1. C1q and HMGB1 reciprocally regulate human macrophage polarization
by Myoungsun Son, Amit Porat, Mingzhu He, Jolien Suurmond, Frances Santiago-Schwarz, Ulf Andersson, Thomas R. Coleman, Bruce T. Volpe, Kevin J. Tracey, Yousef Al-Abed, and Betty Diamond
Blood
Volume 128(18):
November 3, 2016
©2016 by American Society of Hematology

2. C1q inhibits expression of HMGB1-induced IFN-α, MX1, and inflammatory cytokines by human monocytes.
C1q inhibits expression of HMGB1-induced IFN-α, MX1, and inflammatory cytokines by human monocytes. (A) Messenger RNA (mRNA) levels for IFN-α, MX1, IL-6, and TNF-α in monocytes stimulated with or without HMGB1 (3 μg/mL) and various concentrations of C1q (μg/mL) for 6 hours in serum-free medium (mean ± standard deviation [SD] of triplicates, N = 3). (B) HMGB1 (3 μg/mL) -induced IL-6 and TNF-α secretion were reduced in the presence of various concentrations of C1q, assessed at 24 hours (mean ± SD of duplicates, N = 3). (C) IFN-α, MX1, IL-6, and TNF-α mRNA expression by monocytes transfected with control siRNA or LAIR siRNA and treated with C1q (25 μg/mL), HMGB1 (3 μg/mL) for 6 hours (mean ± SD of triplicates, N = 3). Data are expressed as fold induction relative to controls. R.E., relative expression. ns, not significant. *P < .05; **P < .01; ***P < .001 (one-way ANOVA with Bonferroni’s correction for multiple comparisons).
Myoungsun Son et al. Blood 2016;128:
©2016 by American Society of Hematology

3. C1q inhibits HMGB1 internalization by human monocytes.
C1q inhibits HMGB1 internalization by human monocytes. (A) Human monocytes were incubated with AF555-labeled HMGB1 with (+C1q) or without C1q (−C1q) for 30 minutes at 4°C or 37°C and stained with fluorescein isothiocyanate–anti-CD14 antibody. Imaging flow cytometry was performed on an ImageStreamX Mark II. Intensity of AF555-HMGB1 of a CD14+ population in each condition is shown. (B) The Internalization Wizard following membrane masking was used to quantify intracellular HMGB1. The median fluorescent intensity from 3 independent experiments is shown. ***P < (C) Representative images of monocytes that internalized HMGB1 (red) without C1q (left) and in the presence of C1q (right).
Myoungsun Son et al. Blood 2016;128:
©2016 by American Society of Hematology

4. C1q, HMGB1, RAGE, and LAIR-1 form a multimolecular complex in lipid rafts.
C1q, HMGB1, RAGE, and LAIR-1 form a multimolecular complex in lipid rafts. (A) SPR assay of C1q and RAGE binding; KD = 855 nM. Experiments were repeated 3 times. (B) SPR assay of RAGE-C1q-HMGB1 trimolecular complex. RAGE (500 nM) was immobilized onto a CM5 chip and the first analyte (C1q, 200 nM) was added to saturation. HMGB1 (500 nM) was added to the RAGE-C1q complex in multiple pulses (left). Alternatively, HMGB1 was added to immobilized RAGE until the chip was saturated, followed by C1q in multiple pulses (right). N = 3. (C) SPR assay for HMGB1 and C1q binding; KD = 200 nM. N = 4. (D) SPR assay of HMGB1 in different redox states and C1q binding. C1q (200 nM) was immobilized onto a CM5 chip, and disulfide-, all thiol-, or oxidized-HMGB1 (500 nM) was added. N = 3. (E) C1q-coated beads were incubated with saturating amounts of HMGB1, followed by 250 ng or 500 ng of RAGE. Complexes were analyzed by western blot using anti-RAGE, anti-Cbp tag antibody for HMGB1 or infrared-labeled streptavidin for biotinylated-C1q. N = 3. ***P < .001 (one-way ANOVA). (F) Human monocytes were treated with HMGB1 (3 μg/mL) in the absence or presence of C1q (25 μg/mL) for 15 minutes at 37°C. Colocalization of LAIR-1 and RAGE on the plasma membrane was assessed by PLA. Red dots (PLA positive), representing colocalization of RAGE and LAIR-1, are only seen in the presence of C1q, with or without HMGB1. Percentage of PLA-positive cells in total cells was counted in several random fields in 3 independent experiments. (G) PLA assay using C1 (25 μg/mL) complex instead of C1q. N = 3. (H) Lipid raft fractions from monocytes treated with HMGB1 (3 μg/mL), in the absence or presence of C1q (25 μg/mL), for 15 minutes at 37°C were concentrated and analyzed by western blot for LAIR-1, RAGE, HMGB1, C1q, or Flotillin1 as a lipid raft marker. N = 3. DAPI, 4′,6-diamidino-2-phenylindole.
Myoungsun Son et al. Blood 2016;128:
©2016 by American Society of Hematology

