The ability of Ag-pulsed dendritic cells (DC) to induce primary immune responses has led them to be used for vaccination purposes. However, irrelevant Ags (e.g., FCS) can also be taken up by DC during their isolation and culture and then presented in vivo. To circumvent this, we have established a serum-free (SF) culture system. Murine epidermal cell (EC) suspensions were prepared with and without FCS and cultured for 3 days either in SF or FCS-containing medium. In spite of the lower Langerhans cell (LC) yields under SF conditions, both SF- and FCS-cultured LC (SF-cLC, FCS-cLC) underwent a similar maturation process, as evidenced by a similar increase in the cell surface expression of MHC class II and of costimulatory molecules. The further observation that SF-EC cultures elaborated comparable amounts of granulocyte-macrophage (GM)-CSF as FCS-cultured EC, but were relatively impaired in their IL-1α and TNF-α production, supports the role of GM-CSF in LC maturation and, less so, in LC survival. Functionally, freshly isolated SF-LC compared with FCS-LC in their Ag-processing capacity. Three-day-cultured SF-LC were as potent stimulators of polyclonal T cell responses and of the primary allogeneic MLR as FCS-cLC, but were relatively poor activators of naive, syngeneic CD4+ T cells. In vivo, hapten-modified SF-cLC induced a contact hypersensitivity response similar in magnitude and kinetics to that evoked by FCS-cLC. Our data show that, in the absence of serum and exogenous cytokines, LC mature into potent activators of T cell responses and could thus be a valuable cellular source for DC-based immunotherapy.

Dendritic cells (DC)3 are bone marrow-derived APC that are present in small numbers in both lymphoid and nonlymphoid tissues. One of their key functions in the regulation of immune responses is their ability to prime naive T cells (1). Langerhans cells (LC) are specialized skin DC residing within the epidermis and are crucial for the induction of immune responses against Ags introduced or newly generated in the skin (2, 3). After in vitro and in vivo perturbation of their microenvironment, LC undergo a process of maturation, as evidenced by changes in their phenotype and function. They down-regulate CD32 and F4/80, up-regulate MHC class I, MHC class II, and CD44 Ags, and de novo express CD80 and CD86 costimulatory molecules. Functionally, they reduce their Ag-processing potency and increase their ability to stimulate naive T cells (4, 5, 6). This strong immunostimulatory capacity makes cultured LC (cLC) attractive candidates for vaccination purposes. In fact, evidence exists that LC are capable of inducing protective tumor (7, 8, 9, 10) and microbial (11, 12, 13) immunity. However, while LC in culture process and present the relevant immunizing Ag, they also have the ability to take up other undefined Ags being present in the culture medium, such as those of FCS. In vivo, this may result in the presentation of irrelevant peptides, causing activation of nonspecific and/or potentially autoreactive T cells and, thus, may preclude the use of LC for immunotherapy.

In this study, we have made an attempt to establish culture conditions that would allow the generation of immunostimulatory LC in the absence of serum and, by doing so, to gain further information about the factors mediating and securing LC maturation and survival.

BALB/c (H-2d), C3H/HeN (C3H) (H-2k), and C57BL/6 (H-2b) inbred mice of both sexes (6–12 wk old) were obtained from Charles River Wiga, Sulzfeld, Germany and bred and maintained in the Vienna International Research Cooperation Center animal facilities, Vienna, Austria. All animal procedures were approved by the Austrian Ministry of Science, Transportation, and Art (GZ 66.009/160-Pr/4/95).

RPMI 1640 supplemented with 25 mM HEPES, 50 mg/ml gentamicin, 2 mM l-glutamine, 0.1 mM nonessential amino acids, 1 mM sodium pyruvate, 50 mM 2-ME, and 1× antibiotic-antimycotic solution (Life Technologies, Grand Island, NY), either serum-free (SF-CM) or containing 10% heat-inactivated FCS (PAA Laboratories, Linz, Austria) (FCS-CM), was used. In some experiments, SF-CM was supplemented with 2% of the serum replacement product TCM (ICN Biomedicals, Eschwege, Germany) (TCM-CM), which is characterized by a low and defined protein content and the absence of growth factors, cytokines, and steroid hormones.

Ears from BALB/c and C3H mice were washed in 70% ethanol, dried, split, placed dermal side down on a 1% trypsin-PBS solution, and incubated for 35 (thin ear side) or 45 min (cartilage ear side) at 37°C. Epidermal sheets were peeled from the underlying dermis, floated in RPMI 1640 medium (with or without FCS) containing DNase (200 μg/ml DNase I; Sigma, St. Louis, MO) and mechanically agitated over a stainless steel mesh. Resulting cell suspensions were filtered through a cell strainer (70 μm; Falcon, Lincoln Park, NJ), washed three times, and then cultured (1.5 × 106 cells/ml) in SF- or FCS-CM in 75-cm2 flasks (Costar, Cambridge, MA) at 37°C. After 3 days of culture, nonadherent EC were harvested by vigorous pipetting, and dead cells were largely eliminated by density gradient centrifugation (Lympholyte-M; Cedarlane Laboratories, Hornby, Ontario, Canada). Viability of both SF- and FCS-EC was >80% (n = 13) as determined by trypan blue exclusion. After phenotypic analysis of small aliquots, the two populations of cultured EC were adjusted to equal numbers of LC (SF-cLC, FCS-cLC) and used as stimulator cells of in vitro and in vivo immune responses.

The CTLL-2 cell line obtained from the American Type Culture Collection (ATCC, Manassas, VA; TIB 214) was maintained in FCS-CM supplemented with 1% culture supernatant of the IL-2-producing murine myeloma cell line X63Ag8–653, kindly provided by Dr. F. Melchers (Basel Institute of Immunology, Basel, Switzerland). The OVA-specific T cell hybridoma E8 was provided by Dr. N. Romani (Department of Dermatology, University of Innsbruck Medical School, Innsbruck, Austria).

