| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Noda Institute for Scientific Research, Chiba-ken, Japan;
Section of Allergy and Clinical Immunology, Department of Internal Medicine, Yale University School of Medicine, New Haven, CT 06520;
Clinical Immunology Section, National Institute on Aging, National Institutes of Health, Geriatric Research Centre, Baltimore, MD 21224;
§
Department of Immunology, Scripps Institute, La Jolla, CA 92037; and
¶
Children’s Hospital, Perlmutter Laboratory, Harvard Medical School, Boston, MA 02115
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
mRNA and protein
expression, as markers of T cell arrival and activation, began by
4 h after Ag challenge. In contrast to early C5a chemotactic
activity, late chemotactic activity 24 h after Ag challenge was
unaffected by anti-C5, was active on
C5aR-/- macrophages, was T cell-dependent,
and by ELISA appeared largely due to chemokines
(macrophage-inflammatory protein-1
and -1ß, IFN--inducible
protein-10, and monocyte chemoattractant protein-1). Importantly, early
generation of C5a was required for T cell recruitment because
C5aR-/- mice had absent 24-h CS. Taken
together, these findings indicate an important linkage of C5a as a
component of early activated innate immunity that is required for later
elicitation of acquired T cell immunity, probably by facilitating the
initial recruitment of T cells into the Ag-challenged local site in CS
responses. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
is a key cytokine in local expression
of this 24-h inflammation in DTH (3). Because T cells are
essential and very few Ag-specific recruited T cells can mediate DTH
(4), an important question for understanding and possibly
manipulating CS and similar T cell DTH responses concerns how the
few essential T cells themselves are recruited into local Ag-challenged
sites. There are several possible pathways for this initial T cell
recruitment to elicit CS. Traditionally, T cells are thought to survey
the circulation and tissues until they encounter Ag peptides on APC
(5). Accordingly, T cells would randomly migrate into
tissues, but it was clearly demonstrated recently that activation of
endothelial cells is needed for movement of T cells into extravascular
tissues at the site of Ag challenge (6).
Endothelial activation needed for T cell recruitment in CS involves
both Ag-nonspecific events and Ag-specific events. These processes act
in an early and required CS-initiating phase that follows soon after
local Ag challenge to elicit CS. Nonspecific local irritation by the
chemically reactive hapten Ag was suggested to recruit T cells in CS by
activating keratinocytes to produce endothelial-activating cytokines,
such as IL-1 and TNF- (5, 7). However, in contrast, we
demonstrated an Ag-specific early cascade reaction that was required
for subsequent elicitation of 24-h CS in mice challenged 4 days after
sensitization (8, 9). This Ag-specific CS-initiating
cascade was activated just within 2 h after Ag challenge
(10). Participating components in this Ag-specific CS
initiation process were mast cells (10), platelets
(11, 12), their release of serotonin (5-HT)
(10, 11, 12, 13), which is the crucial vasoactive amine in mice
(14), and also TNF- (15, 16), leading to
local activation of the endothelium. Thus, it was shown recently in
murine CS that early Ag-specific release of TNF- induced expression
of VCAM-1 and ICAM-1 on the luminal surface of local endothelium,
beginning at 4 h after Ag challenge (17). Prior
release of TNF- by skin mast cells probably mediated this local
expression of vascular adhesion molecules by 4 h, because skin
injection of TNF- itself caused expression of these adhesion
molecules within 2 h (17).
Recently, we showed that complement (C) also is involved in CS initiation (18, 19). C is known to be a major component of acquired B cell effector immunity, to play a role in regulation of B cell immunity (20, 21), and to participate in innate immunity (20). Although C has received little attention in acquired T cell immunity, our recent data suggest that C could participate in CS and DTH (18, 19). We showed that congenitally C5-deficient mice have impaired CS, which was restored by reconstitution with C5 (18), and that anti-C treatment with anti-C5 mAb or with C-inhibitory soluble C receptor-1 (sCR1), inhibited CS and DTH responses. Inhibition of CS by these C-inhibiting agents was demonstrated by decreased ear swelling, cellular infiltration, and impaired elaboration of chemotactic activity in 24-h CS ear extracts. Inhibition of CS required that C-inhibitors were given before Ag challenge (19), whereas they were ineffective if given just after the early component of CS (19), suggesting that C5 was activated early in elicitation of CS. Also, we showed that B cells were involved in CS, because CS was impaired in B cell-deficient µMT mice (19). Taken together, these findings suggested that early after Ag challenge, C could be activated locally via Abs to produce a proinflammatory mediator, such as C5a, that might directly or indirectly lead to recruitment of T cells. Thus, the major question of the current study about CS-initiating events early during elicitation of CS was whether local activation of C5 to elaborate C5a could play a role in early T cell recruitment.
To answer these questions we needed a way to directly demonstrate activation of C at the CS challenge site early after elicitation with Ag. Commonly used approaches to similar questions, such as knock-out mice, transgenic mice, mAb treatment, drug blocking, or correlative in vitro assays, were not sufficient to answer these questions. These techniques only allow measure of the final outcome of a CS response, which usually is evaluated 24–48 h after Ag challenge by macroscopic ear swelling and by microscopic cell infiltration. In contrast, it is hard to directly study the early CS initiation processes that are needed for eliciting subsequent 24–48 h CS, because the early phase has no cell infiltration (22) and is present as a smaller and transient ear swelling that peaks at 2 h (8, 9, 10). Thus, we previously had identified separate components of the early CS-initiating phase by determining ear swelling at 2 h after Ag challenge or indirectly by measuring the resulting 24-h ear swelling (10, 11, 12, 13).
