Cultured Epidermal Langerhans Cells A'ctivate Effector T Cells for Contact Sensitivity

Conrad Hauser Department of Dermatology, Hopital Cantonal U niversitaire, Geneva, Switzerland

To investigate whether Langerhans cells are capable of in­ducing contact sensitivity effector T cells, we incubated puri­fied T cells from naive mice with syngeneic cultured trini­trophenyl-modified Langerhans cells for 4 - 5 d. The cells were then expanded in interleukin-2 and fresh medium for another 6 - 9 d and injected intravenously into naive syngen­eic recipient mice. After the ears of recipient mice were painted with 1 % trinitrochlorobenzene, we observed an ear­swelling response peaking at 24 - 48 h. The ear-swelling re­sponse was hapten specific. C08- but not C04-depleted T cells mediated strong contact sensitivity. Systemic adoptive

Contact sensitivity (CS) is a cell-mediated inflamma­tory response of delayed onset (27 - 72 h) to the epi­cutaneous application of some highly reactive simple chemical compounds (haptens). In mice, CS can be induced by a single application of compounds such as

dinitrochlorobenzene, trinitrochlorobenzene (TNCB), or oxazo­lone. The inductive phase of CS has extensively been studied in the last few years and Langerhans cells (LC) - major histocompatibility complex (MHe) class II positive, intraepidermal dendritic antigen­presenting cells (APe) -are thought to playa crucial role in this reaction.

These conclusions are drawn from a large number of experiments performed by different groups of investigators. In one kind of ex­periment, the failure to induce CS by application of haptens to LC-deficient skin (tai l skin, hamster cheek pouch) or skin with altered LC ultraviolet B [UVB]- or glucocorticoid-treated skin) was taken as evidence for the role of LC in the induction of CS [1 - 4].

Manuscript received January 29, 1990; accepted for publication May 23, 1990.

This work was supported by a grant from the Swiss National Foundation for Scientific Research (3.115-0.88).

Previously presented in part at the Annual Meetin ~ of the Society of Investigative Dermatology, April 26-30, 1989, WashIngton, D.C.

Reprint requests to Dr. Conrad Hauser, Clinique de Dermatologie, Hopi­tal Cantonal Universitaire, 1211 Geneve 4, SWltzeriand.

Abbreviations: APC: antigen-presenting cells CS: contact sensitivity cLC: cultured Langerhans cells FlTC: fluorescein isothiocyanate IL-2: interleukin-2 1L-4: interleukin-4 LC: Langcrhans cell MHC: major histocompatibility complex MoAb: monoclonal antibody TNCB: trinitrochlorobenzene TNP: trinitrophenyl-UVB: ultraviolet B

transfer into nude mice also lead to a hapten-specific delayed ear-swelling response. However, this response was less prO­tracted than in euthymic animals, suggesting the participa­tion of the recipient (non-immunized) T cells in the ear­swelling response of the euthymic mice. Lymphokine analysis of in vitro primed and restimulated T cells revealed predominant production of interleukin-2 but little or no in­terleukin-4. These in vitro primed cells therefore resemble type 1 or inflammatory T helper cell clones.] Invest Damatol 95:436-440,1990

The latter experiments seemed particularly conclusive because a phenotypic alteration of LC was associated with the failure to in­duce CS . In another line of experiments, hapten-modified LC-con­taining epidermal cells or LC-enriched epidermal cells were in­jected into naive animals. The successful immunization of rnest animals was taken as further evidence for the critical role of LC for the induction of CS [5 - 7) .. The failure to induce CS with LC-ddi­cient epidermal cells underlined the significance of these e».-peri­ments.

However, other experimental data speak against a major role Q '

LC in the induction of CS. It was found that low-dose UYE impain the induction of CS only in some mouse strains [8]. The susceptibil­ity to low-dose UVE was later shown to be controlled by multiplt genetic loci [9]. In addition, a decrease in LC markers after 10w-dOst UVB was observed in UVB non-suspceptible strains [10]. Experi. ments with high-dose UV irradiation [11-14J were not conclusive for a role ofLC in CS, because this treatment led to immunosuppres­sion, which was accompanied by a systemic defect of APC function. Furthermore, it was found that guinea pigs were sensitized to hap­ten through tape-stripped skin, i.e., skin with decreased LC density [15). In the guinea pig system, injection of hapten-modified L reportedly failed to sensitize the animals [16), whereas injection hapten-modified peritoneal exudate cells and LC-depleted epider. mal cells successfully induced CS [16,17]. Finally, the recent report that the injection into naive mice of harten-modified dermal MH class II+ cells is capable of inducing CS 18] leaves open the possibil­ity that skin-derived dermal APC but not epidermal LC could ~ responsible for the induction of CS by the transplantation of hap­ten-modified skin [19].

