American journal ol Pathology, Vol. 147, No. 4, October 1995Copyright ) American Societ for Inavestigative Pathology

Transforming Growth Factor-f3 in HumanCutaneous Leishmaniasis

Aldina Barral,* Mariana Teixeira,*Patricia Reis,* Vera Vinhas,* Jackson Costa,tHelio Lessa,* Achilea L.Bittencourt,*Steve Reed,* Edgar M. Carvalho,* andManoel Barral-Netto*§From the Faculdade de Medicina, Servi9o de Imunologiaand Servi7o de Anatomia Patologica,* da UniversidadeFederal da Bahia, Salvador-Bahia, Brazil; UniversidadeFederal do Maranhao,t Maranhao, Brazil; SeattleBiomedical Researcb Institute,* Seattle, Washington; andLIMI-Centro de Pesquisas Gon7alo Moniz,t FIOCRUZ,Salvador-Bahia, Brazil

Transforming growthfactor (TGF)-,8 has severaldownregulatoryfunctions on the immune systenminhibition ofinterleukin-2 receptor induction, de-crease of interferon-yinduced class II antigenexpression, inhibition of macrophage activa-tion, as weUl as cytotoxic and lymphokine-activated kilter cell generation. TGF-(8 has alsobeen recognized as an important immunoregula-tor in murine leishmaniasis, for which it in-creases susceptibility to disease. In the presentstudy we evaluate the involvement of TGF-j8 inhuman leishmaniasis in vitro and in patientswith cutaneous leishmaniasis. Human macro-phages produce active TGF-,3 after infection byLeishmania amazonensis (480 ± 44.7 pg/mlmean ± SEM), L. donovani chagasi (295 + 7.6pg/ml), or L. braziliensis (196 ± 15.7 pg/ml).When TGF-g3 was added to cultures of humanmacrophages infected with L. braziliensis it led toan increase of approximately 50% in parasitenumbers as compared with untreated cultures.Exogenous TGF- 1 added to macrophage cultureswas able to reverse the effect of interferon-y incontroUing Leishmaniagrowth. Even at 100 IU/mlinterferon-y the presence ofTGF- 8 increases thenumber of intracellular parasites. On the otherhand, TNF-a at high concentration (100 IU/ml)totaly blunts the suppressive effect of TGF-4Immunostaining for TGF-13 was observed in thedermis, produced by fibroblasts and occasion-

ally by inflammatory cells in the biopsies fromhuman leishmaniasis lesions, being present inmost of the biopsies taken from patients withearly cutaneous leishmaniasis (less than 2months of ulcer development) and in cases ofactive mucosal leishmaniasis. Taken togetherthese observations suggest an important roleforTGF-18 in human leishmaniasis, with its produc-tion by infected macrophages being probably re-lated to parasite establishment in the earlystages of the disease. (Am J Pathol 1995,147:947-954)

Leishmaniasis is an important public health prob-lem in several areas of the world, including theAmericas. Leishmania are generally divided intoparasites that cause tegumentary disease (L. bra-ziliensis, L. amazonensis, and L. mexicana in theNew World and L. tropica and L. aethiopica in theOld World) and those agents of visceral disease(L. donovani and L. donovani chagasi). After inocu-lation in the skin, the flagellated promastigote pen-etrates into the macrophage, transforms intoamastigotes, and multiplies inside the parasito-phorous vacuoles. The cutaneous lesions begin atthe site of parasite entrance as a small papule thatdevelops into a nodule that ulcerates in the center.Lesions can take on aspects of papules, nodules,ulcers, tubercles, or infiltrated plaques.1 Clinicalmanifestations of tegumentary leishmaniasis in-clude the self-healing cutaneous disease (CL) andthe destructive and hyper-responsive mucosalform (MCL). More rarely, diffuse cutaneous leish-maniasis (DCL) is observed. The most frequentaspect observed in CL cases is an ulcer with

Supported by grant A130639 from the National Institutes of Health.MT and PR were recipients of Scientific Initiation Fellowships, andALB and MBN are Senior Investigators of the Brazilian NationalResearch Council (CNPq).Accepted for publication July 6, 1995.

Address reprint requests to Dr. Aldina Barral, ServiQo de Imuno-logia-Hosp, Univ. Prof. Edgard Santos, Rua Joao das Botas, s/n,40.110-040 Salvador-Bahia, Brazil.

