Role of CD4 T lymphocytes and interleukin-5 in antigen ... - CiteSeerX

Role of CD4

T lymphocytes

and interleukin-5

antigen-induced eosinophil recruitment cutaneous late-phase reaction in mice Itsuo

Iwamoto,

Second Institute

Department for Medical

Abstract: recruitment (LPR) was this study,

Sanae

Tomoe,

Hisao

Tomioka,

of Internal Medicine, Chiba University Immunology, Kumamoto University

Previous studies suggested that the eosinophil into the site of cutaneous late-phase reaction dependent on IgE antibody and mast cells. In we determined the role of CD4 T cells and

sites

Key

J.

Words:

CD8

T cells

LPR 52: .

mast

is mediated 572-578; cells

.

by 1992.

mast

the

are

late-phase associated

reactions with

(LPRs) provoked intense eosinophil

572

was

Journal

an increase

of Leukocyte

in lymphocytes

Biology

infiltrating

Volume

52,

November

is

Moreover,

of LPR

not

yet

*Department

there

in the

known

is an

skin

of Biology,

increase

[6] and

whether

T

in CD4 [7].

airways

lymphocytes

and

AND

and

METHODS

immunization

Reagents anti-L3T4

mAb produced

factor

into

sites

Yoshida

and

BALB/c mice were purchased from Charles River Laboratonies (Shizuoka, Japan) and were used at 6-8 weeks of age. The mice were immunized intrapenitoneally twice with 1 g of ovalbumin (OVA) (Sigma Chemical Co., St. Louis, MO) in alum at a 2-week interval. At 10 to 14 days after the second immunization, OVA was administered subcutaneously as described below. The titer of anti-OVA IgE antibody in sera of immunized mice assessed by passive cutaneous anaphylaxis reaction was 80- to 160-fold dilution positive.

mAb

53.6-72

GK1.S

(IgG2a) the

from

(IgG2b)

[11] and

[12] were hybnidomas

rat

anti-Lyt-2

purified from by ammonium

ascites sulfate

precipitation. Rat anti--munine IL-S mAb was obtained from the hybnidoma NC17 and prepared as described previously [13]. Purified rat IgG was purchased fromJackson Immunoresearch Laboratories (West Grove, PA). Disodium cromoglycate was obtained from Fujisawa Pharmaceutical

aninto

Co. line

site

there

it

Mice

of antigen administration [1-3]. LPR occurs 3 to 8 h after antigen administration and persists for up to 48 h in a variety of tissues including the skin and airways. There is increasing evidence that the infiltrating eosinophils cause the tissue damage in LPR by releasing the cytotoxic granules and membrane products [4]. Previous studies suggested that the eosinophil infiltration in cutaneous LPR was dependent on immunoglobulin E (IgE) antibody and mast cells [1-3]. In addition to the infiltration of eosinophils, it was shown that

[2, 3, 5].

at the

MATERIALS

Rat

by specific infiltration

Sho

and neutralizing IL-S in vivo with their corresponding monoclonal antibodies (mAbs). We provide direct evidence that CD4 T cells and IL-S mediate the second peak of antigen-induced eosinophil recruitment of cutaneous LPR. In contrast, we have also demonstrated that the first peak of antigen-induced eosinophil recruitment of cutaneous LPR is mediated by mast cells and platelet-activating factor (PAF).

INTRODUCTION Allergic tigens

ofLPR

However,

cells

platelet-activating

and

their cytokines, especially interleukin-5 (IL-5) [8-10], mcdiate antigen-induced eosinophil recruitment into the site of LPR. To solve this issue, we studied the role ofCD4 T cells, CD8 T cells, and IL-S in antigen-induced eosinophil infiltration into mouse skin by depleting both T cell subsets

decreased the first peak of OVA-induced cutaneous eosinophilia in the mouse. These results indicate that CD4 T cells, but not CD8 T cells, cause the second peak of antigen-induced eosinophil recruitment of cutaneous LPR and that IL-S mediates this eosinophil recruitment. In contrast, the first peak of antigen-induced eosinophil of cutaneous Leukoc. Biol.

