Bacterial DNA and immunostimulatory CpG oligonucleotides trigger maturation and activation of murine...

0014-2980/98/0606-2045$17.50+.50/0© WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998

Bacterial DNA and immunostimulatory CpGoligonucleotides trigger maturation and activationof murine dendritic cells

Tim Sparwasser1, Eva-Sophie Koch1, Ramunas M. Vabulas1, Klaus Heeg1, Grayson B.Lipford1, Joachim W. Ellwart2 and Hermann Wagner1

1 Institute of Medical Microbiology, Immunology and Hygiene, Technical University of Munich,Munich, Germany

2 Institute of Experimental Haematology, GSF-National Research Center for Environment andHealth, Munich, Germany

Bacterial DNA and immunostimulatory (i.s.) synthetic CpG-oligodeoxynucleotides (ODN) actas adjuvants for Th1 responses and cytotoxic T cell responses to proteinaceous antigens.Dendritic cells (DC) can be referred to as “nature’s adjuvant” since they display the uniquecapacity to sensitize naive T cells. Here, we demonstrate that bacterial DNA or i.s. CpG-ODN cause simultaneous maturation of immature DC and activation of mature DC to pro-duce cytokines. These events are associated with the acquisition of professional antigen-presenting cell (APC) function. Unfractionated murine bone marrow-derived DC and FACS®-fractionated MHC class IIlow (termed immature DC) or MHC class IIhigh populations (termedmature DC) were stimulated with bacterial DNA or i.s. CpG-ODN. Similar to lipopolysaccha-ride, i.s. CpG-ODN caused up-regulation of MHC class II, CD40 and CD86, but not CD80 onimmature and mature DC. In parallel both DC subsets were activated to produce largeamounts of IL-12, IL-6 and TNF- § . CpG-ODN-activated DC displayed professional APCfunction in allogeneic mixed lymphocyte reaction and in staphylococcal enterotoxin B-drivennaive T cell responses. We interpret these findings to mean that bacterial DNA and i.s. CpG-ODN cause maturation (first step) and activation (second step) of DC to bring about conver-sion of immature DC into professional APC.

Key words: Bacterial DNA / CpG motif / Murine dendritic cell / Cytokine / Surface marker

Received 3/3/98Accepted 2/4/98

[I 18106]

The first two authors contributed equally.

Abbreviations: ODN: Oligodeoxynucleotide i.s.: Immu-nostimulatory BMDDC: Bone marrow-derived dendriticcell I -gal: g -Galactosidase

1 Introduction

Dendritic cells (DC) guide via their antigen presentationmachinery the antigen-specific activation of the adaptiveimmune system [1, 2]. The decision of the adaptiveimmune system whether to ignore presented antigen orto respond to it through tolerance or the induction ofeffector cells likely depends on the maturation state ofantigen-presenting DC [1]. It is only at the mature stagethat DC express the full complement of co-stimulatorymolecules, adhesion molecules and antigen-presentingMHC molecules required for productive T cell activation[3–8]. Understanding the signals driving maturation of

antigen-presenting DC may explain why purified proteinantigens are prone to induce peripheral tolerance ratherthan immunity [9] and why addition of bacterial productssuch as LPS facilitates productive adoptive immuneresponses [10].

Bone marrow-derived DC (BMDDC), which are ubiqui-tously distributed throughout nonlymphoid tissue, areimmature as assessed by expression of MHC class II andco-stimulatory molecules but are efficient in antigenuptake and processing [11, 12]. Upon maturation DCappear to migrate into T cell-dependent areas of sec-ondary lymphoid tissues where they act as professionalAPC activating naive T cells [11, 12]. One signal pathwaytriggering these maturation events appears to functionvia microbial stimuli such as LPS [13, 14]. However, theimmune responses to LPS-negative pathogens areeffective as well, and it thus becomes mandatory to ana-lyze whether additional pathogen-associated ligandsexist which are able to trigger maturation of immatureDC.

Eur. J. Immunol. 1998. 28: 2045–2054 Bacterial DNA triggers maturation and activation of murine DC 2045

Figure 1. CpG-ODN serve as adjuvants for CTL induction invivo. The primary CTL response after injection of C57BL/6mice with liposome-entrapped g -gal using ODN 1668 asadjuvant is shown. The target cells were EL4 cells (opensymbols) or g -gal-transfected EL4 cells (closed symbols)and specific lysis was measured by 51Cr release. Resultsfrom two independent experiments, experiment a (triangles)and b (squares) are shown.

Figure 2. Bacterial DNA and CpG-ODN trigger up-regulation of co-stimulatory molecules in BMDDC. MHCclass II-positive cells (black bar, mean fluorescence G 10)were gated and expression of CD80, CD86 and CD40depicted. Synthetic CpG-oligonucleotides and bacterialDNA induce up-regulation of MHC class II, CD86, and CD40.(A) ODN 1668 (solid histograms), ODN 1720 (black line), sol-vent control (dotted line); (B) ODN IL12p40 (solid histog-rams), ODN AP1 (black line), solvent control (dotted line); (C)E. coli DNA (solid histograms), DNase-treated E. coli DNA(black line), solvent control (dotted line). The data are repre-sentative of three independent experiments.