Recently we employed a new approach to determine the biochemical events
in CS by preparing extracts from ears undergoing CS responses, and we
identified and quantified molecules by sensitive in vitro assays
(19). Thus, we analyzed the ear extracts for their
chemotactic activity to attract the J744A.1 macrophage cell line that
is rich in C5a receptors (C5aR; CD88), because they migrated strongly
to C5a and had no chemotactic activity to C5a-deficient
zymosan-activated mouse serum (ZAMS), which was prepared from
C5-deficient mice. In this current study, we established that the early
phase of CS is mainly C5a-dependent and that C5a is an essential
molecule in early CS initiation that is needed for early recruitment of
the CS effector Th1 cells, to then be activated to produce
IFN-. The exact role of C5a in the CS initiation process is
uncertain. It is possible that C5a may act directly as a chemotactic
factor for the T cells (23) or to activate endothelial
C5aR (24, 25). However, the current data combined with
prior studies (10, 11, 12, 13, 14, 17) favor that C5a elaborated early
after challenge in CS ears likely triggers C5aR on local mast cells to
release TNF- (26, 27) and also activates C5aR on
platelets and mast cells to release 5-HT (28, 29, 30).
Together, these C5a-released vasoactive mediators probably lead to
local endothelial activation and expression of adhesion molecules,
which is crucial for local early T cell recruitment in CS (17, 22).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Specific pathogen-free male CBA/J, DBA/1, DBA/2, B10.D2/n, and
B10.D2/o mice (6–8 wk old) were obtained from The Jackson Laboratory
(Bar Harbor, ME) and were rested for at least 1 wk before use. The
DBA/2 and B10.D2/o mice were congenitally C5-deficient.
ßTCR-deficient
(TCR-/-; bred >10
generations with BALB/c), 
TCR-deficient
(TCR-/-; bred >10
generations with C57BL/6) (both from Adrian Hayday, Yale University,
New Haven, CT), and C5aR (CD88)-deficient
(C5aR-/-) mice
(31) and 129/B6 controls were bred and maintained in
filter-topped microisolator cages in a bioclean room and were fed
autoclaved food and water. All experiments were conducted according to
guidelines of the Animal Care Committee of Yale University School of
Medicine.
Reagents
Picryl chloride (PCl; Nacalai Tesque, Kyoto, Japan) was
recrystallized twice and stored protected from light. Human recombinant
C5a was purchased from Sigma (St. Louis, MO). Both
4-ethoxymethylene-2-phenyl-2-oxazolin-5-one (OX) and Plummer’s reagent
(DL-2-mercaptomethyl-3-guanidinoethylthiopropanoic acid)
were obtained from Aldrich (Milwaukee, WI) and Calbiochem (La Jolla,
CA). Rabbit anti-murine macrophage-inflammatory protein (MIP)-1,
MIP-1ß, and monocyte chemoattractant protein-1 (MCP-1) Abs and rabbit
anti-human IFN--inducible protein-10 (IP-10) Abs were prepared
by multiple-site immunization of New Zealand White rabbits using
recombinant murine chemokines in IFA. Polyclonal Abs were titered by
ELISA, and specificity was verified by examining cross-reactivity with
other murine and human chemokines. Anti-human IP-10 antiserum was
cross-reactive with murine IP-10.
Immunization, elicitation, and in vivo evaluation of CS responses
Mice were contact sensitized with 100 µl of 5% PCl or 3% OX in absolute ethanol and acetone (4:1) on the shaved chest and abdomen. Four days later, CS responses were elicited by painting both ears with a topical application of 10 µl of 0.4% or 0.8% PCl in acetone and olive oil (1:1). In some experiments, a lower dose of 0.4% was chosen, compared with the higher conventional eliciting dose of 0.8% PCl, because effects of C alterations on CS responses usually were observed with a moderate challenge dose (18). Mice on a C57BL/6 and C57BL/10 background produce inferior responses to the above PCl contact sensitization procedure, which is optimal for CBA and BALB/c. Thus, B6 background C5aR-/-, 129/B6, B10.D2/n, and B10.D2/o mice produced suboptimal sensitization with our standard procedure. Therefore, to achieve more optimal CS response in these strains, they were sensitized twice on days 0 and 1 and were challenged with 0.8% PCl in acetone and olive oil (4:1) on day 7. Resulting thickness of the Ag-challenged ears in all strains was measured with a dial caliper (Ozaki, Tokyo, Japan) before challenge and then at 2 h and 24 h after challenge. Increases in ear thickness in groups of 4–6 mice were expressed as the mean ± SE x 10-3 cm.
Soluble TNF-R (sTNFR), a fusion protein of dimeric human TNF-R
p80 and non-complement-fixing human IgG1 Fc,
which binds murine TNF-, was used to treat mice either before or
3 h after PCl ear challenge, to see whether TNF- was involved
in the early or the late phase of elicited CS responses. The sTNFR (rhu
TNFR:Fc; Immunex, Seattle, WA) was injected i.p. at a dose of 250 µg
per mouse 30 min before or 3 h after 0.8% PCl challenge on the
ears of previously 5% PCl immunized CBA/J mice. Controls received
saline containing 250 µg human IgG1
(Sigma).
