunosuppression regimens. In agreement with our data, Chan et al8 found that unmodified allografts did not produce Th2 cytokines; however in vivo depletion of CD8+ T cells resulted in allograft Th2 cytokine production. They suggest that donor-reactive CD8+ T cells inhibit intragraft production of Th2 cytokines.8
The information made available by cell-surface staining for CD4 and CD8
highlights one great advantage of the ICCS method, that is, the ability
to clearly identify the source of cytokine protein production. Using
directly chromophore-conjugated anti-cytokine antibodies, it is
possible to stain simultaneously for two intracellular products, such
as IFN- and IL-4, as well as identify the cytokine-secreting cell
using a surface marker of a third color. This technique could be used
to explore the role of Th0 cells in acute rejection and the kinetics of
Th0 differentiation. This technique might also serve to define the
cytokine contributions of every cell type present in a rejecting
allograft and could be adapted to examine other intracellular products
of interest, such as inducible nitric oxide synthase production by
macrophages.
The present experiments establish flow cytometric analysis of intracellular cytokine staining as a useful tool in the evaluation of solid organ rejection; this approach allows quantification of cellular cytokine protein production and concurrent identification of the secreting cell type. This technique should be of great utility in the study of transplantation immunology. The methodology could also be used to examine cytokine production in other immunologically mediated or inflammatory processes, such as response to pathogens, autoimmune disease, atherosclerosis, or tumors. As the roles of Th1 and Th2 cytokines in allograft rejection are clarified, intracellular cytokine staining with flow cytometric analysis could conceivably be performed on tissue from human biopsy samples to monitor the success of immunosuppression and for diagnostic and prognostic purposes in the clinical pathology laboratory.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported in part by National Institutes of Health Grant RO1 HL-43364.
Accepted for publication July 31, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
deficiency prevents coronary arteriosclerosis but not myocardial rejection in transplanted mouse hearts. J Clin Invest 1997, 100:550-557[Medline]
. Am J Pathol 1998, 152:1187-1197[Abstract]
+ T cells. J Immunol 1996, 156:232-237[Abstract]
and interleukin-4 in response to Th1- and Th2-stimulating pathogens by T cells in vivo. Nature 1995, 373:255-257[Medline]
production by T cells from HIV-infected individuals. J Immunol 1996, 157:2712-2718[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  A. I. Skaro, R. S. Liwski, J. Zhou, E. L. Vessie, T. D.G. Lee, and G. M. Hirsch CD8+ T cells mediate aortic allograft vasculopathy by direct killing and an interferon-{gamma}-dependent indirect pathway Cardiovasc Res, January 1, 2005; 65(1): 283 - 291. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Koga, J.-i. Suzuki, H. Kosuge, G. Haraguchi, Y. Onai, H. Futamatsu, Y. Maejima, R. Gotoh, H. Saiki, F. Tsushima, et al. Blockade of the Interaction Between PD-1 and PD-L1 Accelerates Graft Arterial Disease in Cardiac Allografts Arterioscler. Thromb. Vasc. Biol., November 1, 2004; 24(11): 2057 - 2062. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Furukawa, S. E Cole, R. V Shah, Y. Fukumoto, P. Libby, and R. N Mitchell Wild-type but not interferon-{gamma}-deficient T cells induce graft arterial disease in the absence of B cells Cardiovasc Res, August 1, 2004; 63(2): 347 - 356. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Afanasyeva, D. Georgakopoulos, D. F. Belardi, A. C. Ramsundar, J. G. Barin, D. A. Kass, and N. R. Rose Quantitative Analysis of Myocardial Inflammation by Flow Cytometry in Murine Autoimmune Myocarditis: Correlation with Cardiac Function Am. J. Pathol., March 1, 2004; 164(3): 807 - 815. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Futamatsu, J.-i. Suzuki, H. Kosuge, O. Yokoseki, M. Kamada, H. Ito, M. Inobe, M. Isobe, and T. Uede Attenuation of experimental autoimmune myocarditis by blocking activated T cells through inducible costimulatory molecule pathway Cardiovasc Res, July 1, 2003; 59(1): 95 - 104. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. J. Pinderski, M. P. Fischbein, G. Subbanagounder, M. C. Fishbein, N. Kubo, H. Cheroutre, L. K. Curtiss, J. A. Berliner, and W. A. Boisvert Overexpression of Interleukin-10 by Activated T Lymphocytes Inhibits Atherosclerosis in LDL Receptor-Deficient Mice by Altering Lymphocyte and Macrophage Phenotypes Circ. Res., May 31, 2002; 90(10): 1064 - 1071. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. M. Spriewald, M. Hara, A. Bushell, S. Jenkins, P. J. Morris, and K. J. Wood Differential Role for Competitive Reverse Transcriptase-Polymerase Chain Reaction and Intracellular Cytokine Staining as Diagnostic Tools for the Assessment of Intragraft Cytokine Profiles in Rejecting and Nonrejecting Heart Allografts Am. J. Pathol., November 1, 2000; 157(5): 1453 - 1458. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Shimizu, U. Schonbeck, F. Mach, P. Libby, and R. N. Mitchell Host CD40 Ligand Deficiency Induces Long-Term Allograft Survival and Donor-Specific Tolerance in Mouse Cardiac Transplantation But Does Not Prevent Graft Arteriosclerosis J. Immunol., September 15, 2000; 165(6): 3506 - 3518. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Furukawa, D. A. Mandelbrot, P. Libby, A. H. Sharpe, and R. N. Mitchell Association of B7-1 Co-Stimulation with the Development of Graft Arterial Disease : Studies Using Mice Lacking B7-1, B7-2, or B7-1/B7-2 Am. J. Pathol., August 1, 2000; 157(2): 473 - 484. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Furukawa, G. Becker, J. L. Stinn, K. Shimizu, P. Libby, and R. N. Mitchell Interleukin-10 (IL-10) Augments Allograft Arterial Disease : Paradoxical Effects of IL-10 in Vivo Am. J. Pathol., December 1, 1999; 155(6): 1929 - 1939. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||
