CD8 dendritic cells preferentially cross-present Saccharomyces cerevisiae antigens



Mouse splenic dendritic cell (DC) subsets possess distinct antigen-presentation abilities. CD8+ DC are specialized in cross-presentation of antigens to CD8+ T cells, whereas CD8 DC are more efficient in antigen presentation to CD4+ T cells. In this study, we examined the capacity of CD8+ and CD8 DC subsets to present fungal antigens in MHC class I and II molecules to CD8+ and CD4+ T cells, respectively. We used ovalbumin-expressing Saccharomyces cerevisiae (yeast-OVA) as a fungal model system. Both CD8+ and CD8 DC subsets phagocytosed yeast in equal amounts and uptake was mediated via dectin-1. In addition, both DC subsets induced similar OVA-specific CD4+ T cell proliferation after incubation with yeast-OVA. However, the induction of OVA-specific CD8+ T cell activation was largely restricted to the CD8 DC subset. Furthermore, only CD8 DC produced cytokines such as IL-10 and TNF-α and increased IL-23p19 and IL-23p40 mRNA levels in response to yeast. Our results strongly suggest that DC subsets have different functions in the elicitation of adaptive immune responses in vivo.




DC-associated C-type lectin-1


complement receptor 3


mannose receptor


ICAM-3-grabbing nonintegrin (DC-SIGN) homolog, SIGN-related 1


Fungal infections with pathogenic yeasts such as Candida albicans, Aspergillus fumigatus, Histoplasma capsulatum and Cryptococcus neoformans are a major problem in immunocompromised persons. These opportunistic pathogens usually do not pose problems in healthy individuals, but frequently become invasive in patients suffering from HIV, patients with malignancies or transplant recipients, resulting in life-threatening diseases 13.

Both innate and adaptive immune responses are essential to induce protection against fungi. Dendritic cells (DC) play a central role in the adaptive immune response by capturing fungi and presenting their antigens in major histocompatibility complex (MHC) class I and MHC class II molecules to T cells 4, 5. DC express a panel of microbial recognition receptors such as Toll-like receptors (TLR) and C-type lectin receptors that are essential for the elicitation of adaptive immune responses 6. Of these receptors, TLR2 and the DC-associated C-type lectin (dectin-1) are specifically involved in recognition of fungal components 79. Zymosan and fungal pathogens like C. albicans and A. fumigatus are sensed by TLR2 that signals via myeloid differentiation factor 88 (MyD88), resulting in NF-κB activation and the production of cytokines 1013. The importance of this signaling cascade can be inferred from the fact that MyD88-knockout mice show an increased susceptibility to fungal infections 14, 15.

Dectin-1 is expressed on myeloid cells like macrophages, neutrophils, monocytes, and even a subset of T cells 16. Dectin-1 mediates fungal recognition and uptake by binding β-glucans 7, 9, 17, 18, which are conserved cell wall components of fungi and some bacteria. β-Glucans are hidden in a layer of mannosylated proteins and chitin, but they can be exposed on the surface of fungi in budscars and following heat-inactivation of the yeasts 19. Dectin-1 recognizes zymosan and Saccharomyces cerevisiae, but also pathogenic fungi like C. albicans, Pneumocystis carinii and A. fumigatus9, 17, 2022. Recent studies with dectin-1-deficient mice show the importance of dectin-1 for the immune response towards fungal infections in vivo23, 24.

Signaling through an immunoreceptor tyrosine-based activation motif (ITAM)-like motif allows dectin-1 to generate inflammatory responses in addition to its role as a phagocytosis receptor. Ligand binding results in phosphorylation of the ITAM motif, leading to the activation of the downstream tyrosine kinase Syk, CARD9 and initiation of phagocytosis, respiratory burst and production of the cytokines IL-2 and IL-10 25, 26.

In addition to dectin-1 and TLR2, complement receptors such as CR3 have also been shown to bind directly to non-opsonized zymosan via their lectin domains 27, 28. In addition, SIGNR1 29 and the mannose-receptor (MR) can also bind fungal components, but these receptors are mainly expressed by macrophages 11, 30, 31.

Although phagocytosis of fungi by macrophages has been studied extensively, the receptors involved in the uptake and presentation of fungal antigens by DC are less well known, nor is it known whether different types of DC are equally capable of recognizing and presenting fungal antigens. Murine splenic CD11c+ DC can be divided into different subsets, based on the differential expression of CD4, CD8 and other phenotypic markers, and it has been proposed that DC subsets have specialized functions in the induction and regulation of immunity and tolerance 32. Recently, CD8+ DC in the T cell areas of the spleen have been shown to preferentially activate CD8+ T cells, while CD8 DC, which are located in the marginal zone of the spleen, are more efficient in MHC class II presentation and activation of CD4+ T cells 33, 34. These results are consistent with the observations that CD8+ DC are specialized in cross-presentation of antigens derived from apoptotic cells to CD8+ T cells and that CD8+ DC can cross-present antigens from immune complexes in the absence of stimulatory signals 35, 36.

