IL-6 increases B-cell IgG production in a feed-forward proinflammatory mechanism to skew hematopoiesis and elevate myeloid production

Abstract

Src homology 2 domain-containing inositol 5-phosphatase (SHIP(-/-)) animals display an age-related increase in interleukin-6 (IL-6), a decrease in B lymphopoiesis, and an elevation in myelopoiesis. We investigated the origin of the IL-6 production and show that it is largely produced by peritoneal and splenic macrophages. IL-6 production by these macrophages is not a direct result of the loss of SHIP: IL-6 production is not spontaneous, is absent from bone marrow-derived macrophages, declines with prolonged culture of macrophages, and requires a stimulus present in vivo. The IL-6-rich peritoneal cavity of SHIP(-/-) mice shows more than 700-fold more immunoglobulin G (IgG) than wild-type, approximately 20% of which is aggregated or in an immune complex and contains B220(+) cells that secrete IgG. The SHIP-deficient peritoneal macrophages show evidence of IgG receptor stimulation. Animals lacking both the signal-transducing gamma-chain of IgG receptors and SHIP or Ig and SHIP produce less IL-6. The data indicate a feed-forward process in which peripheral macrophages, responding through IgG receptors to secreted IgG, produce IL-6, to support further B-cell production of IgG. Because of the proinflammatory phenotype of SHIP(-/-) animals, these findings emphasize the importance of IL-6-neutralizing strategies in autoimmune and proinflammatory diseases.

Figure 1

Serum IL-6 level increases with age in SHIP^−/− mice. Serum IL-6 levels determined by ELISA of individual WT and SHIP^−/− mice at the age of 4 and 8 weeks are shown. ND indicates not detected.

Figure 2

Peritoneal macrophages from SHIP^−/− mice produce IL-6. (A) IL-6 production of freshly isolated and purified macrophages from bone marrow, spleen, lung, and peritoneum of WT and SHIP^−/− mice, analyzed by ELISA after a 6-hour culture. Data are mean ± SE of 3 identical experiments. (B) Subpopulations of peritoneal cells were identified by flow cytometry as macrophages (Mac-1⁺, F4/80⁺), monocytes (Mac-1⁺, F4/80⁻, Ly-6C^hi), myeloid dendritic cells (CD11c⁺, Ly-6C⁺), and neutrophils (Mac-1⁺, Gr-1⁺, Ly-6C⁺) and analyzed for intracellular IL-6. Shown is the total number of each population that expressed intracellular IL-6. Data are mean ± SE of 3 animals. (C) IL-6 histogram of peritoneal macrophages (defined as Mac-1⁺, F4/80⁺) from WT (dashed line) or SHIP^−/− (solid line) mice. Unstained cells are shown in gray. (D) Peritoneal macrophages identified with anti–Mac-1 and anti-F4/80 antibodies. (E) Total cells from 7 separate animals.

Figure 3

SHIP^−/− mice had a high level of total and aggregated IgG in peritoneum. (A) Total IgG obtained by ELISA in peritoneal lavages of WT and SHIP^−/− mice. (B) The lavage fluid was centrifuged 55 000g, and IgG (defined as aggregated IgG) was measured in the pellet. (A-B) The data were obtained from 6 individual mice, and each point represents a single animal. (C) IgG production from peritoneal cells was determined by ELISA. The F4/80⁺ macrophages were removed, and the remaining cells were cultured for 24 hours. Cells were collected, and secreted IgG was measured by ELISA from 9 individual mice.

Figure 4

SHIP negatively regulated IL-6 production caused by FcγR stimulation. (A) Peritoneal macrophages were purified from WT, SHIP^−/−, or γ-chain^−/− mice by adherence and cultured 24 hours. The cells were stimulated with nothing (no stimulus) or with 10 μg/mL of heat-aggregated IgG (ΔIgG) and incubated for an additional 12 hours before supernatants were collected. IL-6 production stimulated by FcγR engagement of peritoneal macrophages was determined by ELISA. Results are from 3 separate experiments and are the mean ± SE of IL-6 in nanograms per milliliter. N.S. indicates not significant. (B) Peritoneal lavage fluid was collected from SHIP^−/− mice and concentrated 10-fold by ultrafiltration (to 250 μL). The concentrated lavage material was incubated with nothing or with Sepharose-conjugated protein A (PA) or GST as indicated. Bone marrow–derived macrophages of SHIP^−/− or γ-chain^−/− mice were then treated with 50 μL of the concentrated lavage fluid and incubated for 12 hours. The supernatants were measured by IL-6 ELISA. Results are from 2 separate experiments and are the mean ± SE of IL-6 in nanograms per milliliter. (Inset) Western blot of 50 μL of lavaged material probed with rabbit anti–mouse Ig.

Figure 5

Signal transduction proteins of the FcγR pathway are activated in ex vivo peritoneal macrophages of SHIP^−/− mice. Freshly purified peritoneal macrophages were lysed without any treatment. Western blots were used to determine levels of tyrosine-phosphorylated Vav (pVav) and total Vav (A), serine-phosphorylated Akt (pAkt) and total Akt (B), and serine- and tyrosine-phosphorylated SAPK/JNK (pJNK) and total JNK (C). (D) The ratio of the signal of phosphoprotein to total of the SHIP-deficient sample to the WT sample. Data are representative of 3 independent experiments. (E) Freshly isolated peritoneal macrophages from SHIP^−/− mice were cultured with the PI3K inhibitor (LY294002) or the JNK inhibitor (SP60025) or dimethyl sulfoxide for 6 hours. Supernatants were collected, and IL-6 levels were determined by ELISA. Results from 3 separate experiments are mean ± SE of IL-6 in picograms per milliliter.

Figure 6

The loss of Fc γ-chain reduces the IL-6 level in the sera of SHIP^−/− mice. Serum IL-6 levels of SHIP^−/−, SHIP, μMT, and SHIP, γ-chain double-knockout mice at the age of 8 weeks. The serum was obtained by cardiac puncture, and IL-6 level was determined by ELISA. Results from 8 mice of SHIP^−/− and SHIP, γ-chain double-knockout and 3 mice of SHIP, μMT double-knockout each strain are indicated as pictograms per milliliter.