INTERNATIONAL JOIIRNAL OF LEPROSY^ Volume 61, Number 4

Printed in tlw

Differential Production of Interleukin 1 (IL-1),IL-6, Tumor Necrosis Factor, and IL-1 Receptor

Antagonist by Human Monocytes Stimulated withMycobacterium leprae and M. bovis BCG1

Katsuya Suzuki, Yasuo Fukutomi, Masanori Matsuoka,Keiko Torii, Hidetoshi Hayashi, Takemasa Takii,

Yasukazu Oomoto, and Kikuo Onozaki2

Leprosy and tuberculosis, caused by in-fection with Mycobacterium leprae and M.tuberculosis, respectively, are chronic infec-tious diseases from which more than 20 mil-lion patients are suffering worldwide 24).Although these two mycobacteria are com-mon in several of their constituents, suchas peptidoglycan and the arabinogalactan-mycolic acid complex 23), the clinical as-pects of the two diseases are complicated.Based on the clinical features as well as im-munologic responses, leprosy patients canbe grouped into four types or categories:lepromatous (L), tuberculoid (T), borderline(B) and indeterminatc (I). Patients with tu-berculoid leprosy manifest a strong cell-me-diated immunity (CM!) response to M. lep-rae but produce a relatively low levei ofantibody. In contrast, in lepromatous pa-tients there is a prominent humoral anti-body response but an anergy in CMI to Al.leprae (I").

Previous studies have suggested that theseimmunological deviations result from theaccumulation of a particular subset of T cells

' Received for publication on 22 June 1993; ac-cepted for publication in revised form on 9 September1993.

= K. Suzuki, M.Sc.; K. Torii, B.S.; H. Hayashi, Ph.D,Assistant Professor; T. Takii, M.Sc., Assistant Profes-sor; K. Onozaki, Ph.D., Professor, Department of Hy-gienic Chemistry, Faculty of Pharmaceutical Sciences,Nagoya City University, Mizuho, Nagoya 467, Japan.Y. Fukutomi, Ph.D., Senior lnvestigator; M. Matsu-oka, D.V.Sc., Room Chief of Laboratory 4, NationalInstitute for Leprosy Rcsearch, Kiyosc, Tokyo 204,Japan. Y. Oomoto, Ph.D., Research Scientist, De-partment oflmmunology, Otsuka Pharmaceutical Co.,Ltd., Tokushima, Tokushima 770, Japan.

Reprint requcsts to Dr. Onozaki.

in the lesions, namely, tuberculoid lesionscontam n a predominance of CD4+ T cells■,vhereas lepromatous lesions contamn main-ly CD8+ T cells ('")• It also has been re-ported that in tuberculoid lesions interleu-kin 2 (IL-2) and interferon-gamma (IFN--y)mRNA, products of the Th 1 -type helper Tcells, were being expressed while in lepro-matous lesions IL-4, IL-5, and IL-10mRNA, products of the Th2-type helper Tcells, were exprcssed (1'). Immunologicaldysfunctions of macrophages also have beenreported. Macrophages of lepromatous pa-tients suppress T-cell responses and pro-duce minor or only low leveis of IL-1 andtumor necrosis factor (TNF) (2. 2'. 10). Sincemacrophages function as antigen-presentingcells to T cells and produce a variety ofimmunoregulatory cytokincs, it is possiblethat the primary interaction between mac-rophages and mycobacteria results in theimmunological deviation.

IL-1, IL-6 and TNF are major cytokinesproduced by macrophages, and these cyto-kines stimulate immunological and inflam-matory reactions (4). In contrast, IL-1 re-ceptor antagonist (IL-ira), which is alsoproduced by macrophages, inhibits IL- 1 ac-tivity as well as IL-1-triggered chain reac-tions by competitively binding to the IL-1receptors (5). IL-10 also inhibits macro-phage functions and influences the subse-quent macrophage/T-cell interaction (7)•Therefore, macrophage function is regulat-ed by both immunostimulative and im-munosuppressive cytokines.

In this paper, we studied the productionof cytokines, IL-1, IL-6, TNF and IL-ira,by human monocytes stimulated with liveor killed Al. lepraeor BCG, and showed that

609

610^ International.Iournal of Lepras .) ,^1993

M. leprae is a very poor inducer of immu-nostimulatory cytokines compared to BCG.However, a substantial amount of IL-1 racan be induced by stimulation with A1. lep-rae even when no other cytokines are in-duced. In addition, the phagocytosis of A1.leprae and cytokine production appeared todepend partially on serum factor(s).

MATERIALS AND METHODSReagent. Human recombinant IL-1a (2

x 10' U/ml) was provided by Dr. M. Ya-mada, Dainippon Pharmaceutical Co., Osa-ka, Japan, and human rccombinant IL-2 byShionogi Co., Osaka. Concentrated buffycoat from healthy donors was supplied byAichi Red Cross Blood Center, Aichi, Ja-pan. RPMI 1640 and polymyxin 13 werepurchased from Sigma Chemical Co., St.Louis, Missouri, U.S.A., and FBS was ob-tained from Bocknek, Toronto, Canada.