947

948 Barral et alAJP October 1995, Vol. 147, No. 4

elevated borders and sharp crater. CL patientsexhibit anti-leishmania cell-mediated immunity(CMI). Mucosal involvement, probably caused byhematogenous spread from the primary lesion, oc-curs in approximately 3% of patients infected by L.braziliensis and is related to multiple or extensiveskin lesions above the belt and inadequate ther-apy of the primary lesion.2 Initial involvement of thenasal mucosa can spread to the hard and softpalate, uvula, pharynx, gums, and upper lip. MCLis considered the hyper-responsive pole of thedisease because of the potent anti-leishmania CMIresponses observed in these patients. The initiallesion of DCL resembles those of CL without ulcer-ation. The lesions are erythematous and appear aspapules, nodules, tuberous lesions, infiltratedplaques, and diffuse infiltration of a part of thebody. DCL, similar to lepromatous leprosy, occursin the absence of anti-parasite CMI.3'4

Details of the host or parasite properties respon-sible for the variations in healing and severity ofdisease are not known. CMI responses are consid-ered elemental for the resistance to and recoveryfrom leishmanial infection. Leishmanicidal activity isprobably a result of the increased capacity of themacrophages to produce toxic oxygen5 and nitro-gen radicals6 in response to interferon (IFN)-y.7Upon its entry, Leishmania induces the macrophageproduction of tumor necrosis factor (TNF)-a,8 whichpotentiates the action of IFN-y and promotes macro-phage activation,9'10 and transforming growth factor(TGF)-f3,11-13 linked to macrophage deactivationand inhibition of IFN-y action.11'14'15 Initial survival ofLeishmania inside the macrophage may critically de-pend on which of these opposing cytokines predom-inate in the microenvironment of infection.

TGF-4 is a molecule with proinflammatory andimmunoregulatory activities, most notably IFN-y-induced class 11 antigen expression16 and macro-phage activation,17 " as well as cytotoxic andlymphokine-activated killer cell generation. 19.20 Theimportant role of TGF-,B in murine leishmanial infec-tion has been recently reviewed.21 In vitro infection ofmouse peritoneal macrophages with L. amazonen-sis,12 L. braziliensis,11 or L. major13 leads to TGF-f3production. There is also an increase in local TGF-13production in mouse footpads after Leishmania infec-tion, as shown by immunohistochemistry.13 Addition-ally, exogenous TGF-3 leads to increased parasiteload both in vitro1 1-13 and in vivo 11,12

All of these observations prompted us to investi-gate the role of TGF-,B in human leishmaniasis, bothin in vitro macrophage cultures and in lesions of

patients with different clinical presentations (muco-sal, cutaneous, or diffuse forms) of CL.

Material and MethodsPatients

We have studied lesions from 12 patients with CL forless than 2 months (early CL); 5 patients with CL withulcerations for more than 2 months (late CL); and 8patients with MCL, with a single ulcerated lesion inthe nasal or oral mucosa. Mucosal lesions were clas-sified as active when represented by a lesion witherythema, edema, ulceration, or hyperemic tissueexhibiting ulcerated borders. Microscopically activelesions were characterized by intense inflammatoryinfiltrate with abundant macrophages and lympho-cytes and occasionally by the presence of theparasite. MCL without activity was characterizedmacroscopically by pale, shallow, non-ulceratedcicatricial lesions and histologically by a light chronicinflammatory infiltrate and fibrosis. We have alsoevaluated 12 biopsies (before and after treatment) of6 patients with DCL. All patients with CL or MCL arefrom the endemic region of Corte de Pedra, BahiaState (Northeast Brazil). Patients with DCL are fromMaranhao State (North Brazil).A detailed evaluation including a complete phys-

ical examination was performed on each patient. Theclinical diagnosis was confirmed by at least onelaboratory test (serology, skin test, or parasite cul-ture). Anti-Leishmania antibodies were detected by astandard alkaline-phosphatase enzyme-linked im-munosorbent assay procedure as described else-where.22 The reaction was considered positive whenthe O.D.402 nm was greater than 0.05 (at a 1/100serum dilution). Montenegro's skin reaction was per-formed with 20 ,gg of protein of crude leishmanialantigen as previously described.23 After 48 hours,the largest diameter of the indurated area was mea-sured and considered positive if larger than 5 mm.All patients received conventional antimonial (Glu-cantime) treatment consisting of 20 mg/kg/day SbVfor 20 days. Patients with MCL and DCL had re-ceived several courses of treatment at the time ofstudy.