In this study we investigated the capacity of CD8+ and CD8 DC subsets to present fungal antigens to CD8+ and CD4+ T cells. Therefore, we used a model system in which we transfected S. cerevisiae with ovalbumin (yeast-OVA). Even though S. cerevisiae is not considered a pathogenic yeast, it is frequently used as a model system for pathogenic yeasts such as candida37. Using yeast-OVA we assessed the contribution of MR, SIGNR1, dectin-1, CR3 (CD11b/CD18) and TLR in the uptake and processing of fungal antigens by mouse DC. We detected uptake of S. cerevisiae by both CD8+ and CD8 DC subsets, which was mediated by dectin-1. Following yeast uptake, both DC subsets stimulated CD4+ T cell responses. Surprisingly, CD8 DC were significantly more efficient than CD8+ DC in cross-presentation of yeast antigens to CD8+ T cells and in cytokine production in response to yeast. This indicates that yeast antigens are processed differently in CD8+ and CD8 DC. Our results further support the hypothesis that DC subsets posses different roles in the priming of CD8+ T cell responses depending on the type of pathogen.


Yeast immunization results in activation of CD8+ and CD4+ T cells in vivo

S. cerevisiae was transfected with ovalbumin (yeast-OVA) to generate a model system in which we could investigate antigen processing and presentation to T cells. Western blot analysis showed an expression of 20 ng OVA protein/106 yeast cells (Fig. 1). We then analyzed whether yeast-OVA that was incubated with splenic CD11c+ DC would activate OVA-specific T cells in vitro. Yeast-OVA induced strong proliferation of OVA-specific TCR transgenic OT-I and OT-II cells in vitro (Fig. 2A and B). When compared with soluble OVA protein, yeast-OVA was at least 1000 times more efficient in stimulating OT-I cells. Yeast-OVA was approximately 4 times more efficient than OVA-loaded apoptotic cells that contain 2 ng/106 cells (unpublished observation, J. M. M. den Haan). This illustrates that yeast-OVA is very efficiently taken up, processed and presented in MHC class I to CD8+ T cells. In addition, presentation of yeast-OVA to CD4+ T cells was similarly strong (Fig. 2B).

Figure 1.

Immunoblot analysis of yeast-OVA. Cell lysates of 106 wild-type and OVA transfected yeast were run on 4–12% gel and blotted with polyclonal rabbit anti-OVA IgG. Indicated amounts of OVA protein were loaded as reference for OVA quantification.

Figure 2.

Yeast-OVA primes OVA-specific OT-I and OT-II T cells in vitro and in vivo. The capacity of splenic, CD11c-purified DC to induce T cell proliferation after incubation with yeast-OVA, OVA-loaded β2m–/– spleen cells or soluble OVA protein was compared in an in vitro T cell proliferation assay. For this, DC were cultured with a titration of indicated antigen sources together with purified OT-I (A) or OT-II (B) T cells. Black bars: OVA protein, 1:3 titration series, ranging from 270 μg/mL to 0.1 μg/mL. Gray bars: yeast-OVA, 1:3 titration series, ranging from 9 × 105 to 412 yeast cells per well. White bars: OVA-loaded β2m–/– cells, 1:3 titration series, ranging from 9 × 105 to 412 cells per well. After 2 days, cultures were pulsed with 1 μCi [3H]thymidine. T cell proliferation was measured after an additional co-culture of 16 h. (C, D) CFSE-labeled OT-I or OT-II OVA-transgenic T cells were transferred into C57BL/6 mice and 24 h later, yeast-OVA was injected intravenously. To determine proliferative expansion of transferred T cells, CFSE intensity of spleen cells was measured after 3 days by staining cells for Vα2 and CD8 or CD4. Plots are gated on CD8+Vα2+ cells for OT-I transferred mice (C) or CD4+Vα2+ cells for OT-II transferred mice (D). These are representative FACS figures of three mice per group analyzed in two independent experiments.

To determine whether CD8+ T cells and CD4+ T cell activation occurred in vivo following yeast immunization, we transferred CFSE-labeled OT-I and OT-II cells into B6 recipient mice. At 72 h after intravenous administration of heat-inactivated yeast-OVA, we determined the proliferation of OT-I and OT-II cells by analyzing the level of CFSE dye intensity (Fig. 2C and D). Whereas transferred T cells did not proliferate in the absence of immunization, clear proliferation of both OT-I and OT-II cells was observed after priming with yeast-OVA. In addition, transferred OT-I cells produced IFN-γ (data not shown). These data indicate that immunization with yeast resulted in the activation of both CD8+ and CD4+ T cells in vivo.