Mycobacteria. A1. leprae strain Thai-53were grown in the foot pads of nude mice('). Mouse foot pads were aseptically re-moved, minced with scissors, and homog-enized with 7H12 medium. After centrifug-ing the homogenate for 10 min at 100 x g,the supernatants were obtained and againcentrifuged for 20 min at 3500 x g. Theprecipitates were resuspended with 7H12medium, and the bacillary number was de-termined by the method of Shepard andMcRae ( 27 ). The bacillary number was con-sistent with that counted under microscopywith a hematocytometer. Freeze-diied ^ í.bovis BCG were obtained from Japan BCGCompany, Tokyo, Japan. The BCG weresuspended with phosphate buffered saline(PBS). These mycobacteria were homoge-nized by mild sonication. The bacillarynumber was counted under microscopy witha hematocytometer. Heat-killed mycobac-teria were obtained by treating them at 120°Cfor 15 min.

Supernatants of monocytes stimulatedwith mycobacteria. The buffy coat fromhealthy donors was diluted 1:3 in Hanks'balanced salt solution (HBBS). Mononucle-ar cens (MNC) were separated over Ficoll-Hypaque, washed twice in HBBS, and sus-pended in RPMI 1640 medium supple-mented with 100 U/ml of penicillin G, 100µg/m1 of streptomycin and 15 mM HEPES.The number of monocytes was estimatedby incubating the cell suspension in a he-

matocytometer at 37°C for 3 min in air con-taining 5% CO, and then counting thespreading cens. The spreading ccll numberwas consistent with that of cells adhering tothe tissue culture plate. One ml of a MNCsuspension containing monocytes (1 x 10 6cens/ml) was added to each well of a 24-well plate (Falcon, Lincoln, New Jersey,U.S.A.). After 2 hr of culturing at 37°C inair containing 5% CO,, the cens were washedtwice with HBBS. More than 90% of theadherent cclls were monocytes as deter-mined by morphological criteria withGiemsa staining and the ability to phago-cytose latex beads. To the adherent mono-cytes, 1 ml of RPMI 1640 supplementedwith 1% fetal bovine serum (FBS) or humanserum (HS), untreated or heat inactivatedat 56°C for 30 min, containing mycobacteriawere added, and then the cells were culturedat 37°C. Although the medium was endo-toxin-free according to the Limulus ame-bocyte assay (sensitivity limit of 0.1 ng/ml),we usually added polymyxin 13 (5 lig/mi) tothe culture to inhibit the effect of a smallamount ofendotoxin. After each culture pe-riod, the supernatants were obtained by cen-trifugation. These antibiotics (peniciilin G,streptomycin and polymyxin B) did not af-fect the viability of A1. leprae or BCG for 1week or 24 hr, respectively, as determinedby the metabolism of palmitic acid ( 8 ).

Assay for IL-1 activity. 1L-1 activity wasdetermined by a proliferation assay with anIL-1-dependent mouse T-cell line, Dl ON4M,which was provided by Dr. S. J. Hopkins,University of Manchester, Manchester,U.K.(' 2 ). In brief, cens were cultured inRPMI 1640, 10 mM HEPES, antibiotics, 5x 10 -5 M 2-mercaptoethanol, 10% FBS,concanavalin A (3 ag/mi), IL-2 (40 U/ml)and standard IL-la or test samples. Cells (1x 10 4 ) were cultured in wells of flat-bottommicrotiter plates at 37°C in 5% CO, in air.After 3 days of culture, ccll proliferationactivity was assessed by the 3-[4,5-dimeth-ylthiazol-2-yl]-2,5-diphenyltetrazoli umbromide or thiazolyl blue (MTT) method.After solubilization of the formozan with20% SDS and 50% DMF (dimethyl for-mamide) in water, the absorbance at 595nm was measured on an ELISA autoreader(Bio-Rad). IL-1 activity was expressed asunir equivalent to standard recombinant IL-1a.

61, 4^Suzuki, et al.: Monoeyte Cytokincs atter M. leprae and BCG^

611

Assay for 11,-6 activity. The biologicalactivity of 1L-6 was measureci by its prolif-erative action on the I L-6-depenclent m u-rine hybridoma clone MI-I60.BSF2 (provid-ed by Dr. T. Hirano, Osaka University,Osaka, Japan) (15). Proliferation was mea-sured b■,' the MTT method (2"). One unit ofIL-6 activity was defined as the reciprocaiof the dilution ofsamples that exhibited 50%of maximum response.

Assay for TNF activity. The activity ofTNF was determined by a L929 libroblastcell lytic assay (27). 13riefly, 100 mi of a sus-pension of TNF-sensitive mouse L929 fi-broblast cells (5 x 109 Mis/mi) was culturedwith serially diluted test samples in wells ofa flat-bottom microtiter plate at 37°C for 18hr in air containing 5% CO in the presenceof actinomycin D (1 lig/mi). After culture,the plates were washed, and cell lysis wascletermined by staining the plates with crys-tal violei (0.5%) in metlizinol-water (1:25,v/v). After the dye-stained cells were solu-bilized with 0.1 ml of 0.1% SDS, the dyeuptake was calculated by an ELISA auto-reader. One unit of TNF activity was de-fined as the reciprocai of the dilution ofsamples that lysed 50% of the L929 cens.