FITC-conjugated mAbs RM4-4 (anti-CD4), M1/69 (anti-CD24), 7D4 (anti-CD25), 3/23 (anti-CD40), IM7.8.1 (anti-CD44), 30F11.1 (anti-CD45), RA3-6B2 (anti-CD45R/B220), 3E2 (anti-CD54), 16-10A1 (anti-CD80), and GL1 (anti-CD86) were purchased from PharMingen (San Diego, CA). Tissue culture supernatant of the hybridoma NLDC-145 (anti-DEC-205) was obtained from Serotec (Oxford, U.K.). Hybridomas GK1.5 (anti-CD4, TIB 207), 53-6.72 (anti-CD8, TIB 105), J11d.2 (anti-CD24, TIB 183), IM7.8.1 (anti-CD44, TIB 235), M5/114.15.2 (anti-I-Ab,d,q&I-Ed,k, TIB 120), and RA3-3A1/6.1 (anti-CD45R/B220, TIB 146) were obtained from ATCC. Hybridomas 3C7 (anti-CD25) and 2B6.2D8 (anti-CD4) were kindly provided by Dr. E. M. Shevach (National Institute of Allergy and Infectious Diseases, Bethesda, MD), 3.168.81 (anti-CD8) by Dr. H. R. MacDonald (Ludwig Institute for Cancer Research, Epalinges, Switzerland), and 145-2C11 (anti-CD3ε) by Dr. J. A. Bluestone (University of Chicago, Chicago, IL). Anti-cytokeratin 14 mAb (anti-K14, clone LL002) was obtained from BioGenex (San Ramon, CA), phycoerythrin-conjugated M5/114.15.2 (anti-I-Ab,d,q&I-Ed,k) mAb was purchased from Boehringer Mannheim, Vienna, Austria. For the detection of nonlabeled mAb anti-K14, the FITC-conjugated second step reagent goat F(ab′)2 anti-mouse IgG (H+L) (Caltag, San Francisco, CA) was used. Some of the mAb used were purified from supernatants of the corresponding hybridomas, and protein concentrations were adjusted to 1 mg/ml before they were FITC conjugated.

For two-color analyses, cells (3 × 105/sample) were resuspended in cold PBS/1%FCS/0.1%NaN3 and serially incubated with FITC-conjugated mAb directed against selected mouse Ags and phycoerythrin-conjugated anti-MHC class II mAb (30 min, 4°C). For control purposes, cells were incubated with isotype-matched mAbs. Stained cells were analyzed using a FACScan flow cytometer equipped with LYSYS II software (Becton Dickinson, Mountain View, CA). Dead cells were excluded by 7-aminoactinomycin D (Sigma) uptake.

For the analyses of cytoplasmic Ags, cells were fixed with paraformaldehyde (0.25% in PBS) for 10 min at room temperature, washed and permeabilized with saponin (0.3% saponin in PBS/1%FCS, Sigma) for 10 min at room temperature. After washing, cells were incubated with the nonconjugated first-step Ab, washed again, and exposed to an appropriately diluted FITC-conjugated second-step reagent.

T cells were prepared from mesenteric lymph nodes using Ab- and C-mediated lysis as described previously (14). Briefly, cell suspensions were passed through nylon wool columns, and nonadherent cells were treated with a mixture of the following mAbs: 3C7, 7D4, J11d.2, M5/114.15.2, IM7.8.1, RA3-3A1/6.1 for 30 min at 4°C. Subsequently, cells were incubated with Low-tox-M rabbit C (Cedarlane Laboratories) for 45 min at 37°C. Of these cells, 97 to 99% were CD3+ as determined by flow cytometry. To purify CD4+ or CD8+ T cells, mAb 3.168.81 or GK1.5 plus 2B6.2D8 were added to the Ab mixture. The purity of both CD4+ and CD8+ T cells was 97 to 98%. All proliferation assays were performed in the presence of FCS-CM in 96-well round-bottom microtiter plates. SF- and FCS-cLC (103–104/well) were cultured with either purified unfractionated T cells, CD4+ T cells, or CD8+ T cells (105/well) and, for control purposes, alone. In some experiments, SF- and FCS-cLC were cocultured with naive, syngeneic CD4+ or CD8+ T cells in the presence of either Con A (Pharmacia Biotech, Uppsala, Sweden; 2.5 μg/ml and 5 μg/ml for CD4+ and CD8+ T cells, respectively) or soluble anti-CD3ε mAb (clone 145-2C11, 0.1 μg/ml and 1 μg/ml for CD4+ and CD8+ T cells, respectively). Proliferation was assessed by [3H]TdR (Amersham, Arlington Heights, IL) incorporation (37 kBq/well) added during the final 12 h of culture. Data are expressed as mean cpm ± SD of triplicate cultures.

For secondary T cell proliferation assays, naive T cells (H-2b, 2 × 106/well) were first stimulated with allogeneic FCS-cLC (H-2d, 105/well) in 24-well culture plates. Four days later, cells were harvested, washed and rested for 2 days in FCS-CM. Following density gradient (Lympholyte-M) centrifugation, dead cells were removed, and viable cells (105/well) were restimulated with SF- or FCS-cLC (H-2d, 103/well). Proliferation was determined after 2, 3 and 4 days of culture.

The processing activity of SF- and FCS-LC (H-2k) was measured using the OVA-specific, MHC class II (I-Ek)-restricted T cell hybridoma E8 (15). The activation of hybridoma cells was determined by measuring IL-2 production in a bioassay using the IL-2-dependent CTLL-2 cell line (15, 16). Briefly, freshly prepared SF- and FCS-EC (1.5 × 106 cells/ml) were pulsed with OVA (0–2 mg/ml, Sigma) for 20 h in 75-cm2 culture flasks. Thereafter, nonadherent cells were harvested, extensively washed, and enriched for LC by density gradient centrifugation. The number of stimulator cells in both SF- and FCS-EC was adjusted to equal numbers of LC (SF-LC, FCS-LC). OVA-pulsed and, for control purposes, unpulsed SF- and FCS-LC (5 × 103/well) were cultured either with E8 hybridoma cells (105/well) or alone in 96-well flat-bottom microtiter plates. Supernatants were harvested after 24 h and added to the CTLL-2 cells (6 × 103/well) at a final dilution of 1:3. After incubation for 36 h at 37°C, [3H]TdR incorporation was determined as described for T cell proliferation assays. In other experiments, SF- and FCS-cLC (5 × 103/well) were cocultured with E8 hybridoma cells (105/well) in the presence of graded concentrations of OVA (0–1250 μg/ml). Supernatants were harvested after 24 h and the IL-2 content was assessed as described.

SF- and FCS-EC were prepared and cultured as described above. At 10, 24, 48, and 72 h, supernatants were harvested and stored at −20°C until use. Cytokine concentrations were determined by ELISA according to the manufacturer’s instructions for GM-CSF, IL-1α (Endogen, Cambridge, MA), and TNF-α (Genzyme, Cambridge, MA).

