Human interdigitating dendritic cells directly stimulate CD40-activated naive B cells

1266 P. Bjorck et al. Eur. J. Immunol. 1997.27: 1266-1274

Pia Bjorck,

Yong-Jun Liu

Human interdigitating dendritic cells directly stimulate CD40-activated naive B cells Leopoldo Flores-Romo and

Schering-Plough, Laboratory for ~mmuno~ogica~ Research, dill^, France

Human interdigitating dendritic cells (IDC) were isolated from tonsils based on their CD40' lineage-negative expression in situ. Isolated IDC displayed a phe- notypic profile similar to that of IDC in tonsils and spleen in situ, characterized by high-level expression of major histocompatibility complex class 11, the co- stimulatory molecules B7.1 (CD80) and B7.2 (CD86), expression of the late DC maturation marker CD83, and no expression of CDla, CD13, or CD33. IDC also showed weak nonspecific esterase staining and had the ability to induce an allogeneic mixed lymphocyte reaction. In this study, we further show that in the presence of surrogate activated T cells in the form of CD40 ligation and IL-2, IDC enhance the proliferation of naive B cells and induce their differentiation into plasma cells producing IgM. Evidence for the anatomical co-localization of naive B cells and IDC in the T cell area together with the data obtained in vitro implies a role for IDC in the initiation of the extrafollicular reaction.

1 Introduction

Interdigitating dendritic cells (IDC) represent a subset of bone marrow-derived dendritic cells (DC) that are local- ized in the T cell areas of peripheral lymphoid organs [ l , 21. These cells can be found within the periarteriolar lymphoid sheaths (PALS) of spleen that originate from the blood precursors [3, 41, or within the paracortex of lymph nodes, and are derived from skin Langerhans cells migrat- ing as veiled cells via the afferent lymph [5]. DC isolated either from peripheral lymphoid tissues or generated from their precursors all display a remarkable capacity to stimulate naive T cells both in vivo and in vitro [2]. Several lines of evidence suggest that IDC within the T cell areas may play key roles in the initiation of T cell-dependent activation and differentiation of naive B cells, including the primary antigen-specific B cell activation that takes place in the T cell zones in close association with T cells and IDC [6], the observation that antigen-pulsed DC induce strong primary and secondary antibody responses upon injection in mice [7], and primary antibody responses generated from cell suspensions of mouse spleen cultures containing specific antigen, in which clustering of DC, helper T cells, and histocompatible B cells were observed [8]. Further- more, DC proved to be essential for single germinal center B cells from Peyer's patches to produce IgA in the presence of helper T cells [9].

[I 165221

Received December 4, 1996; in revised form February 11, 1997; accepted February 17, 1997.

Present address: L. Flores-Romo, M. D. Anderson Cancer Cen- ter, Houston, Texas, USA.

Correspondence: Pia Bjorck, Immunobiology and Cancer, Okla- homa Medical Research Foundation, 825 Northeast 13th Street, Oklahoma City, OK 73104, USA Fax: + 1-405-271 85 68; e-mail: [email protected]

Abbreviations: IDC: Interdigitating dendritic cells PALS: Periarteriolar lymphoid sheath

Key words: Interdigitating dendritic cells I CD40 I IgD-positive B cells

Thus we asked whether extrafollicular B cells can be activated by culturing B cells, clearly defined populations of IDC, and T cells or T cell signals; and whether IDC can directly interact with B cells, since it has been found that DC/B cell clusters can be isolated from rat thoracic duct lymph in vitro, and that these cells form aggregates within 2 h of culture (N. Kushnirn and G. MacPherson, personal communication).

To address these questions, we thoroughly characterized the phenotype of human DC subsets in situ in different compartments of tonsils, and further looked for similar subsets in spleen, lymph nodes, thymus, and skin. We show here that in contrast to the DC found in epidermal and dermal areas of skin, tonsils, and those found in the thymic medulla, IDC within the T cell areas of tonsils and the PALS of spleen co-express all important stimulatory molecules including CD40, B7.1 (CDSO), B7.2 (CD86), and the late DC maturation marker CD83 [lo]. Based on the immunohistological observations in situ, tonsillar IDC were isolated by FACS sorting. These IDC expressed no myeloid markers; thus, the phenotype of the IDC isolated here is distinct from the tonsillar DC described earlier [ll]. Upon CD40 activation, IDC further up-regulated all co- stimulatory molecules and were induced to secrete macrophage inflammatory protein (MIP)-la, MIP-1P, IL-8, and IL-12. In addition to stimulation of allogeneic CD4' T cells, IDC enhanced the proliferation and differentiation of CD40-activated naive B cells.

Our present study directly demonstrates that IDC isolated from the T cell areas of peripheral lymphoid organs play an important role in T cell-dependent primary B cell activation.

