Microbial stimulation fully differentiates monocytes to DC-SIGN/CD209(+) dendritic cells for immune T cell areas

Abstract

Dendritic cells (DCs), critical antigen-presenting cells for immune control, normally derive from bone marrow precursors distinct from monocytes. It is not yet established if the large reservoir of monocytes can develop into cells with critical features of DCs in vivo. We now show that fully differentiated monocyte-derived DCs (Mo-DCs) develop in mice and DC-SIGN/CD209a marks the cells. Mo-DCs are recruited from blood monocytes into lymph nodes by lipopolysaccharide and live or dead gram-negative bacteria. Mobilization requires TLR4 and its CD14 coreceptor and Trif. When tested for antigen-presenting function, Mo-DCs are as active as classical DCs, including cross-presentation of proteins and live gram-negative bacteria on MHC I in vivo. Fully differentiated Mo-DCs acquire DC morphology and localize to T cell areas via L-selectin and CCR7. Thus the blood monocyte reservoir becomes the dominant presenting cell in response to select microbes, yielding DC-SIGN(+) cells with critical functions of DCs.

Figure 1. DCs derived from marrow monocytes express DC-SIGN and are potent APCs

(A) Marrow monocytes (Figure S1) were cultured in GM-CSF and IL-4 4–7 days. (Left) DIC image with typical dendritic morphology. (Right) Western blot with rabbit polyclonal αDC-SIGN and mAb KL295 αMHC II. (B) As in A, showing MHC II, CD11c, and DC-SIGN Alexa 647-MMD3 (or isotype control, middle panel) on Mo-DCs. (C) Surface markers on freshly isolated monocytes, GM-CSF/IL-4 induced Mo-DCs, and fresh spleen populations. (D) Presentation of CSP or OVA, 40 μg/ml, to TCR transgenic T cells by graded doses of Mo-DCs or CD11c^hi DEC-205⁺ and DEC-205⁻ DCs from spleen. Gating strategy for CFSE^lo T cells is in Figure S1D. (E) Presentation of stably transduced, irradiated CHO-OVA cells by graded doses of different populations of DCs cultured from bone marrow (DC:T cell ratio on the x-axis), including the equivalents of CD8⁺ and CD8⁻ classical DCs from flt-3L expanded marrow cultures (Figure S1E). Representative of 2–3 experiments in triplicate or quadruplicate cultures Error bar=SD.(D-E).

Figure 2. Mobilization of DC-SIGN⁺ Mo-DCs to the T cell areas of lymph node

A) TLR4 competent (C3H/HeN) or TLR4 mutant (C3H/HeJ) mice were injected with 5 μg of LPS i.v. After 24 h, lymph node cells were stained intracellularly with Alexa 647 MMD3 α-DC-SIGN and Alexa 488 α-MMR/CD206 mAbs. (B) Mice were injected with 10 μg of α-DC-SIGN-Alexa 647 mAb and 5 μg of TLR agonist i.v. (C) Labeling of frozen sections with the indicated mAb 12–24 h after PBS, 5 μg LPS i.v. or 5 × 10⁶ heat killed E. coli or B. subtilis i.v. Alexa 647 B220 mAb marks B cell areas (blue). 100× (D-E) Lymph node sections from PBS or LPS treated mice were stained with the indicated mAb. 400×.

Figure 3. DC-SIGN⁺ Mo-DCs are induced upon treatment with LPS or LPS-bacteria

(A) 30 μg Alexa 488 MMD3 α-DC-SIGN or control mAb were injected i.v. with LPS into WT or DC-SIGN^−/− mice. 12 h later, lymph node sections were fixed and stained with rabbit α-Alexa 488 to visualize the injected mAb in green. α-MMR/CD206 (red) identifies Mo-DCs, and A647 B220 mAb (blue) B cells. 400× (B) Separation of 3 lymph node DC populations 12 h after injecting 10 μg of Alexa-647 MMD3 α-DC-SIGN mAb plus 5 μg of LPS i.v. Skin draining lymph node cells were stained for lymphocyte lineage markers (CD3, CD19, NK1-1 (or DX-5)), CD11c, and DEC-205. Live, lineage⁻ CD11c⁺ cells were gated and 3 populations defined (Pop#1, #2, #3). Isotypes for DC-SIGN and DEC-205 are mouse IgG2c and rat IgG2a respectively. (C) Expression of maturation markers on 3 DC populations. (D) Representative morphology (DIC images) of DC-SIGN⁺ cells sorted from lymph nodes of LPS treated mice as in B. 600x (E) Three DC populations as in B were sorted and stained with PE-mAbs. (F) As in B, but FACS analyses and total numbers of lineage⁻ CD11c⁺ cells from mice 12 h after i.v. injection of MMD3 α-DC-SIGN mAb plus killed E. coli or B. subtilis (data with live organisms are in Figure S3C).