Determination of ra. I L- 1ra contentwas determined by an ELISA using mousemonoclonal antibody (IgG) and rabbit poly-clonal antibody (IgG) against human re-combinant IL-ira.

Phagocytosis by monocytes of M. lepraeand BCG. Monocytes (5 x 106 cens) werecultured in RPMI 1640 medium supple-mented with I% untreated or heat-inacti-vated FBS or HS with killed M. leprae orkilled BCG (1.5 x 10') on coverslips in 24-well culture plates. After culture for 24 hr,the coverslips were washed and stained withacid-fast staining. Phagocytosis of the my-cobacteria by monocytes was determinedunder microscopy.

RESULTSCytokine production by monocytes stim-

ulated with M. leprae and BCG. In miei-to determine the difference of cytokine pro-duction by monocytes stimulated with AI.lepra(' and BCG, human monocytes (1 x106cells) were treated with varying numbersof killed AI. leprae, live or killed BCG. Be-cause cytokine induction by a humanmonocytic cell line with lipopolysaccharide

requires at least 1% FBS (17), in this studywe first conducted the experiment in thepresence of 1% heat-inactivated FBS. TNF'ias determined after 6 hr ofculture and IL-1, IL-6, and IL-ira were determined after24 hr of culture. As shown in Figure 1, upto 3 x 106 of AI. leprac induced no detect-able IL-1, IL-6 or TNF. At I x 10 theyinduced only very low leveis of these threecytokines. In contrast, BCG, either live orkilled, induced these cytokines at more than3 x 104 or 105 edis; as shown 109-106 wasthe optimal. Live BCG induced largeramounts of cytokines than killed BCG atlower numbers of bacteria. Thus, BCG ap-peared to be far more potent than M. leprae.In contrast to IL-1, IL-6 and TNF. IL-irawas induced by AI. /eprae-stimulatedmonocytes in a dose-dependent manner.BCG also induced IL- 1 ra, but its produc-tion was inversely related to the number ofbacteria. Although the data are not shown,104 live or killed BCG induced more IL-irathan 105 BCG. A control culture with poly-myxin B alone did not induce detectableleveis of any of the cytokines.

Kinetics of cytokine production by mono-cytes. The differences in cytokine produc-tion might have resulted from differentkinetics. Therefore, monocytes were stim-ulated. with 10' mycobacteria, and kineticsstudies of cytokine production were con-ducted. IL-1 and IL-6 production by mono-cytes stimulated with either BCG or Al. lep-ra(' increased with the duration of theincubation period and peaked at 18-24 hr.TN F production by BCG-stimulatedmonocytes peaked ai 12-18 hr. M. lepraecontinuously induced IL-1ra up to 44 hr,but BCG did so only up to 12.5 hr. There-fore, the different cytokine production bymonocytes stimulated with Al. leprae andBCG did not appear to result from differentkinetics.

FBS and HS and the effect of heat-inac-tivation. The experiments in Figures 1 and2 were conducted in the presence of heat-inactivated FBS. Since it is reported thatphagocytosis of AI. leprac by human mono-eles depends on complement C3 (25), wedetermined the effect of heat-inactivationof the serum on cytokine production bymonocytes stimulated with killed AI. lepracor BCG. A comparison of FBS and HS wasalso made. As demonstrated in The Table,

1993612^ International Journal of Lepro.ti_t'

A2000

000E

4000

10000

8 0 00

6000

2000

o

^ M.Ieprac ( killed )® OCO

^ OCO ( laled )

3x10 13.104^10'^3x10 5^106^3x10'^10' 10 6^3x10 6^10i106^3x10

ii

1 0 5 10 610 ^3x10 ^106^3x10 6^10

Number of mycobacteria / monocytes 10 6 cells

3.10'

Number of mycobacteria / monocytes 10 cells

C

30

E

o

2

o

E

z20

10

L7 M.leprae (killed)

® OCO^ OCO (kled)

FIG. I. Dose response of cytokine production by M. leprae- or BCG-stimulated monocytes. Human mono-cytes (10' cells) were cultured in medium supplemented with 1% hcat-inactivatcd fetal bovine serum with varyingnumbers of killed .t/. leprae, live or killed BCG. After culture for 24 hr, amounts of IL-1, IL-6, and IL-Ira inculture supernatants were dctcrmined as described in Materiais and Methods. Amount of tumor necrosis factor(TNF) was determined aftcr a 6-hr culture. Mean ± of S.D. of triplicate cultures is shown. A = IL-I; B = IL-6; C = TNF; D = IL-Ira.

monocytes in untrcated HS produced thehighest leveis of cytokines in response toeither mycobacteria. Heat-inactivated FBSexhibited a lower effect. Although the dataare not shown, there was no difference be-tween untreated and heat-inactivated FBS.The monocytes in heat-inactivated HS pro-duced the lowest leveis of cytokine. In thereexperiments, however, BCG was again morepotent than M. leprae.