Reagents

TGF-,3 and anti-TGF-f3 polyclonal antibody, pro-duced in rabbit, were gifts from Dr. D. Twardzik(Bristol-Meyers Squibb Pharmaceutical Research In-stitute, Seattle, WA). IFN-y was provided by Dr.Maud Brandely (Institute Roussell Uclaf, Paris,

TGF-,B in Human Leishmaniasis 949AJP October 1995, Vol. 147, No. 4

France) and TNF-a was provided by Dr. K. Grabstein(Immunex Corp., Seattle, WA). All cytokines werehuman recombinant products. Anti-TGF-,3 monoclo-nal antibody (MAb; 1 D11.16) was kindly provided byDr. L. Ellingsworth (Celtrix Corp., Palo Alto, CA).

ParasitesL. braziliensis (MHOM/BR/88/Ba92), L. chagasi (MHOM/BR/90/Ba307), and L. amazonensis (MHOM/BR/89/Ba276) were used in this study. All of them have beentyped by serodeme with a panel of MAbs and byisoenzymes (courtesy of Drs. Gabriel Grimaldi and H.Momem, Funda,ao Oswaldo Cruz, Rio de Janeiro,Brazil). Serodeme analysis was performed by usingpanels of anti-Leishmania MAbs specific for membersof the L. braziliensis, L. amazonensis, or L. donovani cha-gasi complexes already described24-26 in an indirectradioimmune binding assay.24

Macrophage Cultures

Macrophage cultures were performed with periph-eral blood mononuclear cells from normal volun-teers. Cultures were performed in round coverslipsplaced inside 24-well plates. Briefly, 1 x 106 periph-eral blood mononuclear cells in RPMI 1640 withoutserum were plated overnight in a volume of 0.5 ml/well and cultured at 37°C in a humid atmospherewith 5% CO2. Nonadherent cells were removed bywashing with warm phosphate-buffered saline (PBS)and medium replaced for RPMI 1640 with 10%normal human AB serum (RPM110H). Culture wasmaintained for 5 to 7 days and then infected withstationary-phase promastigotes at 106 parasites/well, during 3 hours at 35°C. Extracellular parasiteswere removed by extensive washings with warm me-dium, followed by additional incubation for up to 72hours in RPM110H, at 370C in a humid atmospherewith 5% CO2. Monolayers (in duplicate) werewashed with PBS, fixed in acetone-methanol, andstained with Giemsa. The percentage of infectedmacrophages and the number of amastigotes per100 macrophages were determined by microscopicexamination.

TGF-,B Production and Assay

Cell-free supernatants from macrophage cultureswere obtained 72 hours after infection and from un-

infected cultures kept for a similar duration under thesame conditions. Supernatants (with and withoutprevious acid treatment) were assayed for TGF-f3 bya CCL-64 mink lung epithelial cell bioassay as de-

scribed.27 Cells were plated at 105 cells/well into96-well microtiter plates for 1 hour at 370C in a vol-ume of 100 jpl of Dulbecco's minimal essential me-dium containing 1% heat-inactivated fetal calf se-rum. After 1 hour, TGF-3 standards ranging from 2ng to 0.49 pg or test supernatants were added atappropriate dilutions to a final volume of 200 ,tl/well.The concentration of TGF-,B in each sample wasdetermined by comparison with a curve generatedfrom TGF-1 standards. The specificity of the assayswas confirmed by inhibition with 1 D 1.16 anti-TGF-,BMAb.

Immunohistochemistry for TGF-,BWe used the technique as previously described, withslight modifications.12 Tissues obtained from biop-sies from lesions or normal skin were used afterfixation in buffered formaldehyde and embedded inparaffin. Sections were treated with hyalunoridase (1mg/ml; Sigma Chemical Co., St. Louis, MO) for 10minutes at room temperature, blocked with 1.5%normal goat serum in PBS containing 5% bovineserum albumin for 1 hour, and incubated with anti-TGF-f rabbit MAb (15 ,ug/ml) overnight at 40C. Anincubation with biotinylated goat anti-rabbit IgG (Sig-ma Chemical Co.) at room temperature for 30 min-utes was followed by avidin-alkaline phosphatase(Biogenex, San Ramon, CA; 1:10 for 20 minutes).The substrate Fast Red (1 mg/ml; Sigma ChemicalCo.) in Tris-HCI buffer (pH 8.2) with 0.02% naphtholAS-MX phosphate (Sigma Chemical Co.) was ap-plied for 20 minutes, and the slides were counter-stained with Mayer's hematoxylin. Controls for theimmunohistochemistry included the use of no pri-mary antisera or the use of normal rabbit IgG in placeof the primary antibody and the use of normal humanskin. All three controls always gave negative results.