2 Materials and methods

2.1 Reagents

Antibodies to CDla (IgG1) was from Immunotech (Mar- seille Luminy, France) and CDla (IgG2b) from Becton Dickinson (Mountain View, CA), MHC class I1 (DR, IgG2b) from Immunotech. The CD40 antibody mAb89 (IgG1) was produced in our laboratory [12], as well as the

00 14-298019710505- 1266$17.50 + .5010 0 VCH Verlagsgesellschaft mbH, D-69451 Weinheim, 1997

Eur. J. Immunol. 1997.27: 1266-1274 Human IDC activate naive B cells 1267

CD80 (B7.1) mAb 104 [13] and an mAb to MHC class I1 (IgG1). Antibodies to CD86 (B7.2) were purchased from Ancell Corp. (Bayport, MN). The CD83 antibody HB15 (IgG2b) [ 101 was a kind gift from Dr. Thomas Tedder. Anti- bodies to human IgD (biotinylated) were from Sigma (St. Louis, MO). Recombinant IL-2 was from Amgen (Thousand Oaks, CA), and PHA from Murex Diagnostics (Dartfort, GB) was used at 0.1 pg/ml.

2.2 Tissue staining

Tissues from tonsil, spleen, lymph nodes, skin, and thymus were snap-frozen in liquid nitrogen and cryostat sections (8 pm) were cut and fixed in acetone for 20 min at 4°C. Sections were air-dried and stored at -80°C until used. Staining was performed on consecutive sections, which were incubated at room temperature for 1 h with primary antibodies diluted in PBS (Gibco, Paisley, Scotland). Sec- tions were washed and secondary antibodies (mouse anti- human IgGl or biotin-labeled anti-human IgG2b; Binding Site, Birmingham, GB) were added. After further incubation and washing, mouse alkaline phosphatase-anti- alkaline phosphatase complex (APAAP; Dako ah, Den- mark), or extravidin-peroxidase (Sigma) were added. Staining with the secondary antibodies was repeated followed by mouse APAAP and extravidin-peroxidase. APAAP was developed by Fast Blue (Sigma) and finally, 3-amino-9-ethylcarbazole (Sigma) was used to develop the extravidin-peroxidase. Slides were examined under a light microscope. At least 200 cells from each section were counted.

Staining for CD80 and CD86 were performed by a method developed by Munro et al. [14]. Briefly, tissue sections were incubated with antibodies for 1 h at room temperature, washed, and incubated with a horse biotinylated anti- mouse IgG (y chain-specific) mAb (Vector, Burlingame, CA) in 2 % human AB+ serum. After washing, biotin- avidin horseradish peroxidase complex in PBS was added (1: 150 dilution, Vector Elite). Finally, sections were washed and incubated with diaminobenzidine (DAB) tab- lets (Sigma) diluted in 0.1 mM Tris-buffered saline pH 7.6 including 1 mM imidazole (Merck, Darmstadt, Germany), 0.05% sodium azide and 0.3 YO hydrogen peroxide.

2.3 Purification of IDC from tonsil

IDC were purified from human tonsils obtained from pa- tients undergoing tonsillectomy. Tonsils were cut into frag- ments and further digested with DNase I and collagenase IV (Sigma) to obtain a single-cell suspension. Enzymes were added and the cell suspension was incubated with stirring for 20 min at 37 “C. The cells were resuspended in RPMI supplemented with gentamycin (gentilline; Schering-Plough) and big tissue pieces were allowed to set- tle to the bottom of the flask. The supernatant was collected and the nondigested material was subject to a sec- ond treatment with enzymes. Free cells were washed once in RPMI, layered on a 50 % Percoll (Pharmacia, Uppsala, Sweden) gradient and centrifuged at 900 x g for 20 min. Cells in the interphase were collected, washed, and depleted of B cells and macrophages by using Dynabeads (Dynal, AS, Norway): cells were incubated with antibo-

dies to CD14 and CD19 for 20 min at 4”C, washed and further incubated with Dynabeads for 15 min. Cells that bound to beads were removed with a Dynamagnet. The depletion step was repeated once, and the negative cell fraction was recovered and further purified by FACS sorting.

2.4 Cell sorting

Human IDC were sorted based on their expression of the CD40 antigen as revealed by tissue staining ([15] and here). Cells were labeled with a biotinylated antibody to CD40 (mAb89-biotin) and directly FITC-conjugated antibodies to CD3, CD19 (Immunotech), and CD20 (Becton Dickinson). Cells were further incubated with tricolor- labeled streptavidin (Caltag, San Francisco, CA) and sorted with a FACStar (Becton Dickinson) equipped with a 2-W argon laser.