Effect of serum on phagocytosis of my-cobacteria by monocytes. In order to de-termine the effect of serum on the phago-cytosis of there mycobacteria by monocytes,

the monocytes were cultured in mediumsupplemented with untreated or heat-inac-tivated FI3S or HS with killed .%í. leprae orkilled BCG. Monocytes in untreated HSwere most active in the phagocytosis of AI.leprae. Monocytes in other cultures exhib-ited lower leveis of phagocytosis. There wasno difference between heat-inactivatcd HS,untrcated and heat-inactivatcd FBS. Incontrast to AI. leprae, there was no differ-ence in the phagocytosis of BCG betweenuntreated and hcat-inactivated HS. Mono-cytes in FBS either untreated or heat-inac-tivated exhibited lower leveis of phagocy-

4.5 6.5 12 .5 18 24 44

18 2 42^7.5^6.5^12.5

Incubation Time ( h )

61, 4^Suzuki, et al.: Monoc}ate Ci'tokincs n¡tei M. leprae and BCG^613

6000 - B

5000 -

4000 -

3000 -

2000 -

1000 -

o

^ M.leprac (killed)

• BCG

^ OCO (kaled)

4.5^12.5^24

Incubation Time (h )

18

2Eo,9.

c

A4000

^ M.leprae ( killed )

• BCG

^ BCG ( killed )

3000

E2000

1000

1000

800

600E

400

200

FIG. 2. Kinetics of cytokine production by monocytes. Monocytes (10" cells) were cultured in mediumsupplemented with 1% heat-inactivated fetal bovine serum with killed M. leprae, live or killed BCG (10') forperiods indicated. After culture, amounts of IL-1, IL-6, tumor necrosis factor (TNF) and IL-Ira in culturesupernatants were determined. Mcan ± S.D. of triplicate cultures is shown. A = IL-1; B = IL-6; C = TNF; D= IL- Ira.

tosis (data not shown). However, under thesame experimental conditions, BCG wereusually phagocytosed more than M. leprae.

Live and killed M. leprae in induction ofcytokines by monocytes. Untreated HS ap-peared to be the most efficient constituentof the medium for monocytes to producecytokines in response to mycobacteria. Themonocytes then were cultured in the pres-ence of untreated HS, and their ability toproduce cytokines in response to live orkilled M. leprae and live BCG was deter-mined. As shown in Figure 3, significant butlow leveis of IL-1 and IL-6 were inducedby a small number of ALI. leprae. No differ-ence was observed between live or killed M.leprae. In contrast to IL-1 and IL-6, TNF

was not induced by up to 106 Al. leprae.BCG was more potent than M. leprae. IL-1 ra was equally produced by 10 5 to 10' M.leprae. Again, no difference was observedbetween live and killed A. leprae. In con-trast to the results in FBS, BCG induced IL-1 ra in a dose-dependent manner.

DISCUSSIONIn this study, we demonstrated the dif-

ference between M. leprae and BCG in re-gard to their ability to induce cytokine pro-duction from human monocytes. The efI'ectsof live and killed mycobacteria and thoseof heat-inactivation of serum were also de-termined. BCG has been used as a vaccinefor almost 50 years. Macrophages from mice

6 I 4^ International fournal of Leprosy^ 1993

THE TABLE. Effects of sentiu and heat inactivation on IL - 1 and IL - 6 production bynionocyles."

Iluman serum Fetal bovine serumSlim ula^t

Untreated Ileat-inactIvated Ileat-inactivated

IL-1 (U/m1)None 9.4 ± 8.9 ND 4.7 ± 1.5

leprae 67.1 ± 8.9 8.8 ± 3.3 48.3 ± 7.2BCG 1669.8 ± 159.4 479.1 ± 38.2 890.7 ± 80.9

IL-6 (Vim!)None 80.6 ± 67.5 ND NDA/. /cprae 713.5 ±^119.5 29.0 ± 2.0 124.7 ± 22.5BCG 2884.3 ± 136.2 1712.0± 195.5 1097.3 ± 141.5

Monocytes (5 x 10" cells) were cultureel in mettium supplemented with I% untreated or heat-inactivatedhurnan serum or heat-inactivated fetal bovine semi]) with killed A/. hprae or killed BCG (1.5 x 10'). Afterculture for 24 hr, the amounts of IL-1 and 1L-6 in the culture supernatants were determined. Mean ± S.D. oftriplicate cultures is shown. ND = Not detected (< 3 U/ml).

administered BCG became tumoricidal andthosc from a normal mouse produced IL-6in response to BCG in vitro (14,22,.) There-fore, it is expected that BCG is a potentinducer of cytokines from animal and hu-man macrophages. The components of M.tuberculosis muramyl dipeptide (MDP), li-poarabinomannan (LAM), and proteinshave been shown to induce cytokine pro-duction by human monocytes or macro-phages 21' 29)• AI. tuberculosis also is re-ported to induce IL-1 and TNF by humanmonocytes (=v). However, as far as we knowit is not clear what kind of cytokines BCGinduce in human monocytcs and that if theydo, how BCG differs from virulent M. tu-berculosis. In addition, an immunosuppres-sive cytokine IL-ira has not been investi-gated in relation to its production bymacrophages/monocytes stimulated withmycobacteria. In this study BCG, live orkilled, induced production of IL- 1, TNF,IL-6 and IL-ira by human monocytes.Therefore, BCG, similar to M. tubercidosis,appeared to be a potent inducer of cyto-kines. IL-1, IL-6 and TNF were induced ina dose-dependent manner at small bacterialnumbers and decreased at higher numbers.However, IL- Ira production was inverselyrelated to the bacillary number, suggestingthat a very small number ofBCG can induceIL-ira even when other cytokines are notinduced. It is known that immunologicaltolerance can be induced by repeated injec-tion of a very small amount or a very large

