5- (and 6-) carboxyfluorescein diacetate succinimidyl ester
In the early phase of leishmaniasis three types of potential antigen-presenting cells, including epidermal Langerhans cells (LC), dermal dendritic cells (DC) and inflammatory DC, are localizedat the site of infection. Therefore, it has been a central question which cell type is responsible for the initiation of a protective immune response. In the early stage of an anti-Leishmania immune response, detectable Leishmania major antigen was localized in the paracortex of the draining lymph nodes (LN). Characterization of antigen-positive cells showed that L. major co-localized with DC of a CD11c+ CD8α– Langerin– phenotype. To determine the area of antigen uptake, dermis or epidermis, and to further define the type of antigen-transporting cells, L. major was inoculated subcutaneously and concurrently LC were mobilized with fluorescein isothiocyanate (FITC). After 3 days, DC carrying L. major antigenwere always FITC–, indicating a dermal and not an epidermal origin. Moreover, addition of L. major antigen to ex vivo isolated CD8α– and CD8α+ DC fromthe draining LN of L. major-infected C57BL/6 mice demonstrated that both DC subpopulations were able to stimulate antigen-specific T cell proliferation in vitro. Without addition of exogenous antigen only the CD8α– Langerin– DC were capable of stimulating antigen-specific T cell proliferation. Thus, we demonstrate that CD8α– Langerin– DC and not LC are the basis of the protective immune response to intracellular L. major parasites in vivo.
The protozoan parasites Leishmania spp. are known to infect a variety of mammalian hosts and causes diseases ranging from localized cutaneous to systemic visceral forms 1. Experimental cutaneous mouse leishmaniasis is widely used as a model for this parasitic disease. After the subcutaneous (s.c.) infection of mice with L. major, the obligatory intracellular parasites reside within phagocytic cells, mostly monocytes/macrophages and DC 1, 2. It has been demonstrated that the course of mouse leishmaniasis strongly depends on factors determined by the genetic background 1–3. The resistant mouse strain C57BL/6 shows an early induction of IL-12 and an expansion of parasite-specific IFN-γ- and IL-2-producing Th1 cells 4. These mice are able to control the pathogen predominantly at the site of infection and to resolve the lesion. In contrast, susceptible BALB/c mice show a progressive and fatal disease 1, 2. This course of disease is due to an expansion of IL-4-, IL-5-, and IL-10-producing Th2 cells 3 propagated by the early presence of large amounts of IL-4 5.
DC are central for antigen presentation and the priming of naive T cells. They capture antigen and migrate to lymphoid organs, where they induce a specific cellular immune response 6, 7. In experimental leishmaniasis, L. major amastigotes interact with 8 or reside within epidermal Langerhans cells (LC) 9. Fetal skin-derived DC, a cell line with the characteristics of immature DC, take up amastigotes and respond to this challenge with the production of IL-12, the cytokine that ultimately determines the type of the anti-leishmanial immune response 10. Moreover, it has been demonstrated that isolated epidermal LC pulsed with L. major antigen in vitro and adoptively transferred to an infected host, have the capacity to transport and to present L. major antigen to T cells and to establish immunity 11.
It has not yet been attempted to isolate and characterize L. major-presenting DC directly isolated ex vivo because, after migration to the draining lymph node (LN), it was not possible to distinguish between LC and dermal DC. However, a set of recently defined marker molecules enables this differentiation. Recent studies have indicated that LC up-regulate CD8α upon migration to the draining LN, whereas dermal DC remain CD8α–12–14. Additionally, LC strongly express the C-type lectin Langerin 15 and are still positive for this molecule after they have reached the LN 16. Therefore, to solve the central question, which DC subtype is transporting L. major antigen and inducing protective immunity in vivo, we analyzed the phenotype of L. major-harboring DC using these novel markers. Furthermore, we separated DC subpopulations from the draining LN of L. major-infected mice and compared the potential of CD8α– DC and CD8α+ LC to induce a Leishmania-specific T cell proliferation.