To investigate whether LC can give rise to CS effector celJs, " set up an in vitro system that would produce CS effector cells usin" hapten-modified syngeneic cultured LC (cLe). Since Moorh~ demonstrated that T cells are capable of transfering CS [20], we t advantage of an earlier described in vitro system that allows the '. vitro generation of T cells with desired specificity from T cell ' non-immunized animals [21]. In this paper, we report on the -. cessful expression of CS by the systemic transfer of T-cell popu! tions that were cultured with hapten-modified cleo

0022-202X!90jS03.50 Copyright ©1990 by The Society for Investigative Dermatology, Inc.

436

VOL. 95, NO.4 OCTOBER 1990

MATERIALS AND METHODS

LC-T Cell Cultures cLC were prepared and hapten modined as previously described [21 J. The cell suspension contained 30 - 50% cLC as assessed by fluorescent microscopy and flow cytometry using antibodies to MHC class II determinants. The remaining cells were mainly keratinocytes [21,30J. T cells were prepared by passing splenocytes and lymph node cells of non-immunized female Balb/c mice over nylon wool columns. The cells were then incubated with either anti-class II monoclonal antibody (MoAb) (M5/114.15.2, ATCC, Rockville, MD) or anti-class II plus either anti-CD4 MoAb (GK 1.5, ATCC) or anti-COB MoAb (53-6.72, ATCC) for 30 min on ice. The washed cells were subsequently incubated for 45 min with low toxic rabbit complement (1: 10, Cedarlane, Hornby, Ontario, Canada) and a mouse MoAb to rat immunoglobulin kappa chain (MAR 18.5, ATCC). The purified T cells were cultured with hap.ten-modined cLC at a ratio of about 100 T cells to 1 cLC at ?/r c in a humidified atmosphere. The culture media consisted of RPM11640 or Iscove's modified Dulbecco's medium supplemented with FCS (10%), penicillin (100IU/ml), streptomycin (100 j.lg/ml), glutamine (2mM), 2-mercaptoethanol (2 X 10-5 M, all from GIBCO, Basel, Switzerland). After 4- 5 d of culture fresh media containing 50 U /ml recombinant human interleukin-2 (Cetus, Emeryville, CA) were added. The cultures were fed with fresh media and IL-2 every 2-3 d. After 11-13 d, the cells were washed and prepared either for adoptive transfer or in vitro restimu­Wion. For in vitro restimulation, 0.5 X 106 T cells, that had been stnSitized with trinitrophenyl (TNP)-modified cLC, were incu­bated for 2 - 3 d with lOS TNP-modified cLC. The resulting culture ropematant was then tested for lymphokine activity .

Freshly prepared cells were analyzed by flow cytometry using Guorochrome-conjugated MoAb to CD4 and CDB (Becton Dick­inson, Mountainview, CA). Cultured cells were stained with the same MoAb and 1000 cells scored by fluorescent light microscopy.

r,ymphokine Assay [22] To obtain a measure for IL-2 and inter­/cukin-4 (IL-4) production, Iymphokine-containing T helper cell culture supernatants were incubated in serial dilutions with CTLL cells, a kind gift of Dr. E.M. Shevach (LI, NIAID, NIH). These CTLL cells respond to both IL-2 and IL-4. To these cultures either S4B6 [23J, a neutralizing MoAb to mouse IL-2 (kindly provided by {)r.T . Mosmann, DNAX, Palo Alto, CA), !lIBll [24J, a neutraliz­ing MoAb to mouse IL-4 (kindly provided by Dr. W.E. Paul, LI, NIAlD, NIH), both MoAb or no antibody were added. Internal 5Undards for IL-2 and IL-4 were incubated in parallel. Proliferation fJas assessed by 3H-methyl-thymidine incorporation after pulsing me cultures (0.5 tiCi/well) for the final 6-8 h of the 24 h culture period.

Transfer Experiments 5 X 107 in vitro sensitized T cells were resuspended in Hank's balanced salt solution (108 cells/ml) and 10 U/mJ heparin. The cells were then injecred intravenously through the lateral tail vein or the retro-orbital venous plexus. The ears of the animals were immediately painted with 1 % TNCB (Poly­~nces, Warrington, PA; in olive oil) or 0.5% fTuorescein.isothio­cyanate (FITC) (Sigma, St. Louis, MO, in acetone/dibutyl-phrha­Iuc, 1 : 1). After the indicated time period following transfer, ear thickness was assessed by an ear caliper (Interapid, Rolle, Switzer­Iznd). Results are expressed as follows: ear swelling (mm X t(J-2) = ear thickness after transfer (mm X 10-2) minus ear thick­ntSS before transfer (mm X 10-2). In some experiments, the rt/evant hapten was painted on one ear and the control hapten was painted on the other ear of the same animal. Significance levels were akulated using the Wilcoxon-Mann-Whitney rank sum test.