In vitro measurement of chemotactic activity in CS ear extracts
Chemotactic activity was measured in extracts of three 4-mm punch biopsies per ear collected from the distal site of CS ear responses (19). The biopsies were frozen together rapidly in liquid N2 and were subsequently thawed and extracted in 300 µl cold PBS on ice with a tissue microhomogenizer (Biospec Products, Racine, WI). Separate supernatants from four to five mice per group were diluted two to four times in RPMI 1640 with 1% gelatin without serum (RPMI-gelatin) and placed in lower 96 chemotaxis wells (Neuro Probe, Cabin John, MD). Target J774A.1 macrophages were suspended in RPMI-gelatin at 2–5 x 106 cells/ml, and 50 µl was added to upper wells, allowing migration through a polyvinyl pyrrolidone-free polycarbonate filter with 5- or 8-µm pores at 37°C for 4 h under 5% CO2. In experiments with C5aR+/+ vs C5aR-/- mice, peritoneal exudate macrophages were induced by i.p. injection of 2 ml thyoglycolate broth. Peritoneal lavage on day 3 consisted of about 95% macrophages (32). Migrated macrophages attached to the lower surface of the filter were fixed, stained with Diff Quick (Kokusai Chemicals, Tokyo, Japan) and counted at five different filter spots from each well, or each filter was extracted with 4 M urea before measurement of absorbance at 650 nm. J774A.1 migration was not due to chemokinesis because addition of ear extracts to upper wells, where the cells were loaded, caused diminished migration.
Quantitative sandwich ELISA for IFN- and chemokines
Two specific anti-IFN- mAbs (PharMingen, San Diego, CA)
were employed (33). Briefly, wells were coated overnight
with 2 µg/ml capture anti-IFN- mAb (R4-6A2) in 0.1 M
NaHCO3 (pH 8.3) at 4°C. After blocking with 1%
BSA in PBS at 25°C for 2 h, ear samples and recombinant mouse
IFN- (Genzyme, Cambridge, MA) were added and incubated for 1 h
at room temperature. Then, 1 µg/ml of another biotinylated
anti-IFN- mAb (XMG1.2) and 1:3000 diluted HRP-conjugated
streptavidin (Vector, Burlingame, CA) were added to detect IFN-.
Then, tetramethylbenzidine (TMB), peroxidase substrate, and TMB one
component stop solution (Kirkegaard & Perry, Gaithersburg, MD) were
used for color development at 450 nm.
The levels of murine MIP-1, MIP-1ß, MCP-1, and IP-10 in ear
extracts were measured by specific ELISA as described
(34). Similar ELISAs are not yet available for mouse C5a.
Briefly, microwells were coated with 1 µg/well of affinity purified
rabbit anti-mouse MIP-1, MIP-1ß, MCP-1, or IP-10 Abs for
18 h at 4°C. After blocking with 3% BSA in PBS for 120 min at
37°C, 50 µl of ear extracts and recombinant chemokines (R&D
Systems, Minneapolis, MN) was added and incubated for 1 h at
37°C. Then, 50 µl of biotinylated polyclonal rabbit anti-mouse
chemokine Ab and 50 µl of a peroxidase-conjugated streptavidin
(Bio-Rad, Richmond, CA) were added to detect chemokines. Chromogen
substrate (Bio-Rad) was added and incubated at room temperature until
optimal development was observed. The reaction was terminated with 50
µl of 2 M H2SO4 solution,
and the plates were read at 490 nm on a Bio-Rad ELISA reader.
RNA isolation
Naive, vehicle-painted, and PCl-sensitized CBA/J mice had ears removed at various times after local PCl challenge that were flash-frozen in liquid N2 and stored at -70°C until use. For ear RNA extraction, Trizol reagent (Life Technologies, Rockville, MD) was added to the frozen ears pooled from three mice/group, and tissue homogenate was obtained using Polytron electric homogenizer. RNA was purified per manufacturer’s instructions as a modification of the single-step RNA isolation method (35) and then was dissolved in diethyl pyrocarbonate-treated, autoclaved, double-distilled H2O and stored at -70°C.
RT-PCR and gel electrophoresis
For first strand synthesis of cDNA, 1 µg total RNA was reverse
transcribed in a reaction mixture (25 µl of: 50 mM Tris-HCl (pH 8.3);
75 mM KCl; 3 mM MgCl2; 10 mM DTT; 0.5 mM each of
dATP, dCTP, dGTP, and dTTP (Pharmacia, Piscataway, NJ); 20 µg/ml
oligo d(T)15 (Promega, Madison, WI); 1 U/ml
RNasin (Promega); and 8 U/ml M-MLV Reverse Transcriptase (Promega)) at
37°C for 45 min and then at 99°C for 1 min for one cycle before
soaking at 4°C and using a Perkin-Elmer (Norwalk, CT) Thermocycler.
Amplification of IFN--specific cDNA was conducted on a 1/10 aliquot
of the ss-cDNA (25 µl of 50 mM KCl, 10 mM Tris-HCl (pH 9.0), 0.1%
Triton X-100, 2.5 mM MgCl2, 250 µM dNTPs, 0.8
mM each of forward and reverse primers (human, mouse, and rat IFN-
specific amplimer set; Clontech Laboratories, Palo Alto, CA), and 1
U/ml Taq polymerase in storage buffer A (Promega)).
