Induction of leukotriene production before antigen challenge enhances antibody affinity in genetically selected mice

CEILI

LAR IMMUNOLOGY

136, 173-184 (1991)

Induction of Leukotriene Production before Antigen Challenge Enhances Antibody Affinity in Genetically Selected Mice CATHERINE PHILLIPS London

School

qf Hygiene

Received

and Tropical

November

Medicine.

16. 1990; accepted

London, March

Uniled II,

Kingdom

1991

Mice genetically selected for their incapacity to produce high-affinity antibody to protein antigens in adjuvant (nonmaturing (NM) mice) were treated with indomethacin, an inhibitor of the cyclooxygenase pathway of arachidonic acid metabolism. Pretreatment with indomethacin significantly enhanced the affinity of antibodies produced 21 days after immunization with human serum albumin (HSA). Blockage of the cyclooxygenase pathway in this way was shown to induce the production of leukotrienes via the lipoxygenase pathway. The production of leukotrienes may well be responsible for the enhanced antibody affinity, since blockage of the lipoxygenase pathway in addition to the cyclooxygenase pathway reversed the effect. In an attempt to elucidate the mechanisms involved, IL-l production and Ia expression by macrophages were examined. Ia expression by peritoneal cells from untreated NM mice was significantly lower than that by their high-affinity-producing counterparts 3 days after immunization. Indomethacin pretreatment raised inducible la antigen levels on macrophages of NM mice to those seen on cells from untreated high-affinity mice. Indomethacin treatment alone induced the production of IL-l by macrophages in NM mice. However, 3 days after immunization and the withdrawal of indomethacin in NM mice. IL-l production was significantly lower than the response of NM mice given antigen alone, suggestive of the induction of a feedback mechanism. Thus indomethacin pretreatment results in a cascade of events in macrophages which produce a decrease in IL-1 production and an increase in inducible Ia expression 3 days after antigen challenge. 0 1991 Academic PRESS. hc.

INTRODUCTION Two lines of mice which produce either low-affinity (LO) or high-affinity (HI) antibodies to protein antigens in saline have been derived from the originally outbred TO mouse strain (1). When antigen is given in Freund’s complete adjuvant (FCA) LO mice respond by either producing antibodies of high affinity or continuing to produce low affinity antibodies. Affinity nonmaturing (NM) mice were derived from the LO line by interbreeding of those mice which failed to produce high-affinity antibody to antigen in adjuvant (2). During the breeding program, brother-sister matings were avoided in order to preserve the outbred nature of the lines and to maximize the possibility of selecting solely for factors controlling antibody affinity. The availability of these mice, therefore, provides a unique opportunity to study the control of antibody affinity maturation as any consistent differences between the lines are likely to be directly relevant to this process. It has been shown that spleen adherent cells from HI line mice produce large amounts of prostaglandin E2 spontaneously in vitro whereas those from NM mice produce almost no PGEZ. LO mice produce intermediate levels (3). In addition, administration l?3 0008-8749/9 1 $3.00 Copyright0 1991 by Academic

F’res., Inc. All rights of reproduction in any form rewwd.

174

I‘N'HERINE PHILLIPS

of indomethdcin to HI mice during the first week after immunization significantly decreased the affinity of their antibody response (3) demonstrating an important role of arachidonic acid metabolites in the control of antibody affinity maturation. PGE2 is a major arachidonic acid metabolite of the mouse macrophage (4), although these cells also produce leukotriene C4 (5). PGE2 has been implicated in the control of a large number of cellular functions involved in the immune response, presumably resulting from elevation in intracellular CAMP (6-12). It is involved in a feedback induction loop with interleukin-1 (IL-I) (13, 14) down-regulation of Ia expression ( 15) and the inhibition of the production of IL-2 (16). The precise effect of PGE2 upon B cells is yet to be clarified. PGE2 has been shown to potentiate differentiation into immunoglobulin-secreting cells if added early to cuhures but inhibit terminal immunoglobulin secretion if added late ( 17). On the other hand, arachidonic acid metabolites produced by the lipoxygenase pathway have also been implicated in the induction of a number of lymphokines, including IL- 1 (18) IL-2 ( 19) and y interferon (20). Thus in this study the control of arachidonic acid breakdown in the affinity nonmaturing mice and its relationship to possible mechanisms of antibody affinity maturation was explored in order to elucidate the nature of the defect in these mice. MATERIALS

AND

METHODS

High- and low-affinity and NM mice were obtained from an initially randomly bred population of TO mice by selective breeding as described previously ( 1, 2). Male and female mice were used at 8 to 12 weeks of age.