RESULTS

r Cells Sensitized In Vitro with eLC Mediate CS in N on-Im­munized Recipient Mice T cells from non-immunized mice 1lJ)Sirized in vitro with hapten-modified cLC and expanded in vitro with IL-2 were injected intravenously into two groups of syngeneic lIOIl-immunized recipient mice (5 X 107 cells per animal). One

LANGERHANS CELLS ACTIVATE EFFECTOR T CELLS 437

16

)\ ....,. N

0 ,... x

I 1" E .5 12 w

~ (/) z 0 a. (/) ! w II: 8

" z ::i ...J W ~ (/)

II: 4 « w

4 24 48 72 96 120 144

TIME AFTER CHALLENGE (h)

Figure 1. Ear-swelling response of mice receiving T cells sensitized in vitro to TNP. T cells (5 X 107 cells per animal) from non-immunized Balblc mice sensitized in vitro with cultured TNP-modilied syngeneic cLC were injected into non-immunized syngeneic mice. Ears were immediately chal­lenged with either TNCB or FITC and ear-sweIling responses assessed after the indicated time period. (solid circle) TNCB challenged mice (n = 4); (opetl circle) FITC challenged mice (n = 4); all p < 0.01 between 24 and 96 h. Error bars, .1 SEM.

group was immediately challenged with the hapten used for in vitro sensitization and the other grOl1p was painted with an irrelevant hapten. The group of mice challenged with the homologous hapten TNCB developed delayed ear swelling peaking at 48 h (Fig 1). The group of mice receiving the irrelevant hapten FITC did not show 'lny signincant ear swelling. Similarly, a delayed ear-swelling re­sponse to FITC challenge was observed when the T cells were sensitized in vitro with FITC-modified cLC (data not shown) . Mice receiving 2 X 107 cells sensitized with TNP-modined cLC and challenged with TNCB did not show any significant ear swelling. These experiments demonstrate that hapten-specific CS-mediating T-cell populations-can be generated-by-cultnrirrg-hapten~m0dined ~ cLC with syngeneic non-immunized T cells.

CD4+ T Cells Confer' CS In Vivo After In Vitro Activation with eLC In order to find out whether CD4+ or CD8+ T cells are capable of conferring CS in vivo after primary in vitro activation with hapten-modified cLC and subsequent expansion, we treated the freshly isolated T cells with MoAb to either CD4 (GK 1.5) or CD8 (53-6.72), followed by mouse anti-rat immunoglobulin MoAb (MAR 18.5) and complement. The cells were then sensitized in vitro, expanded, and finally injected into recipient mice. Data in Fig 2 show clearly that CD8-depleted T cells conferred CS in vivo. This suggests that CD4+ T cells were responsible for the transfer of CS, although a contribution from CD4-CD8- cells cannot be excluded. Little or no CS was elicited in mice receiving in vitro activated and expanded CD4-depleted T cells (Fig 2). Aliquots of cells from experiment 2 were analyzed by immunofluorescence mi­croscopy before transfer. The CD4-depleted cell fraction contained 3.0% CD4+ cells, the CD8--depleted fraction less than 3% CD8+ cells. Experiments with depletion of CD4+ or CDB+ cells after in vitro sensitization and expansion but before transfer were not per­formed.

438 HAUSER

Expt 1

Expl2

SPECIFICITY HAPTEN FO R AND SUBSET CHALLENGE

GROUP r------,--­

TNP. whole T

TNP. COB dept. T

TNP. CD. depl. T

TNP. coa depL T

TNP. CD' depl. T

EAR SWELLING (mm x 102)

>0 15 20

Figure 2. Ear-swelling responses of mice receiving either CD8- or CD4-depleted T cells sensitized to TNP. RaIble T cells depleted of CD8+ or CD4+ cells and sensitized in vitro with TNP-modified cLC were injected into naive Balb/c mice and immediately challenged with TNCB on the right ear and with FrTC on the left ear. Peak ear-swelling responses are shown. Group 1, n = 2; group 2, n = 4; group 3, n = 8; group 4, n = 7; group 5, n = 7. TNCB induced ear-swelling responses: group 2 vs 3, P < 0.01 ; group 4 vs 5, P < 0.001. Error bars, 1 SEM.