Amplification of the housekeeping G3PDH cDNA was similar, using 1/25
aliquot of the ss-cDNA and Clontech’s mouse G3PDH-specific amplimer
set. Amplification cycles were as follows: one cycle of 94°C for 2
min; then 30 cycles of 94°C for 40 s, 60°C for 20 s, and
72°C for 25 s; and then one cycle of 72°C for 5 min. Finally,
samples were soaked at 4°C. The amplified products were
electrophoresed on a 1.8% agarose gel in Tris-acetate-EDTA buffer and
viewed by ethidium bromide staining.
Statistics
Statistics were performed using the paired two-tailed Student t test, and p < 0.05 was taken as the level of significance.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
in CS ear extracts is a marker for the onset of T cell
recruitment
Th1 and Tc1 cells mediate cellular immunity in DTH by producing
IFN- locally (3, 36). To biochemically identify
CS-initiating events occurring before the onset of the late 24-h T cell
component of CS, we performed a study to detect the time course of
IFN- levels in ear extracts. A large amount of IFN- was found by
ELISA at 24 h (Fig. 1
a).
In addition, a small amount of IFN- was first detectable as early as
4 h after elicitation of CS by painting the PCl Ag on the ears of
mice that were contact-sensitized with PCl 4 days before. In contrast,
no increased IFN- was found in ear extracts of similarly
PCl-challenged control mice that were sham-immunized previously by
painting on the abdomen with the vehicle alone (Fig. 1a). In
addition, assay of IFN- mRNA, determined by RT-PCR on the ear
extracts, confirmed that IFN- was detectable in sensitized and
challenged mice at 4 h through 24 h after local application
of PCl, but not at 2 h and not in sham-sensitized controls that
were challenged similarly with PCl (Fig. 1b).
|
begins to
induce expression of endothelial adhesion molecules (17),
which likely is needed for local T cell recruitment across the vessels
to then be activated by Ag/APC in the tissues to produce the IFN-.
Fig. 1
c confirms that TNF- probably was involved in this
early initiating phase of CS. Treatment of PCl-sensitized CBA/J mice
with a sTNFR just before PCl ear challenge significantly reduced the
2-h and 24-h components (group D vs group C). In contrast, similar
sTNFR treatment beginning at 3 h after challenge, and thus just
after the early initiating component of CS, had no effect on the 24-h
component (group E vs group C). These results suggest that TNF- was
released during the early initiating phase of CS, because treatment
with the sTNFR after the 2-h peak of macroscopic CS initiation had no
effect.
Elaboration of IFN- mRNA in CS ear extracts at 4 and 24 h
required ß-T cells because sensitized and challenged
TCR-/- mice, compared
with sensitized and challenged TCR+/+ mice, had
no IFN- detected by RT-PCR at 4 h (Fig. 1d) and no
IFN- detected by ELISA of CS ear extracts at 24 h (Fig. 1e). Because prior studies indicated that
CD4+ MHC class II-restricted T cells mediated CS
in the PCl model of CBA mice that we employed (37), we
concluded that local IFN- levels at 24 h could be used as a
marker for the recruitment and subsequent local activation of effector
ßTCR+ Th1 cells in CS ears. In addition, the
data indicated that early release of TNF- led to recruitment and
then activation of Th1 cells, such that local measurements made in the
ears prior to 4 h were part of CS initiation and therefore
occurred before and did not depend on T cells or on their elaborated
IFN-.
Early generation of local chemotactic activity in CS responses
We analyzed chemotactic factors in ear extracts from the early
phase of CS that might eventually lead to T cell recruitment to produce
IFN- in the late phase. We previously have shown that elaboration of
chemotactic activity at 24 h in CS ears was C- and C5-dependent,
because injection of C-inhibitory sCR1 and anti-C5 mAb before Ag
challenge inhibited the 24-h chemotactic activity in CS ears
(19). Importantly, these prior studies suggested that C
activation actually occurred early after local Ag challenge (within
3 h), because when sCR1 was injected systemically just 3 h
after Ag challenge, there was no inhibition of 24-h CS
(19). Because CS was decreased in C5-deficient mice
(18) and late phase IFN- was blocked by anti-C5 mAb
(19), we hypothesized that local chemotactic activity was
a result of C5-derived C5a, which is a powerful chemotactic factor.
Thus, we employed a chemotactic assay with J77.4A1 indicator cells that
highly express C5aR and thus detect C5a by its chemotactic activity.
Fig. 2 shows a time course analysis of
chemotactic activity in CS ear extracts after Ag challenge. Chemotactic
activity was present as early as 1 h after Ag challenge of
sensitized mice, clearly near the onset of the early phase. Fig. 2 also
shows chemotactic activity at each time point after 0.8% PCl challenge
in separate groups of mice that previously were PCl-immunized or were
sham-immunized with vehicle. At all times after Ag challenge, i.e.,
from 1 h through 24 h, chemotactic activity was elaborated
into PCl-challenged CS ears of PCl-immunized mice and was dependent on
immunization and not just on challenge. Thus, at each time point,
chemotactic activity in ear extracts of immunized mice (Fig. 2, 
)
was greater than in those of control vehicle-immunized mice (Fig. 2, 
) when both were challenged equally. Furthermore, in another
experiment, 24-h ear extract chemotactic activity again was
significantly greater in PCl-immunized and PCl-challenged mice,
compared with similarly challenged vehicle-immunized controls (Fig. 2, group K vs group L) and, PCl-challenged totally
naive mice (groups M and N), and naive and
unchallenged mice (group O). Taken together, these results
suggest that by 4 days after contact sensitization, a particular aspect
of acquired immunity was induced, such that subsequent local Ag
challenge on the ears triggered production of this very early local
chemotactic activity, which was not elaborated by similar challenge of
control vehicle-immune or naive mice.