Recently we [15–17] and others [18] described that bac-terial DNA and immunostimulatory (i.s.) CpG oligodeoxy-nucleotides (ODN) activate macrophages in vivo and invitro to express activation markers, to translocate NF O B-and to secrete pro-inflammatory cytokines such as TNF-§ , IL-1, IL-6 and IL-12. While the induction of potentially

useful cytokines such as IL-12 or potentially harmfulcytokines such as TNF- § appears to be dependent onthe sequence of the CpG-ODN used [17], a hallmark ofbacterial DNA and i.s. CpG-ODN is its profound adju-vanticity for the induction of murine Th1 responses[19–23]. The use of CpG-ODN as adjuvants allows invivo the induction of cytotoxic T cell responses to thesoluble protein OVA [19]. In view of these findings we for-mulated the hypothesis that bacterial DNA and CpG-ODN in vivo trigger immature DC to mature into profes-sional APC. Here we report results substantiating thisprediction.

2 Results

2.1 CpG-ODN adjuvanticity allows in vivo CTLinduction to I -gal protein

Induction of T cell-mediated cytotoxic immune respon-ses to soluble protein antigens in vivo is poor unlessaided by the use of adjuvants. Even though the mode of

adjuvant action is ill defined, it is likely that adjuvantsactivate professional APC [24, 25]. Recently wedescribed that bacterial DNA or i.s. CpG-ODN displaypowerful adjuvanticity for the induction of CTL to OVA invivo [19]. The results detailed in Fig. 1 extend thisobservation to the protein g -galactosidase ( g -gal). g -galentrapped in liposomes without adjuvant was an inef-fective inducer of CTL activity (data not shown). Co-injection of i.s. ODN 1668 resulted in a substantial pri-mary CTL response. These findings suggest that bacte-rial DNA or defined CpG-ODN adjuvants are naturalstimuli now available to test the prediction that they tar-get DC and trigger their maturation into professionalAPC. We therefore evaluated the effect of CpG-ODN onDC differentiated in vitro from murine BM cells (BMDDC).

2.2 CpG-ODN up-regulate MHC class II, B7 andCD40

BM cells cultured in granulocyte-macrophage colony-stimulating factor (GM-CSF)-conditioned mediuminclude different cell populations such as myeloid

2046 T. Sparwasser et al. Eur. J. Immunol. 1998. 28: 2045–2054

Figure 3. CpG-ODN induce up-regulation of co-stimulatorymolecules in MHC class II-positive BMDDC. BMDDC weresorted into MHC class IIlow and high populations (upper panel)and stimulated overnight with CpG-ODN 1668 and LPS.Solid histograms show ODN 1668 (left panel) or LPS (rightpanel) stimulated BMDDC, white histograms (black line)show the solvent control. Dotted lines indicate isotype con-trols. CpG-ODN and LPS trigger up-regulation of MHC classII, CD86 and CD40 surface molecules in MHC class IIlow

(immature) DC. Maturated MHC class IIhigh BMDDC showup-regulation of CD86 and CD40 co-stimulatory molecules.

progenitors, granulocytes, macrophages and DC. After8 days about 30–40 % of the nonadherent cellsexpressed high levels of MHC class II molecules andabout 10–20 % expressed intermediate levels of MHCclass II molecules. The portion of contaminating B orT cells was about 5 % (B220+ or CD3+), and about 10 %were GR-1 positive. Giemsa staining revealed sheet-likeprocesses and large veils that extended in several dire-ctions from the cell body. These features were typical of

“low” MHC class II-expressing cells and were more obvi-ous in “high” MHC class II-expressing cells. As previ-ously described, MHC class IIlow cells were classified“immature” DC and MHC class IIhigh cells were termed“mature” DC [2].

In a first set of experiments we observed that i.s. CpG-ODN 1668 (Fig. 2A), ODN IL12p40 (Fig. 2B) or bacterialDNA from E. coli (Fig. 2C) caused the mixture of imma-ture and mature BMDDC to up-regulate co-stimulatorymolecules and MHC class II expression. On the otherhand, non-i.s. CpG-ODN such as ODN 1720 (inverted“CG” [27], Fig. 2A) or AP1 (without “CG-motif”, Fig. 2B)[17] were ineffective. DNase-treated bacterial DNA wasineffective as well (Fig. 2C).

To analyze whether i.s. CpG-ODN up-regulate MHCclass II, CD80 (B7.1), CD86 (B7.2) and CD40 on imma-ture DC, DC cultures were sorted into MHC class IInegative,MHC class IIlow and MHC class IIhigh cells (Fig. 3, upperpanel). Aliquots of the sorted cells were cultured over-night (20 h) in the presence of i.s. CpG-ODN 1668 orLPS. Thereafter the expression of MHC class II, CD80,CD86 and CD40 was analyzed by FACS® (Fig. 3). Neitheri.s. CpG-ODN 1668 nor LPS caused up-regulation ofCD80, CD86 and CD40 on MHC class IInegative cells (datanot shown). However, CpG-ODN 1668 triggered the sub-set of immature DC to display increased levels of MHCclass II molecules, CD86 and CD40 but not CD80 (Fig. 3,left panel, MHC-IIlow cells). Essentially the same resultswere obtained with LPS (Fig. 3, right panel). Of note, theeffect of i.s. CpG-ODN 1668 or LPS on the cell surfacephenotype of mature DC was less pronounced (Fig. 3,MHC-IIhigh cells). We conclude that i.s. CpG-ODN triggera subset of immature DC to enhance expression ofCD86, CD40 and MHC class II molecules, as does LPS.As a consequence, from the mixture of immature andmature DC a uniform DC population is generated whichis rich in co-stimulatory and MHC class II molecules.