SF-, TCM- and FCS-cLC were washed twice in HBSS without Ca2+ and Mg2+ and then incubated (107 cells/ml) with 2,4,5-trinitrobenzenesulfonic acid (TNBS, Sigma) at a concentration of 5 mM (pH 7.2) for 10 min at 37°C. Cells were then washed three times in a 10-fold volume excess of SF and 10% FCS-containing RPMI 1640 medium for SF-cLC/TCM-cLC and FCS-cLC, respectively. Trypan blue exclusion revealed >70% viable cells (n = 5).

BALB/c mice were anesthetized i.p. with tribromoethanol (2.5% in isotonic saline, 350 μl per mouse, Aldrich, Steinheim, Germany). As a positive control, mice were painted on dry shaved abdominal skin with 50 μl of 2% trinitrochlorobenzene (TNCB; TCI, Tokyo Kasei, Tokyo, Japan) in a 4:1 acetone and olive oil carrier solution (epicutaneous control group). Other groups of mice were injected with syngeneic, trinitrophenyl (TNP)-modified or unmodified SF-, TCM- and FCS-cLC (graded numbers: 103-104 LC/mouse) in 100 μl HBSS with Ca2+ and Mg2+ s.c. into the tail base. Five days later, all groups of mice including a naive group (mice receiving ear challenge only) were challenged on the dorsal and ventral surfaces of their right ears with 20 μl of 0.5% TNCB. Ear thickness was measured before challenge and at 24, 48, and 72 h after challenge with an engineer’s micrometer (Hahn und Kolb, Stuttgart, Germany). The value measured before challenge was subtracted from the value assessed after challenge. Experimental groups consisted of four mice each. Ordinary one-way ANOVA was used to compare experimental and control groups.

Single cell suspensions, containing 2 to 3% MHC class II+ cells, were prepared from mouse ear epidermis and cultured for 3 days either in SF- or FCS-containing medium. In both cultures, cells with cytoplasmic extensions were already detectable at 24 h, and their dendritic morphology was more pronounced after 48 h of culture (Fig. 1). After 3 days, nonadherent EC were harvested from both cultures, and cell numbers and viabilities were assessed and compared. While the cellular recoveries from SF and FCS cultures were 48% and 71% (BALB/c) or 66% and 79% (C3H), respectively, the numbers of viable EC were essentially equal under both culture conditions in both mouse strains. Among viable EC recovered, the percentage of dendritically shaped cells in SF cultures was approximately half of that encountered in FCS cultures (Table I). Immunofluorescence staining revealed that these dendritically shaped cells, but not other EC, expressed high levels of MHC class II molecules and, thus, represent LC (data not shown). When we cultured the EC in RPMI 1640 supplemented with the serum replacement product TCM, recovery and viability of both LC and keratinocytes were not different from those observed in SF cultures, suggesting that TCM does not provide the survival factor(s) for LC or keratinocytes (data not shown).

FIGURE 1.

Morphology of LC cultured with and without FCS. EC from C3H mice were prepared as described in Materials and Methods and cultured for 48 h under SF (A) and FCS (B) conditions. Cells with a dendritic morphology (LC) are indicated by arrows. Magnification ×100

FIGURE 1.

Morphology of LC cultured with and without FCS. EC from C3H mice were prepared as described in Materials and Methods and cultured for 48 h under SF (A) and FCS (B) conditions. Cells with a dendritic morphology (LC) are indicated by arrows. Magnification ×100

Close modal
Table I.

Cellular recoveries from SF- and FCS-EC cultures after 3 daysa

MiceCulture ConditionsNumber of Isolated Cells/EarEC Numbers Before CultureCellular Recovery After 3 Days of CultureNumbers of Viable EC After 3 Days of CultureNumbers of Viable LC After 3 Days of Culture
BALB SF 1.4–1.8 × 106 31 × 106 (100%) 14.9 × 106 (48%) 4.9 × 106 (15.8%) 4 × 105 (1.3%) 
 FCS 1.8–2.1 × 106 31 × 106 (100%) 22.1 × 106 (71%) 5.2 × 106 (16.8%) 9 × 105 (2.9%) 
C3H SF 1.2–1.5 × 106 30 × 106 (100%) 19.9 × 106 (66%) 3.6 × 106 (12.0%) 4 × 105 (1.3%) 
 FCS 1.5–1.9 × 106 30 × 106 (100%) 23.6 × 106 (79%) 3.8 × 106 (12.7%) 8 × 105 (2.7%) 
MiceCulture ConditionsNumber of Isolated Cells/EarEC Numbers Before CultureCellular Recovery After 3 Days of CultureNumbers of Viable EC After 3 Days of CultureNumbers of Viable LC After 3 Days of Culture
BALB SF 1.4–1.8 × 106 31 × 106 (100%) 14.9 × 106 (48%) 4.9 × 106 (15.8%) 4 × 105 (1.3%) 
 FCS 1.8–2.1 × 106 31 × 106 (100%) 22.1 × 106 (71%) 5.2 × 106 (16.8%) 9 × 105 (2.9%) 
C3H SF 1.2–1.5 × 106 30 × 106 (100%) 19.9 × 106 (66%) 3.6 × 106 (12.0%) 4 × 105 (1.3%) 
 FCS 1.5–1.9 × 106 30 × 106 (100%) 23.6 × 106 (79%) 3.8 × 106 (12.7%) 8 × 105 (2.7%) 
a

SF- and FCS-EC were prepared from BALB/c and C3H mice and cultured for 3 days. Thereafter, nonadherent EC were harvested, washed, and counted in a hemocytometer. Viability of EC was determined with trypan blue. LC were enumerated by their dendritic morphology, and their numbers were calculated from the numbers of viable EC. Viability of both SF- and FCS-cLC was >98%. Results are representative of three experiments for each mouse strain.