Antigen presentation of yeast by DC is dependent on β-glucan recognition but not on MyD88-mediated TLR signaling

The results described above prompted us to study the processing and presentation of yeast antigens by DC in more detail. Different receptors on DC have been described to be involved in yeast phagocytosis. To address the role of β-glucan binding receptors in yeast recognition, DC were co-incubated with laminarin, consisting of soluble β-glucans derived from seaweed Laminaria digitata, in the in vitro cross-presentation assay. Laminarin inhibited presentation of yeast-OVA by DC (Fig. 3A), showing that DC recognize yeast in a β-glucan-specific manner 9. In contrast, incubation with mannan did not influence the antigen-presenting capacities of DC for both MHC class I and class II, indicating that the mannose receptor and SIGNR1 are not essential for the uptake of yeast by DC (Fig. 3A and B).

Figure 3.

DC present yeast-OVA in MHC class I and class II after β-glucan-mediated yeast uptake. Splenic CD11c+ DC from B6 mice were co-incubated for 10 min with 200 μg/mL laminarin or 200 μg/mL mannan and then incubated with purified OT-I CD8+ T cells (A) or OT-II CD4+ T cells (B) and a 1:3 titration of yeast-OVA. After 2 days, cultures were pulsed with 1 μCi [3H]thymidine. T cell proliferation was measured after an additional co-culture of 16 h. DC isolated from B6 and MyD88 ko mice were co-cultured with yeast-OVA and OT-I CD8+ T cells (C) or OT-II CD4+ T cells (D), respectively. Error bars indicate SEM of triplicate wells. Data are representative of two independent experiments. ***p<0.001 versus medium control. p values were calculated by one-way ANOVA with Bonferroni's correction (GraphPad Prism 4 software).

To address the role of TLR2 in presentation of yeast antigens, we isolated splenic DC from B6 and MyD88–/– mice and studied their ability to present yeast-OVA to CD8+ and CD4+ T cells. MyD88 is an adaptor molecule required for signaling of the majority of TLR, including TLR2. No significant differences in OT-I priming (Fig. 3C) and OT-II priming (Fig. 3D) were observed between DC from wild-type mice and MyD88-deficient mice, showing that TLR2 is dispensable for cross-presentation of yeast-OVA in vitro.

In all thymidine-incorporation experiments, DC and T cells were incubated with wild-type S. cerevisiae not expressing OVA. This resulted in background proliferation of less than 4500 cpm.

Both CD8+ and CD8DC subsets express dectin-1

Given that β-glucan recognition of S. cerevisiae by DC was necessary for yeast antigen presentation, we focused on the expression of dectin-1 on mouse DC subsets. Dectin-1 is the most important receptor described to interact with β-glucans of different types of fungi 9, 17, 2022. Whereas neutrophils and macrophages show heterogeneous but strong expression of dectin-1 16, 38, dectin-1 expression on DC subsets is not well defined. Initially, subpopulations of both CD8+ and CD8 DC were described to express dectin-1 38, whereas others detected dectin-1 only on a fraction of CD8 DC 39. We observed low staining with anti-dectin-1 antibody on both DC subsets and both CD8 and CD8+ DC contained comparable numbers of dectin-1-positive cells (Fig. 4A and B).

Figure 4.

Both CD8+ DC and CD8 express dectin-1. (A) DC were gated on high CD11c expression with exclusion of autofluorescent cells. DC subsets were stained for the expression of dectin-1 and CD8α and analyzed by flow cytometry. Numbers indicate percentage of cells per quadrant. (B) Histograms show plots after gating on CD11chighCD8 (left) or CD11chighCD8+ (right) cells. Filled histograms show dectin staining on DC and gray lines show level of background staining with control IgG2b. Numbers indicate percentage of cells with specific staining above background. Data are representative of five independent experiments

Phagocytosis of GFP-yeast by both DC subsets is mediated by dectin-1

To determine the ability of splenic DC to take up yeast, we set up an in vitro phagocytosis assay using GFP-expressing S. cerevisiae. In this FACS assay we gated on CD11chigh cells to select DC (Fig. 5A). The fluorescence of endocytosed GFP-yeast (upper right quadrant) was lower than that of free yeast (upper left quadrant), possibly due to the acidic environment in the endosomes 40. To discriminate between extracellular binding and phagocytosis, the tests were performed at 4°C and 37°C: 22.3 ± 1.6% of total splenic CD11chigh DC showed GFP fluorescence at 37°C, as compared to 9.0 ± 2.4% at 4°C. Adding anti-dectin-1 antibodies inhibited phagocytosis of GFP-yeast by CD11c+ DC cells to background levels (13.3 ± 1.5%) (Fig. 5A).

Figure 5.