2.1 L. major parasites are located in the paracortex of the draining LN
We analyzed sections of the central part of draining popliteal LN of infected mice at days 3 and 6 after infection for the presence of L. major antigen. At both of time points after infection, L. major-specific signals could be detected in the paracortical area of the LN (Fig. 1A). At day 6 after infection, the LN was substantially enlarged and a highly significant increase of the detectable L. major antigen in the paracortical area per LN section was visible (Fig. 1B). To quantify the number of L. major-positive signals per LN section the signals were counted in five 10-μm median sections of four to six different LN from infected mice at both time points and found to increase from 40 signals at day 3 to 180 signals at day 6 (p<0.001, Fig. 1C). The number of signals detected by immunofluorescence microscopy was consistent with results of limiting dilution assays published earlier that showed a similar order of magnitude of viable parasites in the draining LN in the early phase of infection 17, 18.
2.2 Characterization of the cell type harboring L. major antigen
Popliteal LN sections of infected mice were stained with a serum specific for L. major antigen and mAb against either CD11c, F4/80, CD8α, or CD11b. Analysis of sections with anti-L. major serum in combination with anti-CD11c mAb indicated a co-localization of parasite antigen with CD11c+ DC (Fig. 2A). In contrast, a combination of anti-L. major serum with anti-F4/80 (macrophages, Fig. 2B) or anti-CD8α mAb (T cells and LC, Fig. 2C) revealed only distinct L. major-negative cells. Finally, the application of anti-L. major serum in combination with an anti-CD11b mAb (monocytes and some subpopulations of DC) showed in some cells a co-localization (Fig. 2D, E). To exclude an extracellular localization of L. major parasites, a triple staining with phalloidin (f-actin and c-actin 19), anti-L. major serum, and 4′,6-diamidino-2-phenylindole (cell nucleus) was performed. This staining confirmed that the parasites resided in very close proximity to or within cells (data not shown).
To further analyze the phenotype of L. major-positive DC populations, CD8α+ and CD8α– DC subpopulations were isolated from the draining LN of infected animals (sorting purity >95%) and analyzed using reverse transcription (RT)-PCR. CD8α+ and CD8α– DC were mature and expressed both CC chemokine receptor 7 and IL-12 (p40) mRNA (Fig. 2F). The CD8α+ DC population was L. major mRNA-negative and expressed the LC-specific molecule Langerin (Fig. 2F). In contrast, CD8α– DC carried L. major mRNA and where Langerin– (Fig. 2F).
To exclude that the route of infection (i.d. versus s.c.) influenced which DC subpopulation transports L. major to the LN, we inoculated C57BL/6 mice i.d. in the ear and analyzed the phenotype of L. major-positive DC reaching the draining LN 6 days after infection. After i.d. injection L. major antigen co-localized exclusively with DC of a CD11c+ CD8α– Langerin– phenotype in the draining LN (data not shown).
2.3 L. major antigen is transported to the draining LN by DC, but not by LC
To further analyze the DC population that transports L. major antigen, promastigote parasites were inoculated s.c. Simultaneously, the site of infection was painted with FITC, to mobilize and monitor the migration of epidermal LC to the draining LN 20, 21. This combined insult resulted in migration of dermal and epidermal DC to the paracortical area of the popliteal LN. Three days after infection the LN was analyzed by means of immunohistology. All L. major-positive DC within the paracortex were negative for FITC (Fig. 3A). The phenotype of FITC+ cells was analyzed with CD8α as a marker for epidermal LC. Staining with mAb specific for CD11c and CD8α disclosed that FITC+ cells were CD11c+ and CD8α+ (Fig. 3B), demonstrating that these cells were of epidermal origin and belonged to the LC population 14. This confirms that not CD8α+ FITC+ LC but CD8α– FITC– DC transport L. major to the draining LN.