Sorted cells were stained with mAb directly labeled with FITC and examined by a FACScan. The following antibodies were used: CD4, CD13 CD23 from Immunotech (Marseille Luminy, France), CDllb, CDllc, CD18

Table 1. Phenotype of FACS-sorted IDC”’

Expression Clone Source

T cell markers aPTCR - WT13 B D ~ ) CDla - SK9 BD CD2 - S 5.2 BD CD4 f 13B8.2 lmmunotech

CD19 - HD39 Dako kappa ++ polyclonal Ortho lambda ++ polyclonal Ortho

CDllb - Bear1 Dako CD21 - BL13 Immunotech CD35 - E l 1 PharMingen

CD13 k SJlDl Immunotech CD14 - MFP9 BD CD15 - MMA BD CD33 - D3HL60.251 Immunotech

CD16 - NKP1.5 BD CD23 - 9P25 Immunotech CD32 ++ IV-3 Medarex CD64 - 32.2 Medarex

CDllc + +c) KB90 Dako CD18 ++ MHM23 Dako

CD80 ++ BBl Ancell CD86 ++ Bu63 Ancell

CD83 ++ HB15a Dr. T. Tedder MHC class I +++ W6l32 Seralab. MHC class I1 + + + Tu39 PharMingen

B cell markers

Complement receptors

Myeloid markers

Fc receptors

Adhesion receptors

Activation markers

Miscellaneous

a) Re-analysis of IDC after FACS sorting using directly FITC- conjugated antibodies. The expression ranges from (-) negative, (+ weak, (+) modest, (++) intermediate to (+++) strong.

b) BD: Beckton Dickinson. c) On a 50% subset of the IDC.

1268 P. Bjorck et al. Eur. J. Immunol. 1997.27: 1266-1274

(Dako), CDla, CD2, CD14, CD15, CD16 from Becton Dickinson, CD35, anti-MHC class I1 from Pharmingen (San Diego, CA), anti-kappa, anti-lambda from Ortho (Paris, France), CD80 (B7.1) and CD86 (B7.2) from (Ancell, and CD32 from Medarex (Annandale, NJ) . FITC-labeled CD83 was kindly provided by Dr. T. Tedder

B cells (10000/well) were cultured in flat-bottom 96-well tissue culture plates (Falcon) either alone or together with CD40L L cells (2500/well), sorted IDC (10000/well), with or without 20 U/ml IL-2 in a final volume of 200 p1 for 6 days. ["]Thymidine incorporation was measured during the last 18 h of culture.

POI.

2.5 IDC culture

Mouse fibroblasts transfected with both CD32 and the ligand for human CD40 ([16], CD32/CD40L L cells) were used efficiently to trigger the IDC. L cells were irradiated with 7500 rad and seeded at 2500 cells/well in half-area, multiwell tissue culture plates (Costar, Cambridge, MA) in complete medium: RPMI 1640 supplemented with gentilline, 10% fetal calf serum (Gibco), 2 mM glutamine (Eurobio, Les Ulis, France), Hepes (Gibco), and 5 x M 2-mercaptoethanol. Sorted IDC were added at 2.5 x lo5 cells/ml in final volume of 100 pYwell. Cytospins were prepared either directly after FACS sorting or after incubation on L cells for 1 day and stained either with Giemsa or with antibodies to MHC class I1 (y2b subclass), or analyzed by FACS.

2.8 Cytokine ELISA

IDC were cultured on either nontransfected or CD32/ CD40L L cells at a concentration of 25000 IDC on 2500 L cells in 96-well half-area plates and supernatants were harvested after 1 or 3 days of culture. Detection of IFN-y and IL-4 were made as described earlier [17, 181. Plates were read at 405 nm. ELISA kits (R&D Systems, Minnea- polis, MN) were used for the detection of MIP-la, MIP- lp , RANTES, IL-8, and IL-12. Serial dilutions of supernatant or a calibrated standard were added and the ELISA was performed according to the manufacturer's instruc- tion. Plates were read at 450 and 570 nm.

2.9 Induction of differentiation into plasma cells

To examine the effect of IDC on the generation of plasma cells, IgD' B cells (150000/well) were cultured on irradiated CD32/CD40L L cells (50 000/well) with or without IDC (150 000/well) and 20 U/ml IL-2 for 7 days in 24-well plates (Linbro, Maclean, VA) at a final volume of 1 ml. Cells were washed and stained with FITC-labeled CD20 and PE-labeled CD38 for FACS analysis. Plasma Cells

2.6 TceUs

Peripheral blood CD4' T cells were obtained by magnetic bead using a Of mAb to CD8, CD143

CD16, CD197 CD409 CD567 and MHC population was usually 90-95 % CD4' and less than 2 yo

'I. The

positive for CD147 CDI99 and MHC class I1 (not shown). were determined as a CD20-, CD38* cell Dopulation 1161.

Irradiated IDC (3000 rad) were seeded in serial dilutions in round-bottom multiwell tissue culture plates (Nunc, Denmark). CD4'T cells were added at a concentration of 50 000 cells/well in a final volume of 200 pVwell. For deter- mination of proliferation, plates were incubated for 6 days and [3H]thymidine (Amersham, 1 yCi/well, 18.0 Ci/mmol) was added during the last 18 h of culture.