Both CD8+ DC and CD8 DC phagocytose yeast in a dectin-1-dependent manner. (A) Phagocytosis of GFP recombinant yeast by total DC was studied by co-cultures of GFP-yeast and DC. After 2 h, GFP fluorescence in CD11c+ cells was determined by FACS. To determine nonspecific interactions of yeast with DC, control samples were incubated at 4°C. Dectin-1 was blocked by co-incubation with anti-dectin-1 antibodies. (B) CD8+ and CD8 DC subsets were analyzed for their capacity to take up GFP-yeast. Cells were gated on high CD11c expression with exclusion of autofluorescent cells. (C) To block phagocytosis of GFP-yeast indicated mAb, curdlan, mannan or laminarin were added to the cultures and the percentage of GFP+ DC of both CD8+ and CD8 DC was calculated. Error bars indicate SEM of triplicate wells. ***p<0.001 versus medium control. Data are representative of four independent experiments. p values were calculated by one-way ANOVA with Bonferroni's correction (GraphPad Prism 4 software).

Next we studied the phagocytic capacities of the different DC subsets. Analysis of GFP in gated CD11chighCD8 or CD11chighCD8+ DC revealed that both CD8 and CD8+ DC subsets phagocytosed GFP-yeast. The fluorescence intensity of GFP+ CD8 and CD8+ DC was similar, indicating that both DC types phagocytosed similar numbers of GFP-yeast (Fig. 5B). We calculated the percentage of GFP+ DC for both DC subsets (Fig. 5C): 24.2 ± 2.1% of CD8 DC and 27.2 ± 5.0% of CD8+ DC contained GFP-yeast, which also correlated with the percentage of CD8 and CD8+ DC expressing dectin-1 on the surface (Fig. 4).

To study whether CD8+ and CD8 DC subsets potentially used distinct recognition receptors for yeast uptake receptors, we analyzed phagocytosis of yeast by DC subsets in the presence of blocking antibodies to CD11b and dectin-1 (Fig. 5C). We also tested the blocking effects of laminarin, curdlan and mannan in the phagocytosis assay (Fig. 5C). Curdlan is a dectin-1-specific agonistic β-glucan preparation. The internalization of yeast was inhibited to background levels after blocking dectin-1 with anti-dectin-1 antibody, soluble laminarin as well as curdlan. These inhibitory effects were similar for both CD8+ and CD8 DC subsets. Blocking CR3 with anti-CD11b antibodies did not influence phagocytosis by either DC subset, indicating no function for CR3 in the uptake of S. cerevisiaein vitro by DC. In addition, blocking the mannose receptor and SIGNR-1 with mannan did not inhibit uptake of yeast by either DC subset. These data suggest that similar percentages of both CD8+ and CD8 DC subsets recognize and phagocytose yeast in a dectin-1-dependent way.

Presentation of yeast antigens by CD8+ and CD8 DC subsets

Since both DC subsets phagocytosed yeast to the same degree, we studied antigen-presentation capacities of CD8+ and CD8 DC subsets. For this, we cultured sorted CD8+ and CD8 DC with yeast-OVA and with OT-I and OT-II cells. The capacity to cross-present yeast antigens in MHC class I molecules differed remarkably between DC subsets. As depicted in Fig. 6A, CD8 DC were very efficient in stimulating OT-I T cells after co-incubation with yeast-OVA. The cross-presenting capacity of this DC subset was more efficient than that of the initial total DC preparation. In contrast, the CD8+ DC subset stimulated OT-I T cells to a significantly lesser extent. The DC subsets did not differ in their capacities to stimulate OT-II T cells in an MHC class II-restricted manner (Fig. 6B). Moreover, both DC subsets were able to prime OT-I and OT-II T cells to the same extent when directly pulsed with the relevant OVA-peptide (data not shown). When sorted CD8+ and CD8 DC were co-incubated with apoptotic spleen cells from β2-microglobulin (β2m)-deficient mice that were shocked with OVA, only the CD8+ DC subset could cross-present cell-associated OVA antigen, as described previously 35, showing that sorted CD8+ DC were still functional (data not shown). The results clearly show that cross-presentation of yeast antigens is predominantly performed by CD8 DC and suggest that CD8+ and CD8 DC differentially process yeast antigens.

Figure 6.

Preferential cross-presentation of yeast antigens by CD8 DC. CD11chigh CD8 and CD8+ DC were FACS sorted and tested for their ability to stimulate naive OT-I (A) and OT-II cells (B) after co-incubation with yeast-OVA. After 2 days, cultures were pulsed with 1 μCi [3H]thymidine and T cell proliferation was measured after an additional co-culture of 16 h. The different bars indicate the number of sorted DC in the well. One of four experiments with similar results is shown. Error bars indicate SEM of triplicate wells. ***p<0.001 CD8+ DC versus CD8 DC. p values were calculated by one-way ANOVA with Bonferroni's correction (GraphPad Prism 4 software). (C, D) Sorted DC subsets were cultured for 48 h or 3 h (E, F) in the presence or absence of yeast cells. Supernatants were analyzed for IL-10 (C) and TNF-α (D), or DC were harvested after 3 h and mRNA was extracted and subjected to RT-qPCR analysis to determine mRNA levels of IL-12p40 (E) and IL-23p19 (F) (normalized to HPRT). Error bars indicate SEM of triplicate wells. Data are representative of two independent experiments.