One milligram of human serum albumin (HSA) (Miles Scientific) was injected in FCA (Difco Laboratories, Detroit. MI) by the intraperitoneal route (i.p.) in 0.2 ml phosphate-buffered saline (PBS). A stock solution of indomethacin (Sigma) (5 mg/ml in ethanol) was prepared and stored at - 70°C. A l/25 dilution in PBS was given at I mg/kg every 24 hr. The 5-lipoxygenase inhibitor, L 65 1 392, (Merck, Sharpe, and Dohme) was given i,p. at 4 mg/kg in PEG 400 (Sigma) every 24 hr. The dual pathway inhibitor, BW755c. (Wellcome) was given i.p. at 50 mg/kg in water every 12 hr. Radiolabeled Immztne C.hmpie,wr

Radiolabeled immune complexes were prepared by mixing [““I]-labeled HSA with sera in the proportions found to give maximum precipitation. These were incubated at 37°C for 30 min and at 4’C for 30 min and centrifuged at lOOC@ The resulting pellets were then washed once in PBS, resuspended to I mg/ml HSA in PBS, and stored at 4°C until use (maximum 5 days).

HI mice express Ia‘ exclusively while LO and NM mice expressed la‘ and la”“’ antigens in the approximate proportions 35%:65% (G. Holland, personal communication). Therefore a mouse anti-mouse Iakrfsantibody (10.36: IgG2a. American Type

LEUKOTRIENE

PRODUCTION

Culture Collection), which binds both essential medium (MEM) containing buffer and the antibody was purified mouse (Fc) FITC was obtained from Antibody Afinity

IN

SELECTED

MICE

175

specificities, was used. This was grown in minimal 10% fetal calf serum (FCS) and 4% bicarbonate on a protein A-Sepharose column. Goat antiNordic.

Measurement

The relative affinity (Kr, liters/mole) of antibody to HSA was measured by a double isotope ammonium sulfate globulin precipitation method which incorporated a 22Na volume marker (2 1). Free and antibody-bound antigen was determined at equilibrium over a range of antigen concentrations in antigen excess. The total level of antibody (Abt, picomoles of binding sites per 10 ~1 serum) was obtained by extrapolation of a Langmuir plot of the reciprocal of the bound antigen versus the reciprocal of free antigen, 1lb = l/K * l/c * l/Abt + l/Abt, where b = bound antigen, c = free antigen, and K = affinity. A curvilinear plot was obtained and antibody affinity calculated by linear regression analysis using data points where less than 50% of the total antigen was bound. Affinity was expressed as the reciprocal of the free antigen concentration when 50% of the total antibody binding sites was bound to antigen. Phagocytosis and Oxidative Burst by Macrophages

Zymosan (Sigma) was boiled three times and the number of particles adjusted to 108/ml. A l/20 dilution of this was found to give maximum reduction of nitro-blue tetrazolium and was used subsequently. Spleen cells were allowed to adhere to coverslips for 90 min at 37°C and nonadherent cells washed off. Nitro-blue tetrazolium (Sigma) (5 mg/ml final concentration) was added to the adherent cells with the zymosan and the coverslips were incubated for 20 min at 37°C. They were then put on ice and inverted onto slides and cells were examined under the microscope for reactive oxygen production (blue-black color) and the number of particles was phagocytosed. Degradation

of Immune Complexes by Macrophages

Peritoneal cells were obtained by peritoneal lavage with 5 ml cold RPM1 medium containing 2% FCS and 20 mM Hepes. After washing, cell numbers were adjusted to 106/ml in RPM1 containing 10% FCS, 20 mM Hepes, and 5 X lop5 M 2-mercaptoethanol and 200~~1 aliquots plated into 96-well tissue culture plates. Duplicate plates of peritoneal cells were fixed in 4% formal saline for 20 min and washed and fresh medium was added. Chloroquine (Sigma) (0.1 mM final concentration) and/or indomethacin (Sigma) (1 fig/ml final concentration) were added to live cells and the plates were incubated for 1 hr at 37°C to allow incorporation of the drugs. Twenty microliters of fresh autologous serum containing 2 ~1 of radiolabeled immune complexes were added to peritoneal cells and the plates were incubated for a further 6 hr at 37°C. Two hundred microliters of supernatant was removed from each well and the protein precipitated with trichloroacetic acid. ‘25I in both the pellet and supematant was counted on a gamma counter. Ia Antigen Expression on Macrophages