2.7 B cells

B cells were prepared from tonsils obtained after tonsillectomy. Briefly, tonsils were cut into pieces and passed through a stainless steel sieve to obtain a single-cell suspension. Cells were centrifuged over Ficoll-Paque (Phar- macia) and the interphase was collected. Negative selection of B cells was made by incubation with antibodies to CD2, CD3, CD4, CD8, CD14, CD16, CD56, and glyco- phorin for 30 min at 4°C. After washing, cells were further incubated with Dynabeads. B cells obtained after negative selection were more than 95% pure as determined by FACS analysis using antibodies to C19 and CD20 (not shown). IgD' B cells were obtained by incubation of total B cells with a biotinylated anti-IgD mAb (Sigma) for 15 min at 4°C followed by further incubation with streptavidin-conjugated MACS beads. The B cell suspension was passed twice through a MACS column. Positively selected IgD' B cells were recovered and were 98 YO pure as determined by flow cytometry (not shown).

Cytospins were made from the same cultires and fixid (n acetone. Antibodies to kappa and lambda light chains (both IgGl subclass) and MHC class I1 (IgG2b) were added and slides were incubated for 1 h at room temperature. After washing, anti-mouse IgGl and IgG2b (Binding Site) were added. After incubation and washing, mouse APAAP and streptavidin-labeled peroxidase were added. Finally, APAAP was developed by Fast Blue substrate (Sigma, yielding a blue color) and the peroxidase was developed by DAB (brown color).

2.10 Immunoglobulin production

IgD' B cells were cultured for 14 days on CD32/CD40L L cells together with IDC with or without 20 U/ml IL-2 and supernatants were harvested and frozen at -20°C until use. Detection of IgM and IgG was performed as described [19]. Briefly, microtiter plates (Nunc Maxisorp, PolyLabo, France) were coated with anti-IgM or -1gG mAb (Behring, Germany) overnight at 4°C in carbonate buffer pH 9.6. After washing, serial dilutions of supernatants or a standard were added, and plates were incubated overnight at 4 "C. Alkaline phosphatase-conjugated anti- IgM (Biosys, Compikge, France) or anti-IgG (Dako) were added and plates were further incubated overnight at 4 "C. Finally, p-nitrophenylphosphate substrate was added and plates were read at 405 nm.

Eur. J. Immunol. 1997.27: 1266-1274 Human IDC activate naive B cells 1269

Figure 1. DC subsets in tissues from tonsil, spleen, and thymus shown by double staining of sections using the APAAP and peroxidase method. (a-c) Antibo- dies are directed against CD40 (blue) and MHC class I1 (red). Cells expressing both markers show a purple color. In the T cell area of tonsils, most IDC co-express both markers (a) 2 0 0 ~ and (b) 400~ magnification. (c) In contrast, in the subepithelial area of the tonsils, D C express MHC class 11, whereas CD40 is weak or absent (1000~ magnification). (d) Ton- sillar IDC express high levels of B7.1 (CD80) and (e) B7.2 (CD86). (f-g) show double staining with antibodies to CD40 (blue) and CD83 (red). (f) In the T cell area of tonsil, IDC expressing only CD40 (red) or both CD40 and CD83 (purple) could be found. (8) The same subsets were found in the PALS of the spleen. (h) Thymic D C express strongly the CD40 antigen (blue) but little or no expression of CD83 (red) (GC: germinal center, MZ: mar- ginal zone, c: cortex, and M: medulla).

thus confirming previous reports [15]. In contrast, epidermal CDla+ Langerhans cells and CDla- dermal DC of tonsillar mucosa and of skin express low or no detectable CD40 (Fig. lc). Consistent with a previous report, thymic medullar DC also express high levels of CD40.

3 Results

3.1 Characterization of IDC and other DC subsets in s i tu in human lymphoid tissues

A common feature of virtually all cell types within the DC lincage is their high level of MHC class I1 expression, including epidermal Langerhans cells, dermal DC, migrat- ing veiled cells in the lymph, thymic DC, or IDC in the T cell zone. To specifically isolate IDC, it is necessary to distinguish them from other types of DC.

Double staining with anti-MHC class I1 and CD40 showed that IDC within the tonsillar T cell area (Fig. la, b) and the splenic PALS all co-expressed high levels of CD40,

B7.1 (CD80) has been shown to be highly expressed on human IDC [14, 201 and B7.2 (CD86) to be strongly expressed on mouse IDC [21]. In agreement with these reports, both B7.1 (CD80) and B7.2 (CD86) were found to be highly expressed on tonsillar (Fig. Id. e) and splenic IDC as well as on germinal center B cells. By contrast, the expression of these two co-stimulatory molecules were found to be negativenow on Langerhans cells and dermal DC from tonsillar mucosa and skin as well as on thymic DC (not shown).