2.3 Bacterial DNA and i.s. CpG-ODN activate DCto produce IL-12, IL-6 and TNF- >

The cytokine profile potentially produced by DC present-ing antigen to naive T cells is likely to influence product-ive T cell activation [14]. Therefore we analyzed whetherbacterial DNA or i.s. CpG-ODN can activate DC to pro-duce IL-12. As shown in Fig. 4, DC generated in GM-CSF-conditioned medium produced IL-12 when acti-vated by either LPS or E. coli DNA. The activation ofBMDDC is mediated by bacterial DNA since pretreat-ment with DNase abolished its biological activity. In addi-tion to IL-12, high amounts of other pro-inflammatorycytokines such as TNF- § and IL-6 were also detectable(data not shown).


Figure 4. E. coli DNA triggers IL-12-release in BMDDC.BMDDC (1 × 106/ml) were stimulated at day 7 of culture with0.1 ? g/ml LPS or 1 ? g/ml E. coli DNA and total IL-12 levelsof 20-h supernatants were measured by ELISA. BacterialDNA induced IL-12 release from BMDDC. DNase-treatedbacterial DNA and solvent served as negative controls.

Figure 5. CpG-ODN induce IL-12, IL-6 and TNF- § release inBMDDC. Cytokine release of BMDDC (1 × 106/ml) afterstimulation with synthetic ODN and LPS determined byELISA in duplicate measurement is shown. CpG-ODN 1668and LPS triggered secretion of IL-12, IL-6 and TNF- § . ODNIL12p40 induced comparable levels of IL-12 in the absenceof TNF- § . ODN AP1 and solvent served as negative control.Mean values of two independent experiments are shown.

Figure 6. Cytokine release of FACS-sorted BMDDC. BMcells at day 7 of culture were FACS-sorted into MHC classIIhigh (++), low (+) and negative (−) cells (as detailed in Fig. 3, upperpanel). Cells of each population (1 × 106/ml) were stimulatedwith ODN 1668 or LPS in vitro and cytokines detected in20-h supernatants by ELISA. CpG-ODN 1668 and LPStriggered secretion of IL-12, IL-6 and TNF- § in MHCclass IIhigh (++) and low (+) cells.

CpG-ODN-mediated immunostimulation mimicking theeffects of bacterial DNA depends on unmethylated CpG

motifs and is controlled by neighboring nucleotidesequences [16, 17, 27–30]. For example, the controlODN AP1 is ineffective, the CpG-ODN IL12p40 has thepropensity to trigger IL-12 release but little TNF- § [17],whereas CpG-ODN 1668 appears as “pluripotent” in thatit is mitogenic for B cells [27] and activates macrophagesto produce an array of cytokines such as IL-12, TNF- § ,IL-1, IL-6 and GM-CSF ([16, 17, 19] and unpublisheddata). As such, CpG-ODN 1668 is potentially harmfulsince like bacterial DNA it triggers in vivo TNF- § -mediated septic shock [16]. On the other hand, CpG-ODN IL12p40 is potentially useful since it induces in vivoIL-12 production and strong Th1 responses in theabsence of TNF- § [17, 19]. This information in mind, wecompared the cytokine induction profile of these CpG-ODN on BMDDC. As shown in Fig. 5, the control ODNAP1 was ineffective, whereas CpG-ODN 1668 effectivelytriggered IL-12, IL-6 and TNF- § release, as does LPS.The CpG-ODN IL12p40, however, failed to induce TNF-§ , while induction of IL-12 and IL-6 was strong. We con-

clude that the sequence-dependent immunostimulatoryeffects of CpG-ODN observed in vivo [17] can be repro-duced in vitro using DC.


Figure 7. CpG-ODN induce functional maturation ofBMDDC. (A) MLR: Graded numbers of stimulated (closedsymbols) and mock-treated (open symbols) BMDDC as APCwere co-cultured with lymph node responder T cells fromallogeneic (triangles, squares, circles) or syngeneic (dia-monds) mice. ODN 1668 (squares) or IL12p40 (circles)served as stimulus for BMDDC. CpG-ODN 1668- andIL12p40-treated DC displayed improved APC function forthe activation of allogeneic T cells. (B) SEB-triggered lym-phocyte proliferation: For the SEB-triggered proliferationassay 1 ? g/ml SEB was added to co-cultures of BMDDCand syngeneic lymphocytes. ODN IL12p40 served as APCstimulus at 10 ? M (closed squares) and 1 ? M (closed circles)concentration, respectively. ODN AP1 (closed triangles,1 ? M) served as additional negative control. The solventcontrol is depicted as open diamonds. ODN IL12p40-stimulated BMDDC triggered SEB-dependent T cell prolifer-ation.

To identify the cellular source of cytokines generated, wequantified the cytokine contents of culture supernatantsfrom fractionated (Fig. 3, upper panel) and stimulatedDC. Fig. 6 demonstrates cytokine release from FACS-sorted MHC class IInegative, low and high BMDDC after over-night stimulation with LPS or ODN 1668. In comparingthe data of Fig. 3 and Fig. 6 we conclude that parallel tomaturation of DC (as defined by enhanced MHC class II,CD40 and CD86 expression), i.s. CpG-ODN 1668 trig-gered brisk IL-12, IL-6 and TNF- § production in MHCclass II-positive cells, as did LPS (Fig. 6).