2.4 Competence of ex vivo isolated CD8α+ and CD8α– DC to present L. major antigen to T cells
In the draining LN of infected mice the number of CD8α– DC had increased 20-fold from 2,500±600 cells/LN to 50,000±12,500 cells/LN (mean ± SEM), and the number of CD8α+ DC sevenfold from 1,700±400 cells/LN to 12,500±3,100 cells/LN (mean ± SEM) at day 6 after infection. To test the antigen-presenting capacity of CD8α+ and CD8α– DC populations, both subpopulations of DC were isolated ex vivo from LN of L. major-infected C57BL/6 mice at day 6 after infection (sorting purity >95%). In parallel, CD4+ T cells were isolated from the same LN, labeled with 5-(and 6-) carboxyfluorescein diacetate succinimidyl ester (CFSE) and mixed with either of the isolated DC subpopulations with or without adding further antigen (Fig. 4A). The addition of exogenous L. major antigen or Con A to all combinations of cells always led to strong proliferation (Fig. 4A). This indicates that the cells were healthy and reactive (Con A) and that the antigen-specific CD4+ T cells were able to proliferate in the presence of abundant L. major antigen presented by CD8α+ (34% proliferating CD4+ T cells) or CD8α– (37% proliferating CD4+ T cells) DC populations equally well.
However, in the absence of added antigen only CD8α– DC were able to induce an antigen-specific T cell proliferation (21% proliferating CD4+ T cells, Fig. 4A) whereas T cell proliferation in the presence of CD8α+ DC was not substantially above background (9% proliferating CD4+ T cells, Fig. 4A). This is also shown by the proliferation index, which summarizes the proliferation in the absence of exogenous antigen of three independent experiments (Fig. 4B) and demonstrates that the CD8α– DC subpopulation carries L. major antigen and can induce an immune response. Furthermore, this result is consistent with our histological data (Fig. 1, 2) and PCR results (Fig. 2F).
We investigated the phenotype of DC presenting L. major antigen in the draining LN in vivo and studied the potential of DC subpopulations isolated ex vivo to induce T cell proliferation. Our results show that at days 3 and 6 after infection with L. major, DC of the dermal compartment and not LC that reside in the epidermis are responsible for the transport of L. major antigen to the draining LN. Recent observations have made this differentiation possible. It has been demonstrated that LC up-regulate the CD8α chain after migration to the draining LN 12–14, 16, 22, 23. This has been used, together with the differential expression of the surface antigens DEC 205, CD11b, and CD4, to define at least five different DC subpopulations in the skin draining LN 16. Furthermore, only LC of the epidermal compartment are strongly positive for the C-type lectin Langerin 15 and are still positive for this molecule after they have reached the LN. This marker molecule has made it possible to clearly identify the LC population 16.
Our results show that after i.d. or s.c. injection the L. major antigen that is detectable in the paracortical area of the LN is associated with mature CD11c+, CD8α–, F4/80–, and Langerin– DC and not with CD8α+ Langerin+ LC. Furthermore, only the CD8α– DC population can present endogenous antigen and stimulate T cell proliferation in this model of leishmaniasis. The L. major infection with a concurrent skin painting that labels epidermal LC also demonstrates that only DC of the dermal compartment, but not LC, take up L. major antigen. This interaction between antigen-presenting DC of one skin compartment and a pathogen occurs with a specificity that was not anticipated. The finding that CD8α– DC present antigen argues against cross-presentation as the underlying mechanism of antigen presentation. The uptake of dying DC in the LN by LN-resident DC and the subsequent presentation of their antigenic contents to T cells has only been described for CD8α+ DC 24, 25.
It has been reported that after an inflammatory stimulus the presentation of antigen from the skin in the draining LN is carried out sequentially by different subpopulations of DC 26. In the very early stage of the inflammatory response (3 h after transfer of antigen), both DC of dermal origin and LC that had migrated to the LN shortly before the immunization, presented antigen that had been swept to the LN passively. This first wave of antigen presentation was based on a small amount of antigen, but induced T cell priming and a T cell response in a transfer system of transgenic T cells 26. Later, inflammatory DC transported antigen and presented it. The efficiency of this antigen presentation was significantly higher 26.