In the next series of experiments, we investigated whether the isolation and culture of EC in the absence of FCS influences the phenotype of LC. Flow cytometry analyses revealed high expression of CD24, modest levels of CD45 and MHC class II, but no CD80 or CD86 Ags on LC freshly isolated in the presence (FCS-LC) or in the absence (SF-LC) of FCS (Fig. 2,A). After 24 h of culture, both SF- and FCS-LC up-regulated MHC class II Ags and expressed moderate levels of CD40, CD80, and CD86 Ags (data not shown). After 3 days of culture, EC were enriched for LC by density gradient centrifugation, resulting in 20 to 38% (SF conditions) and 23 to 50% (FCS conditions) of cells exhibiting the morphologic features of LC (n = 13). By flow cytometry, both SF- and FCS-cLC exhibited a similar increase in the cell surface expression of MHC class II, CD25, and of the costimulatory molecules CD40, CD54, CD80, CD86, as well as a similar decrease in the expression of CD24 and CD45 molecules (Fig. 2,B). Identical phenotypic changes were observed in 3-day-cultured SF- and FCS-LC from BALB/c mice (data not shown). In contrast to FCS-cLC, SF-cLC expressed slightly less of DEC-205 and CD44 molecules in C3H mice (Fig. 2 B) and considerably less of these molecules in BALB/c mice (data not shown). We conclude that serum-free EC culture conditions promote a phenotypic metamorphosis in the LC population, similar to that induced by a serum-containing milieu, but are somewhat impaired in their capacity to maintain LC viability.

FIGURE 2.

Phenotypic comparison of LC isolated and cultured under SF and FCS conditions. A, Freshly prepared EC (C3H) were stained and analyzed by flow cytometry as described in Materials and Methods. B, After 3 days of culture, nonadherent EC were enriched for LC by density gradient, washed, stained, and analyzed. In all experiments, stained cells were gated for viable cells, and 7,000 events per sample were acquired. Dot blots represent results of two (A) or three (B) independent experiments.

FIGURE 2.

Phenotypic comparison of LC isolated and cultured under SF and FCS conditions. A, Freshly prepared EC (C3H) were stained and analyzed by flow cytometry as described in Materials and Methods. B, After 3 days of culture, nonadherent EC were enriched for LC by density gradient, washed, stained, and analyzed. In all experiments, stained cells were gated for viable cells, and 7,000 events per sample were acquired. Dot blots represent results of two (A) or three (B) independent experiments.

Close modal

Using the allogeneic MLR as read-out system, we first investigated whether SF-cLC are capable of inducing a primary T cell response. Repeated experiments showed that, over the entire range of stimulator cell concentrations (102–104 LC/well), SF-LC isolated and cultured from two different mouse strains (C3H, BALB/c) induced a vigorous T cell response comparable to that evoked by FCS-cLC (Fig. 3, A and B). The primary T cell alloresponse induced by both SF- and FCS-cLC peaked on days 3 to 4 and markedly declined on day 5 of culture (data not shown). We further observed that SF-cLC are as potent stimulators of allogeneic T cells in a secondary MLR as are FCS-cLC, both inducing a peak of the response on days 2 to 3 (data not shown).

FIGURE 3.

Comparative assessment of the immunostimulatory potential of SF- and FCS-cLC in vitro. A and B, SF- and FCS-cLC (A, 103/well; B, graded numbers) from C3H (A) and BALB/c (B) mice were cultured with naive, allogeneic T cells (C57BL/6, 105/well) in 96-well round-bottom microtiter plates. Proliferative responses were determined on days 3 (A) and 4 (A and B) of culture. Data are representative of three experiments. T cells as well as LC did not proliferate when cultured with medium alone (<400 cpm). C and D, SF- and FCS-cLC (C, 5 × 103/well; D, graded doses) from C3H mice were cultured with naive, syngeneic CD4+ T cells (105/well) in 96-well round-bottom microtiter plates. In C, proliferative responses were determined on days 4 to 6 and in D, on day 5 of culture. CD4+ T cells and cLC incubated with medium alone did not proliferate (<500 cpm). Results are representative of three experiments. E and F, SF- and FCS-cLC (103/well) from C3H mice were cocultured with naive, syngeneic CD4+ or CD8+ T cells (105/well) in the presence of either Con A (2.5 μg/ml and 5 μg/ml for CD4+ and CD8+ T cells, respectively) or soluble anti-CD3ε mAb (clone 145-2C11, 0.1 μg/ml and 1 μg/ml for CD4+ and CD8+ T cells, respectively) in 96-well round-bottom microtiter plates. Proliferative responses were determined on days 2 to 4 of culture. Data shown are representative of two experiments. CD4+ and CD8+ T cells did not respond to any of the stimuli when cultured in the absence of LC (CD4+ T cells + Con A = 126 ± 12 cpm, CD4+ T cells + anti-CD3 = 269 ± 50 cpm, CD8+ T cells + Con A = 119 ± 34 cpm, CD8+ T cells + anti-CD3 = 415 ± 187 cpm).

FIGURE 3.

Comparative assessment of the immunostimulatory potential of SF- and FCS-cLC in vitro. A and B, SF- and FCS-cLC (A, 103/well; B, graded numbers) from C3H (A) and BALB/c (B) mice were cultured with naive, allogeneic T cells (C57BL/6, 105/well) in 96-well round-bottom microtiter plates. Proliferative responses were determined on days 3 (A) and 4 (A and B) of culture. Data are representative of three experiments. T cells as well as LC did not proliferate when cultured with medium alone (<400 cpm). C and D, SF- and FCS-cLC (C, 5 × 103/well; D, graded doses) from C3H mice were cultured with naive, syngeneic CD4+ T cells (105/well) in 96-well round-bottom microtiter plates. In C, proliferative responses were determined on days 4 to 6 and in D, on day 5 of culture. CD4+ T cells and cLC incubated with medium alone did not proliferate (<500 cpm). Results are representative of three experiments. E and F, SF- and FCS-cLC (103/well) from C3H mice were cocultured with naive, syngeneic CD4+ or CD8+ T cells (105/well) in the presence of either Con A (2.5 μg/ml and 5 μg/ml for CD4+ and CD8+ T cells, respectively) or soluble anti-CD3ε mAb (clone 145-2C11, 0.1 μg/ml and 1 μg/ml for CD4+ and CD8+ T cells, respectively) in 96-well round-bottom microtiter plates. Proliferative responses were determined on days 2 to 4 of culture. Data shown are representative of two experiments. CD4+ and CD8+ T cells did not respond to any of the stimuli when cultured in the absence of LC (CD4+ T cells + Con A = 126 ± 12 cpm, CD4+ T cells + anti-CD3 = 269 ± 50 cpm, CD8+ T cells + Con A = 119 ± 34 cpm, CD8+ T cells + anti-CD3 = 415 ± 187 cpm).

Close modal

In contrast to the equal allostimulatory potency of SF- and FCS-cLC, the ability of SF-cLC to stimulate proliferation of naive, syngeneic CD4+ T cells was weaker than that of FCS-cLC (Fig. 3, C and D). The differences in the proliferative responses elicited by SF-cLC vs FCS-cLC were particularly pronounced at high stimulator cell numbers (Fig. 3,D). The T cell proliferation peaked on days 4 to 5, for both SF- and FCS-cLC, and decreased on day 6 (Fig. 3 C).