2.4 CpG-ODN induce functional maturation ofBMDDC

One of the hallmarks of mature DC is their unique capac-ity to sensitize naive T cells [1, 2]. To achieve this DCmust differentiate from immature precursor cells, whichare endowed with high pinocytic and antigen processingactivity, to mature DC, characterized by high levels ofMHC class II, adhesion and co-stimulatory molecules,extensive surface area and high motility [5, 6]. To testwhether i.s. CpG-ODN-mediated DC maturation/activa-tion is associated with functional transition to profes-sional APC, BMDDC were first stimulated with i.s. CpG-ODN for 20 h and subsequently used as APC in an allo-geneic MLR, or as APC for the presentation of staphylo-coccal enterotoxin B (SEB) to syngeneic naive T cells. Asshown in Fig. 7A, i.s. CpG-ODN matured and activatedDC to display professional APC function for the activa-tion of T cells in the MLR. The superantigen SEB doesnot require processing and is best presented by DC [31].Fig. 7B demonstrates that CpG-ODN-stimulated DCtrigger naive T cells to proliferate in a SEB-dependentmanner. The control ODN AP1 was ineffective (Fig. 7B).Of note, BMDDC, although containing a MHC class IIhigh

subset, were ineffective as APC unless stimulated withi.s. ODN.

3 Discussion

Here we describe the functional consequences imposedby bacterial DNA and CpG-ODN on DC from murinebone marrow cells grown in GM-CSF-conditionedmedium. These stimuli were shown to trigger in paralleltwo events. First, they caused maturation of immatureDC to mature DC as defined by up-regulation of MHCclass II, CD86 and CD40 (Figs. 2, 3). Second, they acti-vated mature and “maturing” DC to produce highamounts of cytokines such as IL-12, IL-6 and TNF- §(Figs. 5, 6). The mature stage was associated with theacquisition of professional APC function (Fig. 7). We fur-ther show that the effects elicited by the bacterial stimu-lus LPS on DC are qualitatively and quantitatively nearlyidentical to i.s. CpG-ODN, in that both stimuli causedmaturation/activation of DC in an apparent CD40-CD40ligand (DC40L)-independent fashion.

The capacity of these microbial stimuli to cause simul-taneously DC maturation/activation argues that DC needto function both as APC and cytokine-producing cells tobe effective APC for the initiation of T cell-mediatedresponses. These results lead to a two-step maturation/activation model as described in Fig. 8. According to thismodel, immature DC first mature to APC rich in MHCclass II, adhesion and co-stimulatory molecules, exten-


Figure 8. Two-step model of DC activation/maturation. DC first mature to APC rich in MHC class II, adhesion and co-stimulatorymolecules, extensive surface area and high motility. For production of crucial cytokines activation of mature DC is needed. Thismodel postulates that it is this activation step which confers robust professional APC function to mature DC. As a possible mech-anism of action, i.s. CpG-ODN function in vivo by converting immature DC into mature DC which are rich in co-stimulatory mole-cules and simultaneously produce cytokines such as IL-12, TNF and IL-6.

sive surface area and high motility. However, activationof mature DC is needed for the production of crucialcytokines. The model predicts that it is this activationstep which confers robust professional APC function tomature DC. This prediction is substantiated by ourobservation that only upon CpG-ODN activation themature DC generated in GM-CSF-conditioned mediumdisplay improved professional APC function (Fig. 7).

In mice bacterial DNA and i.s. CpG-ODN exhibit uniqueadjuvanticity for the development of strong Th1 respon-ses to proteinaceous antigens and the induction of CTL([19–21, 23], Fig. 1). The two-step model for conversionof DC into professional APC (Fig. 8) predicts that i.s.CpG-ODN function by converting in vivo immature DCinto mature DC which are rich in co-stimulatory mole-cules and simultaneously produce cytokines such as IL-12. Since IL-12 is likely to be the key cytokine for theinduction of protective Th1 cell-mediated immunity tointracellular pathogens [14], IL-12-producing antigen-presenting DC would be ideally equipped for this task.

The i.s. CpG-ODN-induced maturation of DC implicatestherefore a potential “therapeutic” use.

Previously it has been postulated that DC maturation iseffected by cross-signaling of T cells and DC via CD40L-CD40 [32, 33]. In vitro [34] as well as in vivo [13] LPScauses DC maturation/activation for cytokine produc-tion. We thus hypothesize that microbial DNA and i.s.CpG-ODN induce in vivo CD40L-independent DCmaturation, activation for IL-12 production as well as DCredistribution to T cell areas as shown for LPS [13]. Webelieve that the in vitro data reported here explain theprofound adjuvanticity of CpG-ODN for Th1 responses[19, 22] in vivo. Although bacterial CpG-ODN stronglyup-regulated CD40 and CD86 on immature DC and onmature DC similar to LPS (Figs. 2, 3) CD80 expressionwas unaffected. At present it is not clear whether diffe-rential expression of CD80 versus CD86 conveys conse-quences on APC function [35–37]. Of note, the myeloidDC used here were generated in the absence of IL-4,avoiding up-regulation of CD86 [38]. We also addressed


the question whether IL-12-, IL-6- or TNFR p55-mediated signaling was responsible for up-regulation ofCD40 and CD86. Therefore BMDDC derived from therespective knockout mice were stimulated with CpG-ODN. In essence similar results as depicted in Fig. 2were obtained, negating a role of an autocrine feedback(unpublished data).