5. C1q dephosphorylates RAGE and recruits SHP-1 to LAIR-1.
C1q dephosphorylates RAGE and recruits SHP-1 to LAIR-1. Human monocytes were treated with C1q (25 μg/mL) and/or HMGB1 (3 μg/mL) for 15 minutes at 37°C. (A) Total cell lysates were subjected to immunoprecipitation with antibodies to RAGE followed by immunoblotting with antibodies specific for phosphoserine (top) or RAGE (bottom). Numbers indicate the relative signal intensity. N = 3. (B) Total cell lysates were subjected to immunophosphorylation array to observe the phosphorylation of LAIR-1 ITIM motifs. Relative quantification of phosphorylation of LAIR-1 was normalized to control spots (relative fold induction) from 4 independent experiments. (C) Human monocytes were treated with C1 complex (25 μg/mL) and/or HMGB1 (3 μg/mL) for 15 minutes at 37°C. Immunophosphorylation array was performed to observe the phosphorylation of LAIR-1 ITIM motifs. N = 3. *P < .05; **P < .01; ***P < .001 (one-way ANOVA). (D) Total cell lysates were immunoprecipitated with antibodies to LAIR-1 followed by immunoblotting with antibodies for SHP-1 (top) or LAIR-1 (bottom). N = 4. (E) Human monocytes were treated with HMGB1 and/or C1q for 3 hours, and cell lysates were subjected to immunoblotting with antibodies for activated IKKα (P-IKKα, top), p65 (P-p65, middle), or β-actin. N = 3.
Myoungsun Son et al. Blood 2016;128:
©2016 by American Society of Hematology

6. HMGB1 and C1q induce anti-inflammatory molecules and promote an anti-inflammatory macrophage phenotype.
HMGB1 and C1q induce anti-inflammatory molecules and promote an anti-inflammatory macrophage phenotype. Human monocytes, treated with C1q (25 μg/mL) or C1q tail (53 μg/mL) and/or HMGB1 (3 μg/mL) for 24 hours, were processed for mRNA and protein. (A) Mer tyrosine kinase as assessed by quantitative polymerase chain reaction (qPCR) (left). R.E., relative expression. N = 3. (B) PD-L1 was assessed by qPCR. N = 4. (C) IL-10 was assessed by qPCR (left) and enzyme-linked immunosorbent assay (right). N = 4. (D) CD163 was assessed by qPCR (left) and flow cytometry (right). N = 3 for left; N = 3 for right. (E) HLA-DR (MHCII) was determined by flow cytometry. N = 3. All data represent the mean ± standard error of the mean (SEM) of independent experiments in which each condition was tested in triplicate. Statistical analysis was performed by one-way ANOVA. ns, not significant; *P < .05; **P < .01; ***P < .001.
Myoungsun Son et al. Blood 2016;128:
©2016 by American Society of Hematology

7. HMGB1 and C1q inhibit plasticity of macrophages.
HMGB1 and C1q inhibit plasticity of macrophages. DC differentiation induced by GM-CSF and IL-4 was assessed by flow cytometry. Monocytes were treated with C1q and/or HMGB1 for 24 hours (day 0) and then further cultured with GM-CSF and IL-4 for 2 days (day 2). (A) High levels of CD14 and LAIR-1 denote suppression of DC differentiation. Data represent the mean ± SEM of 3 independent experiments. Statistical analysis was performed by one-way ANOVA. ns, not significant. *P < .05; **P < .01; ***P < (B) Transcription of Mer and CD163 was measured by qPCR. Data represent the mean ± SEM of 3 independent experiments in which each condition was tested in triplicate. Statistical analysis was performed by one-way ANOVA. ns, not significant. *P < .05; **P < .01; ***P < (C) Mixed lymphocyte reaction. Monocytes were exposed to HMGB1 (3 μg/mL), C1q (25 μg/mL), or both for 24 hours in X-Vivo 15 serum-free medium, washed, and further incubated for 2 days in X-Vivo 15 medium. Cell Trace Violet–stained allogeneic primary human CD4 T cells and anti-CD3 (1 μg/mL) were added (2:1). T cells alone in the presence of anti-CD3 antibody or T cells with antibodies to CD3/CD28 were negative and positive controls, respectively. After 4 days, the nonadherent cells, which stained with anti-CD4 antibodies, were assessed by flow cytometry. Live CD4+ T cells were analyzed. Bar graphs represent the mean ± SEM of percentage having undergone division (left) and division index (right). N = 3. *P < .05; **P < .01; ***P < .001 (one-way ANOVA).
Myoungsun Son et al. Blood 2016;128:
©2016 by American Society of Hematology

8. Model showing how C1q utilizes a natural pathway to dampen inflammation.
Model showing how C1q utilizes a natural pathway to dampen inflammation. (A) DNA/RNA-associated HMGB1 is internalized and activates endosomal TLRs, leading to downstream activation of NFκB to induce proinflammatory macrophages. (B) In the presence of C1q without inflammation (basal levels of HMGB1), C1q and LAIR-1 signaling prevents proinflammatory cytokine production. (C) In inflammation with high levels of HMGB1, C1q mediates differentiation of anti-inflammatory macrophages by crosslinking RAGE and LAIR-1 in lipid rafts to facilitate SHP-1 binding to LAIR-1 via phosphorylated ITIMs. Activated SHP-1 may suppress directly or indirectly the pathways downstream of RAGE activation.
Myoungsun Son et al. Blood 2016;128:
©2016 by American Society of Hematology