|
To further study this early CS-initiating phase, we chose to
examine the 2-h time point after Ag challenge, which is before arrival
and activation of T cells (Fig. 1, a, b, and
e), and compared it to the 24-h time point, which instead
reflected Th1 cell-dependent aspects of CS. Twenty-four-hour
chemotactic activity was reduced strongly in two C5-deficient strains,
DBA/2 (Fig. 3b) and B10.D2/o
(Fig. 3c). These strains were challenged on the ears
with low doses of 0.4% PCl to more effectively demonstrate differences
in CS of C-deficient strains, as described previously
(18). This finding of decreased 24-h chemotactic activity
in CS responses of C5-deficient mice was consistent with our prior
findings that anti-C treatment decreased 24-h chemotactic activity
(19). Importantly, elaboration of early chemotactic
activity at 2 h also was dependent on C, because 2-h CS ear
chemotactic activity was strongly reduced in C5-deficient DBA/2 mice
compared with C5-normal DBA/1 mice (Fig. 3a). Therefore,
both early and late CS chemotactic activity appeared to be dependent on
C5. In contrast,
TCR-/- as well as
TCR-/- mice had normal
elaboration of early 2-h chemotactic activity in CS responses (Fig. 4a and b,
left panel). As expected,
TCR-/- mice had no
late 24-h chemotactic activity because this should be dependent on
Ag-specific ß-T cells (Fig. 4a, right
panel). In contrast, the late 24-h chemotactic response in CS ear
extracts was intact in
TCR-/- mice vs C57BL/6
controls (Fig. 4b, right panel). Thus, although
immunization was strictly required to elaborate both early and late
chemotactic activity in CS, only the late-phase chemotactic activity
was ß-T cell-dependent and notably the early phase was T
cell-independent.
|
|
,
upper left panel), or OX-sensitized and
OX-challenged, but not PCl-challenged mice (Fig. 5, upper right
panel), had significant 2-h chemotactic activity in ear extracts.
As expected, chemotactic activity in 24-h CS ears also was Ag-specific
(Fig. 5, lower panels), because Ag-specific late-acting T
cells likely were required. These findings confirmed that early
elaborated chemotactic activity, like 2-h early ear swelling
(38), was a result of Ag-specific acquired immunity that
was not mediated by T cells.
|
Further study showed that the 2-h chemotactic activity probably
was a result of C5a, because anti-rat C5a serum (39)
neutralized the chemotactic activity detected in CS ear extracts (Fig. 6a, middle
columns). The anti-serum was validated by its ability to
neutralize chemotaxis mediated by ZAMS, which is due to C5a (Fig. 6a, left columns). In contrast, chemotactic
activity in 24-h CS extracts was minimally inhibited by anti-C5a
(Fig. 6a, right columns). We concluded that C5a
was the dominant chemotactic factor present in early 2-h CS ear
extracts, whereas other chemotaxins active on J774A.1 macrophages were
involved in the elicitation of late 24-h CS.
|
b,
right panel). Importantly, chemotaxis of
C5aR-/- macrophages
against the 2-h CS ear extracts was impaired significantly, compared
with C5aR+/+ macrophages (Fig. 6b,
left panel). In contrast, chemotaxis of the
C5aR-/- macrophages
against the 24-h ear extracts was not inhibited (Fig. 6b,
left panel). We ascribed the nearly intact chemotactic
activity of the late extracts on
C5a-/- macrophages to the
predominance of chemokines in the 24-h CS extracts (Fig. 7) that act on chemokine receptors rather
than on C5aR. Again, as with anti-C5a (Fig. 6a), these
findings indicated that C5a was dominant in early 2-h CS extracts and
similarly showed that the chemotactic factors that predominated in late
24-h CS ear extracts were not C5a.
|
shows that we detected
Th1-associated MCP-1, MIP-1, MIP-1ß, and IP-10 in CS ear extracts
at 24 h by specific ELISAs. These chemokines, like early C5a, also
were elaborated in an immunization-dependent fashion at 2 h but
were present in much greater amounts at 24 h after specific Ag
challenge (Fig. 7). Thus, the four chemokines measured were detected in
greater amounts in immunized mice than in nonimmunized mice, when both
were equally PCl-challenged on the ears (Fig. 7). Because substantial
amounts of chemokines were detected at 24 h, when C5a was found
not to be responsible for chemotaxis (Fig. 6), the chemokines probably
were responsible for the C5a-independent chemotactic activity noted in
the 24-h CS ear extracts (Fig. 6). Overall, these results clearly
indicated that C5a, which was generated via Ag-specific (Fig. 5)
activation of C early (1–2 h; Fig. 2) after Ag challenge to elicit CS,
accounted for early generated chemotactic activity. This early
elaborated C5a likely participates in the initial recruitment of the T
cells, which are then activated in the tissues starting at 4 h via
Ag/MHC on APC to produce IFN- (Fig. 1). The even later elaboration
of chemokines detected at 24 h (Fig. 7), which probably are
largely derived from local tissue cells and are dependent on T cell
cytokines like IFN- (40), would then lead to
recruitment of nonspecific leukocytes that are found later at 24
h, such as monocytes. Local early elaboration of C5a is crucial for later elicitation of 24 h CS
Use of C5aR-/- mice
gave us the opportunity to determine the absolute requirement for C5a
and C5aR in CS responses. Measurement of early 2-h and late 24-h
chemotactic activity in extracts of CS ears from
C5aR-/- mice produced an
interesting result. The immunized and PCl-challenged
C5aR-/- mice had greatly
augmented C5a-dependent early 2-h chemotactic activity in CS ear
extracts, compared with that of C5aR+/+ mice
(Fig. 8a). This striking
finding may have been because of absent clearance of C5a elaborated
locally early in CS, which may depend on binding to the C5aR, which are
absent in these mice. Thus, the absence of C5aR may have led to the
enhanced local levels of C5a that were just found in the 2-h CS ear
extracts. This finding lends further support to the idea that C5a is
elaborated early after Ag challenge in CS reactions. On the other hand,
chemokine-mediated chemotactic activity of late 24-h CS ear extracts
was greatly reduced in
C5aR-/- mice (Fig. 8b, right panel). This finding suggested that
impaired T cell recruitment that occurred in CS responses of
C5aR-/- mice likely was a
result of the absence of an early-phase C5a interaction with C5aR.