Since in vivo IL-12 and TNF- § production by macro-phages in response to microbial products appears torequire prior activation of the cells [39, 40], we assumethat the early (1–2 h) serum values of IL-12, IL-6 andTNF- § in CpG-ODN-challenged mice [16, 17] may beproduced by DC. Indeed there is evidence that in vivo themicrobial stimulus LPS triggers DC rather than macro-phages to produce IL-12 [14].

The signal pathways through which bacterial DNA andi.s. CpG-ODN cause maturation/activation of immatureDC are ill defined. Preliminary data suggest that, as in thecase of macrophage cell lines, bacterial DNA and CpG-ODN is taken up by DC and transported, via adsorptivepinocytosis, into acidic endosomal organelles. Subse-quently and within 10 min JNK kinases become trig-gered via the stress kinase pathway associated withtranscriptional activity of AP1 and NF O B (H. Häcker et al.,manuscript in preparation). The finding that i.s. CpG-ODN 1668 and CpG-ODN IL12p40 differentially inducecytokines (Fig. 5) may indicate that sequence-dependentdistinct signal pathways become triggered.

Even though the maturation/activation of DC by bacterialDNA and CpG-ODN can clearly be observed in vitro, thephysiological relevance of our findings at a site of infec-tion needs to be evaluated. In particular we need to ana-lyze whether in vivo DC take up sufficient amounts ofpathogen-derived DNA (or plasmids) enabling theirmaturation/activation. In vitro i.s. CpG-ODN activate DCat less than 0.1 ? M (unpublished data).

In conclusion, our in vitro data demonstrate that bacterialDNA and i.s. CpG-ODN cause maturation/activation ofmyeloid DC grown in GM-CSF-conditioned medium. Thedata support the hypothesis that bacterial DNA and syn-thetic CpG-ODN cause DC maturation/activation andthus convert DC into professional APC, able to initiateproductive T cell immune responses. Operationally simi-lar to LPS, bacterial DNA and i.s. CpG-ODN induce co-stimulatory activity, enhanced MHC class II expressionand trigger cytokine release such as IL-12. We hypothe-size that bacterial DNA and i.s. CpG-ODN representphysiological stimuli able to “switch on” professionalAPC function in DC at sites of infection.

4 Materials and methods

4.1 Experimental mice

Female BALB/c, C57BL/6 and IL-12 knockout (−/−) micewere purchased from Harlan Winkelmann (Borchen, Ger-many), Bomholtgard Breeding and Research Center Ltd.(Ry, Denmark), or Charles River Wiga (Sulzfeld, Germany).TNF-receptor p55−/− mice were a kind gift of Dr. K. Pfefferand were bred in our own animal facility. IL-6−/− mice wereobtained from Dr. M. Kopf (Basel, Switzerland). All animalswere housed under specific pathogen-free conditions andwere used at 8–10 weeks of age.

4.2 Cell culture and generation of BMDDC

Spleen cells, lymph node cells and BM cells were cultured inLowTox Clicks/RPMI (Biochrom, Berlin, Germany) supple-mented with 10 % (v/v) FCS (Hyclone, Lab. Inc., Logan, UT),5 × 10−5 M 2-ME and 100 IU/ml penicillin G and 100 IU/mlstreptomycin sulfate. BMDDC from C57BL/6 mice were pre-pared as described [41]. Briefly, femora of C57BL/6 micewere rinsed with cell culture medium using a syringe with an27-gauge needle. BM cells (3 × 106) were seeded into a 100-mm tissue culture petri dish in complete medium supple-mented with 200 U/ml GM-CSF. Cells were not used beforeday 7 when mature DC (GR-1−, MHC class IIhigh, CD86high)represented 30–40 % of the BM population. GM-CSF wasproduced by a cell line as described elsewhere [42] andadded to in vitro cultures of BM cells. For stimulation withLPS or CpG-DNA, nonadherent BMDDC were washed andplated at 1 × 106/ml in LowTox Medium without GM-CSF.The stimulation period lasted 20 h unless otherwise stated.

4.3 Microbial stimuli and synthetic oligonucleotides

LPS from E. coli was purchased from Sigma and was usedat 0.1 ? g/ml. Gram-negative E. coli DNA came from Sigma(Munich, Germany). DNA was heat denatured (95 °C, 10 min)and sonicated (2 min) before use and used at 1 ? g/ml. Toprove that the biological activity is mediated by DNA, E. coliDNA was digested with DNase I (37 °C, 1 ? g DNase I/10 ? gbacterial DNA, 20 h), subsequently heat denatured (95 °C,10 min) and then tested for induction of cytokine release.DNase I and Buffer L were purchased from Boehringer(Mannheim, Germany). Phosphorothioate-stabilized oligo-nucleotides 1668, 1720, AP1 and IL12p40 were synthesizedby TibMolBiol (Berlin, Germany). ODN 1668 (containing the“CG-motif”: 5'-TCC-ATG-ACG-TTC-CTG-ATG-CT) andODN 1720 (inverted CG: 5'-TCC-ATG-AGC-TTC-CTG-ATG-CT) were taken from [27]. ODN AP1 (5'-GCT-TGA-TGA-CTC-AGC-CGG-AA) and ODN IL12p40 (5'-AGC-TAT-GAC-GTT-CCA-AGG) were described recently [17]. All ODN wereused at 1 ? M unless otherwise stated.