Cytokine production by CD8+ and CD8 DC after stimulation with yeast

To determine whether CD8+ and CD8 DC also differ in cytokine production after interaction with yeast, we cultured sorted CD8+ and CD8 DC subsets with yeast for 48 h and analyzed the supernatants for several cytokines. We could not detect any IL-12p70, MCP-1 or IFN-γ. However, CD8 DC but not CD8+ DC specifically produced IL-10 and TNF-α after stimulation with yeast (Fig. 6C and D). Furthermore, higher production of IL-6 was observed by CD8DC compared to CD8+ DC (97 vs. 19 pg/mL). These results are in line with previous reports 4143, and indicate restricted cytokine production by CD8 DC after recognition of yeast.

Fungal stimulation of DC has been reported to lead to IL-23 production and subsequent activation of Th17 cells 42, 44. We cultured sorted CD8+ and CD8 DC subsets with yeast for 3 h and measured mRNA levels of IL-23p19, IL-12/23p40 and IL-23p35. Yeast stimulation of the CD8 DC subset resulted in increased IL-23p19 and IL-12p40 mRNA levels (Fig. 6E and F). CD8+ DC already had high mRNA levels of IL-12p40 under unstimulated conditions that did not change significantly after yeast stimulation. Stimulation with yeast did not induce the transcription of IL-12p19 in CD8+ DC (Fig. 6F). The second subunit of the IL-12 heterodimer, IL-12p35, could not be detected in either DC subset (data not shown). This indicates that specifically the CD8 DC subset is capable of producing IL-23 in response to yeast.


Little is known about the recognition of fungi and fungal antigen processing and presentation by different types of DC in vivo. Here, we demonstrate that S. cerevisiae antigens are presented very efficiently as compared to other well-studied antigen like soluble OVA and apoptotic cells, resulting activation of both CD4+ and CD8+ T cells in vitro and in vivo. Both murine splenic CD8+ and CD8 DC take up and present S. cerevisiae antigens via dectin-1-mediated recognition of β-glucans. We did not find a role for other receptors such as the mannose receptor, SIGNR1, or CR3 (CD11b/CD18), which is in line with other studies 9, 29, 45. Furthermore, we observed no requirement for MyD88-dependent signaling for uptake and presentation of yeast antigens to T cells, which is consistent with the observations that phagocytosis, respiratory burst and the production of certain cytokines are TLR-independent and mediated by dectin-1 18, 24, 26.

Initial reports stated dectin-1 expression on a fraction of both CD8+ and CD8DC subsets 38. In contrast, more recent studies suggested preferential expression by CD8 DC 34, 39. Although we did not detect strong dectin-1 expression by the DC subsets, approximately 25–35% of both CD8+ and CD8 DC expressed dectin-1 and similar percentages of both DC subsets phagocytosed yeasts in a short-term assay. These latter data are in agreement with the study of Rogers et al.26 in which also comparable uptake of zymosan by both CD8 and CD8+ DC subsets was observed. In addition, both CD8+ and CD8 DC subsets stimulated CD4+ T cell responses after uptake of yeast OVA. These data together indicate that both DC subsets take up and process yeast and that the preferential activation of CD8+ T cells by CD8 DC is not the result of preferential uptake. Blocking experiments showed that the uptake was dependent on the recognition of β-glucans by dectin-1 for both DC subsets, suggesting that the low dectin-1 expression was functional. β-Glucans are conserved cell wall components of fungi and dectin-1 has been shown to recognize many pathogenic fungi like C. albicans, P. carinii and A. fumigatus9, 17, 2022. Therefore, dectin-1 expression on mouse DC would enable them to induce adaptive immune responses against these pathogens, although this has not extensively been investigated in vivo23, 24.