Statistical AnalysisComparison of the effect of different cytokines onleishmanial macrophage infection was done by one-way analysis of variance followed by Scheffe's test.

Results

TGF-,f Production in VitroActive TGF-p was measured in 72-hour supernatantsof human monocyte-derived macrophages, infectedwith different Leishmania strains. Figure 1 shows thatL. amazonensis induced the highest production (480+ 44.7 pg/mI, mean + SEM), L. donovani chagasi

950 Barral et alAJP October 1995, Vol. 147, No. 4

5001/

4001

E

300

200

100

/ I i

CL

_ __ __ __ __ __ __ __ . - __ i, _,_, _,

None

250

0

L-

_@

a

0

CRLd Lb La

Figure 1. TGF-f3 in vitro production by human monocyte-derived mac-rophages upon infection with different leishmania strains. Humanmacrophages were cultured at a concentration of 1 X 106 peripheralblood mononuclear cells/well on 24-well plates for 7 days before in-fection. Infection was performed with 1 X 106 stationary-phase pro-mastigotes per well for 3 hours at 35°C. Supernatants (in triplicate)were harvested 72 hours after infection and TGF-,( measured in a

bioassay with CCL-64 mink lung epithelial cells. Ld, L. donovani; Lb, L.braziliensis; La, L. amazonensis.

induced 295 ± 7.6 pg/ml, and the lowest productionwas induced by L. braziliensis (196 ± 15.7 pg/ml).The experiment was repeated twice with similar re-

sults.

Effect of TGF-,B on Parasite Growth insideMacrophages in Vitro

TGF-, added to cultures of human macrophagesinfected with L. braziliensis led to an increase of ap-

proximately 50% in parasite numbers as comparedwith untreated cultures (Figure 2). By contrast, theaddition of either TNF-a or IFN-,y resulted in a reduc-tion in parasitism (Figure 2). Differences among thethree treatments were statistically significant whentested by analysis of variance (F = 22.26; P <

0.0001). Differences between TGF-,B and IFN-,y(Scheff6, 18.89; P < 0.01) and TGF-,B and TNF-a(Scheffe, 15.20; P < 0.01) were statistically signifi-cant, whereas the difference between IFN-y andTNF-a was not. The experiment was performed on

different occasions with cells from 10 different nor-

mal volunteers as the source of macrophages.

TGF-, Blocks the Effects of IFN-y or TNF-aon Leishmania KillingTo evaluate the capacity of TGF-13 in blocking theeffects of IFN-,y or TNF-a macrophages, parallel cul-tures were set up with or without TGF-)S at a concen-

tration of 10 ng/ml. Both sets of cultures were alsoexposed to different doses of IFN-,y (6 doses in therange of 0 to 100 lU/mI) or TNF-a (4 doses rangingfrom 0 to 100 lU/ml). The suppressive effect of TGF-p3is observed across the whole dose-response curve

200 ;

150

100 ---

50 F

0

TGFB IFNy TNFaFigure 2. Effect of TGF-f3, IFN--y, or TNF-a on in vitro parasite multi-plication inside human macrophages. Human monocyte-derived mac-rophages were cultured and infected with L. braziliensis stationary-phase promastigotes as indicated in Figure 1. Immediately afterinfection cultures were treated with TGF-P (10 ng/ml), IFN--y (50IU/ml), or TNF-a ( 100 IU/mt). Number ofamastigotes per 100 macro-phages were determined 48 hours after infection. Results are shown as

percentage of the number ofparasites counted in cultures left withouttreatment. Each point represent a different volunteer usedfor obtain-ing macrophages. In a typical experiment the number of amastigotesper 100 macrophages was as follows: untreated cultures, 296 + 54,and treatments with TGF-,f, 426 161; TNF-a, 197 + 46; and IFN y,143 + 70.

with IFN-,y. Even at 100 lU/ml of IFN-y, the presenceof TGF-j3 increases the number of intracellular para-sites, although at lower levels than in cultures withoutor with small amounts of IFN-y. TNF-a at high con-

centration (100 lU/ml) totally blunts the suppressiveeffect of TGF-,3.