amount of antigens (13). It is possible, there-fore, that IL-1ra, an immunosuppressivecytokine, may play a role in the low-doseantigen-induced tolerance. IL-1ra produc-tion is induccd by IL-4, granulocyte mono-cyte-colony stimulating factor, T-cell growthfactor-f1, and IL-lO (5. =8). However, becauselL-1ra dccreased with the increase of theother cytokines, there may be a cytokinecapable of inhibiting IL- ira production.

Studies revealcd that killed tubercu-losis or BCG protected animais from tu-berculosis oniy weakly, but live BCG did sofor a long period ())• The same was true witholher bacteria. Mouse macrophages stim-uiated withwith live Listeria inonocvlo genesproduced much more IL-1 than those stim-ulated with killed bacteria (18). In our study,at low bacillary numbers live BCG was moreeffective than killed BCG in the inductionof IL-6 and TNF, but not of IL-1 or IL-ira.At high cell numbers, however, there wasno difference. Therefore, the differentialprotective effect between live and killed BCGcould not be explained by IL-1 induction.

When compared with BCG, AI. leprae in-duced only small amounts oU IL-1, IL-6 andTNF. A kinetic study indicated that the dif-ference had not resulted from the differentkinetics of cytokine production. It was ofnote, however, that IL-lra was induced ina dose-dependent manner. The differentdose-dependent responses between BCG andM. leprae presumably were due to the dif-ferent efficacies of these mycobacteria. Thus,

I

J. 1 I C//rI Y/G

5000 - B

4000

3000E

• 2000

1000

n 105 3x10' 106^3x10'^10 7

3000

^ M.leprae

• M lepra° (kdled)

O ECO

2000

E

J- 1000

3r104^105^3x105^100^3x106^107

73210 4̂ 105^3x10

^10^3x10^10

6Number of mycobacteria / monocytes 10 cells

200 C

v

•

M.leprae

• M.leprae (k7led)

150 ^ECG

E

100

LLF

5¡1

o

É

61, 4^Suzuki, et al.: Monocyte Cytokines uiter M. leprae and BCG^615

10^

10 0^

10 7

6Number of mycobacteria / monocytes 10 cells

FIG. 3. Comparison of live and killcd .tt. Ieprae in induction ofcytokincs. Monocytes (10' cells) were cultured

in medium supplemented with 1% untreated human serum with live or killcd ;t1. leprae or live BCG (10'). After

culture for 24 hr, amounts of IL-1, IL-6 and IL-ira in culture supernatants were dctermined. Amount of tumor

necrosis factor (TNF) was dctermined after a 6-hr culture. Mean ± S.D. of triplicate cultures is shown. A = IL-I; B = IL-6; C = TNF; D = IL- Ira.

the finding indicates that IL-ira is the mostreadily inducible cytokine by these myco-bacteria.

It is reported that complement C3 bindsto phenolic glycolipid-I in AI. leprae, there-by facilitating phagocytosis of M. leprae bymonocytes through complement receptors( 25). Therefore, we examined the effect ofheat inactivation of HS and FBS on phago-cytosis and cytokine production. Indeed,monocytes in fresh HS phagocytosed M.leprae more than those in heat-inactivatedHS or FBS. In contrast to M. leprae, therewas no difference in the phagocytosis of BCGbetween untreated and heat-inactivated HS,

indicating that complement is not involved.Monocytes in FBS, untreated or heat-in-activated, exhibited the smallest levei ofphagocytosis. It was of note, however, thatunder any experimental condition BCG wasphagocytosed more than M. leprae. In par-allel to the phagocytosis, monocytes in freshHS produced more cytokines than those inheat-inactivated HS. However, when wecompared HS and FBS the cytokine pro-duction was not in parallel because mono-cytes in heat-inactivated HS and FBS pha-gocytosed A1. leprae at the same levei.Therefore, phagocytosis is not the sole fac-tor responsible for cytokine induction. It is

616^ International Journal of Leprosy^ 1993

also interesting that HS is more efficient thanheat-inactivated HS in the case of BCG.Since there was no difference between thephagocytosis by these monocytes, heat-la-bile serum components may augment thecapacity of monocytes in cytokine produc-tion.