We next investigated whether SF-cLC can function as stimulators in polyclonal T cell responses. As can be seen in Figure 3, E and F, we found that both SF- and FCS-cLC are equally potent accessory cells in Con A- and anti-CD3-driven T cell activation. Not only the magnitude but also the kinetics of the proliferative response induced by SF-cLC compared with those induced by FCS-cLC, with a peak occurring on day 3 and on days 2 to 3 for CD4+ and CD8+ T cells, respectively. In conclusion, these data show that SF-cLC, similar to FCS-cLC, are exceedingly potent stimulators of naive, allogeneic T cells and function as powerful accessory cells in polyclonal T cell responses. However, the syngeneic MLR induced by SF-cLC is considerably weaker than that evoked by FCS-cLC.

While LC are very efficient in processing native protein Ags when freshly isolated from the epidermis, they lose this ability during in vitro culture (5). To determine whether LC, when isolated and pulsed with a soluble protein Ag in the absence of FCS, would still have the ability to process and present this Ag, we pulsed freshly isolated SF-EC as well as FCS-EC with OVA for 20 h, enriched them for LC by density gradient centrifugation (10–20% MHC class II+ cells of both SF- and FCS-EC, n = 4), and cocultured them with the OVA-specific, MHC-class II-restricted T cell hybridoma E8 for 24 h. Figure 4 shows that both OVA-pulsed SF- and FCS-LC are equally potent stimulators of hybridoma cells, as assessed by their IL-2 production. When 3-day-cultured LC were tested for their Ag-processing capacity, we found that neither SF- nor FCS-OVA-pulsed-LC (concentration range: 0–1250 μg/ml) were able to induce IL-2 production in E8 hybridoma cells (data not shown). We conclude that LC freshly isolated without FCS are able to process native protein Ag as effectively as LC isolated in the presence of FCS. The Ag-processing defect exhibited by both SF- and FCS-cLC strongly suggests that a similar level of LC maturation has occurred under both culture conditions.

FIGURE 4.

Ag-processing capacity of LC isolated under SF conditions. Freshly prepared SF- and FCS-EC (C3H) were cultured in the presence of OVA (1 mg/ml) for 20 h. The nonadherent cells were harvested, washed, and enriched for LC by density gradient centrifugation. SF- and FCS-LC-OVA (5 × 103/well) were cocultured with E8 hybridoma cells (105/well) in 96-well flat-bottom microtiter plates for 24 h. Supernatants were harvested and added to IL-2-dependent CTLL-2 cells at a final dilution of 1:3. CTLL-2 cells were cultured for 36 h, and [3H]TdR was added to the cultures 12 h before harvest. Data are representative of three experiments. E8 hybridoma cells failed to produce IL-2 when cultured with unpulsed LC (E8 + SF-LC = 2374 ± 360 cpm, E8 + FCS-LC = 1853 ± 215 cpm).

FIGURE 4.

Ag-processing capacity of LC isolated under SF conditions. Freshly prepared SF- and FCS-EC (C3H) were cultured in the presence of OVA (1 mg/ml) for 20 h. The nonadherent cells were harvested, washed, and enriched for LC by density gradient centrifugation. SF- and FCS-LC-OVA (5 × 103/well) were cocultured with E8 hybridoma cells (105/well) in 96-well flat-bottom microtiter plates for 24 h. Supernatants were harvested and added to IL-2-dependent CTLL-2 cells at a final dilution of 1:3. CTLL-2 cells were cultured for 36 h, and [3H]TdR was added to the cultures 12 h before harvest. Data are representative of three experiments. E8 hybridoma cells failed to produce IL-2 when cultured with unpulsed LC (E8 + SF-LC = 2374 ± 360 cpm, E8 + FCS-LC = 1853 ± 215 cpm).

Close modal

Although the epidermal environment under SF conditions apparently provides the stimuli necessary for the phenotypic and functional maturation of LC, the yield of LC after 3 days of culture in the absence of serum was quite low. Reasoning that the presence of FCS might influence the viability of EC and, therefore, the quality of the cytokine milieu, we compared the cytokine pattern of SF and FCS cultures by testing EC supernatants for the presence of keratinocyte-derived cytokines known to promote survival and maturation of LC (17, 18, 19, 20). While comparable amounts of GM-CSF were measured in both culture supernatants during the entire culture period, the levels of IL-1α and TNF-α in the supernatants of SF-EC cultures were less than those present in FCS-EC culture fluids (Fig. 5). Because basal keratinocytes are the predominant source of these two cytokines in the epidermis (21, 22), our finding suggested that this particular keratinocyte subpopulation may not survive in the absence of FCS. To investigate this, we stained EC from SF and FCS cultures with a mAb directed against K14, an Ag that is expressed only in basal keratinocytes. We found no difference in K14 expression between freshly prepared (Fig. 6,A) and one-day cultured SF- and FCS-EC (data not shown). By contrast, markedly reduced numbers of K14+ cells were found in both adherent and nonadherent cell populations harvested from SF-EC cultures on day 2 (data not shown) and day 3 (Fig. 6, B and C). These data suggest that the low LC yield obtained under SF culture conditions may be due to a relative loss/decrease of soluble mediators produced by basal keratinocytes.

FIGURE 5.

Comparison of the cytokine production of EC cultured under SF and FCS conditions. SF- and FCS-EC (C3H) were cultured for 10 to 72 h. Supernatants, harvested at 10, 24, 48, and 72 h were screened for the presence of GM-CSF (A), IL-1α (B), and TNF-α (C) by ELISA. Data are expressed in pg/ml as mean concentration ± SD of triplicates and are representative of two experiments.

FIGURE 5.

Comparison of the cytokine production of EC cultured under SF and FCS conditions. SF- and FCS-EC (C3H) were cultured for 10 to 72 h. Supernatants, harvested at 10, 24, 48, and 72 h were screened for the presence of GM-CSF (A), IL-1α (B), and TNF-α (C) by ELISA. Data are expressed in pg/ml as mean concentration ± SD of triplicates and are representative of two experiments.

Close modal
FIGURE 6.