Our study shows that both CD8+ and CD8 DC subsets phagocytose yeast via dectin-1 recognition of β-glucans, and that both types of DC present yeast-derived antigens in an MHC class II-restricted manner. Strikingly, we observed that cross-presentation of yeast antigens in MHC class I is significantly more efficient in CD8 DC, and that only CD8 DC produce IL-10, IL-6 and TNF-α and showed increased mRNA levels of IL-23 in response to yeast. This indicates that even though the recognition receptors are similar for CD8+ and CD8 DC, these DC differ in their downstream signaling pathways and/or antigen-processing pathways. The preferential cross-presentation by CD8 DC is a surprising finding since CD8+ and not CD8 DC are considered to be specialized in cross-presentation of exogenous antigens to CD8+ T cells. This specific ability of CD8+ DC has been demonstrated many times with regard to the cross-presentation of cell-associated antigens, although this could potentially be due to their better capacity to phagocytose dead cells 35, 46, 47. In other model systems in which the amount of phagocytosis was controlled, the CD8+ DC subset also showed enhanced cross-presentation compared to CD8 DC 33, 34. More importantly, the CD8+ DC subset exclusively cross-presents antigens from herpes simplex virus, influenza virus, vaccinia virus, lymphocytic choriomeningitis virus (LCMV) and Listeria monocytogenes after in vivo infection 4850. Apparently, the CD8+ DC subset is the principal DC type to activate CD8+ T cell responses towards intracellular cytosolic pathogens.

Our data indicate that CD8 DC are the major cross-presenting DC for fungal antigens. Previously, Salmonella antigens have been shown to be cross-presented by both CD8 DC and CD8+ DC 51. Both pathogens are phagocytosed and killed by macrophages and neutrophils or evade the immune system by avoiding lysosomal degradation 52.Our data suggest that extracellular pathogens that are located in the phagosomes may preferentially be cross-presented by CD8 DC to CD8+ T cells.

How frequently fungus-specific CD8+ T cell responses occur in vivo and what their relevance is, is not known. We here show that OVA-specific CD8+ T cells proliferate in vivo and in vitro in response to cross-presentation of yeast-OVA by DC, which is consistent with previous studies 5. H. capsulatum infection leads to cross-presentation and a protective CD8+ T cell response in mice 53. Furthermore, CD8+ T cell responses specific for various types of yeast are detected in humans 54. This clearly indicates that cross-presentation of fungi results in the activation of CD8+ T cells in vivo.

CD4+ T cells are the most important anti-fungal effector cells of the adaptive immune system, but in the absence of CD4+ T cells, CD8+ T cell-mediated effector functions can play an important role 5557. Obviously, the CD8+ T cell-mediated anti-fungal response is particularly important for HIV patients. A few studies have reported direct killing of yeasts by CD8+ T cells 58, 59. Alternatively IFN-γ or TNF-α, produced by activated CD8+ T cells, could enhance anti-fungal activity of macrophages and neutrophils 60, since IFN-γ is used as therapy for fungal infections 61.

Previously, we have shown that CD8 DC can cross-present immune complexes 36. In CD8 DC this process is dependent on the signaling FcR γ-chain, whereas cross-presentation by CD8+ DC does not require the FcR γ-chain. We hypothesized that CD8+ DC are better cross-presenting cells in non-activated state, but that CD8 DC can cross-present after certain activation stimuli. Here, we did not find a role for MyD88 signaling in the cross-presentation of yeast-OVA by CD8 DC. However, dectin-1-mediated signaling may play a role in cross-presentation of fungal antigens by CD8 DC. Dectin-1 signaling has also been shown to be essential for IL-2 and IL-10 production in bone marrow-derived DC 26. In addition, dectin-1 cooperates with TLR2 to induce pro-inflammatory cytokines like TNF-α 17, 18. We detected IL-10, IL-6 and TNF-α production by CD8 DC. This suggests that dectin-1 signaling in CD8 DC, but not in CD8+ DC, induces cytokine production.

Interestingly, while both CD8 and CD8+ DC subsets induced CD4+ T cell responses after yeast uptake, only CD8 DC produced IL-6, IL-23 and TNF-α and not IL-12. These cytokines are described to promote the differentiation of IL-17-producing CD4+ T cells and C. albicans stimulation of DC results in the induction of Th17 cells 42, 44, 62, 63. Therefore, we speculate that CD8 DC may play a major role in the generation of these Th17 T cells during fungal infection.

In conclusion, we show that dectin-1 is the major receptor for S. cerevisiae on mouse splenic CD8+ and CD8 DC. While both CD8+ and CD8 DC express dectin-1, take up yeast and present yeast-derived antigens in MHC class II to CD4+ T cells, CD8 DC preferentially cross-present yeast-derived antigens in MHC class I to CD8+ T cells. In addition, only CD8 DC produce cytokines in response to yeast. Our studies indicate that CD8+ and CD8 DC subsets exhibit differences in antigen presentation and cytokine production that depend on the type of pathogen encountered. This suggests that DC subsets have different functions in the activation of adaptive immune responses in vivo.