Takatsu,’

of Medicine, Chiba, Japan School, Kumamoto, Japan

T cells

CD8 T cells in causing antigen-induced eosinophil recruitment of LPR in mouse skin. Eosinophil infiltration into the subcutaneous tissue of ovalbumin (OVA)sensitized BALB/c mice was biphasic, reaching the first peak at 6 h after the subcutaneous challenge with OVA and the second peak at 24 to 48 h. The in vivo depletion of CD4 T cells by pretreatment with anti-L3T4 monoclonal antibody (mAb) significantly decreased the second peak (at 24 h and 48 h), but not the first peak (at 6 h), of OVA-induced eosinophil infiltration into the skin of OVA-sensitized mice. However, the depletion of CD8 T cells by pretreatment with anti-Lyt-2 mAb had no significant effect on either the first peak or second peak of OVA-induced cutaneous eosinophilia. Pretreatment with anti-murine interleukin-5 (IL-5) mAb also decreased the second peak, but not the first peak, of OVAinduced cutaneous eosinophilia. In contrast to the inhibitory effects of depletion of CD4 T cells and of anti-IL-S mAb on the second peak of antigen-induced cutaneous eosinophilia, disodium cromoglycate and a selective antagonist for platelet activating factor (PAF) CV-6209

recruitment and PAF.

into the site of

Kiyoshi School Medical

in

(Osaka, (PBS).

Japan) A selective

Abbreviations: activated phase

cell

DTH, sorter;

reaction;

delayed-type IgE,

mAb,

and dissolved in phosphate-buffered PAF antagonist, CV-6209 (2-[N-acetyl-

hypersensitivity;

immunoglobulin

monoclonal

E;

antibody;

OVA,

activating factor; PBS, phosphate-buffered Reprint requests: Itsuo Iwamoto, Second cine, Chiba University School of Medicine, Chiba

260,

Received

the

1992

4, 1992;

accepted

July

FACS, interleukin; ovalbumin;

fluorescenceLPR, PAF,

saline. Department of Internal 1-8-1 Inohana, Chiba

Japan.

May

IL,

1, 1992.

sa-

late-

plateletMcdi-

City,

N-(2-methoxy-3-octadecyl-carbamoyloxypropoxycarbonyl) aminomethyl]-1-ethylpynidinium chloride) from Takeda Pharmaceutical Co. (Osaka, solved in 0.9% saline.

[14], was Japan)

aminoobtained and dis-

Effect of inhibitor of mast cell degranulation antagonist of platelet-activating factor To

determine

philia

Antigen-induced

eosinophil

infiltration

in mouse

Eosinophil infiltration into the skin was induced cutaneous antigen challenge in sensitized mice, number of eosinophils infiltrating the subcutaneous was measured by the method described by Lawman [15]. Briefly, 0.9 ml of air was slowly injected into cutaneous tissue of the left flank of OVA-sensitized mice and

using then

syringe control,

a 1-ml 0.1 ml

tuberculin of OVA

syringe (50 g/ml)

with

a 26-gauge present in

by and

subthe tissue et al. the subBALB/c needle, same

successively introduced into the air bleb. As a alone was injected into the air bleb. At various intervals after the injection, the thin membrane of the air bleb was carefully excised after sacrifice by cervical dislocation and was stretched onto a glass slide. After drying, the specimen was fixed with methanol and stained with WrightGiemsa solution. The slides were coded and then five fields at a magnification of x400 were examined by one observer (S.T.), who was unaware of the identify of the individual specimens. The number of eosinophils was expressed as the sum of eosinophils in the five fields.

Effect

was PBS

of anti-L3T4

To determine

mAb

whether

and

anti-Lyt-2

T lymphocytes

are

T

cells

and

CD8

T

cells

involved

was

confirmed

by

of CD4

Pretreatment numbers of PBS-pnetreated

T cells and CD8 T cells, respectively. purified rat IgG induced no changes in either T cell population compared to those in mice. Eosinophil infiltration into the skin

with

was evaluated at 6, 24, and 48 h after the challenge in those three groups of mice.