Accordingly, this lack of T cell recruitment led to impaired cytokine
production and reduced expression of chemokines at 24 h (Fig. 8b, right panel). Indeed,
C5aR-/- mice
had severely impaired elicitation of 24-h CS after PCl sensitization
and ear challenge on day 4, compared with that of
C5aR+/+ mice (Fig. 8c).
|
). These
findings strongly suggested that C5a was generated early after Ag
challenge. Furthermore, it is noteworthy that the data in Fig. 9, group D, show that prolongation of early C5a activity in
vivo actually enhanced the 24-h T cell-dependent late aspect of
CS.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We studied how Ag-specific effector T cells are recruited into
Ag-challenged sites during elicitation of CS in previously sensitized
mice. We focused on the early CS-initiating phase that precedes and
leads to T cell recruitment into the local extravascular tissue site to
mediate the classical late delayed phase of CS responses. We
established that C was activated in this CS-initiating phase, leading
to early local generation of C5a, which we demonstrated directly in
early CS ear extracts. Elaboration of C5a was because of an Ag-specific
process, and it occurred as early as 1 h after Ag challenge and
required prior immunization, but unlike the late phase of CS, it was
not dependent on ß-T cells. The data suggest that early
Ag-specific generation of C5a is required for elicitation of CS, by
mediating initial T cell recruitment into the local Ag-challenged site
rather than the elaborated C5a acting predominantly in the late phase
of CS, e.g., to recruit monocytes.
Use of IFN- to differentiate CS initiation at 2 h from T
cell-dependent late aspects of CS
We used the term "CS initiation" to signify the early
processes required for recruitment of T cells into the tissues in CS.
To differentiate early CS initiation from the later T cell-dependent
inflammatory phase, we used IFN- as an in situ marker for the end of
the early phase and beginning of the late phase, thus denoting the
arrival and subsequent Ag activation of recruited T cells into locally
challenged sites. Because Th1 and Tc1 subsets can mediate CS and DTH by
selectively producing IFN- in response to local Ag/MHC stimulation
by APC in the tissues (1, 2, 3, 36, 42), it was reasonable to
measure IFN- as a marker of the 24-h late phase. However, it was
noteworthy that IFN- mRNA and protein first could be detected
locally as early as 4 h after Ag challenge (Fig. 1, a,
b, and d). In addition, we showed that TNF-,
which likely leads to IFN-, acted before 3 h (Fig. 1d). This fits with TNF--dependent expression of VCAM-1
and ICAM-1 on the luminal surface of local endothelial cells in CS by
4 h after challenge (17). This likely enables the
first T cells to undergo transendothelial migration, because direct
injection of TNF- induces expression of vascular adhesion molecules
in 2 h (17). In addition, IFN- was not found at 4
or 24 h in CS ears of
TCR-/- mice (Fig. 1, d and e), confirming the usefulness of measuring
IFN- in evaluation of ß-T cell recruitment. Together these
findings assured that analysis of biochemical events in ear extracts at
just 1–2 h after Ag challenge was a manifestation of CS-initiating
events that precede and are required for subsequent recruitment of the
CS-effector T cells to then be activated by APC to produce IFN-
locally.
Chemotactic activity induced early (1–2 h) after Ag challenge of CS ears was a marker of CS initiation
We previously used J774.A1 macrophages as indicator cells to show
that chemotactic activity was a good molecular marker to evaluate the
24-h aspect of CS responses (19). Also, we showed that C5a
selectivity among C components mediated chemotaxis of J774A.1 cells,
which lacked chemotaxis against ZAMS without C5a, which had been
prepared from C5-deficient mouse sera (our unpublished result). Thus,
our prior findings of decreased chemotactic activity in 24-h CS ears
after anti-C treatments, such as anti-C5 mAb or sCR1
(19), suggested that C5a might be responsible for the
chemotactic activity in CS ear extracts. However, in those prior
studies of 24-h chemotactic activity, there was neither an exact
identification of C5a nor a determination of the point at which various
potential chemotaxins acted in the separate early and late phases that
are needed for elicitation of CS. In contrast, the experiments of the
current study identified that C5a was present early and not late and
determined the possible role of C5a in early occurring CS initiation.