In our infection model of leishmaniasis, we were neither able to detect antigen in the LN nor to show the induction of a T cell response within the first 24 h after infection (U.R. and H.K., unpublished results). Obviously, the threshold of detection in our experiments was much higher since we used for detection CD4+ LN T cells which were not enriched for L. major-specific T cells. The CD11c+, CD8α–, F4/80–, and Langerin– DC that are co-localized with L. major antigen in our experiments are consistent with the antigen-transporting DC that collect antigen at the site of immunization, transport it to the draining LN, and present this antigen highly efficiently 26. Although the antigen that leaks to the draining LN induces a T cell response within hours after immunization, and although we cannot formally rule out a contribution of passively transported antigen that is taken up in the LN by DC to the anti-L. major immune response, the results by Itano and colleagues 26 regarding the time frame of the sequential presentation of antigens suggest that passively transported antigens do not significantly contribute to the data presented.
The origin and phenotype of the DC population in an inflammatory situation in the skin are not yet well defined. Inflammatory dermal DC probably descend from monocytic precursor cells that areattracted to the site of infection by inflammatory mediators and differentiate into dermal monocytic DC while transmigrating from the blood to the inflamed tissue 27–29. The use of fluorochrome-labeled microspheres injected intracutaneously together with FITC-painting pointed to a highly specific use of a CD8α– DC subpopulation. DC that had phagocytosed microspheres were positive for the fluorochrome and expressed CD11c and MHC class II, but lacked CD8α 29. In contrast, a recent study investigating epidermal herpes simplex virus infection demonstrated that CD8α+ DC, but not LC, transport and present antigen 30. The usage of CD8α+ DC in this model could be due to the epidermal localization of the challenge. Further studies that deepen our understanding of the early events in inflammation will be needed to resolve these discrepancies between different models.
In conclusion, our study indicates that dermal DC or inflammatory DC are the natural source of L. major antigen in the experimental model of leishmaniasis. This challenges the view that LC are central in the control of L. major infection 11. However, this notion has been based to a large extent on indirect studies that demonstrated the potential of LCto migrate and to present antigen in vitro8, 9, 11, 31. Meanwhile, different studies demonstrated that an immune response within the skin is based on a range of DC subpopulations residing within, or migrating to the skin, after an inflammatory stimulus but does not depend on skin-resident LC 26, 29. These findings together with our study do not rule out a role for LC in the immune response to skin infections like L. major infection. LC could, for example, be involved in the conditioning of the tissue for an inflammatory response 32, but the role of LC has certainly to be redefined.
4 Materials and methods
Inbred C57BL/6 mice were purchased from Charles River (Sulzfeld, Germany). Mice at an age of 8–16 weeks were used.
4.2 Parasites, preparation of antigen and infection of mice
The cloned virulent L. major isolate (MHOM/IL/81/FE/BNI) 33 was used for infection experiments. The course of disease was monitored daily as described 34. Alternatively, the parasites were subjected to four cycles of rapid freezing and thawing to prepare L. major antigen 34. Mice were infected s.c. in one hind footpad or i.d. in the ear with 3×106 stationary-phase promastigotes/foot pad or ear of the third to seventh in vitro passage in a final volume of 50 μl (ear: 10 μl) 34.
The following antibodies were used for immunofluorescence microscopy and flow cytometry: hamster anti-mouse CD11c (clone HL3; biotinylated or PE-conjugated), rat anti-mouse CD11b (clone M1/70;biotinylated or FITC-conjugated), rat anti-mouse CD8α (clone 53–6.7; unlabeled, PE-, or FITC-conjugated), rat anti-mouse pan-CD45 (clone 30-F11; unlabeled), rat anti-mouse CD45R/B220 (clone RA3682; biotinylated). These mAb were purchased from Becton Dickinson/PharMingen (Heidelberg, Germany). Rat anti-mouse F4/80 (clone CI.A3–1; biotinylated) was obtained from Dianova (Hamburg, Germany).The production of a L. major-specific polyclonal serum has been described 35. The following secondary reagents were used for detection of purified or biotinylated primary mAb: Alexa Fluor 488- and 546-conjugated goat anti-rat Ig and streptavidin-Alexa Fluor 488 (Mobitec, Göttingen, Germany), rhodamine-conjugated donkey anti-rabbit IgG, Cy5-conjugated goat anti-rat IgG, Cy5-conjugated streptavidin (Dianova), PerCP (Becton Dickinson/PharMingen).