Fate of basal keratinocytes in SF- and FCS-EC cultures. Freshly isolated (A) and 3-day-cultured adherent (B) or nonadherent (C) SF- and FCS-EC were washed, permeabilized, and stained with an anti-cytokeratin 14 mAb. Whereas closed profiles illustrate the reactivity with the anti-K14 mAb, open profiles represent an irrelevant, isotype-matched control mAb. Results are representative of three experiments.

FIGURE 6.

Fate of basal keratinocytes in SF- and FCS-EC cultures. Freshly isolated (A) and 3-day-cultured adherent (B) or nonadherent (C) SF- and FCS-EC were washed, permeabilized, and stained with an anti-cytokeratin 14 mAb. Whereas closed profiles illustrate the reactivity with the anti-K14 mAb, open profiles represent an irrelevant, isotype-matched control mAb. Results are representative of three experiments.

Close modal

Because LC play a critical role in the induction of the immune response to reactive haptens (23), we used the CHS model to determine whether SF-cLC are functional in vivo. Naive BALB/c mice were injected s.c. with syngeneic, TNP-modified or unmodified SF-, TCM- and FCS-cLC and challenged with the same hapten 5 days later. Results obtained in three independent experiments showed that TNP-modified SF- and TCM-cLC induced a CHS response similar in magnitude and kinetics to that elicited by TNP-modified FCS-cLC and by epicutaneous sensitization (Fig. 7, A and B). The response peaked at 24 h and declined 48 h after challenge (Fig. 7,B). Remarkably, as few as 103 of TNP-modified SF-cLC were able to immunize for CHS with the same potency as equal numbers of TNP-modified FCS-cLC (Fig. 7 C). Similar results were obtained in C3H and C57BL/6 mice, although the magnitude of the CHS response was considerably lower than in BALB/c mice (data not shown). These data show that LC cultured under SF conditions are potent stimulators of cell-mediated immunity in vivo.

FIGURE 7.

Comparison of the in vivo sensitizing capacity of SF- and FCS-cLC. A, On day 0, BALB/c mice were either injected s.c. with syngeneic, TNP-modified or unmodified SF-, TCM- or FCS-cLC (104 cells/mouse), painted with TNCB, or received no treatment. On day 5, all mice were challenged with TNCB on both sides of the right ear. Ear thickness was measured 24 h later. B, Kinetics of the CHS response induced by 104 TNP-modified-cLC were determined 24, 48, and 72 h after ear challenge. C, Ear swelling responses induced by graded numbers of TNP-modified-cLC (103–104 LC/mouse) were measured 24 h after ear challenge. *, p < 0.05; **, p < 0.01; ***, p < 0.001.

FIGURE 7.

Comparison of the in vivo sensitizing capacity of SF- and FCS-cLC. A, On day 0, BALB/c mice were either injected s.c. with syngeneic, TNP-modified or unmodified SF-, TCM- or FCS-cLC (104 cells/mouse), painted with TNCB, or received no treatment. On day 5, all mice were challenged with TNCB on both sides of the right ear. Ear thickness was measured 24 h later. B, Kinetics of the CHS response induced by 104 TNP-modified-cLC were determined 24, 48, and 72 h after ear challenge. C, Ear swelling responses induced by graded numbers of TNP-modified-cLC (103–104 LC/mouse) were measured 24 h after ear challenge. *, p < 0.05; **, p < 0.01; ***, p < 0.001.

Close modal

In this study, we have shown that LC cultured in the absence of FCS acquire a mature phenotype similar to that achieved under FCS conditions. Whether this phenotype corresponds to the terminally mature, IL-10-resistant phenotype of DC (24, 25) remains to be investigated. Our finding that SF-cLC express slightly lower levels of CD44 and DEC-205 molecules than FCS-cLC may be a consequence of reduced amounts of proinflammatory cytokines detected in SF cultures. Indeed, TNF-α has been shown to significantly up-regulate the expression of panCD44 on cultured LC (26). CD44 and DEC-205 molecules are significant for LC function, the first being important for their migratory and sensitizing capacity (27), and the latter being a receptor involved in their Ag-processing machinery (28). Interestingly, neither the Ag-processing and the immunostimulatory function nor the migratory capacity of SF-cLC were adversely affected by the lower expression of CD44 and DEC-205. It is not clear whether the low levels of these molecules expressed on SF-cLC sufficed to secure these functions or whether other molecules with similar properties became involved.

In unperturbed murine skin in vivo, EC constitutively produce low amounts of cytokines, particularly IL-1α, IL-7, TGF-β, GM-CSF, and TNF-α (29), which presumably contribute to the homeostasis of the skin. When the epidermal microenvironment is perturbed, i.e., when EC are dissociated by proteases and then cultured in serum-containing media, EC rapidly increase their cytokine production (e.g., IL-1α, GM-CSF, and TNF-α) and create a milieu that promotes LC survival and their maturation into potent APC (17, 29). We have shown here that the presence of serum is not necessarily required to fully accomplish LC maturation in vitro but helps to secure LC survival in culture. We also observed that the absence of FCS in EC cultures does not affect GM-CSF production by keratinocytes but leads to a decreased secretion of IL-1α and TNF-α, most likely as a consequence of the observed death of basal keratinocytes being the main source of these two cytokines (21, 22). Thus, our data imply that GM-CSF is an indispensable LC maturation factor but does not suffice to keep LC alive. Concerning the role of IL-1α and TNF-α in LC survival, we found that neither the IL-1α-rich SF supernatants of the basal keratinocyte cell line Pam 212 (30, 31) nor the addition of murine TNF-α (100 pg/ml) to SF-EC cultures was able to improve LC yields (A. Kočiková, unpublished observations). Whether other factors known to promote the differentiation and/or survival of LC/DC, such as TGF-β1 (32, 33, 34) or TNF-related activation-induced cytokine (TRANCE) (35), play a role in our experimental system remains to be investigated.

It is known that FCS consists of cytokines, growth factors, hormones, and vitamins (36), but many of its components are still poorly defined. To avoid the presence of xenogeneic proteins in EC cultures, we have established SF conditions and have shown here that, even in the absence of FCS, LC can mature into potent Ag-presenting and accessory cells. Interestingly, in contrast to their allostimulatory capacity, SF-cLC were less effective stimulators of naive, syngeneic CD4+ T cells than FCS-cLC. One explanation of this finding could be that the FCS constituents may structurally resemble self-proteins and, thus, may induce T cell activation when presented in the context of MHC molecules on LC. The absence of potentially cross-reactive FCS proteins in LC/DC cultures may be a benefit for their clinical applicability, because autoreactive responses could be avoided. To the best of our knowledge, this is the first report on the culture of LC under SF conditions and supports the efforts of other researchers who have used autologous serum/plasma or cytokines instead of FCS in DC culture media (33, 34, 37, 38, 39).