Peritoneal cells were allowed to adhere to coverslips for 1 hr at 37°C. These were then washed in warm medium and Fc receptors were blocked by incubation with 50%

human serum for 10 min at room temperature. The medium was removed and replaced with anti-Iak’” plus 5% human serum and incubated at room temperature for 1 hr. Coverslips were washed twice and covered with goat anti-mouse Ig (Fc) FITC, incubated for a further hour, washed 3 times, and fixed in 4% formal saline for 20 min. Coverslips were washed, mounted on slides, and examined for positive staining by fluorescent microscopy

Spleen cells were allowed to adhere to Wcm- tissue culture tlasks (Nunc) at j?“<’ for 90 min and nonadherent cells were removed by washing. Five milliliters of versenc (2.64 mM EDTA) was then added and the flasks were placed at -20°C for 5 min and then on ice for a further I5 min and cells were detached with vigorous banging. These were then washed in RPM1 and adjusted to 10’ cells/ml in RPM1 containing O.l’T; bovine serum albumin (BSA). Cells were incubated in 12 ‘* 75mm polypropylene round-bottomed tubes (Falcon) with 1 PC1 [“Hlarachidonic acid (195 Ci/mmol 15,6,X,9. I 1.12,14,15-“H. Amersham International plc) per milliliter for 6 hr at 37°C iyith occasional agitation. Tubes were then rapidly cooled by being placed at -20°C for 5 minutes and spun at 10008 for 5 min at 4°C. Supernalants were removed into I.5ml polypropylene tubes and nitrogen gas was bubbled through them. The tubes were then tightly sealed and stored at - 30°C until the day of the HPLC run. Supematants were analyzed by reverse-phase HPLC in the Department of Medicine. Guy’s Hospital, under the kind auspices of Tak Lee. Products from each spleen adherent cell culture were applied to a 1O-pm (118 LJltrasil ODS column (4.5 :< 2SO-mm). This had been standardized and calibrated for the retention times ofthe synthetic standards leukotriene B4 (71 min) and 5-HETE (49 min) at a flow rate of ! ml/min. with a ?solvent program which resolves dihydroxy leukotrienes and monohydroxy fatty acids as described in Ref. (22). One-milliliter fractions were collected and 300 ~1 of each fraction was placed in a scintillation vial (standard Packard) and 3 ml Lumagel (Lumac, Basel) was added to each vial. .After whirlimixing the ‘II content was counted on an LS 9800 beta counter (Beckman i

Peritoneal cclis were incubated at llY’/ml with or without lipopolysaccharide from .%hnondlu mimxwra (LPS) (,Sigma) ( 10 &g/ml) for 24 hr at 37°C. Supernatants were harvested and stored at ---70°C. II.-1 content of supematants was kindly assessed by Nick Smithers of Glaxo Ltd, Greenford. using the Dl0.W 1 T cell line (23). Briefly. the D 10 cells were cultured in RPM1 1640 medium with 2 m&‘glutamine, antibiotics. 10% FCS. 60 m-W 2mercaptoethanol. The cells were fed twice a week with alternate stimuli: tirst. with irradiated mouse ICBA) splenocytes, conalbumin (&Sigma).and rat IL-2 (Sigma). and second. with IL-? only. Cells for IL- I assay were harvested by Ficoll centrifugation from cultures that had received feeder cells at least 7 days previously. Dilutions of putative IL- I -containing samples or of an IL- t standard (recombinant lL- i B. Genryme, Koch-light) along with C’oncanavalin A (0.25 Kg/ml final concentration) were added to Dlcl.G4.! ::ells (2 ‘j: 10” cellsjweil) and incubated for 24 hr. fhesc were pulsed wit11 /‘H]thymidino [or a hither 18 hr. harvested, and counted ‘or ‘H incorporatron. IL- 1 contents of the samples were assessed from the half-maximal t’alue of the recombinant IL.-i

LEUKOTRIENE

PRODUCTION

IN SELECTED

MICE

177

RESULTS Pretreatment with Indomethacin

Increases Antibody Afinity

in NM Mice

Four groups of NM mice were treated daily with indomethacin (1 mg/kg) for 7 days in separate consecutive weekly regimes (Days -7 to 0,O to 7, 7 to 14, and 14 to 21). Mice were challenged on Day 0 with 1 mg HSA in FCA and anti-HSA antibody affinity was measured on Day 21. Indomethacin given at any weekly interval after antigen challenge had no effect on antibody affinity. However, if indomethacin was given before antigen (Days -7 to 0) antibody affinity was significantly enhanced (Fig. 1). In order to determine the precise requirement for indomethacin pretreatment, indomethacin was administered for shorter time periods during the week preceding antigen challenge. A minimum of 4 days indomethacin pretreatment was necessary to significantly enhance antibody affinity above that of the control group (Fig. 2). The Efect of Indomethacin