When we compared live and killed M.leprae there was no difference. In addition,AL leprae appeared to be a very poor in-ducer of immunostimulatory cytokines I L-1, IL-6 and TNF, while BCG stimulatedmuch cytokine production under any ex-perimental condition. In contrast, IL-ira (animmunosuppressive cytokine) was inducedby both mycobacteria. Since macrophagefunction is determined by a balance be-tween stimulatory and suppressive cyto-kines, M. leprae appeared to confer im-munosuppressive effects rather thanimmunostimulant ones. BCG induces localinflammation and systemic immunityagainst Al. tuberculosis ())• However, it isnot known whether primary infection of M.leprae induces any actue symptoms. There-fore, the potent cytokine-inducing ability ofBCG was implicated in their immunostim-ulatory effect. In contrast, this study sug-gests that A/. leprae escape from host de-fenses by inducing the least levei ofimmunostimulatory or proinflammatorycytokine production but inducing substan-tial amounts of the immunosuppressive cy-tokine IL- 1 ra. Macrophages from patientswith leprosy are reported to be defective intheir ability to present M. leprae antigensto sensitized T cells (9' 21). Since IL-1 andIL-6 play important roles in antigen pre-sentation, not onl■; the production of lowlevels of IL-1 and IL-6 but also the pro-duction ofsubstantial amounts of IL- 1 ra areimplicated in the immunodysfunction. Thelow levei production of TNF further favorsthe infection and multiplication of M. Lep-rac in monocytes because TNF enhancesthe production ofreactive nitrogen oxide bymurine macrophages and inhibits myco-bacterial growth in murine and human mac-rophages (3.1. In this regard it may be in-teresting to investigate cytokine productionby patients' monocytes. The cells may pro-duce more IL-ira and less IL-1, IL-6 andTN F in response to AL leprae than those ofhealthy individuais. It might also be inter-esting to investigate the component of M.

leprae which is responsible for IL-1raThrough genetic engineering the IL-

1 ra-inducing component could be deletedfrom Al. leprae. Such mutated AL lepra('may be able to induce more cytokines, thusbecoming a good vaccine.

SUMNIARYHuman blood monocytes cultured in var-

ious serum conditions were stimulated withycobacterium lepmeor AL bovis BCG and

their cytokine-inducing abilities were com-pared. BCG, either live or killed, inducedproduction of interleukin 1 (I L-1), IL-6, tu-mor necrosis factor (TNF), and IL-1 recep-tor antagonist (IL- Ira). Live BCG at a lowerbacterial number was more potent thankilled BCG in the induction of IL-6 andTNF. In contrast to BCG, killed M. lepraeinduced few cytokines except for IL- Ira.Similar results were obtained when mono-cytes were cultured in the presence of un-treated or heat-inactivated fetal bovine se-rum (FBS). When FBS and human serum(HS) were compared and the effect of heatinactivation was investigated, monocytes inHS produced the most cytokines, then thosein FBS, irrespective of heat inactivation, andthose in heat-inactivated HS produced theleast cytokines. There were no differencesbetween live and killed AL leprae, and BCGwere far more potent than AL leprae in aliof our experimental conditions, indicatingthat the poor cytokine (IL- 1, IL-6 and TNF)-inducing ability ofM. leprae was not due totheir viability. Cytokine production waspartially in parallel with the phagocytosis ofthe mycobacteria. These results suggest thatAI. leprae favor their infection by evokinglittle host reaction through the induction ofonly low leveis of immunostimulatory orproinflammatory cytokines but a substan-tial amount of immunosuppressive cyto-kine.

RESUMENSe investigó la producción de citocinas por los mo-

nocitos de sangre periférica mantenidos bajo diversascondiciones de cultivo en respuesta a la estimulacióncon Alycobacterium lepra(' o con ti. bons. El BCG,vivo o muerto, indujo la producción de interleucina-1(IL-1), IL-6, factor necrosante de tumores (TNF), y clantagonista del receptor para la IL-1 (IL-Ira). El BCGvivo fue más potente que cl BCG muerto en la induc-ción de IL-6 y TNF. En contraste con cl BCG, cl .ti.leprae muerto sói() indujo la producción de algunas

61, 4^Suzuki, et al.: Alonocptc C't'tokines afiei M. leprae and BCG^617

citocinas, entre cilas cl IL-ira. Se obtuvieron los mis-mos resultados cuando los monocitos se cultivaron enpresencia de suero fetal de bovino (SFB) fresco o inac-tivado por calor. Cuando se compararon cl SFB y clsuero humano (SH) y se investig• el erecto de su inac-tivación por calor, se encontró que los monocitos enSH produjeron más citocinas que las producidas porlos MN en SFB, independientemente de si cl SFI3 es-tuvo o no inactivado por calor. Los MN en SH inac-tivado por calor produjeron la menor cantidad de ci-tocinas. No 'tubo diferencia entre el .il. lepra' vivo ycl M. leprae muerto por calor; cl BCG fue ms potenteque el ;tl. leprae bajo todas las condiciones ensayadas,indicando que la pobre capacidad inductora dc cito-cinas (IL-1, IL-6 y TNF) dcl .t/. leprae no estuvo re-lacionada con su viabilidad. La producción de citoci-nas estuvo parcialmente en paralelo con la fagocitosisde micobacicrias. Estos resultados sugieren que At lep-rae favorece su implantación por evocar una mínimareacción dei huésped, induciendo en él la producciónde bajos niveles de citocinas inmunoestimulatorias oproeinllamatorias y una cantidad substancial dc cito-cinas inmunosupresoras.