1270 P. Bjorck et al. Eur. J . Immunol. 1997.27: 1266-1274

CD40

rl

C

8 CD83

subepithelial trafficking paracortical subepithelial trafficking paracortical subepithelial trafficking paracurtical area area area area area area area area area

Figure 2. Characterization of DC from different parts of the tonsil: double staining with antibodies to MHC class I1 together with mAb to CDla, CD40, or CD83 was performed on cryostat sections from tonsil; the number of DC in different parts of the tonsil were counted. Using the staining with MHC class I1 mAb as a reference, the number of DC positive for CDla, CD40, or CD83 was calcu- lated. At least 200 DC from each part of the section were counted. Data are the mean of calculations from three different tonsils. In (a), CDla is expressed on Langerhans DC in the subepithelial area. Fewer DC in the trafficking zone express this marker, and in the T cell area, CDla seems to be lost. (b) CD40 was virtually absent from Langerhans DC, but is expressed on most IDC in the T cell area. (c) CD83 was absent on DC in the subepithelial area, but was up-regulated on DC in the paracortical area. Only a fraction of the mature IDC express CD83.

a b C

before sorting after sorting

U v) v)

CD35 CD19-, CDZO-FITC

Figure 3. FACS sorting of IDC using FACStar Plus (Becton Dick- inson). Cells were sorted according to CD40 expression and lack of CD3, CD19, and CD20. (a) Size and (b) staining of IDC before sorting are shown. Gates were set as indicated. (c) Sorted IDC had a purity more than 97 % .

CD83 represents a late DC maturation marker [lo]. Dou- ble staining with anti-CD40 and anti-CD83 shows that a subset of IDC in tonsils (47-54 %, n = 3, Fig. If) and in spleen (Fig. lg) express CD83. By contrast, CD83 was not detectable on fetal thymic DC (Fig. lh) or other types of DC insitu.

To dissect further the phenotype of tonsillar DC, double stainings with anti-MHC class I1 and anti-CDla, CD40, or CD83 mAb were made and the number of single-versus

double-positive DC were estimated. Fig. 2 shows that DC in the subepithelial area are mainly CDla', but CD40- and CD83-. However, mature IDC in the T cell area showed no CDla but expresed CD40 and about 50% of these DC expressed CD83.

In summary, tonsillar and splenic IDC are characterized by their high-level expression of all three co-stimulatory molecules, CD40, B7.1 (CD80), and B7.2 (CD86). Only a subset of IDC express the late DC maturation marker CD83.

3.2 Isolation of tonsillar IDC

According to the immunohistological analysis, tonsillar IDC can be distinguished from epidermal Langerhans cells and dermal DC by their high CD40 expression. Thus, a multistep method was developed to purify tonsillar IDC: low-density tonsillar cells were prepared by centrifugation through a 50% Percoll gradient, depletion of CD19+ B cells and CD14+ monocytes by a magnetic bead method, and cell sorting by flow cytometry for CD40' CD3- CD19- CD20- cells (over 97% purity, Fig. 3). By MHC class I1 staining, the isolated cells displayed long, fine dendrites (Fig. 4a, b) and other features of typical mature DC such as the absence of nonspecific esterase activity (Fig. 4c) compared to a monocyte control (Fig. 4d) and high motil- ity (Fig. 4e, f), suggesting that the isolated cells were indeed mature DC.

Table 2. Isolated IDC produce chemokines and cytokines that can be further enhanced by stimulation through CD40")

Source

(pg/ml)

IL-4 IFN-y MIP-la MIP-@ RANTES IL-8 IL-12

IDC, 1 day on CD32/CD40L L cells < 19 4 156 600 360 431.2 5 930 (7.8 IDC, 3 days on CD32/CD40L L cells 419 4 156 2880 425 2510 23 540 125

IDC, 3 days on nontransfected L cells ND ND 300 <15.6 400 8 800 <7.8

L cells done 4 19 < 156 (31.2 <31.2 <7.8 415.6 ~ 3 1 . 2

a) ND: not done; data are representative of two to four different experiments.

Eur. J. Immunol. 1997.27: 1266-1274

FACS analysis revealed that the isolated CD40' cells all expressed the other two co-stimulatory molecules B7.1 (CD80) and B7.2 (CD86), and half of them express the late DC maturation marker CD83, thus corresponding to the phenotype of IDC identified on tissue sections (Ta- ble 1). Lack of expression of CD14, CD19, and ap TCR further indicated that the isolated population was pure and free of contaminating monocytes, B cells, or T cells (Ta- ble 1).