Transferred CS Effector T Cells Recruit Recipient T Cells for the Elicitation Reaction When we transferred in vitro sen­sitized CD8-depleted T cells into syngeneic nude recipients, an ear-swelling response was observed (Fig 3). At 24 h, the ear-swel­ling response was weaker than in euthymic mice. At 48 h, the response of nude mice was near background, whereas the euthymic mice showed maximal response. No significanc ear swelling was observed in nude mice that were challenged with TNCB but did not receive any cells (data not shown) . These findings demonstrate that the 24-h ear-swelling response is partially independent of recipient T cells, but that the 48-h response in normal Balb/c mice depends on non-immune T cells of the recipient animal.

In Vitro Sensitized CD8-Depleted T Cells Produce IL-2 After In Vitro Restimulation To investigate whether in vitro sensitized CD8-depleted T cells produce IL-2 or IL-4, we restimu­lated th ese cells in vitro with hapten-modified cLC. The culture supernatant was removed after 3 d and tested on CTLL cells, which

SPECIFICITY HAPTEN FOR EAR SWELLING (mm x 102) OF T CELLS. CHALLENGE RECIPIENTS

GROUP 10 >5 20

,---,-------,---

Figure 3. Ear-swelling responses of nude Or euthymic BalbI c nUce receiv­ing CDS-depleted T cells s\!nsitized to TNP. Balb/e T cells depleted of CDS+ cel ls and sensitized in vitro to TNP were injected into naive nude (n = 5) or euthymic (n = 4) Balb/c mice. The animals were immediately challenged with TNBC on the right ear and with FITC on the left ear. Ear-swelling responses were assessed at 24 and 48 h. TNCB induced ear­swelling responses: group 1 vs 2, P > 0.5; group 3 vs 4, P < 0.01 . E rror bars, 1 SEM.

c::;-o 60 >< E g. z o i= 40 <t cr:: o a. a: o ()

~ 20 ±

C"l

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

IL2 IL4 TNP-cLC + ThTNP (ConA (X63/0

SN 1/30) SN 1/50)

y o

1 0 30 100 300

RECIPROCAL OF SUPERNATANT DILUTION

Figure 4: Specific bioassay for IL-2 and IL-4 of culture supernatants frem in vitro restimulated CD8-depleted T cells mediating CS in vivo. CD8-d~ pleted T cells sensitized in vitro to TNP and which conferred CS in vi were restimulated in vitro with TNP-modified cLC. The 3-d culture super_ natant was filtered and incubated with CTLL cells that respond to both Il. and IL-4 . T ests were c arried out in the presence of MoAb to IL-2 (tria llgl~ S4B6) and to 1L-4 (rnverted tnallgle, llBl1), both With (star) or withOut MoAb (open circle). 3H-ineorporacion was assessed after 24 h and a labelin period of S h. Concanavalin A-stimulated spleen cell supernatant (ConK SN) and supernatant of X 63/0 cells transfected with a plasmid containin the mouse 1L-4 gene ([48]; X 63/0 SN) were used as internal IL-2 and Il~ standards, respectively.

respond to both IL-2 and IL-4. The CTLL-growth - promotin activity contained in the culture supernatant was inhibited by ~ MoAb to IL-2 (S4B6) but not by a MoAb to IL-4 (1IBll) (Fig 4). Furthermore, culture supernatants to which both MoAb had been added showed the same activity as culture supernatants to which only anti-IL-2 MoAb were added. In other experiments (data no shown), small amounts of IL-4 were detected in the culture super. natants using the same assay. Experiments in which anti-IL-2 and anti- IL-4 antibodies were directly added to in vitro sensitiz CD8-depleted T cells and TNP-modified cLC showed inhibitiol\ with the antibody to IL-2 but no significant inhibition with the antibody to IL-4 (data not shown). These results suggest that the 1: cells restimulated · in vitro with hapten-modified cLC produc mainly IL-2 and little or no IL-4.

DISCUSSION

Our results give the first direct evidence for the ability of cLC l\) induce CS effector T cells in an adoptive transfer system. In contrast to the system described here, the APC was poorly defined in fonnet studies on the induction of DTH effector cells in vitro [25,26]. The in vitro induction protocol for CS effector cells described by Kn P and colleagues [27,28) used bone marrow -derived plastic adherent cells grown in CSF-l containing medium. It is not excluded that these cultures contained small amounts of dendritic cells. Lymphoi dendritic cells are considered to be the APC of primary T-cell re­sponses [29J. cLC and lympoid dendritic cells share the functional ac tivity of being potent for primary T-cell responses. Althou h dendritic cells are capable of activating primary TNP-dependent T-cell proliferation [30J, we do not know whether they are capabl of inducing CS effector T cells in the described system. We ha, previously used cLC to produce hapten-, al1o-, and protein­antigen-specific CD4+ T-cell lines [21 ,22) . These cell lines Wen'

produced by repetitive stimulation with cLC and secreted IL- .