Our results show that freshly isolated SF- and FCS-LC are equally potent in their capacity to process a soluble protein Ag. Thus, our SF system should allow us to investigate the particular steps of the DC/LC Ag-processing machinery in vitro without any influence of foreign proteins/peptides. This may provide a clue about how to preferentially introduce peptides into either the MHC class I or class II processing pathway in the APC and, thus, may be useful in the development of new immunizing strategies. We have preliminary data indicating that, at higher doses of OVA, the SF-LC are even better Ag-processing cells than FCS-LC, perhaps because the MHC molecules of LC are more accessible for loading with specific peptides under SF conditions (A. Kočiková, unpublished observations). As a consequence, one may assume that the Ag dosing can be more accurately accomplished under SF than under FCS conditions. This is important in view of the findings that high doses of soluble Ag lead to the differentiation into Th1-like cells producing predominantly IFN-γ, while low doses of the same Ag can induce differentiation into Th2-like cells producing mainly IL-4 (40, 41). Thus, by modulating the amount of Ag under SF conditions, immune responses could be directed toward a Th1- or Th2-type in vivo. This might be useful for the therapy of diseases in which a particular T cell subpopulation is known to play a pathogenetically important role.

We thank Dr. M. Epstein for critically reading the manuscript, S. Olt for excellent technical assistance, and E. Berger for animal care.

1

This work was supported by a grant to A.E.B. from the Austrian Science Foundation (P10797-MED), and a grant from Bender and Co., Vienna, Austria.

3

Abbreviations used in this paper: DC, dendritic cell(s); C3H, C3H/HeN; CHS, contact hypersensitivity; CM, culture medium; EC, epidermal cell(s); GM-CSF, granulocyte-macrophage CSF; cLC, cultured Langerhans cell(s); SF, serum-free; TNCB, trinitrochlorobenzene; TNP, trinitrophenyl.