Human IDC activate naive B cells 1271

Figure 4. Characterization of isolated IDC: staining with antibodies to MHC class I1 of cyto- spin of IDC (a) directly after sorting and (b) after overnight culture on CD3UCD40L L cells. Cultured cells display long, delicate dendrites with beaded processes. (c) IDC were negative for nonspecific esterase (d) a monocyte control is strongly positive. (e, f) Cultured IDC move their dendrites quickly as shown by time-lapse photography ( 1 5 s interval).

In addition, isolated tonsillar IDC display several interest- ing surface features: they appear to express less CD4 compared to CD4' blood DC and germinal center DC [22,23]. Since CD4 was found to be down-regulated on blood DC upon their spontaneous maturation in vitro [24], it suggests that the isolated IDC are more mature DC. Further, they expressed the Fcy receptor CD32 (but not other Fc or complement receptors) and Ig x and h light chains. It suggests that IDC may carry immune complexes mainly by Fcy receptors. In contrast, follicular DC (FDC) mainly use complement receptors CRl, CR2 and CD3, and to a lesser


3.3 Cross-talk between IDC and T cells 8o 1

I \r

10.0 2.50 0.62 0.16 0.04 0.01

stimulators x 1 ~ 3

Figure 5. Sorted IDC can induce an allogeneic MLR. Peripheral blood CD4' T cells (50000/well) were seeded onto graded num- bers of IDC. Tonsillar B cells were used as a control. Cells were incubated for 6 days and proliferation was measured by ['Hlthy- midine incorporation during the last 18 h of culture. A stimulation index of (SI) of 600 for IDC at lo4 celldwell was observed, while for B cells, an SI of only 0.6 was seen. One experiment of a t least three is shown.

extent Fc receptors CD23 and CD32 to retain immune complexes. Unlike DC generated in vitro from monocytes cultured with granulocyte/macrophage (GM)-CSF and IL- 4, or CD4'CDllc' blood DC and germinal center DC, IDC express low or no myeloid markers CD13 and CD33, thus suggesting that IDC may represent lymphoid-derived nc- UL.

Like all other DC, tonsillar IDC induce strong proliferation of allogeneic CD4' T cells (Fig. 5). On the other hand, recent studies have shown that T cell signals, such as IL-3, IL-4, and CD40L also induce DC maturation and activation [25, 261. Consistent with these findings, CD32/ CD40L-transfected L cells strongly up-regulated the expression of MHC class 11, B7.1 (CD80) and B7.2 (CD86) on the IDC (Fig. 6). In addition, CD40 ligation also up-regulated the late maturation marker CD83, but down-regulated CD4. CD40 triggering also induced IDC to secrete large amounts of IL-12 and chemokines such as MIP-la, MIP-1P, RANTES, and IL-8 (Table 2).

3.4 IDC directly enhance the proliferation and differentiation of CD40-activated naive B cells

To determine whether isolated IDC contribute directly to primary T cell-dependent B cell activation that occurs in the extrafollicular T cell areas, double staining with anti- CD40 and anti-IgD mAb on human tonsillar sections indeed revealed the presence of IgD+ naive B cells within the T cell zones in proximity to CD40' IDC (Fig. 7a). Based on this observation in situ, isolated IgD+ tonsillar B cells and IDC were co-cultured on CD32/CD40L- transfected L cells that act as surrogate activated T cells. As shown in Fig. 7b, IDC were consistently able to enhance the [3H]thymidine uptake by CD40-activated naive B cells. This effect was marginally potentiated by the addition of IL-2. No proliferation was seen when B cells were cultured with IDC in the absence of a CD40 signal.

In parallel experiments, the effect of IDC on the differentiation of CD40-activated naive B cells were analyzed. After 7 days of culture, the percentage of generated CD20- CD38' plasma cells versus CD20+ CD38- B blasts was determined. Fig. 7c shows that using CD40L alone, only 0.5 % of plasma cells were generated. Further addi-

MHC class 11 CD83 CD80 CD86 CD4

day 0

day 1

1

1 0 1 1 1 2 1 3 L 1 1

1 b 1 * 1 2 1 '

1

1 0 1 1 1 2 1 ' b 1

Figure 6. FACS profiles of anti- gens expressed on sorted IDC: cells were stained with FITC- conjugated antibodies to MHC class 11, CD83, CD80, CD86, or CD4 either directly after sorting or after overnight culture on CD32/CD40L L cells. One representative experiment of three with similar results is shown. logjuoresence intensity

Eur. J. Immunol. 1997.27: 1266-1274

A)

Human IDC activate naive B cells 1273

5 2 0 -

10 -

I I

C W L CDNL+IL-2 CDNL+IDC

CD2O-FITC - E)