Effect

of anti-murine

IL-5

we

examined

cutaneous

eosino-

the

of diso-

effect

cnornoglycate, an inhibitor of mast cell degranulation [16-18], on OVA-induced eosinophil infiltration in mouse skin. We also evaluated mast cell degnanulation according to the criteria described by Wershil et al. [19] by examining 200 mast cells in the dermis under a light microscope at a magnification of x 1000. In these experiments, disodium cromoglycate (10 tg) or vehicle (PBS) was administered subcutaneously

air

with

OVA

(S g;

total

volume

0.1

ml)

into

the

bleb.

To determine whether PAF mediates antigen-induced eosinophil infiltration into the skin, we examined the effect of a PAF antagonist, CV-6209 [14], on OVA-induced eosinophil infiltration in mouse skin. CV-6209 (1 jg/g) or vehicle (saline) was administered intrapenitoneally 30 mm before the subcutaneous injection of OVA (S pg). We chose this dose of CV-6209 because CV-6209 (1 g/g) completely protected mice against PAF-induced anaphylactic death [14]. Eosinophil infiltration was then evaluated at 6 and 24 h after the injection of OVA. The effect of CV-6209 was also examined at 48 h after the injection of OVA. In these expeniments, CV-6209 (1 g/g) was administered intraperitoneally 30 mm before and 24 h after the subcutaneous injection of

Data

in causing

fluorescence-activated cell sorter (FACS) analysis using spleen cells prepared from the mice 2 days after treatment with the anti-L3T4 mAb or anti-Lyt-2 mAb. The FACS analysis by use of another anti-L3T4 mAb YTS 191.1 and anti-Lyt-2 mAb YTS 169.4 revealed that the pretreatment with anti-L3T4 mAb and anti-Lyt-2 mAb caused complete depletion

antigen-induced

dependent,

OVA.

mAb

antigen-induced eosinophil infiltration into the tissue, we cxamined the effect of the in vivo depletion of CD4 T cells on CD8 T cells on antigen-induced cutaneous eosinophilia in sensitized mice. OVA-sensitized BALB/c mice were injected intrapenitoneally once with 1 mg ofanti-L3T4 mAb (GK1.5) [11] 2 days before the subcutaneous challenge with OVA (5 pg). Other OVA-sensitized mice were given 1 mg of anti-Lyt-2 mAb (53.6-72) [12] intrapenitoneally twice at a 2-day interval and then challenged subcutaneously with OVA 2 days later. As a control, OVA-sensitized mice were injected intnapenitoneally with purified nat IgG (1 mg) 2 days before the subcutaneous OVA challenge. The depletion of

CD4

cell

dium

skin

the

whether

is mast

and

subcutaneous

OVA

Analysis

Data are summarized as mean ± SD. Inhibition infiltration by a given treatment was expressed 1 [(OVA-induced response with the treatment induced nesponse)/(OVA-induced response responding vehicle PBS-induced response)]. analysis of the results was performed by analysis using Fisher’s least significant difference test comparisons [20]. P values < .05 were considered

RESULTS

Kinetics mouse

after after course

Since IL-S has been shown to be chemotactic for eosinophils in mice [10], the effect ofanti-munine IL-S mAb on antigeninduced cutaneous eosinophilia was examined in sensitized mice. An anti-murine IL-S mAb NC17 (1 mg) [13] on purified rat IgG (1 mg) was injected intrapenitoneally 24 h before the subcutaneous challenge with OVA (S sg). Eosinophil infiltration into the skin was then evaluated at 6, 24, and 48 h after the subcutaneous OVA challenge.