Immunohistochemistry for TGF-f3 in Lesions

The presence of TGF-13 was observed diffusely in thedermis and occasionally in epidermis appendages.We focused our attention, however, on TGF-f3 ob-served in association with the inflammatory infiltrate.Immunostaining for TGF-f3 was observed in the der-mis, produced by fibroblasts and occasionally byinflammatory cells in the biopsies from human leish-maniasis lesions (Figure 4A). The presence of TGF-,Bwas observed in the majority (approximately two-thirds) of lesions obtained from patients with diseasefor less than 2 months (Table 1). The rate of positivitydropped in the lesions from cutaneous patients withlonger disease periods and was present in approx-

imately one-half of mucocutaneous lesions (Table 1).The mucocutaneous lesions with detectable TGF-,Bwere obtained from cases with active disease.TGF-,3 was observed in three cases of DCL even

after treatment (Table 2). Although DCL is character-

00

8CC)co0 08

-0

0

TGF-,B in Human Leishmaniasis 951AJP October 1995, Vol. 147, No. 4

150

100

50

0 1 10 20 50 100IFN-y (IU/ml)

0 10 50

TNF-a (IU/ml)

Table 1. Immunohistochemistry to TGF-f3 in Lesionsfrom Human Cuitaneous and MucosalLeishmaniasis

Disease

Form

Early cutaneous

Late cutaneous

Mucosal

100

Figure 3. Ejfect o/ TGF- on in vitro intra macrophage Leishmaniakillin,g induced by, dyfifrent doseJs oJ IFN- y (upper panel) or TNF-a(lower panel). Hliman mnonocyte-derived niacrophages u'ere ciultuiredand inifected nith L. braziliensis stationary-phase promastigotes as

indicated in Fignre 1. Inwnic-diatelyajterinfictioni cudtures were treatedwith TGF- (10 nig/lnt), IFV-y ( 50 1(1mln), 7NF-a (100 IY1mt). or a

mi7ixtture of TGF-, ( 10 ngl/m) pluis the indicated amnounts of eitherIFN- y or TNF-a. Number of amnastigotes per 100 macrophages were

cleterniined 48 hbor', alter infection.

ized by intense parasite growth, TGF-p was detectedwith high intensity in only one of six cases examined(Table 2). Immunohistochemistry for TGF-f3 in normalskin was negative.

Anti-TGF-f reaction

Duration Intensity Positivity

15 d15 d15 d15 d

1 m1 m1 m1 m2 m2 m2 m2 m4 m5 m5 m8 m3y6 m8 m1 y2 y3y12 y15 y18 y

+

++

Neg

Neg

Neg

++

±+Neg

Neg

Neg

NegNeg

Neg

+

Neg

8/12

2 /5

4 /8

Intensity of staining was graded on a scale of 0 to ++++. d,days; m, months; y, years; Neg, negative.

Discussion

The production of active TGF-p by monocyte-derived human macrophages after infection byLeishmania extends to the human system previousobservations made with mice.'1113 Latent TGF-,3may be produced by uninfected macrophages, butthe presence of the active molecule may bear impor-tance for parasite survival and infectivity. Leishma-

Figure 4. Immunohistochemical demonstration ofTGF- /3in human cutaneous leishmaniasis lesion, obtainedfrom a CL case uith 15 days ofdurationanid considered +++. Left panel: Positive staini (red) for TGF-,/ in the dermis of a tissue section from a leishmanial skin lesion treated uvithanti-TGF- polYclonal antibody. Right panel: Negative control, shoning absence of stain in a section treated with normal rabbit IgG.