Materials and methods


Female C57BL/6-J (B6) mice, 8–16 weeks of age, were obtained from Charles River (L'Arbresle, France). OT-I, OT-II and MyD88–/– mice (generated by Dr. Akira, Osaka University) were bred at the animal facility of the VU University Medical Center (Amsterdam, The Netherlands). OT-I and OT-II mice have transgenic Vα2Vβ5 T cell receptors that recognize the OVA257–264 peptide in the context of H2-Kb and the OVA323–339 peptide in the context of I-Ab, respectively. All mice were kept under specific pathogen-free conditions and used in accordance of local animal experimentation guidelines.


The multi-copy OVA-HA expression vector pFvL210 was generated by cloning of a C-terminally HA-tagged version of the chicken OVA cDNA from pAc-neo-OVA1 64 into the BamHI site of pTCG 65. The plasmid was transformed into yeast strain UCC7164 (a derivative of S288C) 66 and transformed cells were grown in synthetic media lacking tryptophan to select for the plasmid. OVA expression was induced by growth in media containing galactose. A derivative of S288C containing a GFP-tagged PGK1 gene 67 was purchased from Invitrogen. Cells were grown to late log phase, washed with cold PBS, and heat-inactivated by a 30-min incubation at 65ºC. For immunoblot analysis yeast cells were lysed in SUME buffer (10 mM MOPS, 1% SDS, 8 M urea, 10 mM EDTA, 1 mM DTT, pH 6.8) and run on a 4–12% SDS-PAGE gel. The blot was probed with polyclonal rabbit anti-OVA antibody (ICN Pharmaceuticals) followed by chemiluminescence detection.

Media, reagents and mAbs

Culture medium was Iscove's modified Dulbecco's medium (IMDM; Gibco, Paisley, UK) supplemented with 10% heat-inactivated FCS, 200 U/mL penicillin, 200 μg/mL streptomycin (BioWhittaker), 4 mM L-glutamine (BioWhittaker) and 50 μM 2-mercaptoethanol (Fluka) (IMDM/DC). For DC isolation and MACS purification, 10 mM EDTA/20 mM HEPES was added (IMDM/HE). mAb were used for flow cytometry and in blocking experiments. Anti-CD8α (clone 53–6.7), anti-CD11c (clone N418) and anti-CD11b (clone M1/70) were obtained from eBioscience (San Diego, CA). Unlabeled antibody against dectin-1 (clone 2A11, kindly provided by Dr. G. D. Brown) were detected by anti-rat IgG. Unlabeled anti-dectin-1, anti-CD11b and rat control IgG were used in blocking of yeast phagocytosis and in vitro antigen presentation. In these assays mannan (from S. cerevisiae), laminarin (from Laminaria digitata) and curdlan (from Alcaligenes faecalis) (all Sigma-Aldrich) were also used in concentrations as indicated.

Isolation of DC

Spleens of 5–15 B6 mice were cut into grain size pieces and incubated in 1 mL/spleen of 1 WU/mL (wunsch unit/mL) Liberase I (Roche Diagnostics, Mannheim, Germany) and 50 µg/mL DNase I (Roche) in medium without FCS and 2-mercaptoethanol with continuous stirring at 37°C for 30 min or until digested. EDTA was added to a 10 mM final concentration, and the cell suspension was incubated for an additional 5 min at 4°C. Red blood cells were lysed with ACK lysis buffer. Cells were washed once with IMDM/HE and undigested material was removed by filtration. DC were purified by positive selection with anti-CD11c MACS microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany) following manufacturer's protocol. Cells were stained for CD11c and CD8 and then subsequently cell sorting was performed with Moflow (Dako Cytomation) in HBSS with 25 mM HEPES by gating on high expression of CD11c and the absence or presence of CD8. Autofluorescent cells were excluded. Purity and viability of the sorted samples was determined by re-analysis on the cell sorter. Sorting for CD8 and CD8+ DC populations resulted in 85–90% pure populations.

Isolation of T cells

CD8+ T cells and CD4+ T cells were isolated from spleen and lymph nodes (LN) from OT-I and OT-II transgenic mice, respectively, and single-cell suspensions were obtained. Specific T cells were purified by negative depletion using bead-based T cell isolation kits (Dynal Biotec ASA, Oslo, Norway) following manufacturer's protocols. Purity of T cell preparations was between 80% and 90%.

Generation of OVA-loaded β2m–/– cells

Single-cell suspensions were prepared in serum-free medium from spleens from β2m-deficient mice. Cells were loaded with OVA by osmotic shock as described previously 64. In brief, ∼15 × 107 cells were incubated in 1 mL hypertonic medium (0.5 M sucrose, 10% polyethylene glycol 1000, and 10mM HEPES in RPMI 1640, pH 7.2) containing 10 mg/mL OVA (Calbiochem) for 10 min at 37°C. Pre-warmed hypotonic medium (13 mL; 40% H2O, 60% RPMI 1640) was added and the cells were incubated for an additional 2 min at 37°C. The cells were centrifuged, washed twice with cold PBS, and irradiated (1500 rad).