4.4 Immunofluorescence microscopy
Tissue blocks from LN were embedded in optimal cutting temperature compound (OCT; Diatec, Hallstadt/Bamberg, Germany) and stored at –70°C. Tissue sections (10 μm) were thawed onto gelatin-coated glass slides, air-dried, fixed in acetone (for 10–15 min, at –20°C) and rehydrated with PBS and 0.01% Tween-20 (Sigma). Nonspecific binding sites were blocked for 30 min at room temperature with PBS containing 1% BSA and 10% mouse serum. The sections were first stained with rabbit anti-L. major serum in PBS/BSA and 0.1% saponin (Roth, Karlsruhe, Germany). Rabbit anti-L. major serum was revealed with a rhodamine-conjugated donkey anti-rabbit IgG. A biotinylated mAb was used for a second layer (either anti-CD11c, anti-CD11b, anti-CD45R/B220, or anti-CD8α) followed by Alexa 488- or Cy5-conjugated streptavidin. Alternatively, unlabeled F4/80 was used, followed by incubation with FITC anti-rat IgG. Each incubation step lasted 30 min at room temperature alternating with five washing steps in PBS/Tween (0.01%, 10 min, room temperature). Sections were analyzed using an immunofluorescence microscope (Zeiss, Jena, Germany) equipped with high-sensitivity gray scaledigital camera (Openlab System; Improvision, Heidelberg, Germany). Separate color images were collected for each section, analyzed and merged afterwards. Final image processing was performed using Corel Photo-Paint software (Corel Corporation, Ottawa, Canada).
4.5 Isolation of cells from LN and skin
Draining popliteal LN and foot pads were harvested and digested with collagenase D (1 mg/ml; Boehringer Mannheim, Mannheim, Germany) in Hanks' balanced salt solution (plus Ca2+, Mg2+) for 30 min at 37°C. The reaction was stopped with 5 mM EDTA for 10 min at 37°C. Digested tissues were passed through a stainless steel sieve and subsequently filtered through a 70-μm cell strainer (Becton Dickinson, Heidelberg, Germany) and washed twice with FACS buffer containing PBS, 0.1% BSA, and 0.02% sodium azide. The number of viable cells was determined by Trypan blue exclusion.
4.6 Flow cytometry and flow cytometric cell sorting
Multi-color flow cytometry was performed as described 36. Cells isolated from the skin and LN (see above) were washed and resuspended in PBS containing 0.1% BSA, sodium azide and stained directly or indirectly with fluorochrome-conjugated mAb. Cells were analyzed by flow cytometry. To exclude cell debris, cells were electronically gated according to light scatter properties and the hemopoietic marker CD45. Data were collected using a FACSCalibur flow cytometer (Becton Dickinson). Analysis was performed using CellQuest software (Becton Dickinson). Alternatively, cells isolated from skin and draining LN were labeled and subjected to flow cytometric cell sorting using a MoFlo highspeed cell sorter (Cytomation Bioinstruments, Freiburg, Germany). After sorting, purity of cells was controlled with the FACSCalibur.