200

aa5Sa

SS

og

E

2

0

0

E

952 Barral et alAJP October 1995, Vol. 147, No. 4

Table 2. Immunobistochemistry to TGF- in Pre- andPost-Treatment Lesions from Human DiffuseCutaneous Leisbmaniasis

Patient

1

2

3

4

5

6

Clinical status

PretreatmentPost-treatmentPretreatmentPost-treatmentPretreatmentPost-treatmentPretreatmentPost-treatmentPretreatmentPost-treatmentPretreatmentPost-treatment

Intensity

NegNeg

NegNegNegNeg

+

+

Intensity of staining was graded on a scale of 0 to ++++.Neg, negative.

nia-induced production of TGF-,B by murine macro-

phages has been observed with L. amazonensis12 or

L. braziliensis.11 In the case of L. braziliensis a higherproduction of active TGF-,B was obtained with a more

virulent isolate as compared with a less virulentone.11 It is premature to draw firm conclusions relat-ing quantifiable differences in TGF-13 production andLeishmania virulence, as it is very difficult to comparethe severity of disease caused by different isolates inman.

When evaluated by immunohistochemistry in hu-man tissues from leishmanial lesions, the presence

of TGF-,B was observed in 8 of 12 early (2 months or

less) human cutaneous lesions and in active muco-

sal leishmaniasis. These data are consistent with a

role for TGF-,B in the initiation of Leishmania infection.It is of importance that TGF-,B was still detectable

in several cases of DCL even after therapy. Shortlyafter treatment, DCL patients exhibit transient peri-ods of marked clinical improvement, followed byrelapses; post-treatment biopsies were taken duringthis period. TGF-f3 has been shown in the epidermisof human viral infection26 and may indicate that localdepression of immune mechanisms are important inallowing the pathogen to grow, as observed forLeishmania in the present study. In human neoplasticdiseases, the presence of TGF-f3 in lesions has beencorrelated with the advancing edge of primary tu-mors29 or depth of tumor invasion,30 also indicating

the role of TGF-3 in lesion progression. A recentstudy demonstrated a higher expression of mRNAfor TGF-,3 in human CL lesions with more than 4months of disease than in earlier lesions.31 It is pos-

sible that the continued production of TGF-,B in somecases is implicated in the persistence of the Leish-mania and in disease progression.

The presence of Leishmania inside the macro-phage interferes with antigen presentation32 andmay influence the course of the disease. As TGF-f3allows greater Leishmania proliferation inside themacrophage, it is possible that this is one of themechanisms that explain the disease-enhancing ef-fect of TGF-f3 in leishmaniasis.

In several immunologically mediated skin dis-eases, major histocompatibility complex (MHC)class 11-positive keratinocytes are observed associ-ated with IFN-y production by the surrounding Tcells. Besides antigen presentation, those cellsseem to be able to produce cytokines and can main-tain clonal expansion of activated lymphocytes.MHC class 11-positive keratinocytes have been ob-served in CL but not in DCL lesions, suggesting alower IFN-y production in DCL lesions.33 TGF-,B in-hibits IFN-y-induced MHC class 11 expression16 andmay be implicated in decreased keratinocyte induc-tion of CMI responses in leishmanial lesions.

Another possible action of TGF-,B in leishmaniasisis through its inhibitory effect over cytotoxiccells.19,20 CD8+ T cells are present in both CL andMCL human lesions but not in delayed-type hyper-sensitivity reactions elicited by Leishmania antigen,suggesting a role for these cells in the destruction ofinfected cells.34 We have shown the presence ofcytotoxic cells in the peripheral blood of CL and MCLpatients and its down-modulation by TGF-f or inter-leukin-10.35 We recently developed an assay withautologous infected macrophages both as antigen-presenting and target cells. In this system, MCLpatients exhibited cytotoxic activity mediated byCD8+ and/or NK cells, and this response is down-modulated by TGF-,3 (manuscript in preparation).

After Leishmania penetration, the human macro-phage produces not only TGF-,B, as shown in thisstudy, but also other cytokines such as TNF-a,jwhich promotes macrophage activation for Leishma-nia killing.910 Other products, such as interleukin-12,that are produced early after Leishmania infectionprobably also participate in this balance.36 The out-come of the initial infection will likely depend on thepredominance of host-protective or deleterious cyto-kines; however, what determines the predominantproduction on any opposing group remains to beinvestigated.

AcknowledgmentsWe acknowledge Dr. Elenita Yong (Seattle Biomed-ical Research Institute) for the TGF-f3 determinationsand Drs. R. Twardzik, M. Brandely, K. Grabstein, and

TGF-,B in Human Leishmaniasis 953AJP October 1995, Vol. 147, No. 4

L. Ellingsworth for their gifts of reagents. Dr. John Hogave invaluable suggestions. We also thank Mr.Jackson Lemos for secretarial help.