Figure 4. Monocyte origin of DC-SIGN⁺ Mo-DCs

(A) CD45.2⁺ marrow monocytes were transferred i.v. into CD45.1⁺ hosts. 24 h later, PBS or 5 μg of LPS was injected i.v. with 10 μg Alexa647-MMD3 α-DC-SIGN and 24 h later, DC-SIGN⁺ CD206⁺ DCs of CD45.2 origin were enumerated. One of 3 similar experiments. (B) WT and LysMCre×iDTR mice were injected with DT and 12 h later, blood monocytes (Ly6G⁻ CD115⁺ CD11b⁺ Ly6C^hi/^lo) were analyzed (left panels). 24 h after DT, 5 μg of LPS plus 10 μg of MMD3-Alexa 647 mAb were given i.v. and 12 h later, skin draining lymph node cells were analyzed as CD19/CD3/NK1.1⁻ and CD11c ^high and segregated into 3 DC populations (right) to look for DC-SIGN⁺ Mo-DCs (arrow). (C) Lymph node sections were stained for Mo-DCs with α-DC-SIGN (BMD10, green), resident DCs with α-DEC-205 (NLDC145, red), and B cells with α-B220 (blue) at 100×. (D) WT and Flt3^−/− mice were injected with 5 μg of LPS and 10 μg of MMD3-Alexa 647 α-DC SIGN mAb to enumerate Mo-DCs expressing DC-SIGN (blue) or MMR/CD206 (red) 24h later. Cells/10⁶ lymph node cells from two independent experiments with 2 mice/group.

Figure 5. L-selectin and CCR7 dependent trafficking of DC-SIGN⁺ Mo-DCs

(A-B) MAb to block L-selectin (MEL-14, 100 μg i.v) was given 1 h before injection of LPS and α-DC-SIGN mAb. After 24h, lymph nodes were analyzed by staining at 100X (A) or FACS (B). (C) Chemokine receptor KO mice were injected with LPS i.v. and 24 h later, lymph node cells were stained for intracellular DC-SIGN and MMR/CD206. Systemic injection of LPS was confirmed by CD86 up-regulation on spleen DCs (right). (D) Blood chimerism 6 wks after lethal-irradiation and reconstitution with CD45.1 (WT) and CD45.2 (CCR7^−/−) in CD45.1 hosts (left). 12 h after LPS, blood monocytes had largely disappeared (middle). Lymph nodes from these same animals were stained for CD45.1, CD45.2, DC-SIGN, and MHC II to show that Mo-DCs were WT in origin.

Figure 6. Presentation of malaria CS and OVA proteins by 3 types of DCs

(A). C57BL/6 or B6×BALB/c F1 mice were injected i.v. with 5 μg LPS 12 h to isolate 3 DC fractions (>95% purity) as in Fig. 3B. Graded doses were added with 40 μg/ml protein to 50,000 CFSE-labeled T cells and 3–4 d later, CFSE^lo T cells were counted. An MLR was also run to verify DC activity. (B) As in A, but mice received 5 μg of LPS i.v. for 12 h, as well as 50 μg of CSP or OVA protein s.c. in each paw for 2 h before DC and B cell isolation. Representative data of 2 experiments in triplicate or quadruplicate cultures. Error bar = SD. (C) As in B, but ELISA was used to measure the indicated cytokines in the medium of cocultures in which different types of lymph node DCs were used to present OVA to OTII CD4⁺ T cells. Error bar=SD (D) 5–10 ×10⁶ E. coli-OVA or control E. coli, were injected s.c. with 10 μg MMD3 mAb. 12 h later, 3 populations of lymph node DCs were isolated and used to stimulate OT I CD8⁺ T cells. Representative of 2 experiments in triplicate. Error bar=SD

Figure 7. MoDCs selectively express CD14, a required coreceptor for their mobilization

(A) Quantitative PCR to assess expression of mRNA for several TLRs and CD14 in marrow monocytes and Mo-DCs. Error bar=SD. (B) DC-SIGN⁺ MoDCs colabel for CD14 expression. (C) CD14^−/− mice, fail to mobilize Mo-DCs in response to LPS. (D) DC-SIGN⁺ MoDCs are mobilized in MyD88^−/− but not MyD88^−/− × Trif^−/− mice. The number of DC-SIGN⁺ Mo-DCs per million lymph node cells are on the panels. (E) Kinetics of formation and disappearance of Mo-DCs in lymph nodes from LPS treated mice, monitored by in vivo labeling of lymphocyte negative, CD11c^hi DCs with MMD3 anti-DC-SIGN mAb or by ex vivo labeling for CD14. MoDC’s numbers are averages of two mice each per time point. (F) Mice were injected with LPS for 12 h, and 2 h prior to isolation CSP was injected. Mo-DCs were labeled either with MMD3 mAb in vivo, or anti-CD14 ex vivo, and used to stimulate CSP-specific CD8⁺ T cells. Error Bar = SD. Representative data of at least 2 independent experiments (A-F).