4.4 Flow cytometry

Cells (5 × 105–106) were washed in PBS containing 2 % FCSand incubated for 10 min at 4 °C with anti-Fc + RII/III antibodyfrom PharMingen (Hamburg, Germany) to block unspecificbinding of the following antibody reagents. mAb against MHCclass II (2G9), CD80, CD86, CD40, B220, CD3 and GR-1 wereused at 5–20 ? g/ml. FITC- and PE-labeled Ab were pur-chased from PharMingen. Isotype controls included purifiedrat IgG2a, rat IgG2b, and hamster IgG (all from PharMingen).Between all incubation steps (30 min, 4 °C), cells werewashed with PBS/FCS. FACS analysis was performed on aCoulter Epics XL cytofluorometer (Krefeld, Germany), acquir-ing 10 000 events. FACS data were analyzed using WinMDI2.6 FACS software. For FACS-sorted preparations cells werewashed in LowTox cell culture medium, stained for MHCclass II (2G9, 30 min, 4 °C) and separated into a MHC classIInegative, low and high population on a Becton Dickinson FACStarPLUS cytofluorometer as detailed in Fig. 3 (upper panel).

4.5 Immunization protocols

For CTL induction, liposomes containing 2.5 mg/ml g -galwere prepared as described previously [26]. i.s. ODN(10 nmol ODN 1668) was co-injected with liposome-entrapped g -gal (0.25 mg/mouse) into the hind footpads ofC57BL/6 mice. After 4 days cells of the draining lymphnodes were harvested and cultured for additional 4 days inthe presence of 10 U/ml rIL-2. CTL assays were performedas described elsewhere [26]. EL4 cells expressing g -galserved as targets for CTL.

4.6 Plasmid construction and transfections

The E. coli g -gal gene was removed from the CMV- g -BglIIplasmid (provided by Dr. H. Häcker) by digestion with BglIIand NotI. The fragment was then cloned into the polylinkerof pcDNA3 (Invitrogen, Leek, The Netherlands). Electropora-tions were performed in 400 ? l Click’s medium containing20 % FCS with 20 ? g plasmid and 5 × 106 EL4 cells at 240 V.Selection in G418 (Life Technologies, Eggenstein, Germany)at 600 ? g/ml was begun 24 h after electroporation.

4.7 In vitro lymphocyte proliferation assays

ODN-stimulated (20 h) and mock-treated murine BMDDCwere irradiated (20 Gy), washed and used as APC in MLR orSEB proliferation assays. Graded numbers of BMDDC wereco-cultured for 5 days (SEB stimulus) or 6 days (MLR) with1 × 105 lymph node cells. Proliferation was measured by[3H]thymidine incorporation over the last 15 h. For MLR allo-geneic-(BALB/c) leukocytes (axillary and inguinal lymphnodes) served as responder T cells. For SEB-triggered lym-phocyte proliferation 1 ? g/ml SEB (Toxin Technology, Sara-sota, USA) was added to the co-culture of BMDDC and syn-geneic lymph node cells.

4.8 Determination of cytokines

Cytokine levels in culture supernatants after overnight stim-ulation (20 h) were determined by commercially availableELISA kits (TNF- § and total IL-12-Duoset and InterTest-6XIL-6-ELISA from Genzyme, Rüsselsheim, Germany). Theassay was performed as described by the manufacturer.Determination of cytokines was done in duplicates.

Acknowledgment: We are grateful to Drs. G. Schuler andM. Lutz (Erlangen, Germany) for teaching us the art of culti-vating murine BMDDC. We thank Dr. K. Pfeffer for providingthe TNF-receptor p55−/− mice. We also are grateful to Dr. M.Kopf (Basel, Switzerland) for providing IL-6−/− mice. Thiswork was supported by SFB 391 Project C-4 and BMFTgrant 11675.

5 References

1 Steinman, R. M., The dendritic cell system and its role inimmunogenicity. Annu. Rev. Immunol. 1991. 9: 271–296.

2 Steinman, R. M., Pack, M. and Inaba, K., Dendriticcells in the T-cell areas of lymphoid organs. Immunol.Rev. 1997. 156: 25–37.

3 Steinbrink, K., Wolfl, M., Jonuleit, H., Knop, J. andEnk, A. H., Induction of tolerance by IL-10-treated den-dritic cells. J. Immunol. 1997. 159: 4772–4780.

4 Fu, F., Li, Y., Qian, S., Lu, L., Chambers, F., Starzl, T. E.,Fung, J. J. and Thomson, A. W., Costimulatorymolecule-deficient dendritic cell progenitors (MHC classII+, CD80dim, CD86−) prolong cardiac allograft survivalin nonimmunosuppressed recipients. Transplantation1996. 62: 659–665.

5 Romani, N., Koide, S., Crowley, M., Witmer Pack, M.,Livingstone, A. M., Fathman, C. G., Inaba, K. andSteinman, R. M., Presentation of exogenous proteinantigens by dendritic cells to T cell clones. Intact proteinis presented best by immature, epidermal Langerhanscells. J. Exp. Med. 1989. 169: 1169–1178.