We first performed a time course analysis of chemotactic activity in CS
ears after Ag challenge of contact-sensitized mice and found that
chemotactic activity was induced as early as 1 h after Ag
challenge of sensitized mice (Fig. 2). We examined the C-dependency of
early 2-h vs late 24-h chemotactic activity in CS ears of C5-deficient
mice and found that elaboration of both the early and the late
chemotactic activity was C5-dependent (Fig. 3). Interestingly, early
2-h chemotactic activity was elaborated normally in ß-T cell- and
-T cell-deficient mice (Fig. 4), but the late 24-h chemotactic
activity required the presence and activation of ß-T cells (Fig. 4a). These findings suggested that early chemotactic
activity was not dependent on T cells, but that late chemotaxis was
quite different and was dependent on ß-T cells, which probably
depended on early C5a for local recruitment in CS.
C5a and chemokines, respectively, were largely responsible for 2-h and 24-h chemotactic activity
We employed Ab neutralization of C5a activity (Fig. 6a)
and C5aR-deficient macrophages as migrating indicator cells in vitro
(Fig. 6b) to confirm that the chemotactic activity detected
early in CS ears was due to C5a. Although early 2-h chemotactic
activity was diminished employing both methods, late 24-h chemotactic
activity was not affected by either method, suggesting that C5a was
mostly responsible for early 2-h chemotactic activity and that other
chemotaxins were involved in the late 24-h CS ear chemotactic activity,
which required C5a recruitment of ß-T cells in order to be
elaborated. These findings suggesting that C5a acted early are in
accord with our previous observation that inhibition of CS with sCR1
required that this C inhibitor be given before Ag challenge, whereas
anti-C treatment given just 3 h after Ag challenge was no
longer effective (19). In addition, early 2-h C5a
chemotactic activity was greatly augmented in C5aR-deficient mice (Fig. 8a, right panel), suggesting that C5a was
elaborated early, and in this instance was not adsorbed out from
tissues, perhaps because of the absence of binding to C5aR. Finally,
neither late 24-h chemotactic activity (Fig. 8b) nor late
24-h ear swelling (Fig. 8c) was elicited in
C5aR-/- mice, confirming
the crucial role of early interactions between C5a and C5aR in early
CS-initiating steps leading to T cell recruitment for elicitation of
late chemokine elaboration and macroscopic CS responses.
The importance of chemokines in immune inflammation recently has been
described (43), and some chemokines were reported to be
mediators in CS responses (44). Because J774A.1 cells also
can respond to some chemokines (our unpublished results), we measured
chemokines in CS ear extracts by specific ELISA. We found that
MIP-1, MIP-1ß, MCP-1, and IP-10 also were produced in 2-h CS ears,
but they were present in much greater amounts in the 24-h CS ear
extracts (Fig. 7). Importantly, like C5a (Fig. 2), elaboration of these
chemokines was not induced by the irritating effect of the hapten Ag
(PCl) used to challenge the ears because large amounts of the
chemokines only were detected in mice that were both previously
immunized and then ear challenged (Fig. 7). Therefore, these results
suggested that late 24-h chemotactic activity in CS principally was due
to local elaboration of chemokines. Thus, we concluded that sCR1 and
anti-C5 inhibition of early activation of C and of C5 needed to
generate C5a led to impaired T cell recruitment, causing reduced
chemokines found in late 24-h CS ear extracts, because elaboration of
late 24-h chemotactic activity was dependent on ß-T cells (Fig. 4a). Furthermore, the chemokines detected in CS fit with the
type of CS-effector T cell in that they were relatively Th1-associated,
particularly IP-10 (44, 45), which is induced by IFN-
in local tissue cells, particularly in the skin by keratinocytes
(40).
Significance of C5a in CS
The requirement for C5a in elicitation of CS was clearly
demonstrated by the strongly impaired 24-h CS responses and 24-h
chemotactic activity in
C5aR-/- mice (Fig. 8).
Also, administration of Plummer’s reagent, which causes enhanced and
prolonged C5a activity by inhibiting the C5a-inactivating enzyme,
carboxypeptidase N, and thus leads to inhibited generation of inactive
C5a-desArg (41), resulted in significant augmentation of
both early 2-h and late 24-h ear swelling responses (Fig. 9). Thus,
augmentation of local C5a levels by Plummer’s reagent led to augmented
elicitation of both early 2-h and late 24-h CS. Taken together, these
genetic and biochemical findings make it likely that C5a, which is
elaborated very early after Ag challenge (Fig. 2), plays a crucial role
in CS by leading eventually to the migration of effector Th1 cells into
extravascular sites of CS elicitation.
It is important to note that early 2-h chemotactic activity, which
likely was because of C5a, was elaborated Ag-specifically (Fig. 5) but
independent of T cells (Fig. 4). This Ag specificity of C-activation to
generate local elaboration of C5a, together with its
immunization-dependency (Fig. 2) and lack of T cell-dependence (Fig. 4), strongly suggested that Ab derived from B cells might be involved
in the early CS-initiating cascade. In fact, we previously demonstrated
that CS was impaired in B cell-deficient µMT mice (19).
Further data we present in another paper suggest that early after
sensitization, the B-1 cell subset of B cells in central tissues
produces circulating IgM Ab to mediate local Ag-specific C activation
and to generate C5a locally (R. F. Tsuji, M. Szcepanik, I. Kawikova, R.
Campos, M. Akahira-Azuma, and P. W. Askenase, manuscript in
preparation).