1
Steinman, R. M..
1991
. The dendritic cell system and its role in immunogenicity.
Annu. Rev. Immunol.
9
:
271
2
Streilein, J. W., P. R. Bergstresser.
1984
. Langerhans cells: antigen-presenting cells of the epidermis.
Immunobiology
168
:
285
3
Stingl, G., C. Hauser, E. Tschachler, V. Groh, K. Wolff.
1989
. The immune functions of epidermal cells. D. A. Norris, ed.
Immune Mechanisms in Cutaneous Disease
3
Marcel Dekker, Inc, New York and Basel.
4
Romani, N., M. Witmer-Pack, M. Crowley, S. Koide, G. Schuler, K. Inaba, R. M. Steinman.
1991
. Langerhans cells as immature dendritic cells. G. Schuler, ed.
Epidermal Langerhans Cells
191
CRC Press, Boca Raton.
5
Romani, N., S. Koide, M. Crowley, M. Witmer-Pack, A. M. Livingstone, C. G. Fathman, K. Inaba, R. M. Steinman.
1989
. Presentation of exogenous protein antigens by dendritic cells to T cell clones: intact protein is presented best by immature, epidermal Langerhans cells.
J. Exp. Med.
169
:
1169
6
Schuler, G., R. M. Steinman.
1985
. Murine epidermal Langerhans cells mature into potent immunostimulatory dendritic cells in vitro.
J. Exp. Med.
161
:
526
7
Grabbe, S., S. Bruvers, R. L. Gallo, T. L. Knisely, R. Nazareno, R. D. Granstein.
1991
. Tumor antigen presentation by murine epidermal cells.
J. Immunol.
146
:
3656
8
Cohen, P. J., P. A. Cohen, S. A. Rosenberg, S. I. Katz, J. J. Mulé.
1994
. Murine epidermal Langerhans cells and splenic dendritic cells present tumor-associated antigens to primed T cells.
Eur. J. Immunol.
24
:
315
9
Cavanagh, L. L., R. Sluyter, K. G. Henderson, R. StC. Barnetson, G. M. Halliday.
1996
. Epidermal Langerhans’ cell induction of immunity against an ultraviolet-induced skin tumour.
Immunology
87
:
475
10
Celluzzi, C. M., L. D. Falo, Jr.
1997
. Epidermal dendritic cells induce potent antigen-specific CTL-mediated immunity.
J. Invest. Dermatol.
108
:
716
11
Moll, H., H. Fuchs, C. Blank, M. Röllinghoff.
1993
. Langerhans cells transport Leishmania major from the infected skin to the draining lymph node for presentation to antigen-specific T cells.
Eur. J. Immunol.
23
:
1595
12
Moll, H., S. Flohé, M. Röllinghoff.
1995
. Dendritic cells in Leishmania major-immune mice harbor persistent parasites and mediate an antigen-specific T cell immune response.
Eur. J. Immunol.
25
:
693
13
Sato, H., H. Kamiya.
1995
. Role of epidermal Langerhans’ cells in the induction of protective immunity to Schistosoma mansoni in guinea-pigs.
Immunology
84
:
233
14
Elbe, A., S. Schleischitz, D. Strunk, G. Stingl.
1994
. Fetal skin-derived MHC class I+, MHC class II dendritic cells stimulate MHC class I-restricted responses of unprimed CD8+ T cells.
J. Immunol.
153
:
2878
15
Gaspari, A. A., S. I. Katz.
1988
. Induction and functional characterization of class II MHC (Ia) antigens on murine keratinocytes.
J. Immunol.
140
:
2956
16
Gillis, S., M. M. Ferm, W. Ou, K. A. Smith.
1978
. T cell growth factor: parameters of production and a quantitative microassay for activity.
J. Immunol.
120
:
2027
17
Luger, T. A., S. Beissert, T. Schwarz.
1997
. The epidermal cytokine network. J. D. Bos, Jr, ed.
Skin Immune System (SIS): Cutaneous Immunology and Immunodermatology
2nd Ed.
271
CRC Press, Boca Raton/New York.
18
Koch, F., C. Heufler, E. Kämpgen, D. Schneeweiss, G. Böck, G. Schuler.
1990
. Tumor necrosis factor α maintains the viability of murine epidermal Langerhans cells in culture, but in contrast to granulocyte/macrophage colony-stimulating factor, without inducing their functional maturation.
J. Exp. Med.
171
:
159
19
Heufler, C., F. Koch, G. Schuler.
1988
. Granulocyte/macrophage colony-stimulating factor and interleukin 1 mediate the maturation of murine Langerhans cells into potent immunostimulatory dendritic cells.
J. Exp. Med.
167
:
700
20
Witmer-Pack, M. D., W. Olivier, J. Valinsky, G. Schuler, R. M. Steinman.
1987
. Granulocyte/macrophage colony-stimulating factor is essential for the viability and function of cultured murine epidermal Langerhans cells.
J. Exp. Med.
166
:
1484
21
Gatto, H., J. Viac, A. Reano, M. Charveron.
1991
. Levels of immunoreactive interleukin 1 α and β in basal, suprabasal and cornified epidermal cell subpopulations of normal human epidermis: comparison with keratinocytes cultured in defined medium.
Eur. J. Dermatol.
1
:
55
22
Piguet, P. F., G. E. Grau, C. Hauser, P. Vassalli.
1991
. Tumor necrosis factor is a critical mediator in hapten-induced irritant and contact hypersensitivity reactions.
J. Exp. Med.
173
:
673
23
Sullivan, S., P. R. Bergstresser, R. E. Tigelaar, J. W. Streilein.
1986
. Induction and regulation of contact hypersensitivity by resident, bone marrow-derived, dendritic epidermal cells: Langerhans cells and Thy-1+ epidermal cells.
J. Immunol.
137
:
2460
24
Enk, A. H., V. L. Angeloni, M. C. Udey, S. I. Katz.
1993
. Inhibition of Langerhans cell antigen-presenting function by IL-10: a role for IL-10 in induction of tolerance.
J. Immunol.
151
:
2390
25
Steinbrink, K., M. Wölfl, H. Jonuleit, J. Knop, A. H. Enk.
1997
. Induction of tolerance by IL-10-treated dendritic cells.
J. Immunol.
159
:
4772
26
Osada, A., H. Nakashima, M. Furue, K. Tamaki.
1995
. Up-regulation of CD44 expression by tumor necrosis factor-α is neutralized by interleukin-10 in Langerhans cells.
J. Invest. Dermatol.
105
:
124
27
Weiss, J. M., J. Sleeman, A. Renkl, H. Dittmar, C. C. Termeer, S. Taxis, N. Howells, M. Hofmann, G. Köhler, E. Schöpf, H. Ponta, P. Herrlich, J. C. Simon.
1997
. An essential role for CD44 variant isoforms in epidermal Langerhans cell and blood dendritic cell function.
J. Cell Biol.
137
:
1137
28
Jiang, W., W. J. Swiggard, C. Heufler, M. Peng, A. Mirza, R. M. Steinman, M. C. Nussenzweig.
1995
. The receptor DEC-205 expressed by dendritic cells and thymic epithelial cells is involved in antigen processing.
Nature
375
:
151
29
Kilgus, O., E. Payer, S. Schreiber, A. Elbe, R. Strohal, G. Stingl.
1993
. In vivo cytokine expression in normal and perturbed murine skin: analysis by competitive quantitative polymerase chain reaction.
J. Invest. Dermatol.
100
:
674
30
Matsue, H., P. D. Cruz, Jr, P. R. Bergstresser, A. Takashima.
1992
. Cytokine expression by epidermal cell subpopulations.
J. Invest. Dermatol.
99
:
42S
31
Marthinuss, J., L. Lawrence, D. Seiberg.
1995
. Apoptosis in Pam 212, an epidermal keratinocyte cell line: a possible role for bcl-2 in epidermal differentiation.
Cell Growth Differ.
6
:
239
32
Borkowski, T. A., J. J. Letterio, A. G. Farr, M. C. Udey.
1996
. A role for endogenous transforming growth factor β1 in Langerhans cell biology: the skin of transforming factor β1 null mice is devoid of epidermal Langerhans cells.
J. Exp. Med.
184
:
2417
33
Strobl, H., E. Riedl, C. Scheinecker, C. Bello-Fernandez, W. F. Pickl, K. Rappersberger, O. Majdic, W. Knapp.
1996
. TGF-β1 promotes in vitro development of dendritic cells from CD34+ hemopoietic progenitors.
J. Immunol.
157
:
1499
34
Riedl, E., H. Strobl, O. Majdic, W. Knapp.
1997
. TGF-β1 promotes in vitro generation of dendritic cells by protecting progenitor cells from apoptosis.
J. Immunol.
158
:
1591
35
Wong, B. R., R. Josien, S. Young Lee, B. Sauter, H.-L. Li, R. M. Steinman, Y. Choi.
1997
. TRANCE (tumor necrosis factor [TNF]-related activation-induced cytokine), a new TNF family member predominantly expressed in T cells, is a dendritic cell-specific survival factor.
J. Exp. Med.
186
:
2075
36
Maurer, H. R..
1992
. Towards serum-free, chemically defined media for mammalian cell culture. R. I. Freshney, Jr, ed.
Animal Cell Culture: A Practical Approach
2nd Ed.
15
IRL Press at Oxford University Press, New York.
37
Bender, A., M. Sapp, G. Schuler, R. M. Steinman, N. Bhardwaj.
1996
. Improved methods for the generation of dendritic cells from nonproliferating progenitors in human blood.
J. Immunol. Methods
196
:
121
38
Romani, N., D. Reider, M. Heuer, S. Ebner, E. Kämpgen, B. Eibl, D. Niederwieser, G. Schuler.
1996
. Generation of mature dendritic cells from human blood: an improved method with special regard to clinical applicability.
J. Immunol. Methods
196
:
137
39
Jonuleit, H., U. Kühn, G. Müller, K. Steinbrink, L. Paragnik, E. Schmitt, J. Knop, A. H. Enk.
1997
. Pro-inflammatory cytokines and prostaglandins induce maturation of potent immunostimulatory dendritic cells under fetal calf serum-free conditions.
Eur. J. Immunol.
27
:
3135
40
Hosken, N. A., K. Shibuya, A. W. Heath, K. M. Murphy, A. O’ Garra.
1995
. The effect of antigen dose on CD4+ T helper cell phenotype development in a T cell receptor-αβ-transgenic model.
J. Exp. Med.
182
:
1579
41
Constant, S., C. Pfeiffer, A. Woodard, T. Pasqualini, K. Bottomly.
1995
. Extent of T cell receptor ligation can determine the functional differentiation of naive CD4+ T cells.
J. Exp. Med.
182
:
1591