6 Sallusto, F., Cella, M., Danieli, C. and Lanzavecchia,A., Dendritic cells use macropinocytosis and the man-nose receptor to concentrate macromolecules in themajor histocompatibility complex class II compartment:downregulation by cytokines and bacterial products [seecomments]. J. Exp. Med. 1995. 182: 389–400.

7 Suss, G. and Shortman, K., A subclass of dendriticcells kills CD4 T cells via Fas/Fas-ligand-induced apo-ptosis. J. Exp. Med. 1996. 183: 1789–1796.

8 Laufer, T. M., DeKoning, J., Markowitz, J. S., Lo, D.and Glimcher, L. H., Unopposed positive selection andautoreactivity in mice expressing class II MHC only onthymic cortex. Nature 1996. 383: 81–85.


9 Weigle, W. O. and Romball, C. G., CD4+ T-cell subsetsand cytokines involved in peripheral tolerance. Immunol.Today 1997. 18: 533–538.

10 Weigle, W. O., Scheuer, W. V., Hobbs, M. V., Morgan,E. L. and Parks, D. E., Modulation of the induction andcircumvention of immunological tolerance to humangamma-globulin by interleukin 1. J. Immunol. 1987. 138:2069–2074.

11 Cella, M., Engering, A., Pinet, V., Pieters, J. and Lan-zavecchia, A., Inflammatory stimuli induce accumula-tion of MHC class II complexes on dendritic cells. Nature1997. 388: 782–787.

12 Schuler, G., Thurner, B. and Romani, N., Dendriticcells: from ignored cells to major players in T-cell-mediated immunity. Int. Arch. Allergy Immunol. 1997.112: 317–322.

13 De Smedt, T., Pajak, B., Muraille, E., Lespagnard, L.,Heinen, E., De Baetselier, P., Urbain, J., Leo, O. andMoser, M., Regulation of dendritic cell numbers andmaturation by lipopolysaccharide in vivo. J. Exp. Med.1996. 184: 1413–1424.

14 Reis e Sousa, C., Hieny, S., Scharton-Kersten, T.,Jankovic, D., Charest, H., Germain, R. N. and Sher, A.,In vivo microbial stimulation induces rapid CD40 ligand-independent production of interleukin 12 by dendriticcells and their redistribution to T cell areas. J. Exp. Med.1997. 186: 1819–1829.

15 Sparwasser, T., Miethke, T., Lipford, G., Borschert, K.,Häcker, H., Heeg, K. and Wagner, H., Bacterial DNAcauses septic shock. Nature 1997. 386: 336–337.

16 Sparwasser, T., Miethke, T., Lipford, G., Erdmann, A.,Häcker, H., Heeg, K. and Wagner, H., Macrophagessense pathogens via DNA motifs: induction of tumornecrosis factor- § -mediated shock. Eur. J. Immunol.1997. 27: 1671–1679.

17 Lipford, G. B., Sparwasser, T., Bauer, M., Zimmer-mann, S., Koch, E.-S., Heeg, K. and Wagner, H.,Immunostimulatory DNA: Sequence dependent produc-tion of potentially harmful or useful cytokines. Eur. J.Immunol. 1997. 27: 3420–3426.

18 Stacey, K. J., Sweet, M. J. and Hume, D. A., Macro-phages ingest and are activated by bacterial DNA. J.Immunol. 1996. 157: 2116–2122.

19 Lipford, G. B., Bauer, M., Blank, C., Reiter, R., Wagner,H. and Heeg, K., CpG-containing synthetic oligonucleo-tides promote B and cytotoxic T cell responses to pro-tein antigen: a new class of vaccine adjuvants. Eur. J.Immunol. 1997. 27: 2340–2344.

20 Roman, M., Martin-Orozco, E., Goodman, J. S.,Nguyen, M.-D., Sato, Y., Ronaghy, A., Kornbluth, R.S., Richman, D. D., Carson, D. A. and Raz, E., Immu-nostimulatory DNA sequences function as Th1-promoting adjuvants. Nat. Med. 1997. 3: 849–854.

21 Chu, R. S., Targoni, O. S., Krieg, A. M., Lehmann, P. V.and Harding, C. V., CpG oligodeoxynucleotides act asadjuvants that switch on T helper (Th1) immunity. J. Exp.Med. 1997. 186: 1623–1631.

22 Zimmermann, S., Egeter, O., Hausmann, S., Lipford,G. B., Röcken, M., Wagner, H. and Heeg, K., CpG oli-gonucleotides trigger curative Th1 responses in lethalmurine leishmaniasis. J. Immunol. 1998. 160:3627–3630.

23 Davis, L. H. and Krieg, A. M., CpG DNA is a potentenhancer of specific immunity in mice immunized withrecombinant hepatitis B surface antigen. J. Immunol.1998. 160: 870–876.

24 Sprent, J., Tough, D. F. and Sun, S., Factors controllingthe turnover of T memory cells. Immunol. Rev. 1997.156: 79–85.

25 Janeway, C. A. J., Approaching the assymptote? Evolu-tion and revolution in immunology. Cold Spring Harb.Symp. Quant. Biol. 1989. 54: 1–13.

26 Lipford, G. B., Wagner, H. and Heeg, K., Vaccinationwith immunodominant peptides encapsulated in Quil A-containing liposomes-induces peptide-specific primaryCD8+ cytotoxic T cells. Vaccine 1994. 12: 73–80.