Overall scheme for elicitation of CS
Our studies have established that early elaboration of C5a is
required for Th1 cell recruitment in CS. An important question concerns
how C5a actually leads to T cell recruitment. Does C5a act directly as
a chemotactic factor for attraction of the CS-effector T cells into the
tissues, or does C5a act indirectly via triggering C5aR on local
mediator cells to release substances that influence the tissue
microenvironment to facilitate local T cell recruitment? Recently, it
was shown that C5a is chemotactic in vitro for activated human T cells
and that C5aR are expressed in vivo on T cells in DTH-related
encephalomyelitis in rats (23). This suggests that C5a may
act directly to recruit recently immunized T cells. Alternatively,
although we assayed C5a in vitro as a chemotactic factor for
macrophages and this also might apply to T cells, it also is likely for
two reasons that C5a indirectly leads to T cell recruitment in CS by
influencing the local microenvironment in CS. First, local mast cells
and platelets, which are established to express functional C5aR
(28, 29, 30), were shown previously to be involved in CS
initiation by local release of 5-HT (10, 11, 12, 13) and TNF-
(17), which activate the endothelium to facilitate local T
cell recruitment across the vessels (17, 22). In the
current study, we confirmed the early involvement of TNF- (Fig. 1d). Second, C5a also can activate endothelium directly
through endothelial cell C5aR (46, 47). Taken together, we
propose, in addition to a possible direct attraction of T cells by C5a,
that T cell recruitment in CS requires C5a-dependent release of TNF-
and 5-HT, which are produced in the early CS-initiating phase by local
C5aR-bearing mediator cells, like mast cells and platelets, and/or that
C5a acts directly to facilitate local T cell recruitment by activating
C5aR on the endothelial cells. Activation of endothelium for
permeability and expression of adhesion molecules (17)
allows T cells to bind to and then migrate across the endothelium into
the tissues.
After extravasation of the recruited T cells into the tissues, C5a may
also facilitate T cell chemotaxis directly and perhaps may contribute
to T cell movement toward conjugation with local APC. In the tissues,
the subset of the recruited T cells that are specific for the
challenging Ag are then selected to bind with specific Ag/MHC complexes
on the local APC, causing T cell activation and release of IFN- to
mark the onset of the late phase and the end of the early CS initiation
phase. Then eventually, T cell cytokines like IFN-, produced by the
Ag/MHC-activated T cells, generate chemokines by activation of local
tissue cells (40, 44, 45). The chemokines, rather than
C5a, most likely are required for later late-phase monocyte recruitment
that is associated with elicitation of 24-h CS ear swelling because
tissue monocytes and ear extract chemokine levels both are increased in
CS ear tissues in the later phase of CS responses, whereas C5a was
decreased in CS ear extracts in this later phase. Thus, although we
employed in vitro monocyte chemotaxis as an assay method to detect C5a
in CS ear extracts, our data suggest that C5a does not act late in CS
as a dominant chemotaxin for monocytes, but is elaborated very early in
the initiation phase of CS responses and then acts early in elicitation
of CS to lead to initial T cell recruitment. In contrast, we found that
monocyte recruitment in CS was more remotely dependent on C5a because
C5a acted indirectly via early attraction of T cells and, after local
Ag/MHC activation of the T cells, led to monocyte recruitment,
predominantly via T cell-dependent chemokines released from tissues
cells.
In summary, our present results suggest that early phase generation of
C5a via Ag-specific C activation is required for important and
underinvestigated crucial initiating aspects of CS effector T cell
responses in vivo. These components of CS initiation were elicited by
ear challenge as soon as day 4 after immunization and occur quite early
(within 1–2 h) after elicitation by local challenge with Ag. Besides
mediating possible T cell chemotaxis, C5a probably triggers a local
CS-initiating cascade in the tissues. Accordingly, C5a likely acts on
mast cell and platelet C5aR, leading to release of mediators,
particularly TNF- and also 5-HT, that activate local endothelia to
facilitate T cell recruitment. In contrast, C5a does not appear to act
dominantly in the late phase of CS to attract monocytes, which is more
likely due to T cell-dependent chemokines. Understanding these newly
delineated initiating aspects of T cell immunity in vivo may be
critical for development of new therapeutic strategies to control
clinically important T cell-mediated disease processes by regulating
these crucial early steps.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Ryohei F. Tsuji, Noda Institute for Scientific Research, 399 Noda, Noda-shi, Chiba-ken 278-0037, Japan.
3 Address correspondence and reprint requests to Dr. Philip W. Askenase, Section of Allergy and Clinical Immunology, Department of Internal Medicine, Yale University School of Medicine, 333 Cedar Street, New Haven, CT 06520-8013.
4 Abbreviations used in this paper: CS, contact sensitivity; DTH, delayed-type hypersensitivity; 5-HT, serotonin; C, complement; sCR1, soluble C receptor-1; ZAMS, zymosan-activated mouse serum; PCl, picryl chloride; OX, 4-ethoxymethylene-2-phenyl-2-oxazolin-5-one; MIP, macrophage-inflammatory protein; MCP, monocyte chemoattractant protein; IP-10, IFN--inducible protein-10; sTNFR, soluble TNF-R.
Received for publication September 7, 1999. Accepted for publication May 11, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
in delayed-type hypersensitivity mediated by Th1 clones. J. Immunol. 143:2887.[Abstract]
. J. Immunol. 162:1648.: a possible initiating role of B cells. J. Exp. Med. 186:1015.