Iwamoto

et aL

of antigen-induced skin

eosinophil

infiltration

into

Subcutaneous administration of OVA caused biphasic infiltrations of eosinophils into the skin of OVA-sensitized mice. Eosinophil infiltration into the subcutaneous tissue of OVA-sensitized BALB/c mice occurred 4 h after the subcutaneous injection of OVA (S pg), reached the first peak 6 h after the injection, and then declined at 12 h (n = S mice at each time point) (Fig. 1). The antigen-induced eosinophil infiltration into mouse skin again increased at 24 and 48 h the

1). In

mAb

of eosinophil as percent of PBSwith a corStatistical of variance for multiple significant.

injection,

addition,

reaching

there

the subcutaneous of cutaneous

was

the

injection eosinophilia

Effect of depletion of CD4 antigen-induced cutaneous The

in vivo depletion of CD4 anti-L3T4 rnAb decreased the but not the first peak (at 6 h), eosinophilia in the mouse. The with anti-L3T4 mAb (GK1.S;

CD4

T cells

and

IL-S

peak (n = 5) (Fig. blood eosinophilia of OVA during the entire in the OVA-sensitized mice. second

no significant

T cells and eosinophilia

CD8

T cells

on

T cells by pretreatment with second peak (at 24 and 48 h), of antigen-induced cutaneous intrapenitoneal preinjection 1 mg) 2 days before the sub-

in antigen-induced

eosinophilia

573

Eosinophils (number/5 fields)

Eosinophils (number/5 fields) 100

120

80

100

80

60

60

40

40 20

20

0 6

12

18

24

Time

30

36

42

0

48

6h

(h) Fig.

Fig.

1.

skin.

OVA

Time

course

(5 g;

subcutaneous sinophils

of antigen-induced

solid

tissue

squares) into

the

Open

circles

sensitized

mice.

The

number

five

fields

SD

for

in

five

the

mice

PBS

of OVA-sensitized

infiltrating

injection. sinophils

or

eosinophil

skin

indicate

at each

was

(open

squares)

BALB/c

mice,

counted

at

OVA-induced

of eosinophils at

was

and

various

eosinophil

was

a magnification

time

infiltration

mouse

injected

the

into

number

intervals

x400.

Data

are

un-

of anti-murine

mice.

Data

on

antigen-induced

anti-L3T4

mAb

before

the

mice

(GKI.5)

(solid

subcutaneous

eosinophil

were

injected

columns)

or

challenge

with

in-

IgG OVA (5

rat

after

the

subcutaneous

are

means

±

mean

SD value

challenge

with

for six mice in each of the corresponding

PBS

in

OVA-sensitized

group. control

Significantly response

(rat

Eosinophils (number/5 fields)

mAb

induced eosinophil infiltrations into the mouse 48 h by 65 and 64%, respectively (n = 6, P < which were similar to the degrees of decrement pretreatment with anti-L3T4 mAb (Fig. 2). mAb

T cells

OVA-sensitized

the simultaneous subcutaneous administration of disodiurn cromoglycate (10 zg) with OVA decreased OVA-induced eosinophil infiltration into the mouse skin at 6 h by 55% (PBS 81.0 ± 12.2 vs. disodium cromoglycate 38.7 ± 6.6 eosinophils/five high-power fields, n = 6, P < .001) (Fig. 5). The administration of disodium cromoglycate also significantly decreased mast cell degranulation induced by subcutaneous OVA challenge at 6 h (Table 1). However, disodium cromoglycate only slightly affected OVA-induced cutaneous eosinophilia at 24 h (n = 6, P < .05) (Fig. 5). The intrapenitoneal pretreatment with the selective PAF antagonist CV-6209 (1 g/g) 30 mm before the subcutaneous injection of OVA also decreased OVA-induced eosinophil

Pretreatment with anti-murine IL-S mAb also decreased the second peak, but not the first peak, of antigen-induced cutaneous eosinophilia in the mouse. The intrapenitoneal pneinjection with anti-munine IL-S mAb (NC17; 1 mg) 24 h before the subcutaneous OVA challenge decreased OVA-

preinjection with anti-IL-S the first peak of OVA-induced skin (Fig. 4).

of CD4’ skin.