VOL. 95, NO. 4 OCTOBER 1990

IL-4 - producing cell lines, however, were reported not to mediate delayed-type hypersensitivity [31]. We were unable to grow our hapten-specific IL-4 producing lines up to cell numbers (5 X 107

per animal) necessary for an adoptive transfer. We did not test freshly isolated LC, because these cells bear a low class II density in our hands and die quickly in cell culture [32,33] . However, in epi­dermal cell culture, LC increase their class II density and become potent APC [32- 34]. Granulocyte-macrophage colony-stimulat­ing factor together with IL-l, which are both produced by keratino­cytes (more than 90% of freshly isolated epidermal cells) in culture, can induce these changes in highly purified fresh LC [35,36]. Our data show that cLC can trigger the generation of CS effector cells but a role for LC in the induction of CS in vivo has not yet been directl y demonstrated, although it is most probable. LC are thought to migrate via lymphatic vessels to the paracortical T-cell areas of the regional lymph nodes after hapten exposure in the epidermis. Indeed, FITC-bearing, dendritically shaped class II+ cells have been observed in the lymph node 24 h after skin painting with FITC [37] . These cells are capable of inducing CS in naive recipients, stimulate the proliferation of non-sensitized T cells, bear the dendritic cell marker 33Dl, and contain Birbeck granules [37 -40]. However, their identity with or their derivation from LC remains unproved. Additionall y, it is possible that dermis-derived cells might also func­tion as APC for the induction of CS.

Transfer experiments revealed that CD8-depleted T cells were responsible for the elicitation of CS in the recipients. The CS re­sponse peaked at 24-48 h (Fig 1). However, persistent ear swelling was found up to 6 d. It is possible that the swelling measured after 48 co 72 h was due to concomitant transfer of hapten-modified cLC still present in the IL-2-expanded T-cell preparation. Injection of hapten-modified LC has been shown to induce delayed ear-swelling responses upon challenging after 5 d [7J. However, the earlier 24-48 h ear-swelling responses cannot be explained by transfer of T cells contaminated with hapten-modified cLC. Our data are in ac­cordance with other experimental systems showing that the elicita­tion of CS is dependent on CD4+ T cells [41]. CD4-depleted T cells which were sensitized in vitro did not confer strong CS in vivo in our system. It has, however, recently been shown that CD8+ T cells playa role in vivo as CS effector cells [42]. It is therefore possible that the cells generated in vitro are functiona lly different from those generated in vivo and that they might need additional factors and/or cells to become functional in vivo. The fact that, for technical reasons, we used lymph nodes and spleens (instead of solely lymph nodes draining the skin) to prepare the large amounts of T cells (2 -4 X 108 cells per experiment) needed for these experiments might be another reason why CD4-depleted T cells did not confer strong ear swelling. Furthermore, it is possible that CD8+ T cells became competent for the adoptive transfer of LS after in vitro sensitization with cLC in the presence of CD4+ T cells, as in the experiment with the whole T-cell fraction, but not when they were sensitized with cLC in the absence of CD4+ T cells. Experiments including separation of CD4+ T cells and CD8+ cells after sensitization and expansion but before transfer would answer this question. However, T-cell numbers required for such experiments (- 109 cells) in our system were not ob­tained.

How could the small response with CD4~depleted T cells be explained? Our protocol for purification ofT-cell subsets routinely yields T cells with no detectable contamination as assessed by flow cytometric analysis (data not shown). However, the analysis was done before culture. In experiment 2 (Fig 2), we also analyzed the cells before transfer but after culture. The CD4-depleted fraction contained 3.0% CD4+ cells. Thus, it cannot be decided whether the small contamination by CD4+ T cells or the CD8+ T cells alone were involved in the small response seen in animals that received CD4-depleted T cells. It is known that CD4/8 phenotype and functio n do not strictly correlate. In addition, in experiments with hapten-modified skin grafts, it has been shown that induction of CS is not restricted only by the class II MHC, but also by the class I MHC [19]. It is therefore possible that class 1- and/or class II-

LANGERHANS CELLS ACTIVATE EFFECTOR T CELLS 439

restricted CD8+ (lymphokine secreting?) T cel ls at low frequencies were involved in the small response observed with CD4-depleted cells. Alternatively, the 3% contamination with CD4+ cells could also be directly or, via help to th e CD8+ cells, indirectly account for this small response.