Pretreatment of NM Mice on Macrophage Function

The finding that indomethacin can enhance antibody affinity when given as a pretreatment but not when given at any time after antigen challenge suggests that indomethacin affects an early immunological process, such as antigen presentation by macrophages. Therefore, the effects of indomethacin on a number of parameters of macrophage function was examined and the results are summarized in Table 1. Stimulation of macrophages via Fc, complement, and mannose receptors generally triggers a respiratory burst that results in the generation of superoxide anion (O*-) and other reactive oxygen intermediates (24). It requires the activation of an NADPH oxidase which is increased in activated cells (25). Nitro-blue tetrazolium (NBT) is a yellow compound which, when reduced by reactive oxygen intermediates, forms a

t CD 0 X

1.0-

a, E 3 i?? c -I c

0.5-

z-

COntrOl

Pretreatment

1st week

2nd week

3rd week

FIG. I. NM mice: the effect of indomethacin on anti-HSA antibody affinity (Kr (liters/mole)) 2 1 days aRer antigen challenge with 1 mg of HSA in FCA. Indomethacin (1 mg/kg daily) was given in separate weekly regimes before and after antigen challenge (5 mice per group, mean f SE). *P -c0.2.

17x

i A’i’lHERINI:

I’IIIL.I.IPS

~-IL,. 2. NM II~ICC‘: the ctlect 01 rndomethacin on anti-HSA after antigen challenge with I mg of HSA in FCA, Indomethacin Day 0 being the day of antigen challenge. The vehicle control mean i- SE). *.P e: O.Oi.

antibody affinity (Kr (Iuersjmole)) 51 days (1 mg/kg) was given on the days indicated. was 4% ethanol in PBS (5 mice per group..

blue/black precipitate. Thus II can be used to measure respiratory burst activity m response to zymosan which is phagocytosed after attachment to mannose receptors (26). A higher proportion of spleen adherent cells from HI mice was found to produce a respiratory burst in response to zymosan than those from NM mice in this assay (Table 1). The number of cells recoverable by peritoneal lavage and the ability of these cells to degrade radiolabeled immune complexes via a chloroquine-sensitive pathway were examined. Neither of these functions differed between cells from HI and NM mice, although indomethacin significantly enhanced both parameters in both lines (Table 1) i ABl.1.

i

Summary of Macrophagc I\lumhcr and Functton in NM and f-11 ,Mice: Reactice Oxygen Production hy Spleen Adherent Cells (Nitroblue Tetrazolium Reduction); The EfTect of Indomethacin Treatment ( 1 mg/kg Daily for 4 Days) on Peritoneal Cell Number (Recoverable by Lavage) and Immune Complex Processing Ability (Chloroquine-Inhibitable Release oFTrichloroacetic Acid-Precipitable Radiolabeled Protein) NM lndomethacin

HI mice

pretreatment:

4ssa> NBT (‘i; positive ‘r St) Degradation of immune complexes (cpm -t SE\ Peritoneal cell numbers ‘; 10 h 5 SE

.Vote.5

mice

mice per group. * P < 0.05. ** P < O.Ol

10.4 t_ 3. i*

3200 f 204 I * 7.0 -i 0.5 mean

I: SE.

n&I.

11.400

f

‘I.0

lY60*

! i.2 I 2.8

12,800

i I.h*

? lY60**

?.(I f I?.T*

n.d.

34.600 i6.h

.’ 7367** 7 3.4*

LEUKOTRIENE

PRODUCTION

IN SELECTED

MICE

179

Peritoneal macrophages constitutively express low levels of Ia antigen (27). However, Ia expression is rapidly increased after antigen challenge (27,28). Although no differences in constitutive Ia expression by cells from HI or NM mice could be found (data not shown), Ia expression 3 days after antigen challenge was significantly greater by cells from HI mice than by those from NM mice (Fig. 3). In addition, indomethacin pretreatment of NM mice significantly increased levels of Ia antigen on their peritoneal cells 3 days after immunization, to those comparable with HI mice (Fig. 3). This effect was still discernible by Day 7 (data not shown). Indomethacin Pretreatment Induces Leukotriene Production in Spleen Adherent Cells from NM Mice Inhibitors of the cyclooxygenase pathway of arachidonic acid metabolism can enhance production of metabolites generated by the 5-lipoxygenase pathway (29). In addition, 5-lipoxygenase products have been shown to have a number of potentiating effects upon the humoral immune response, notably upon IL-l production ( 18). Therefore, the ability of indomethacin to induce lipoxygenase products was examined. This was done using reverse-phase HPLC with a system of buffers designed to resolve dihydroxy leukotrienes and monohydroxy fatty acids (22). Spleen adherent cells from untreated NM mice secreted a very small amount of 5 HETE (47 min) and another unidentified product (34 min) in contrast to those from HI mice, which secreted large amounts of a product with a short retention time, probably PGE2 (3) (Figs. 4A and 4B). The control of the production of different arachidonic acid metabolites is very precisely balanced and subject to feedback control. Thus. after a 4-day treatment with indomethacin in vivo spleen adherent cells were obtained and cultured with or without indomethacin during the assay. When indomethacin was given in vivo and also included in vitro, a large peak was seen at 16 min (Fig. 4C). This was not leukotriene B4 as the standard eluted at 21 min, but the