4.7 Proliferation assay
Six days after infection mice were sacrificed and draining popliteal LN (n=6–8 per group) were removed. Pooled popliteal LN cells were stained for CD11c, CD8α and CD4. CD4+ T cells and antigen-presenting DC were isolated from LN cells according to the following phenotypes: CD11c– and CD4+ (CD4+ T cells), CD11c+ and CD8α– (dermal DC), CD11c+ and CD8α+ (LC). To monitor cell proliferation, CD4+ T cells were incubated with CFSE (Mobitec) at a final concentration of 1 μM in PBS for 10 min at37°C 37. CFSE-labeled T cells (1.5×105) were incubated in microtiter plates in RPMI 1640 supplemented with 10% FCS (Sigma), glutamine, Hepes, and antibiotics (Seromed-Biochrom, Berlin, Germany). DC (either CD8α+ or CD8α–) were added in a ratio of 1:10 and cultures were either stimulated with lysed L. major promastigote total antigen (at a ratio of ten parasites per total number of cells) or with Con A (final concentration 1 μg/ml). Background proliferation was estimated using isolated CD11c+ CD8α– or CD11c+ CD8α+ DC from non-infected C57BL/6 mice together with L. major-specific CD4+ T cells.
After 72 h of incubation, supernatants were removed and stored at –70°C. Cells were harvested and beads (Calibrate Beads; Becton Dickinson; 1.0×104 beads/sample) were added to the cells to quantify the proliferating cells and to determine the proliferation index according to the formula: (Proliferating cells/bead in cultures plus DC) / (Proliferating cells/bead in cultures without DC). Proliferating cells and beads were quantified with a FACSCalibur.
4.8 RNA extraction and RT-PCR
RNA was extracted from sorted cells or tissue using the perfect RNA mini kit (Eppendorf, Hamburg, Germany) according to the manufacturer's instructions. All samples were treated with RNase-free DNase (Promega, Mannheim, Germany) for 15 min followed by chloroform/phenol extraction to remove the enzyme. First-strand cDNA was synthesized from 1–2 μg of total RNA using murine Moloney leukemia virus reverse transcriptase (Promega) and oligo(dT) primer (Gibco Life Technologies, Karlsruhe, Germany). The primers, which were specific for CCR7 (sense: 5′-ATTTCTACAGCCCCCAGAGC-3′, antisense: 5′-TGAGCCTCTTGAAATAGATGTACG-3′), CD8α (sense: 5′-CACGAATAATAAGTACGTTCTCACC-3′, antisense: 5′-ATGTAAATATCACAGGCGAAGTCCA-3′), β-actin (sense: 5′-AATCCTGTGGCATCCATGAAAC-3′, antisense: 5′-CGCAGCTCAGTAACAGTCCG-3′), L. major actin (sense: 5′-TGACAACGAGCAGAGCTCCA-3′, antisense: 5′-CCCACGATCGAAGGGAAAA-3′), and mouse-Langerin (sense: 5′-ACGCACCCCAAAGACCTGGTACAG-3′, antisense 5′-AGACACCCTGATATTGGCACAGTG-3′) 38, were used at a concentration of 200 nM. Amplification (35 cycles; 20 s 94°C, 20 s 58°C, and 50 s 72°C) was performed with 1 U Taq polymerase/reaction (PAN, Aidenbach, Germany).
4.9 Skin painting
Mice were infected with 3×106 L. major promastigotes as described above. Immediately after infection the footpads were painted with 400 μl FITC (5 mg/ml dissolved inequal volumes of dibutylphtalate and acetone). After 3 and 6 days, popliteal LN were stained with biotinylated mAb layer (either anti-CD11c or anti-CD8α), or L. major-specific serum (as described above). Sections were analyzed by immune fluorescence microscopy to determine the phenotype of migrated FITC+ cells.
4.10 Presentation of results and statistical analysis
The results in the figures are expressed as means ± SEM. Differences between experimental groups were tested for statistical significance by Student's t-test for unpaired samples (two-tailed).
The authors thank Prof. Bogdan and Dr. Lutz for critical reading of the manuscript. This work was supported by the Deutsche Forschungsgesellschaft (H.K.; Ko 1315/3–3, Ko 1315/4–1) and by the Federal Ministry of Education and Research (BMBF) and the Interdisciplinary Center for Clinical Research (IZKF) at the University Hospital of the University of Erlangen-Nürnberg, Germany (H.K., IZKF Nachwuchsgruppe 1).