References

1. Barral-Netto M, Machado P, Barral A: Human cutane-ous leishmaniasis: recent advances in physiopathologyand treatment. Eur J Dermatol 1995, 5:104-113

2. Marsden PD: Mucosal leishmaniasis ("espundia" Es-comel, 1911). Trans R Soc Trop Med Hyg 1986, 80:859-875

3. Bryceson ADM: Diffuse cutaneous leishmaniasis inEthiopia. I. The clinical and histological features of thedisease. Trans R Soc Trop Med Hyg 1969, 63:708-737

4. Barral A, Costa JML, Bittencourt AL, Barral-Netto M,Carvalho EM: Polar and sub-polar diffuse cutaneousleishmaniasis in Brazil: clinical and immunopathologi-cal aspects. Int J Dermatol 1995, 34:474-479

5. Nathan CF, Murray HW, Wiebe ME, Rubin BY: Identifi-cation of interferon y as the lymphokine that activateshuman macrophages oxidative metabolism and antimi-crobial activity. J Exp Med 1983, 158:670-689

6. Green SJ, Crawford RM, Hockmeyer JT, Meltzer MS,and Nacy CA: Leishmania major amastigotes initiate theL-arginine-dependent killing mechanism in IFN-y-stimulated macrophages by induction of tumor necro-sis factor-a. J Immunol 1990, 145:4290-4297

7. Corradin SB, Mauel J: Phagocytosis of Leishmania en-hances macrophage activation by IFN-y and lipopoly-saccharide. J Immunol 1991, 146:279-285

8. Barral-Netto M, Badaro R, Barral A, Almeida RP, SantosS, Badaro F, Pedral-Sampaio D, Carvalho EM, FalcoffE, Falcoff R: Tumor necrosis factor (cachectin) in hu-man visceral leishmaniasis. J Infect Dis 1991, 163:853-857

9. Titus RG, Sherry B, Cerami A: Tumor necrosis factorplays a protective role in experimental murine cutane-ous leishmaniasis. J Exp Med 1989, 169:2097-2104

10. Liew FY, Parkinson C, Millot S, Severn A, Carrier MJ:Tumour necrosis factor (TNFa) in leishmaniasis. I.TNFa mediates host-protection against cutaneousleishmaniasis in vivo. Immunology 1990, 69:570-573

11. Barral A, Barral-Netto M, Yong EC, Brownell CE,Twardzik DR, Reed SG: Transforming growth factor-f3as a virulence mechanism for Leishmania braziliensis.Proc Natl Acad Sci USA 1993, 90:3442-3446

12. Barral-Netto M, Barral A, Brownell CE, Skeiky YAW,Ellingsworth LR, Twardzik DR, Reed SG: Transforminggrowth factor-f3 in leishmanial infection: a parasite es-cape mechanism. Science 1992, 257:545-548

13. Nacy CA, Meierovics Al, Green SJ, Nelson BJ: Cyto-kine networks and regulation of macrophage antimicro-bial activities. Prog Leukocyte Biol 1991, 11:271-276

14. Ding A, Nathan CF, Graycar J, Derynck R, Stuehr DJ,Srimal S: Macrophage deactivating factor and trans-forming growth factors-13l, -j32, and -,B3 inhibit induc-

tion of macrophage nitrogen oxide synthesis by IFN--y.J Immunol 1990, 145:940-944

15. Nelson BJ, Ralph P, Green SJ, Nacy CA: Differentialsusceptibility of activated macrophage cytotoxic effec-tor reactions to the suppressive effects of transforminggrowth factor-,B1. J Immunol 1991, 146:1849-1857

16. Czarniecki CW, Chiu HH, Wong GHW, McCabe SM,Palladino MA: Transforming growth factor f-1 modu-lates the expression of class 11 histocompatibility anti-gens on human cells. J Immunol 1988, 140:4217

17. Tsunawaki S, Sporn MB, Ding A, Nathan CF: Deactiva-tion of macrophages by transforming growth factor p.Nature 1988, 334:260

18. Ding A, Nathan CF, Graycar J, Derynck R, Stuehr DJ,Srimal S: Macrophage deactivating factor and trans-forming growth factors-f31, -32, and -,3 inhibit induc-tion of macrophage nitrogen oxide synthesis by IFN--y.J Immunol 1990, 145:940-944