To characterize the in vitro immunized CD4+ T cells, we re­stimulated in vitro aliquots of cells that conferred CS in vivo. The lymphokines identified in the cu lture supernatant and the results from the direct inhibition experiments suggest that the cells pro­duced mainly IL-2 but little or no detectable IL-4 . IL-2 seems to be a crucial mediator in TNP-induced delayed skin-swelling responses because elimination of IL-2 receptor bearing cells inhibited this response [43 ,44] . Our in vivo activated and expanded CD8-de­pleted T cells resemble CD4+ T-cell clones of type 1 as defined by Mossman and colleagues [45,46] in 2 points. Type 1 T helper cells also produce IL-2 but not IL-4 and may mediate a delayed inflamma­tory reaction after adoptive transfer [31 ,47]. However, as defined by Mossman, they are long-term cultured and cloned T-cell lines de­rived from conventionally immunized mice, have been shown to be active in a local adoptive transfer assay and possess a stable pheno­type. On the contrary, our CD8-depleted T cells were sensitized in vitro with cLC and have not been tested in a local adoptive footpad swel ling assay. In addition, the lymphokine profile of our cells is not stable because repeated stimulation leads to th e establishment of CD4+ T cell lines that produce IL-4 and thus resemble type 2 T helper cells [22]. Type 1 T helper cell clones might represent a "frozen" differentiation stage ofT cells that mediate inflammatory phenomena in contrast to our cells, which lose this phenotype when repeatedly stimulated in vitro. The question of whether type-2-like T helper cells may derive from type-l-like T helper cells remains to be answered.

In summary, this paper demonstrates that the co-culture of hap­ten-modified cLC and CD8-depleted T cells from naive mice fol­lowed by expansion with IL-2 leads to type 1-like T helper cells , which produce IL-2 and mediate CS when injected into naive mice. This T -cell phenotype is lost upon subseguent culture. It remains to be shown whether loss of type l-like T helper cells plays a role in the down-regulation of allergic CS.

We thallk F.jall//ill Jor excelle/Jltecilllical assistm/ce, F. Me/chersJor the trallsJected X63/ 0 cell lilies, P.F. Pigllet alld G.E. CrallJor help with the ual/sJer experimellts. C. Sti llgl, C.E. Crall, K.M. Maller, alld f.-H. Sail rat Jor critical readil/g oj (he lltalll/script, G. Stillgl, R.E. Tigelaar, alld S.l. Katz Jor helpJlI1 disClissiolls alld S. Deschamps Jor typillg the lI1allllscript.

REFERENCES

1. Toews CB, Bergstresser PR, Streilein JW: Epidermal Langerhans cell density determines' whether contact hypersensitivity or unrespon­siveness follows skin painting with DNFB. J Immunol 124:445-450,1980

2. Streilein JW, Toews CB, Bergstresser PR: Langerhans cells: func­tional aspects revealed by in vivo grafting studies. J Invest Dermatol 75:17 -21,1980

3. Streilein JW, Bergstresser PR: Langerhans cell function dictates in­duction of contact hypersensitivity or unresponsiveness to DNFB in syrian hamsters. J Invest Dermatol 77:272- 274, 1981

4. Lynch DH, Gurish MF, Daynes RA: Relationship berween epidermal Langerhans cell density, ATPase activity and the induction of con­tact hypersensitivity. J Immunol 126:1892 - 1897, 1981

5. Ptak W, Rozycka D, Askenase PW, Gershon RK: Role of antigen­presenting cells in the development and persistence of contact hyper­sensi tivity. J Exp Med 151 :362-373, 1980

6. Tamaki K, Fujiwara H, Katz SI: The role of epidermal Langerhans cells in the induction and suppression of contact sensitivity. J Invest Dermatol 76:275-280, 1981

7. Sullivan S, Bergstresser PR, Streilein JW: Intravenously injected TNP-derivatized, Langerhans cell-enriched epidermal cells induce

440 HAUSER

contact hypersensitivity in syrian hamsters. ] Invest Dermatol 84:249-253, 1985

8. Sauder DN, Katz SI: Strain variation in the induction of tolerance by epicutaneous application of trinitrochlorobenzene.] Invest Derma­tol 80:383-386, 1983

9. Streilein]W, Bergstresser PR: Genetic basis of ultraviolet-B effects on contact hypersensitivity. Immunogenetics 27:252-259, 1988

