Lymphoid precursors are directed to produce dendritic cells as a result of TLR9 ligation during herpes infection

Abstract

Hematopoietic stem and progenitor cells were previously found to express Toll-like receptors (TLRs), suggesting that bacterial/viral products may influence blood cell formation. We now show that common lymphoid progenitors (CLPs) from mice with active HSV-1 infection are biased to dendritic cell (DC) differentiation, and the phenomenon is largely TLR9 dependent. Similarly, CLPs from mice treated with the TLR9 ligand CpG ODN had little ability to generate CD19+ B lineage cells and had augmented competence to generate DCs. TNFalpha mediates the depletion of late-stage lymphoid progenitors from bone marrow in many inflammatory conditions, but redirection of lymphopoiesis occurred in TNFalpha-/- mice treated with CpG ODN. Increased numbers of DCs with a lymphoid past were identified in Ig gene recombination substrate reporter mice treated with CpG ODN. TLR9 is highly expressed on lymphoid progenitors, and culture studies revealed that those receptors, rather than inflammatory cytokines, accounted for the production of several types of functional DCs. Common myeloid progenitors are normally a good source of DCs, but this potential was reduced by TLR9 ligation. Thus, alternate differentiation pathways may be used to produce innate effector cells in health and disease.

Figure 1

Production of DCs from lymphoid progenitors in HSV-1–infected mice. (A) Total numbers of pro-B/large pre-B, small pre-B, B, pDCs, IKDCs, and cDCs were enumerated from femurs of mice after 7 days of acute infection or 30 days of latent HSV-1 infection. Percentage control values represent BM cellularity in each condition relative to BM cellularity in uninfected controls. The results are representative of 3 independent experiments. CLPs were double sorted from HSV-1–infected and control mice and placed in lymphoid cultures for 8 days. (B) The indicated flow cytometry gates were used to discriminate B220⁺CD19⁻CD11c⁺CD11b⁻ pDC/NK-like IKDCs and B220⁻CD19⁻CD11c⁺CD11b⁺ cDCs. (C,D) Total numbers of recovered cells of each type after acute and latent infection were calculated and expressed as yields per input of common lymphoid progenitor (*significant difference, P < .05 by t test). Data are representative of 3 independent experiments. Error bars represent SEM.

Figure 2

Lymphoid progenitors from TLR-deficient mice are not primed to become DCs during HSV-1 infection. (A) Total numbers of pre-B and B cells were enumerated in BM of TLR^+/+ and TLR^−/− mice on day 7 of acute infection. Data are means plus or minus SEM (*significant difference, P < .05). CLPs were highly purified from HSV-1–infected and control mice and placed in lymphoid cultures for 8 days. (B) The indicated flow cytometry gates were used to discriminate B220⁺CD19⁻CD11c⁺CD11b⁻ pDCs + NK-like IKDCs and B220⁻CD19⁻CD11c⁺CD11b⁺ cDCs. Data are representative of 2 independent experiments with 3 replicates each.

Figure 3

CpG treatment preferentially depletes late-stage B lymphoid precursors in BM and simulates production of some inflammatory cytokines. (A,B) Mice were given single intraperitoneal injections of CpG, and 48 hours later lymphoid progenitors in BM were evaluated with respect to IL-7Rα expression by flow cytometry and RT-PCR. (C) Using stringent flow cytometry gating criteria, numbers of pro-B/large pre-B, small pre-B, and B cells in BM were evaluated. Data are mean plus or minus SEM (*significant difference, P < .05). (D) Levels of proinflammatory cytokines and Flt3-L (FL) in sera were tested by ELISA. Data are representative of more than 3 experiments with each having N = 3.

Figure 4

Reduction in B lineage cells in BM of CpG-treated mice is partially mediated by TNFα, but direction to dendritic fates does not require this cytokine. (A) TNFα^+/+ and TNFα^−/− mice were injected with LAL water (control) or CpG, and total numbers of pro-B/large pre-B, small pre-B, B, pDCs, NK-like IKDCs, and cDCs were enumerated from femurs 48 hours later. Percentage control values represent BM cellularity after CpG treatment relative to BM cellularity in controls. The results are representative of 3 independent experiments. CLPs were sorted in the same experiment and cultured under conditions designed to support B lymphopoiesis. (B) Flow cytometry was used to evaluate these cultures 8 days later. CD19⁻ cells were further resolved on the basis of CD11b and CD11c to reveal conventional and pDC and NK-like IKDCs. (C) Yields per input progenitor were calculated as before. Data are mean plus or minus SEM (*significant difference, P < .05). Data are representative of 2 experiments with each having N = 3. Error bars represent SEM.

Figure 5

DCs generated from CLPs in response to TLR9 ligation are functional. CLPs were sorted from CpG-injected mice and cultured under conditions designed to support B lymphopoiesis. Harvested cells were then evaluated after 8 days by flow cytometry. (A) B220⁺CD19⁻CD11c⁺CD11b⁻ cells were further fractionated into pDC and NK-like IKDC populations by the NK1.1 marker, whereas B220⁻CD19⁻CD11c⁺CD11b⁺ cDCs were NK1.1⁻. (B) The 3 categories of DCs were sorted, and their competence to respond to CpG stimulation by production of cytokines was evaluated.

Figure 6

Transgenic mice with an Ig locus recombination substrate are used to detect redirection of lymphoid progenitors to dendritic fates. (A) CLPs with a history of Ig recombination were sorted on the basis of VEX fluorescence and placed in culture with or without CpG. Flow cytometry performed 8 days later revealed normal lymphopoiesis and continued VEX labeling for control cells (left 2 panels), whereas CpG-treated cells generated several classes of DCs and NK-like IKDCs (right 4 panels). (B) Transgenics were also injected with CpG, and BM cells were evaluated by flow cytometry 1 week later. Percentages of cells with a history of recombination (VEX⁺) are given for each of the indicated cell types. Data are representative of 2 independent experiments (*significant difference, P < .05). Error bars represent SEM.

Figure 7

Lymphoid biased progenitors in BM express high levels of TLR9 and respond to CpG in culture. BM fractions were isolated as described in “Isolation of cell populations and flow cytometry” before analysis by quantitative RT-PCR. (A) The results are representative of 2 independent experiments and are normalized to β-actin and compared with pDC, the subset with highest expression. Normal Lin⁻IL-7Rα⁺c-kit^loSca1⁺ CLPs were sorted and treated in serum-free, stromal-free cultures for 2 or 48 hours with 0.6 μg/mL CpG. The cultures were washed and then incubated for an additional 8 days with stem cell factor, FL, and IL-7 before flow cytometric analyses. Whereas almost pure populations of B220⁺CD19⁺ lymphocytes were present in control cultures, CD19⁻ cells emerged as a result of extended CpG treatment (B, left panel). B220^+/− CD19⁻ cells were gated (B, right panel) and further analyzed to reveal B220⁺CD19⁻CD11c⁺CD11b⁻ pDCs and/or NK-like IKDCs, and B220⁻ CD19⁻CD11c⁺CD11b⁺ cDCs. Sorted CLPs were placed in limiting dilution stromal-cell free, serum-free cultures with and without 0.6 μg/mL of CpG for 48 hours. They were washed and then returned to culture for an additional 8 days before flow cytometry analysis. (C) Individual wells were scored as being positive for CD19⁺ B lineage and/or CD11c⁺ DCs.