CDNL+IDC+ILQ

Figure 7. (a) IDC can interact with naive B cells in situ and in virro. Double staining of theT cell zone from tonsils with CD40 (blue) and IgD (red). A few scattered naive B cells (purple) in close contact with IDC are seen (magnification 400X). (b) Sorted IDC can induce proliferation of CD40-activated naive B cells. Tonsillar IgD' B cells (lOOo/well) were cultured on CD4OL L cells with or without IDC (10000/well), with or without IL-2. Cells were incubated for 6 days and proliferation was measured by [3H]thymidine incorporation during the last 18 h of culture. One representative experiment out of four is shown. (c) Sorted IDC and IL-2 induce differentiation to plasma cells from naive B cells. Tonsillar IgD' B cells (150000/well) were cultured on CD3UCD40L L cells with or without sorted IDC (150000/well) with or without 1L-2 in a final volume of 1 ml in 24-well plates. After 7 days of culture, cells were washed and stained with antibodies to CD20 and CD38. Gates were set to exclude dead cells and debris. At least loo00 cells were counted. Plasma cells express CD38, but lack the CD20 antigen. (d) Cytospins were fixed in acetone, washed, and stained with antibodies to kappa and lambda light chains (both IgGl subclass) and MHC class I1 (IgG2b subclass). After incubation with secondary antibodies, kappa and lambda were developed by APAAP using Fast blue salt (blue) and MHC class I1 by streptavidin-peroxidase using DAB (brown). Plasma cells are found in the cluster of IDC (magnification 1OOOX). (e) Sorted IDC can induce differentiation of CD40-activated naive B cells; Culture conditions are as in (b). Cells were incubated for 14 days and production of IgM and IgG were measured by ELISA. One experiment of three is shown.


tion of IDC o r IL-2 led to only 2 % plasma cell generation. However, addition of both IDC and IL-2 together led 10 % plasma cells, that is, 20-fold higher than the percentage of plasma cells generated from naive B cells cultured with CD40L alone. Staining of cytospins made from the same cultures using mAb to x and h light chains and M H C class I1 revealed plasma cells within the IDC clusters (Fig. 7d). Consistent with the increase in plasma cell generation, I D C significantly enhanced IgM, and to a lesser extent, IgG production by CD40-activated naive B cells in the presence or absence of IL-2. Ig production in cultures with IDC and B cells without CD40 activation was negligible, as was the production when B cells were cultured alone on CD40L L cells (Fig. 7e).

4 Discussion

IDC, which represent the key antigen-presenting cells within the T cell areas of secondary lymphoid tissue, were isolated from human tonsils, according to their CD40hi lineage- phenotype. The isolation was carried out at 4 "C to keep as much as possible their IDC features ex vivo. Indeed, the isolated cells displayed all the features of IDC: they display long, fine and highly motile dendrites, they expressed the co-stimulatory molecules CD40, B7.1 (CD80), and B7.2 (CD86), they did not express the Lan- gerhans cell marker C D l a , they expressed lower levels of CD4' than the CD4+CDl lc+ blood DC and DC in the germinal center d o [22, 311, and many of them expressed the late D C maturation marker CD83.

The positive selection of I D C by anti-CD40 antibodies did not interfere with the signaling through CD40 to activate IDC, as demonstrated by the fact that triggering I D C by FcyR(CD32)/CD40L-transfected L cells induced IDC to up-regulate MHC class 11, B7.1 (CD80), B7.2 (CD86), and CD83, and to secrete certain chemokines as well as IL-12. A similar observation has been made in CD40- activated DC generated either from CD34' progenitor cells [25, 261 or generated from monocytes [27, 281.

Using CD32/CD40L double-transfected L cells as surrogate activated T cells, we demonstrated that IDC isolated from human tonsils directly enhance the activation and differentiation of IgD' naive B cells. The molecular mecha- nisms underlying the IDClB cell interaction is currently unknown, but IL-2 potentiates the IDC effect on B cells. In this context, it is important to note that the surrogate activated Tcells (CD40 triggering) induced IDC to produce large amounts of IL-12, which was recently shown to strongly promote B cell activation in the presence of IL-2 [29].

A decade ago, MacLennan and Gray [30] proposed that early primary B cell activation may take place within the paracortex of lymph nodes or the PALS of the spleen [30]. These sites represent the first place where B cells and T cells meet each other within the interdigitating network of IDC. Indeed, immunohistological studies by labeling specific antigen-binding B cells on tissue sections directly showed that the primary T cell-dependent B cell responses were initiated within the T cell- and IDC-rich areas [6]. The present study provides evidence in favor of this model, demonstrating the signaling from T cells to IDC and from

The authors wish to thank Ms. 1. Durand for excellent assistance with the FACS sorting, Drs. E. Garcia, C. Mueller, H . Martinez- Valdez and fl Kincade for help and critical reading of the manu- script, and Dr. J . Chiller for support.

5 References 1 Hoefsmit, E. C., Duijvestijn, A. M. and Kamperdijk, E. W.,

2 Steinman, R. M., Annu. Rev. Irnrnunol. 1991. 9: 271. 3 Kupiec-Weglinski, J. W., Austyn, J. M. and Morris, P. J., J.

Exp. Med. 1988. 167: 632. 4 O'Doherty, U., Steinman, R. M., Peng, M., Cameron, P. U.,

Geltzer, S., Kopeloff, I., Swiggard, W. J., Pope, M. and Bhardway, N., J. Exp. Med. 1993. 178: 1067.