60 -

NM

HI

FIG. 3. The effect of indomethacin pretreatment (1 mg/kg daily for the 4 days preceding antigen challenge) on la expression (% positive staining with anti-Ia (krfs) and goat anti-mouse (Fc) FITC) by peritoneal cells from NM and HI mice 3 days after antigen challenge. Lightly hatched bars, controls; dark-hatched bars, indomethacin-pretreated mice (5 mice per group, mean f SE). *P < 0.05. +P < 0.05.

I x0

i“,U IlFRINF

PHILLIPS

Mm

-

FIG. 4, Reverse-phase high-pressure liquid chromatograph) of arachldonrc acid metabolitea of spleen adherent cells from NM and HI mice. (A) Untreated NM mice; (B) untreated HI mice: (C)NM mice treated with indomethacin both iw viva and irr I./PO: (D) NM mice treated with indomethacin in viva alone. (“H content (cpm) of I-ml samples (column rate: I mlimin))

unresolved double peak of the 6-transleukotriene B4 diasterioisomers (30) which are the nonenzymatic breakdown products of leukotrienes C4 and D4. When indomethacin was not included in culture after 4 days of in viva treatment large peaks of a range of metabolites of shorter retention times were obtained (Fig. 4D). Thus, indomethacin pretreatment of NM mice induces the production of leukotrienes but the pattern of arachidonic acid metabolites is rapidly altered once indomethacin is withdrawn.

ment qf’Antihod~. .4finity in Indomethacin-Pretreated

NM Micrj

In order to see if the enhancement in antibody affinity m indomethacin-pretreated NM mice is a result of leukotriene synthesis, mice were treated with drugs that block either the cyclooxygenase or lipoxygenase pathway, or both. L651 392 is a specific inhibitor of Slipoxygenase (3 1). It has been designed to functionally mimic the substrate of the enzyme and has been shown to be a potent inhibitor of 5-hpoxygenase activity in vitro (3 1). In addition, it is effective in a number of in viva systems when administered p.o. at a dosage of 5 mg/kg in 5% ‘Tween 80 (3i). In preliminary experiments, this regime was found to enhance antibody affinity in NM mice (data not shown) and this was found to be due to the effects of the vehicle. The use of PEG 400 has been shown to increase the bioavailabihty of the drug (Merck, Sharpe, and Dohme, internal information) and in subsequent experiments the drug was administered i.p. in order to maximize the possibility that it reached the cells involved. PEG 400, alone, did not enhance antibody affinity (Fig. 5). In addition, L65 I 392 in PEG 400 had no effect on antibody affinity (Fig. 5). However, in conjunction with

LEUKOTRIENE

PRODUCTION

IN SELECTED

MICE

181

Pretreatment NOW

/

PEG 400

L651 392

J

t

lndomethacin

--*

I

+

L651 392 plus lndomethacl BW755C

-015 Kr (litreslmole)

10

x 10

-6

FIG. 5. NM mice: the effect of blockage of the cyclooxygenase and/or 54ipoxygenase pathways on antiHSA antibody affinity (Kr, liters/mole). (5 mice per group, mean f SE). *P < 0.05. *P < 0.05.

indomethacin, this drug reversed the enhancement in antibody affinity caused by indomethacin (Fig. 5), suggesting that the increased production of lipoxygenase products induced by indomethacin is responsible for the increase in antibody affinity induced by this agent. L65 1 392 is an experimental drug and may not be effective in viva in all compartments of the body. However, the results above suggest that this drug has been effective in vivo in this system. BW755c is a better-characterized and more widely used drug which has been consistently shown to block both the lipoxygenase and cyclooxygenase pathways in vivo as well as in vitro (32-37). Once again, blockage of the 5-lipoxygenase, in conjunction with the cyclooxygenase pathway, was able to reverse the effect of blockage of the cyclooxygenase pathway alone (Fig. 5). Indomethacin