Flow cytometry

CD11c+ cells were incubated with 2.4G2 mAb (anti-FcγRII/III-CD16/32) on ice for 15 min to prevent nonspecific mAb binding and stained with mAb of interest for 20 min on ice. All procedures were performed using ice-cold PBS containing 1% BSA and 0.02% NaN3. Cells were fixed in 2% paraformaldehyde (PFA) and flow cytometry was performed on a FACSCalibur with the use of CellQuest software (Becton Dickinson, Franklin Lakes, NJ) with autofluorescent cells excluded.

In vivo T cell proliferation

T cells were isolated from spleens and LN from OT-I or OT-II transgenic mice. CD4+ or CD8+ T cells were stained with 5 μM CFSE (Molecular Probes) at 37°C in PBS + 0.1% BSA. After 10 min, cells were washed several times in IMDM with 10% FCS to stop labeling. Subsequently, 4 × 106–5 × 106 CFSE-labeled, OVA-specific Vα2Vβ5 T cells were injected intravenously, followed by yeast-OVA after 24 h. After 3 days, mice were killed and spleen cells were isolated and stained with CD8 or CD4, respectively, and Vα2 TCR. T cell proliferation was measured by CFSE fluorescence dilution.

Cross-presentation assay

Splenic DC were isolated from B6 or MyD88–/– mice. To detect yeast-OVA presentation, DC were used as stimulators for naive T cells in a [3H]thymidine-incorporation assay. For this, 105 CD11c+ DC were co-incubated with varying numbers of yeast-OVA, OVA loaded β2m–/– cells or purified OVA in the presence of 105 purified OT-1 CD8+ T cells, or 105 purified OT-II CD4+ T cells in triplicate in flat-bottom 96-wells plates (Nunc). As a positive control, stimulator cells were coated with 1 µg/mL MHC class I OVA257–264 peptide or 10 μg/mL MHC class II OVA323–339 peptide for 1 h and washed three times. After 48-h incubation in a CO2 incubator, plates were pulsed with 1 µCi [3H]thymidine/well. After an additional 16 h of incubation, [3H]thymidine incorporation was measured on a Wallac-LKB Betaplate 1205 liquid scintillation counter. Data were summarized as the mean and SEM.

Detection of cytokines

For in vitro stimulation, 105 FACS-sorted DC were cultured in 96-well flat-bottom plates in the presence or absence of yeast in IMDM/DC. Culture supernatants were collected at 48 h and assayed for the presence of inflammatory cytokines by Cytometric Bead Array (CBA, Mouse Inflammation kit, BD Bioscience) following the manufacturer's protocol. Data were summarized as the mean and SEM.

Cytokine mRNA analysis

Total RNA was prepared with TRIzol reagent (Invitrogen) from sorted CD8+ and CD8 DC. The cDNA was synthesized from total RNA with random hexamers and M-MuLV reverse transcriptase (Fermentas). Quantitative PCR was determined with PowerSYBRGreen (AB Applied Biosystems, Warrington, UK) incorporation for IL-23p19 (primer sequences: forward, 5′-AATCTCTGCATGCTAGCCTGG-3′, reverse, 5′-GATTCATATGTCCCGCTGGTG-3′) and IL-12p40 (primer sequences: forward, 5′-GAATGTCTGCGTGCAAGCTC-3′, reverse, 5′-ACATGCCCACTTGCTGCAT-3′) on an ABI Prism 7900HT Sequence Detection System (Applied Biosystems). Results were normalized to those obtained with HPRT and results are presented as relative quantities of mRNA.

Phagocytosis assay

To study the uptake of S. cerevisiae, 105 CD11c+ purified DC were stained for CD11c and CD8 and co-incubated in triplicate with GFP-recombinant yeast for 2 h at 37°C (E:T ratio =5). To block uptake, DC were pre-treated for 10 min with mannan, laminarin (both 200 μg/mL), curdlan (100 μg/mL), or the mAb against dectin-1, CD11b or control rat IgG (all 50 μg/mL). Background binding was determined by incubation at 4°C for 2 h. DC were washed with PBS containing 10 mM EDTA, PFA fixed and analyzed on a FACSCalibur flow cytometer. The percentage of DC that had phagocytosed GFP-yeast was calculated as the number of GFP+ cells divided by the total number of DC. This was done for each subset individually. Data were summarized as the mean and SEM.


This work was supported by grant NWO917.46.311 from the Dutch Scientific Research program. We thank the staff of our animal facility for the care of the animals used in this study. We thank T. O'Toole for expert technical assistance during FACS sorting and J. J. García Vallejo for providing RT-qPCR primers.


Conflict of interest: The authors declare no financial or commercial conflict of interest.


  1. 1


  2. 2


  3. 3


  4. 4


  5. 5


  6. 6