RESUMEDes monocytcs dc sang humain cultivés dans du

serum sous diverses conditions ont été stimuiés par duA1t'cobacterium leprae ou du BCG de Al. bons et leurcapacité à induire des cytokines a été comparéc. LcBCG, vivant ou tué, provoquait ia production d'in-terlcukine 1 (IL-l), IL-6, de facteur nécrosant des tu-mcurs (FNT) cd d'antagonistc du rcccptcur d'IL-1 (IL-lra). Le BCG vivant en quantité bactérienne moindreétait plus puissant que le 13C0 tué dans l'inductiond'IL-6 et FNT. En contraste avec le BCG, le ;t/. lepraetué induisait peu dc cytokines, excepté I'IL- ira. Dcsrésultats semblablcs ont été obtenus par des monocytcscultivés en présence dc scrum bovin foctal (SBF) nontraité ou inactiveé par la chalcur. Quand le SBF et lescrum humain (SH) ont été comparés et que I'effet del'inactivation par la chalcur a été examiné, les mono-cytcs cultivés dans lc scrum humain produisaicnt leplus de cytokines, suivis par ceux cultivés dans le SBF,indépendamment de l'inactivation par la chalcur, etccux cultivés dans lc SH inactivé parla chalcur le moinsde cytokines. II n'y avait pas de différence entre le AI.leprae vivant et tué, et le BCG était beaucoup pluspuissant que .tl. leprae dans toutes nos conditions ex-périmentales, indiquant que la faible capacité de M.Ileprae à induire des cytokines (iL-1, IL-6, et FNT)n'était pas dele à sa viabilité. La production de cytokineétait particllement en paralèlle avec la phagocytose desmycobactéries. Ces résultats suggèrent que .tl. lepraefavoris son infection en suscitant peu dc réaction de1'hôte par l'induction de faibles taux de cytokines sti-mulant l'immunité ou l'inflammation, mais un tauxsubstantial dc cytokinc immunosuppressivc.

Acknowledgment. We thank Dr. S. Hida for the iL-lra ELISA. This work was supported in part by a grant

from the U.S.-Japan Cooperative Medical Scicnce Pro-gram. The authors also wish to gratcfully acknowlcdgcthe Sasakawa Memorial Health Foundation for theirfinancial support of this study.

REFERENCES1. BARNES, P. F., CIIATTERJEE, D., ABRAMS, J. S., Lu,

S., WANG, E., YAMAMURA, M., BRENNAN, P. J. andMODLIN, R. L. Cytokinc production induced byMrcobacterium tubercalosis lipoarabinomannan.J. Immunol. 149 (1992) 541-547.

2. BARNES, P. F., CHATTERJEE, D., BRENNAN, P. J.,REA, T. H. and MoDLIN, R. L Tumor necrosisfactor production in patients with lcprosy. Infcct.lmmun. 60 (1992) 1441-1446.

3. I3ERMUDEZ, L. E. M. and YOUNG, L. S. Tumornecrosis factor, alonc or in combination with IL-2, but not IFN-gamma, is associated with mac-rophage killing of.11t'cobacteriam ariuni com plex.J. Immunol. 140 (1988) 3006-3013.

4. DINARELLO, C. A. Interlcukin-1 and its biologi-cally related cytokines. Adv. Immunol. 44 (1989)153-205.

5. DINARELLO, C. A. and THOMPSON, R. C. BlockingIL-1: interleukin 1 receptor antagonist in vivo andir, nitro. Immunol. Today 12 (1991) 404-410.

6. DING, A. H., NATHAN, C. F. and STUEHR, D. L.Release of reactive nitrogen intermediates and re-active oxygen intermediatcs from mouse perito-neal macrophagcs. Comparison of activating cy-tokines and evidcnce for indcpendent production.J. Immunol. 141 (1988) 2407-2412.

7. FIORENTINO, D. F., ZLOTNIK, A., VIEIRA, O.,MOSMANN, T. R., HOWARD, M., MOORE, K. W.and O'GARRA, A. IL-10 acts on the antigen-pre-senting ccll to inhibit cytokinc production by Th ledis. J. Immunol. 146 (1991) 3444-3451.

8. FRANZBLAU, S. G., WIIITE, K. E. and O'SULLIVAN,J. F. Structure-activity relationships oftetrameth-ylpiperidine-substitutcd phenazincs against .ill'-cobacterianr leprae in nitro. Antimicrob. AgentsChcmother. 33 (1989) 2004-2005.

9. FREUND, J. and MIDDLEBROOK, G. Tuberculosis.In: 13acteriologt' and Mrcolic Infection. Dubos, R.J., ed. Philadclphia: Lippincott Co., 1948, pp. 303-320.

10. HASTINGS, R. C., GILLIS, T. P., KRAHENBUHL, J.L. and FRANZBLAU, S. G. Leprosy. Clin. Micro-biol. Rev. 1 (1988) 330-348.

1 L HIRSCHBERG, H. The role of macrophagcs in thelymphoproliferative response to .1l. leprae in litro.Clin. Expl. Immunol 34 (1978) 46-51.

12. HOPKINS, S. J. and HUMEHREYS, M. Simplc, sen-sitivo and spccific bioassay for interleukin-1. J.Immunol. Methods 120 (1989) 271-276.