27 Krieg, A. M., Yi, A. K., Matson, S., Waldschmidt, T. J.,Bishop, G. A., Teasdale, R., Koretzky, G. A. and Klin-man, D. M., CpG motifs in bacterial DNA trigger directB-cell activation. Nature 1995. 374: 546–549.

28 Tokunaga, T., Yano, O., Kuramoto, E., Kimura, Y.,Yamamoto, T., Kataoka, T. and Yamamoto, S.,Synthetic oligonucleotides with particular base sequen-ces from the cDNA encoding proteins of Mycobacteriumbovis BCG induce interferons and activate natural killercells. Microbiol. Immunol. 1992. 36: 55–66.

29 Klinman, D. M., Yi, A. K., Beaucage, S. L., Conover, J.and Krieg, A. M., CpG motifs present in bacteria DNArapidly induce lymphocytes to secrete interleukin 6,interleukin 12, and interferon gamma. Proc. Natl. Acad.Sci. USA 1996. 93: 2879–2883.

30 Sun, S., Beard, C., Jaenisch, R., Jones, P. and Sprent,J., Mitogenicity of DNA from different organisms formurine B cells. J. Immunol. 1997. 159: 3119–3125.

31 Bhardwaj, N., Young, J. W., Nisanian, A. J., Baggers,J. and Steinman, R. M., Small amounts of superanti-gen, when presented on dendritic cells, are sufficient toinitiate T cell responses. J. Exp. Med. 1993. 178:633–642.

32 Cella, M., Scheidegger, D., Palmer Lehmann, K.,Lane, P., Lanzavecchia, A. and Alber, G., Ligation ofCD40 on dendritic cells triggers production of high levelsof interleukin-12 and enhances T cell stimulatory capac-ity: T-T help via APC activation. J. Exp. Med. 1996. 184:747–752.


33 Caux, C., Massacrier, C., Vandervliet, B., Dubois, B.,Van Kooten, C., Durand, I. and Banchereau, J., Activa-tion of human dendritic cells through CD40 cross-linking. J. Exp. Med. 1994. 180: 1263–1272.

34 Verhasselt, V., Buelens, C., Willems, F., De Groote, D.,Haeffner-Cavaillon, N. and Goldman, M., Bacteriallipopolysaccharide stimulates the production of cyto-kines and the expression of costimulatory molecules byhuman peripheral blood dendritic cells: evidence for asoluble CD14-dependent pathway. J. Immunol. 1997.158: 2919–2925.

35 Bluestone, J. A., New perspectives of CD28-B7-mediated T cell costimulation. Immunity 1995. 2:555–559.

36 Freeman, G. J., Boussiotis, V. A., Anumanthan, A.,Bernstein, G. M., Ke, X. Y., Rennert, P. D., Gray, G. S.,Gribben, J. G. and Nadler, L. M., B7-1 and B7-2 do notdeliver identical costimulatory signals, since B7-2 butnot B7-1 preferentially costimulates the initial productionof IL-4. Immunity 1995. 2: 523–532.

37 Thompson, C. B., Distinct roles for the costimulatoryligands B7-1 and B7-2 in T helper cell differentiation?Cell 1995. 81: 979–982.

38 Lu, L., Bonham, C. A., Chambers, F. G., Watkins, S. C.,Hoffmann, R. A., Simmons, R. L. and Thomson, A. W.,Induction of nitric oxide synthase in mouse dendriticcells by IFN-gamma, endotoxin, and interaction withallogeneic T cells: nitric oxide production is associatedwith dendritic cell apoptosis. J. Immunol. 1996. 157:3577–3586.

39 Flesch, I. E., Hess, J. H., Huang, S., Aguet, M., Rothe,J., Bluethmann, H. and Kaufmann, S. H., Early interleu-kin 12 production by macrophages in response to myco-bacterial infection depends on interferon gamma andtumor necrosis factor alpha. J. Exp. Med. 1995. 181:1615–1621.

40 Ma, X., Chow, J. M., Gri, G., Carra, G., Gerosa, F.,Wolf, S. F., Dzialo, R. and Trinchieri, G., The interleukin12 p40 gene promoter is primed by interferon gamma inmonocytic cells. J. Exp. Med. 1996. 183: 147–157.

41 Inaba, K., Inaba, M., Romani, N., Aya, H., Deguchi, M.,Ikehara, S., Muramatsu, S. and Steinman, R. M., Gen-eration of large numbers of dendritic cells from mousebone marrow cultures supplemented with granulocyte/macrophage colony-stimulating factor. J. Exp. Med.1992. 176: 1693–1702.

42 Zal, T., Volkmann, A. and Stockinger, B., Mechanismsof tolerance induction in major histocompatibility com-plex class II-restricted T cells specific for a blood-borneself-antigen. J. Exp. Med. 1994. 180: 2089–2099.

Correspondence: Hermann Wagner, Institute of MedicalMicrobiology, Immunology and Hygiene, Trogerstr. 9,D-81675 Munich, GermanyFax: +49-89-4140-4868e-mail: h.wagner — lrz.tu-muenchen.de


Bacterial DNA and immunostimulatory CpG oligonucleotides trigger maturation and activation of murine...

Documents

Transcript of Bacterial DNA and immunostimulatory CpG oligonucleotides trigger maturation and activation of murine...