10. Orita M: Effects of ultraviolet irradiation on surface marker expression by epidermal immunocompetent cells and contact sensitization to dinitrofluorobenzene in mice. Br J Dermatol 117:721-729, 1987

11. Morison WL, Bucana C, Kripke ML: Systemic suppression of contact hypersensitivity by UVB radiation is unrelated to the UVB-induced alterations in the morphology and number of Langerhans cells. Im­munology 52:299 - 304, 1984

12. Noonan FP, Bucana C, Sauder DN, deFabo EC: Mechanism of sys­temic immune suppressio!} by UV irradiation in vivo. II. The UV effects on number and morphology of epidermal Langerhans cells and the UV -induced suppression of contact hypersensitivity have different wavelength dependencies. ] Immunol 132:2408-2413, 1984

13. Greene MK, Sy MS, Kripke M, Benacerraf B: Impairment of antigen­presenting cell function by ultraviolet radiation. Proc Nat! Acad Sci USA 76:6591-6593, 1979

14. Letvin NL, Greene MI, Benacerraf B, Germain RN: Immunologic effects of whole-body ultraviolet irradiation: selective defect in splenic adherent cell function in vitro. Proc Nat! Acad Sci USA 77:2881-2883,1980

15. Baker D, Parker DD, Turk ]L: Effect of depletion of epidermal den­dritic cells on the induction of contact sensitivity in the guinea-pig. Br J Dermatol 113:285-290, 1985

16. Baker D, Parker D, Healey DG, Turk ]L: Induction of sensitization and tolerance in contact sensitivity with haptenated epidermal cells in the guinea-pig. Immunology 62:659-664,1987

17. Von Blomberg-Van Der Flier M, Scheper R], Boerrigter GH, Polak L: Induction of contact sensitivity to a broad variety of allergens with haptenized macrophages.] Invest DermatoI83:91-94, 1984

18. Tse Y, Cooper KD: Cutaneous dermalla+ cells are capable of initiating delayed-type hypersensitivity responses.] Invest Dermatol 94:267-272, 1990

19. StreileinJW, Wood PJ, Lonsberry LW, Bergstresser PR: Induction of H-2-restricted contact hypersensitivity by hapten-derivatized skin grafts. Transplantation 37:195 - 200, 1984

20. Moorhead JW: Tolerance and contact sensitivity to DNFB in mice. VIII. Identification of distinct T cell subpopulations that mediate in vivo and in vitro manifestations of delayed hypersensitivity. J Im­munoI120:137-145,1978

21. Hauser C, Katz SI: Activation and expansion of hapten- and protein­specific T helper cells from nonsensitized mice. Proc Nat! Acad Sci USA 85:5625 - 5628, 1988

22. Hauser C, Snapper CM, Ohara], Paul WE, Katz SI: T helper cells grown with hapten-modified cultured Langerhans cells produce interleukin 4 and stimulate IgE production by B cells. Eur] Im­munoI19:245-251,1989

23. Mosmann TR, Bond MW, Coffmann RL, Ohara], Paul WE: T-cell and mast cell lines respond to B-cell stimulatory factor I. Proc Nat! Acad Sci USA 83:5654-5657,1986

24. Ohara], Paul WE: Production of monoclonal antibody to and molecu­lar characterization ofB cell stimulating factor-l. Nature 315:333-335, 1985

25. Ramshaw lA, Eidinger 0: The in vitro induction of T cells which mediate delayed-type hypersensitivity toward horse red blood cells. Cell Immunol 42:42-47, 1979

26. Bretscher PA: In vitro induction of delayed-type hypersensitivity. Eur J ImmunoI9:311-316,1979

27. Knop], Stremmer R, Taborski U, Freitag W, de Maeyer-GuignardJ, Macher E: Inhibition of the T suppressor circuit of delayed-type hypersensitivity by interferon. J ImmunoI133:2412-2417, 1984

28. Knop J, Taborski U, de) vlaeyer-Guignard J: Selective inhibition of the generation of T suppressor cells of contact sensitivity in vitro by interferon. J Immunol 138:3684- 3690, 1987

29. Steinman RM, Inaba K, Schuler G, Witmer M: Stimulation of the immune response: contributions of dendritic cells. In: Steinman

THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

RM, North RG (eds.). Host Resistance Mechanisms to Infectious Agents, Tumors and Allografts. The Rockefeller University Press, New York, 1986, pp 71-97

30. Hauser C, Yokoyama W, Katz SI: Characterization of primary l' helper cell activation and T helper cell lines stimulated by hapten­modified, cultured Langerhans cells. ] Invest Dermatol 93 :649~ 655, 1989