13. HowARD, T. G. and MITCHISON, N. A. immu-nological tolcrancc. Prog. Allcrgy 18 (1975) 43-96.

14. HUYGEN, K., VANDENBUSSCHE, P. and HEREMANS,H. Interlcukin-6 production in .tltrobacterium

618^ International .lournal of Leprosy^ 1993

bovis BCG-injected mice. Cell. Immunol. 137(1991) 224-231.

15. KAWANO, M., IIIRANO, T., MATsuoA, T., TAGA,T., HORI, Y., IWATO, K., ASAOKU, H., TANG, 13.,TANABE, O., TANABE, H., KURAMOTO, A. andKISHIMOTO, T. Autocrine generation and requi re-ment of 13SF-2/1L-6 for human rnultiple myelo-mas. Nature 332 (1988) 83-85.

16. MATSUOKA, M., KOHSAKA, K. and DAYANGIIIRANG,A. J. Characterization of leprae Thai-53 strain.(Abstract) Int. J. Lepr. 60 (1992) 722.

17. MATSUSHINIA, K., COPELAND, T. D., ONOZAKI, K.and OPPENHEINI, J. J. Purification and biochern-ical characteristics of two distinct human inter-leukins 1 from the myelomonocytic THP-1 cens.I3iochemistry 3 (1986) 3424-4329.

18. MITSUYAMA, M., IGARASHI, K., KAWAMURA, I.,OHMORI, I. and Nomo-ro, K. Difference in theinduction of macrophage interleukin-1 productionbetween viable and killed cells of Listeria (nono-eytogenes. Infect. Immun. 58 (1990) 1254-1260.

19. MODLIN, R. L., MELANCON-KAPLAN, J., YOUNG,S. M. M., PIRMEZ, C., CONVIT, J., REA, T. 11. andlkoom, Il. R. Learning from lesions: patterns oftissue intlammation in leprosy. Proc. Nati. Acad.Sci. U.S.A. 85 (1988) 1213-1217.

20. MOSSMANN, T. Rapid colorimetric assay for cel-lular growth and survival: application to prolif-eration and cytotoxicity. J. lmmunol. Methods 65(1983) 55-63.

21. NATII, I., VAN ROOD, R. R., MEHRA, N. K. andVAIDYA, M. C. Natural suppressor cells in humanleprosy: the role of IILA-D-identical peripherallymphocytes and macrophages in the in litro mod-ulation oflymphoproliferative responses. Clin. Exp.Immunol. 42 (1980) 203-210.

"". Ruco, L. P. and MELTZER, M. S. Defective tu-moricidal capacity ofmacrophages from C3H/HeJmice. J. Immunol. 120 (1978) 329-334.

23. SANCEAU, J., FALCOFF, R., BERANGER, F, CARTER,D.13. and WIETZERBIN, J. Secretion of interleukin-6

(1L-6) by human monocytes stimulated by mur-amyl dipeptide and tumor necrosis factor alpha.Immunology 69 (1990) 52-56.

24. SANSARRICQ, H. Leprosy in the world today. Lepr.Rev. 52 (1981) 15-31.

25. SCHLESINGER, L. S. and Hokwirz, M. A. Phenolicglycolipid-1 of .11yeabacterilan lepraebinds com-plemcnt C3 in senil]) and mediates phagocytosisby human monocytes. J. Exp. Med. 174 (1991)1031-1038.

26. SHEPARD, C. C. and McRAE, D. II. A method forcounting acid-fast bacteria. Int. J. Lepr. 36 (1967)78-82.

27. TAMATANI, T., KINIURA, S., HASHIMOTO, T. andONOZAKI, K. Purilication of guinea pig tumor ne-crosis factor (TN F): comparison of its antiprolifer-ative and differentiative activities for myeloid leu-kemic cell line with those of recombinant humanTNF. J. 13iochem. 105 (1989) 55-60.

28. VANNIER, E., MILLER, L. C. and DINARELLO, C. A.Coordinated antiinflammatory effects of interleu-kin-4: interleukin 4 suppresses interleukin 1 pro-duction but up-regulates gene expression and syn-thesis of interleukin 1 receptor antagonist. Proc.Nau. Acad. Sei. U.S.A. 89 (1992) 4076-4080.

29. WALLIS, R. S., Amik, T. M. and ELLNER, J. J. In-duction of interleukin 1 and tumor necrosis factorby mycobacterial proteins. Proc. Nati. Acad. Sei.U.S.A. 87 (1990) 3348-3352.

30. WATSON, S., BULLOCK W., NELSON, K., SCHAUF,V., GELBER, R. and JACOBSON, R. Interleukin 1production by peripheral blood mononuclear cellsfrom leprosy patients. lnfect. lmmun. 45 (1984)787-789.

31. YAMAMURA, M., UYEMURA, K., DEANS, R. J.,WEINBERG, K., REA, T. H., 13Loom, 13. R. andMODLIN, R. L. Defining protective responses topathogens: cytokine profiles in leprosy lesions. Sci-ence 254 (1991) 277-279.

Page 1Page 2Page 3Page 4Page 5Page 6Page 7Page 8Page 9Page 10