Pretreatment Induces the Production of IL-l

by NM Peritoneal Cells

As leukotrienes have been found to induce release of IL-l by macrophages (18), supernatants from peritoneal cells from untreated and indomethacin-pretreated NM mice were assayed for IL- 1 levels using the DIO.G4.1 cell line. Indomethacin pretreatment significantly enhanced the spontaneous release of IL- 1 by resident peritoneal cells (Fig. 6). However, 3 days after antigen challenge the effect of indomethacin pretreatment was reversed. At this time cells from indomethacin-pretreated mice produced significantly less IL- 1 in response to LPS than those from control mice (Fig. 6), suggestive of feedback control in indomethacin-pretreated mice. DISCUSSION The major findings of this study are that a minimum of 4 days pretreatment with indomethacin significantly enhanced the affinity of antibodies produced by NM mice (Fig. 2). This regime of indomethacin treatment also induced significant production of leukotrienes (Fig. 4). Furthermore, the enhanced production of leukotrienes induced by indomethacin is likely to be the cause of the enhancement in antibody affinity. This conclusion is based on the results of experiments which showed that blockage of

‘Jnstim /i

i PS 1 p,

‘1

i.?S i,

FIG. 6. NM mice: the rtfect ofmdomethacio pretreatment (1 my/kg daily for 1 days) on IL-i production by peritoneal cells from NM mice before and 3 days after antigen challenge with I mg HSA in FCA. Control: lightly hatched bars: pretreated with indomethacin: dark-hatched bars. !Ilnits/ml of II-.-l, mean ? SE). *P <_ 0.02. ‘P *’ 0,O’.

the lipoxygenase pathway in addition to the cyclooxygenase pathway reversed the enhancement of antibody affinity seen when the cyclooxygenase pathway alone is inhibited (Fig. 5). Macrophage function was examined in both HI and NM mice and a number of differences were found. A greater proportion of spleen adherent cells from HI mice produced reactive oxygen intermediates in response to phagocytosis of zymosan than did cells from NM mice. In addition, a smaller proportion of spleen cells from NM mice was previously found to adhere to plastic than did those from HI mice (3), suggesting that the defective antibody affinity maturation in NM mice may be a result of poor macrophage function. Peritoneal macrophages from HI mice express much higher levels of Ia antigen _i days after immunization than NM mice. However. pretreatment of NM mice with indomethacin induced an increase in levels on these cells comparable with the response seen in HI mice (Fig. 3). The increase in Ia antigen expression on macrophages during the first week after immunization is believed to be mediated by y interferon (38). Leukotrienes have been shown to induce the release of 7 interferon (20). However. leukotrienes, although rapidly induced, decline equally rapidly (39) and may not be present 3 days after the withdrawal of indomethacin. As leukotrienes have been shown to induce the production of lL- i (18). this was measured both before and 3 days after antigen challenge. Whereas IL-1 production by peritoneal macrophages was significantly enhanced immediately following indomethacin treatment, the release of this cytokine was significantly inhibited 3 days after antigen challenge and the cessation of indomethacin treatment (Fig. 6). IL-l induces prostaglandin E2 production (40) and thus presumably its own feedback inhibition (4 1). In addition, increased PGE2 production during the first week after antigen challenge has been shown to be responsible for the high-affinity responses in the HI line mice (3). Thus, it is possible that the enhancement of antibody affinity induced by indomethacin in NM mice is via the induction of a feedback inhibition loop involving leukotrienes and IL- 1, followed by increased PGE2 production. In support of this, we