19. Ranges GE, Figari IS, Espevik T, Palladino MA Jr: In-hibition of cytotoxic T cell development by transforminggrowth factor ,B and reversal by recombinant tumornecrosis factor a. J Exp Med 1987, 166:991-998

20. Mule JJ, Schwartz SL, Roberts AB, Sporn MB, Rosen-berg SA: Transforming growth factor-, inhibits the invitro generation of lymphokine-activated killer cells andcytotoxic T cells. Cancer Immunol Immunother 1988,26:95-100

21. Barral-Netto M, Barral A: Transforming growth factor-,in tegumentary leishmaniasis. Braz J Med Biol Res1994, 27:1-9

22. Badaro R, Reed SG, Barral A, Orge G, Jones TC:Evaluation of the micro enzyme-linked immunosorbentassay (ELISA) for antibodies in American visceralleishmaniasis: antigen selection for detection of infec-tion-specific responses. Am J Trop Med Hyg 1986,35:72-78

23. Reed SG, Badaro R, Masur H, Carvalho EM, Loren,oR, Lisboa A, Teixeira R, Johnson WD, Jones TC: Se-lection of a skin test antigen for American cutaneousleishmaniasis. Am J Trop Med Hyg 1986, 35:79-85

24. Grimaldi G, David JR, McMahon-Pratt D: Identificationand distribution of New World Leishmania speciescharacterized by serodeme analysis using monoclonalantibodies. Am J Trop Med Hyg 1987, 36:187-207

25. McMahon-Pratt D, Bennett E, David JR: Monoclonalantibodies that distinguish subspecies of Leishmaniabraziliensis. J Immunol 1982, 129:929-937

26. Jaffe CL, Bennett E, Grimaldi G, McMahon-Pratt D:Production and characterization of species specificmonoclonal antibodies against Leishmania donovani forimmunodiagnosis. J Immunol 1984, 133:440-447

27. Silva JS, Twardzik DR, Reed SG: Regulation ofTrypanosoma cruzi infections in vitro and in vivo by trans-forming growth factor-,p (TGF-,B). J Exp Med 1991,174:539-545

28. Majewski S, Hunzelmann N, Nischt R, Eckes B, Rud-nicka L, Orth G, Krieg T, Jablonska S: TGF-,B-1 andTNF-a expression in the epidermis of patients with

954 Barral et alAJP October 1995, Vol. 147, No. 4

epidermodysplasia verruciformis. J Invest Dermatol1991, 97:862-867

29. Dalal BI, Keown PA, Greenberg AH: Immunocytochem-ical localization of secreted transforming growth fac-tor-,B1 to the advancing edges of primary tumors and tolymph node metastases of human mammary carci-noma. Am J Pathol 1993, 143:381-389

30. Reed JA, McNutt NS, Prieto VG, Albino AP: Expressionof transforming growth factor-,B2 in malignant mela-noma correlates with the depth of tumor invasion: im-plications for tumor progression. Am J Pathol 1994,145:97-104

31. Melby PC, Andrade-Narvaez FJ, Darnell BJ, Valencia-Pacheco G, Tryon VV, Palomo-Cetina A: Increasedexpression of proinflammatory cytokines in chronic le-sions of human cutaneous leishmaniasis. Infect Immun1994, 62:837-842

32. Fruth U, Solioz N, Louis JA: Leishmania major interferes

with antigen presentation by infected macrophages. JImmunol 1993; 150:1857-1864

33. Pirmez C, Cooper C, Paes-Oliveira M, Schubach A,Torigian VK, Modlin RL: Immunologic responsivenessin American cutaneous leishmaniasis lesions. J Immu-nol 1990, 145:3100-3104

34. Modlin RL, Pirmez C, Hofman F, Torigan V, Uyemura K,Rea T, Bloom B, Brenner M: Lymphocytes bearingantigen specific y-8 T cell receptors accumulate inhuman infectious disease lesions. Nature 1989, 339:544-546

35. Barral-Netto M, Barral A, Brodskyn C, Carvalho EM,Reed SG: Cytotoxicity in human mucosal and cutane-ous leishmaniasis. Parasite Immunol 1995, 17:21-28

36. Afonso LCC, Scharton TM, Vieira LQ, Wysocka M,Trinchieri G, Scott P: The adjuvant effect of interleukin-12in a vaccine against Leishmania major. Science 1994,263:235-237