31 . Cher D], Mosmann TR: Two types of murine helper T cell clones. II. Delayed type hypersensitivity is mediated by T H 1 clones. ] lIn­munol138:3688-3694, 1987

32. Schuler G, Steinman RM: Murine epidermal Langerhans cells ma~ into potent immunostimulatory dendritic cells in vitro. J Exp Mel! 161:526-537,1985

33. Inaba K, Schuler G, Witmer MD, Valinsky], Atassi B, Steinman RM: Immunologic properties of purified epidermal Langerhans cells. Distinct requirements for stimulation of unprimed and sensitized T lymphocytes. J Exp Med 164:605-617, 1986

34. Shimada S, Caughman SW, Sharrow SO, Stephany 0, Katz S1: En­hanced antigen-presenting capacity of cultured Langerhans cells is associated with markedly increased expression of Ia antigen. J lIn­munol 139:2551-2556,1987

35. Witmer-Pack MD, Olivier W, Valinsky], Schuler G, Steinman ru..1: Granulocyte/macrophage colony-stimulating factor is essential for the viability and function of cultured murine epidermal Langerhans cells. ] Exp Med 166: 1484 - 1496, 1987

36. Heufler C, Koch F, Schuler G: Granulocyte/macrophage colony­stimulating factor and interleukin 1 mediate the maturation of mur­ine epidermal Langerhans cells into potent immunostimulatory den­dritic cells.] Exp Med 167:700-705, 1988

37. Macantonia SE, Knight SC, Edwards AJ, Grifhts S, Fryer P: Local.i.za. tion of antigen on lymph node dendritic cells after exposure to th~ contact sensitizer f1uoresceine isothiocyanate. Functional and mor­phological studies.] Exp Med 166:1654-1665, 1987

38. Macantonia SE, Edwards -A], Knight SC: Dendritic cells and the initia­tion of contact sensitivity to fluorescein isothiocyanate. Immunol_ ogy 59:509-515,1986

39. Macantonia SE, Knight SC: Dendritic cells and T cells transfer sensiti­zation for delayed-type hypersensitivity after skin painting with contact sensitizer. Immunology 66:96 -103, 1989

40. Munn CG, Bucana C, Kripke ML: Antigen binding cells from lymph. nodes of mice sensitized epicutaneously with hapten have features of epidermal Langerhans cells.] Invest Dermatol 92:487 - 492, 1989

41. Miller SO, Jenkins MK: In vivo effects ofGK 1.5 (anti-L3T4a) mono­clonal antibody on induction and expression of delayed-type hypet­sensitivity. Cell Immunol 92:414-423, 1985

42. Gocinski BL, Tigelaar RE: In vivo depletions ofL3T4+ and LyT-2+ cells distinguish between delayed-type hypersensitivity and contact sensitivity (abstr). J Invest Dermatol 90:563, 1988

43. Kelley VE, Noar D, Tarcic N, Gaulton GN, Strom TB: Antiinterleu­kin 2 receptor antibody suppresses delayed-type hypersensitivity !U foreign and syngeneic antigens. J Immunol137:2122-2126, 19 6

44. Kelley VE, Bacha P, Pankewycz 0, Nichols ]C, Murphy JR, Strom TB: Interleukin 2-diptheria toxin fusion protein can abolish cell­mediated immunity in vivo. Proc Nat! Acad Sci USA 85:3980-3983, 1988

45. Mosmann TR, Cherwinski H, Bond MW. Giedlin MA, Coffman RL: Two types of murine helper T cell clone. I. Definition according !U profiles of Iymphokine activities and secreted proteins. ] Irnrnunol 136:2348-2353,1986

46. Cherwinski HM, Schumacher KD, Brown KD, Mosmann TR: Tw types of mouse helper T cell clone. Ill. Furcher differences in lym­phokine synthesis between T H 1 clones revealed by RNA hybridiza­tion, functionally monospecific bioassays and monoclonal antibod. ies. ] Exp Med 166: 1229 -1240, 1987

47. Wong RL, Ruddle NH, Padula S], Lingenheil EG, Bergman CM, Rugen RV, Epstein El, Clark RB: Subtypes of helper T cells. Non­inflammatory type 1 helper T cells.] Immunol 141:3329-33 -I, 1988

48. Karasuyama H, Melchers F; Establishment of mouse cell lines which constitutively secrete large quantities ofinterleukin 2,3,4 or 5 usin~ modified eDNA expression vectors. Eur] Immunol 18:97 -10 1988