2 days

different from the IgG), P < .001.

point.

lL-5

with

columns)

infiltrating

±

cutaneous injection ofOVA had no significant effect on OVA (5 sg)-induced eosinophil infiltration into the skin of OVAsensitized mice at 6 h (n = 6) (Fig. 2). In contrast, the intrapenitoneal preinjection with anti-L3T4 mAb significantly decreased OVA-induced eosinophil infiltration into the skin at 24 h by 67% (control rat IgG 113.5 ± 9.8 vs. anti-L3T4 mAb 41.6 ± 6.1 eosinophils/five high-power fields, mean ± SD, n = 6, P < .001) (Fig.2). The intrapenitoneal preinjection with anti-L3T4 mAb also decreased the OVA-induced eosinophil infiltration at 48 h by 68% (control rat IgG 99.3 ± 12.5 vs. anti-L3T4 mAb 35.2 ± 4.5 eosinophils/five high-power fields, n = 6, P < .001) (Fig. 2). However, depletion of CD8 T cells by pretreatment with anti-Lyt-2 mAb (53.6-72; 1 mg, twice) had no significant effect on either the first or second peak of OVA-induced cutaneous eosinophilia in the mouse (Fig. 3).

Effect

mouse

48h

ag). The number of eosinophils infiltrating into the skin was then counted at 6, 24, and 48 h. Open columns indicate the number of eosinophils

of eo-

means

of depletion in

(hatched

the

in

Effect

traperitoneally

of eo-

as the sum

2.

infiltration

the

after

infiltration

expressed

of

into

24h

did not eosinophil

120 100

80

skin at 24 and .001) (Fig. 4), observed for However, the

significantly infiltration

60 40

affect in the

20 0

Effects

of disodium

In contrast

cromoglycate

and

6h

PAF antagonist Fig. 3. Effect of depletion infiltration in mouse

to the

inhibitory effects of anti-L3T4 mAb and anti-IL-S mAb on the second peak of antigen-induced cutaneous eosinophilia, disodium cromoglycate, an inhibitor of mast cell degranulation, decreased the first peak of antigen-induced cutaneous eosinophilia in the mouse. Thus,

574

Journal

of Leukocyte

Biology

Volume

52,

November

traperitoneally

twice

with

24h

of CD8’ T cells skin. OVA-sensitized anti-Lyt-2

mAb

on

antigen-induced mice were

(53.6-72)

2-day interval and then challenged subcutaneously Other experimental protocols and symbols are are means ± SD for six mice in each group.

1992

48h

(solid

with the

eosinophil injected in-

same

columns)

at

OVA

2 days

later.

as

Fig.

Data

in

2.

a

Eosinophils

Eosinophils (number/5

(number/5

fields)

I 20

fields) 120

-

1 00

100

80

80

60

60

40

40

20-

20

0-

0 6h

Fig.

4.

Effect

infiltration

of in

mouse

traperitoneally

IgG

with

(hatched

are

means

the

mean

anti-murine

±

24h

IL-S

skin.

mAb

on

antigen-induced

OVA-sensitized

anti-murine

IL-S

48h

mice

mAb

(NCI7)

were

value

of

the

corresponding

infiltration CV-6209

into the skin 40.1 ± 8.4 n = 5, P < .001) (Fig. not affect OVA-induced at 24 or at 48 h (n =

control

response

injected

(solid

columns) 24 h before the subcutaneous OVA SD for six mice in each group. Significantly (rat

6h

eosinophil columns)

Fig. in-

or

(Fig.

rat

challenge. Data different from IgG), P < .001.

sensitized

.

Injection

of

Di sodium

Cromoglycate

on

mice, at

T cell

and

6 and

24

the

number

h.

involvement

OVA

(S ;Lg)

‘Mast

cell

described

g)

4

PBS

+

Diso dium

degranulation

was

by

et

Wershil

(10 g)

Cromoglycate

classified al.

[191.

morphologically Data

are

means

as extensively ±

SD

for

recruitment our findings

Iwamoto

subcutaneously means the mean < .001.

eosinophil columns)

with infiltrating

±

SD

value

or

OVA into

for six mice of the

induction

of LPR

that

in LPR [6, 7], and they at 24 and 48 h. However,

there

is no significant antigen-induced

T cells

and

skin

6 h (Fig.

Antigen

at

-Induced

in

the

PBS OVA-

skin

in each

was

group.

corresponding

control

is in contrast

Mast

are

to

also consistent with our findings indicate

direct connection between eosinophil infiltration

into

CD4 the

2).