5 Fossum, S., Scand. J. Immunol. 1988.27: 97. 6 Liu, Y.-J., Zhang, J., Lane, I? L. J., Chan, E. Y.-T. and Mac-

7 Inaba, K., Metlay, J. P., Crowley, M. T. and Steinman, R. M.,

8 Inaba, K., Witmer, M. D. and Steinman, R. M., J. Exp. Med.

9 Schrader, C. E., George, A., Kerlin, R. L. and Cebra, J. J.,

10 Zhou, L.-J., Schwarting, R., Smith, H. M. andTedder,T. F., J.

11 Hart, D. N. and McKenzie, J. L., J. Exp. Med. 1988.168: 157. 12 VallC, A., Zuber, C. E., Defrance, T., Djossou, O., de Rie,

M. and Banchereau, J., Eur. J. Irnrnunol. 1989. 19: 1463. 13 VallB, A., Aubry, J.-F'., Durand, I. and Banchereau, J., Int.

Irnmunol. 1991.3: 229. 14 Munro, M., Freedman, A. S., Aster, J. C., Gribben, J . G.,

Lee, N. C., Rhynhart, K. K., Banchereau, J. and Nadler, L. M., Blood 1994. 3: 793.

15 Ling, N. R., MacLennan, I. C. M. and Mason, D., in McMi- chel, A. J. (Ed.), Leukocyte Typing Ill, Oxford University Press, Oxford 1987. pp. 302-335.

16 Arpin, C., DCchanet, J., van Kooten, C., Merville, P., Grou- ard, G., Bribre, E, Banchereau, J. and Liu, Y.-J., Science 1995.268: 720.

17 Favre, C., Wijdenes, J., Cabrillat, H., Djossou, O., Banche- reau, J. and de Vries, J. E., Mol. Imrnunol. 1989.26: 17.

18 Abrahams, J. S., Roncarolo, M.-G., Yssel, H., Anderson, U., Gleich, G. J. and Silver, J. E., Immunol. Rev. 1992.127: 9.

19 Defrance, T., Vanbervliet, B., Pbne, J. and Banchereau, J., J. Irnmunol. 1988.141: 2000.

20 Vandenberghe, P., Delabie, J., de Boer, M., Wolf-Peeters, C. D. and Ceuppens, J. L., Int. Immunol. 1993.5: 317.

21 Inaba, K., Witmer-Pack, M., Inaba, M., Hathcock, K. S., Sakuta, H., Azuma, M., Yagita, H., Okumura, K., Linsley, P. S., Ikehara, S., Muramatsu, S., Rodes, R. J. and Steinman, R. M., J. Exp. Med. 1994. 180: 1849.

22 Weissman, D., Li, Y., Ananworanich, J., Zhou, L.-J., Adels- berger, J., Tedder, T. F., Baseler, M. and Fauci, A. S., Proc. Natl. Acad. Sci. USA 1995. 92: 826.

23 Grouard, G., Durand, I., Filgueira, L., Banchereau, J. and Liu, Y.-J., Nature 1996. 384: 364.

24 Freudenthal, P. S. and Steinman, R. M., Proc. Natl. Acad. Sci. USA 1990. 87: 7698.

25 Caux, C., Massacrier, C., Vanbervliet, B., Dubois, B., van Kooten, C., Durand, I. and Banchereau, J., J. Exp. Med. 1994. 180: 1263.

26 Flores-Romo, L., Bjorck, P., Duvert, V., van Kooten, C., Sae- land, S. and Banchereau, J., J. Exp. Med. 1997. 185: 341.

27 Koch, E, Stanzl, U., Jennewein, P., Janke, K., Heufler, C., Kampgen, E., Romani, N. and Schuler, G., J . Exp. Med. 1996. 184: 741.

28 Cella, M., Scheidegger, D., Palmer-Lehmann, K., Lane, P., Lanzavecchia, A. and Alber, G., J. Exp. Med. 1996. 184: 747.

29 Jelinek, D. F. and Braaten, J. K., J. Imrnunol. 1995.154: 1606. 30 MacLennan, I . C. M. and Gray, D., lrnrnunol. Rev. 1986. 91:

Irnmunobiology 1982. 161: 255.

Lennan, I. C. M., Eur. J. Irnmunol. 1991. 21: 2951.

J . Exp. Med. 1990. 172: 631.

1984. 160: 858.

Int. Irnrnunol. 1990. 2: 563.

Irnmunol. 1992. 149: 735.

I I

IDC to B c e k 61.

Documents

Human interdigitating dendritic cells directly stimulate CD40-activated naive B cells