LEUKOTRIENE

PRODUCTION

IN

SELECTED

183

MICE

have shown in this study that the production of leukotrienes is dependent upon the continued presence of indomethacin, as its withdrawal in vitro results in a change in the pattern of arachidonic acid metabolites to more hydrophobic products (Fig. 4). Early studies examining the involvement of T cell in antibody affinity maturation suggested a limited role as the transfer of thymocytes into T cell-depleted mice did not significantly raise antibody affinity once sufficient T cells had been transferred to induce class switching. In fact, excess primed T cells reduced, whereas adult thymectomy enhanced, antibody affinity (reviewed in Ref. (37)). However, it has been shown that recent thymic emigrant T cells which enter peripherial lymphoid tissue after antigen challenge are quantitatively different from those already in the periphery of the mouse (43). Swain et al. (44) have suggested that the thymectomy-sensitive T cells described by earlier workers (45, 46) are the precursors of the IL-4 (Th2)-producing cells of Mosmann and Coffman (47). In addition, it has been shown that there is a decrease in IL4-producing T cells after adult thymectomy in the rat (48). Evidence presented in this study would indicate that it is the y interferon-producing T cells (Thl) that have a more important role to play in antibody affinity maturation; thus the use of thymocytes in the early studies may have been misleading. In conclusion, it is possible that PGE2 and IL-l may control the balance of lymphokines secreted by T cells which have been shown to be differentially regulated in a number of studies. For example, dexamethasone, which is also a prostaglandin inhibitor, has been shown to enhance IL-4 production (49) and IL-l costimulates in the proliferation of IL4-producing Th2 cell clones but not in the proliferation of Thl clones which produce y interferon (50). In addition, the induction of IL-2 and IL-4 has been shown to be differentially sensitive to an increase in cyclic AMP (5 I, 52). Thus it may be possible that pretreatment of NM mice with indomethacin overcomes a macrophage defect in these mice by inducing a feedback loop of PGE2 and changing the balance of lymphokines in favor of those more effective for antibody affinity maturation. ACKNOWLEDGMENTS I thank bv a grant

Madeleine Devey from the S.E.R.C.

for her help and guidance in facilitating in conjunction with Unilever plc.

this research.

This

work

was funded

REFERENCES Katz, F., and Steward, M. W., Immunology 29, 543, 1975. Devey. M. E., Bleasedale, K., Stanley. C., and Steward, M. W., J. Immunol. 52, 377, 1984. Phillips, C., Cell. Immunol. 119, 382, 1989. Humes, J. L., Bonney, R. J., Pelus, L., Dahlgren, M. E., Saldowski, S.. Kuehl, F. A., Jr., and Davies, P., Nature (London) 274, 261, 1977. 5. Rouzer, C. A., Scott, W. A., Cohn, Z. A., Blackburn. P., and Manning, J. M.. Proc. Natl. Acad. Sri. I. 2 .s. 4.

USA 77,4928, 1980. 6. ‘7. 8. 0. 10. I I, 12. 13. 14.

Zurier, R. B., Dukor, P., and Weissmann, G., Clin. Res. 19, 453, 1971. Weissmann, G., Dukor, P., and Zurier, R. B., Nature New Biol. 231, 131, 197 I. Oppenheim, J. J., Koopman, W. J., Wahl, L. M., and Dougherty, J. F., Cell. Immunol. Goodwin, J. S., Bromberg, S., and Messener, R. P., Cell. Immunol. 49, 64, 1980. Snyder, D. S., Lu, C. Y., and Unanue, E. R., J. Immunol. 128, 1458, 1982. Knudsen, P. J., Dinarello, C. A., and Strom, T. B., J. Immunol. 137, 3189, 1986. Simkin, N. J., Jelinek, D. F., and Lipsky, P. E., J. Immunol. 138, 1074, 1987. Dinarello, C. A., Marnoy, S. 0.. Rossenwasser, L. J., J. Immunol. 130, 890. 1983. Kunkel, S. L., Chensue, S. W., Phan, S. H., J. Immunol. 136, 186, 1986.

60, 298, 1981.

IX4 15. 16. 17. 18. 19. 20. 1 I. 72. 13. 24. 25. 26. 2i. 38. 29. 30. 3 1.

32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 5 I. 52.

( 4 I’HkKlNt:

PHI1 i 11’5

Snyder, D. S.., Lu. C-. k’.. and Linanue. 1:. R.. Vatwt, 299. ih3, !YX.I Stegg, P. S.. Moore. R. N.. Johnson H. M.. and Oppenheim. J. f .f Ey? ll& 129, 2402 iYX3. Jelinek. D. F.. Thompson, P. 4.. and Lipsky, P. E.. .I C/in Invcc: 75, 1339. 1985. Dinarello. C. A.. Bishai, I.. Rossenwasser, I,. J., Coceani. F.. fnt /rnmunclphur7nocir}nl~~,f~ 16, 43. k984 Farrar, W. I_ . and Humen, J. f ._ ,I. Imnwwi 135, 1 153. 1982 Johnson H. M., and Torres. H. A.. .i Irnmcmoi 132. 413. 1Y84 Gaze. S. E., West. N. J.. and Steward. M. W.. .i. Imrnuwc~i M&x!,, 3, 357. I Y73. Lee. C. W.. I,ewis. R. 4.. Tauber. 4 1.. Mehrotra. M. M.. Corq. F. J.. and Austen. f., .i Hloi. ( ‘iw); 25x. 15,004. 1983 Kaye. J.. Gillis, S.. Mirei. 4. l3.. Shevach, i.. M.. !Malek, B R.. Dtnarelio, C *\ 1 1 achmax IL. H.. and Janeway. C. 1.. Jr i .i flnmrnr:i 133, 13.30. 1w: Johnston. R. R.. Jr.. I,wiphol,irw 3. 33. I YX I Nathan, c‘. F.. and Root, R. K.. .I I’xf> .%jtvi. 146, 1648, LY7i Sung. S.-S. ‘T.. Nelson. R. S.. Silverstein, S. C‘ .I C‘C# Bioi 95, 1611. lYX.3 Belier. 1). 1.. Kiely. J.-M , and Unanue. t:. R.. ,i iwrnnnoi 124. 1426. IYXO. Tripp. C. S.. Unanue, I!. Ii.. Yeedleman. P . pr~ji. 4iltl :ltut/. SC,; LrS.4 83, 9655. lY86. Salmon. J A.. Simmons. P M.. and Moncada. S.. .I I’/IUIYIM Phartnclcrd 35, 808. 1983. Williams. J. D. L.ee. T t-1.. f.ewih. R. 4.. and 4usten F.. ./ Irnrnrmol 134, 2624. 1985. Guidon. Y.. Girard. t’.. Maycock. A.. Ford-Hutchinson. A. W.. TV!ai. Irr “Advances in Prostaglandin. rhromboxane and lxukotrienr Research” (B. Samuelsson. R. Paoletti. and P. W. Ramwell. Edsi. Vol. 17. pp. 554-557. Raven Press, Nevi York. 198’7. Higgs, G. A.. Flower. R. J.. and Vane. J. R.. HIIC/~IYI I%arl?iac.~~i. 28, lY5Y. 1’)70. Blackwell, G. J.. and Flower. R. J.. /‘,osiw&rrdrn.\.is 16. 4 17. 193 Burka. J. F.. and Flower. R. j Br .I Phormom/ 65, 35. 1979 Radmark. 0.. Malmstrn. I>.. and Samuelsson. H.. IcF.‘RS 1x/l. 1 JO, ‘13. IYXCI. Randall. R. W.. Eakins. k. I:.. H&s. G 2. Salmon. J A., and ‘l’ateson. J. t.. 31gu7u .A~~rtons 10. 553. I980. Siegel, M. I., McConneil. K. i ., Konser, K. M ‘md Cuatrecasas. P.. il’ru.s/u&r~d~~~s 21, 123, 19X1. Stegg, f’ S.. Moore. R. I\. Johnson. ff. M. and Oppenheim. J. .I,. ,I I-:xp. ?4acl 156. 1780. 1982 Parker, C. W 4nnx Rw IIME~~~o~ 5, 65. 1X07, Dinarello, (‘. .4., Bishai, 1.. and Rossenwasscr. 1. ,I,, .i I~vlm~tn~~/ 130, XYfi. 198.3. Kunkel. S. I... (‘hensue. S W.. and Phan. S. ff .I Iwnrwiol 136, 186. 1986, Steward. M. W.. and Steensgaard. J.. .‘Antibody Afhnitv. Thermodynamic Aspects and Biological Signiiicance.” CRC Press. Rota Katon. FL Iq8’. Koga. Y . Taniguchi K.. and liomotn, ti.. /r,lrn!ivrohrcji~~~~ 166. 3$1. lYX4 Swain. S. L.. McKenzie. D ~!‘. Duttom. R. W.. ‘Tonkonogy. S [ ._ and English. M.. /rri~~irtn(~i &,I. IO?-9 77, lY8P. Raff. M. C and Cantor. Ii “Progrw in Immunology” (R. .Amos, Fd.). Academic Press. New York. 11)-l’ Kapler. .I. VV Huntci-. I‘ i ‘.. Jacobs. D and Lord. L .: Irnn7iiricti. 113, LT. ! 97: Mosmann. T R.. and Coffinan. R. L ., Imrn~x~ To&r 8, 223. 1987 Arthur. R. I’.. and Mason I>.. .I C.x[l ,Ifcvi. 163. 774, 1986. Daynes. R. 1.. and Aranro. B. A.. Blur J /mmunc~i. 19, 2319. 19X?. Kurt-Jones, E. 4.. Hanberg. S.. O’Hara, J.. Paul. W. E.. and Abbas. 4. K.. .i .&\;I. .Med. 168, 543, 19X7 Munoz. M.. Zubiaga. A. M.. Merrow. M.. Saute:, N. I’.. and Huber. B. ‘I,. ./ E,qx 12iled. 172, 95, 1990. Novak. ‘1‘. J.. and Rothenberg. F V.. I’rix, N~rti .l~crd .%I (!LS’,4 87, 9353, 1990.