Cell

Degranulation cell

in

Mouse

degranulation

Skin

(%)

Slight/moderate

Extensive

91.9 47.9

±

4.8

5.7

±

2.1

±

1.3

±

6.5

20.4

±

4.7

31.7

±

5.6

78.2

±

9.2

10.5

±

3.5

11.3

±

4.7

degranulated, six

are

in the

None

(S

(solid

of eosinophils

Data

Mast

PBS

antigen-induced

cromoglycate

administered

different from P < .05,

.

OVA

Disodium

on

a previous notion that the entire LPR is due to IgE and mast cells [1-3]. Frew and Kay [6] found that numerous eosinophil infiltrates were present at 6 h at the site of allergen challenge in human skin of atopic patients and persisted for up to 48 h and that these infiltrates were associated with a significant increase in the infiltration of CD4 T cells at 6 to 48 h. Other indirect evidence of CD4 T cell involvement in antigen-induced eosinophil infiltration of LPR was an increase in both CD4 T cells and eosinophils in bronchoalveolan lavage fluid 48 h after segmental bronchial challenge [7]. These findings suggested that there might be some connection between the infiltrating CD4 T cells and the eosinophil

with

as

skin. was

24h

cromoglycate

of CD4 T cells and IL-S in the second peak of antigeninduced eosinophil recruitment of cutaneous LPR, the first peak is dependent on mast cells and PAF, as indicated by the inhibitory effects of disodium cromoglycate [16-18] and the selective PAF antagonist CV-6209 [14] at 6 h (Figs. S and 6). Our finding of the preventive effect of the depletion of CD4 T cells provides direct evidence that CD4 T cells are involved in the induction of antigen-induced eosinophil recruitment in cutaneous LPR. The present finding of CD4

6).

Effect

mouse

., * Significantly response (PBS),

of cutaneous LPR, as indicated by the of the in vivo depletion of CD4 T cells, but cells, on antigen-induced eosinophil infiltraskin at 24 and 48 h (Figs. 2 and 3). We also mediates this eosinophil recruitment in the LPR (Fig. 4). Finally, in contrast to the role

I.

disodium

columns)

counted

The development of mAb to T cell subsets has made it possible to deplete selectively specific T cell subsets in vivo [21, 22]. This approach has been used successfully in several cxpenimental models to assess the role of T cell subsets in the immunologic control of microbial infections [23], allognaft rejection [24], and autoimmunity [25, 26]. Therefore, since no direct evidence has been provided for the functional importance of CD4 and CD8 T cell subsets in causing antigen-induced eosinophil recruitment into the site of LPR, we have applied this approach to antigen-induced cutaneous eosinophilia in the mouse. In this study, we show that CD4 T cells play an important role in causing the second peak of antigen-induced eo-

TABLE

of

in

(hatched

DISCUSSION

sinophil recruitment preventive effect not of CD8 T tion into mouse show that IL-S second peak of

Effect

infiltration

at 6 h by 56% (saline 87.3 ± 14.4 vs. eosinophils/five high-power fields, 6). Pretreatment with CV-6209 did eosinophil infiltration in mouse skin 5)

5.

mice

et al.

in

slightly

each

CD4

to moderately

degranulated,

2.4

or

none

according

to the

criteria

group.

T cells

and

IL-S

in antigen-induced

eosinophilia

575

Eosinophils (number/5 fields)

cells dominantly Thl-type T cell

120

80 60 40

-

20

-

0-

6h 6. Effect mouse

with 30

a PAF

mm

of

skin.

means mean

and

SD

±

value

By

PAF

antagonist

OVA-sensitized

antagonist

before

CV-6209 24

h after

24h on

antigen-induced

mice

were

(solid

columns)

the

our

results

eosinophil

injected

for

the

“8

“

-?

I U

infiltration

intraperitoneally

or saline

subcutaneous

for five mice in each group. of the corresponding control

contrast,

OVA

twice

(hatched challenge.

Significantly different response (saline), P