Myeloid Src-family kinases are critical for neutrophil-mediated autoinflammation in gout and motheaten models

Abstract

Autoinflammatory diseases include a number of monogenic systemic inflammatory diseases, as well as acquired autoinflammatory diseases such as gout. Here, we show that the myeloid Src-family kinases Hck, Fgr, and Lyn are critical for experimental models of gout, as well as for genetically determined systemic inflammation in the Ptpn6me-v/me-v (motheaten viable) mouse model. The Hck-/-Fgr-/-Lyn-/- mutation abrogated various monosodium urate (MSU) crystal-induced pro-inflammatory responses of neutrophils, and protected mice from the development of gouty arthritis. The Src-family inhibitor dasatinib abrogated MSU crystal-induced responses of human neutrophils and reduced experimental gouty arthritis in mice. The Hck-/-Fgr-/-Lyn-/- mutation also abrogated spontaneous inflammation and prolonged the survival of the Ptpn6me-v/me-v mice. Spontaneous adhesion and superoxide release of Ptpn6me-v/me-v neutrophils were also abolished by the Hck-/-Fgr-/-Lyn-/- mutation. Excessive activation of tyrosine phosphorylation pathways in myeloid cells may characterize a subset of autoinflammatory diseases.

Figure 1.

Myeloid Src-family kinases are required for MSU crystal–induced neutrophil activation. (A–I) WT or Hck^−/−Fgr^−/−Lyn^−/− neutrophils (A–I), along with various single and double knockout neutrophils (C) were stimulated with 1 mg/ml (A and C–I) or the indicated concentrations (B) of MSU crystals. (A–C) Luminometric measurement of reactive oxygen production. (D–I) Cells were incubated for 18 h followed by determination of cytokine, chemokine, and lipid mediator levels from cell-free supernatants using a commercial cytokine array (D) or ELISA assays (G–I). (E) Map of the different analytes on the cytokine array. (F) Quantification of cytokine array results by densitometry. Panels A (mean and SD) and D show results from a representative experiment. Bar graph shows mean and SEM from 4 (B), 3–8 (C), 2 (F), and 5–8 (G–I) independent experiments. Panels C and F show data normalized to WT MSU and WT control samples, respectively. Panel F shows analytes with at least fivefold MSU-induced increase in WT samples. RLU, relative luminescence unit; AUC, area under the curve. Two-way ANOVA genotype × stimulus interaction analysis; *, P < 0.05; **, P < 0.01; ***, P < 0.001. See the text for actual P values. Source data are available for this figure: SourceData F1.

Figure S1.

Analysis of PMA-induced superoxide production. (A and B) WT or Hck^−/−Fgr^−/−Lyn^−/− neutrophils (A) and human neutrophils treated with vehicle or 100 nM dasatinib (B) were stimulated with 100 nM PMA, followed by a spectrophotometric analysis of superoxide release. Kinetic curves (A and B) show mean and SEM of five (A) or three (B) independent experiments. Two-way ANOVA genotype × stimulus (A) or inhibitor × stimulus (B) interaction analysis; n.s., not significant. See the text for actual P values.

Figure S2.

MSU crystal–induced macrophage responses. (A–D) WT or Hck^−/−Fgr^−/−Lyn^−/− bone marrow–derived macrophages were stimulated with 1 mg/ml MSU crystals without (A–D) or with (B–D) overnight priming of the cells with 1 μg/ml ultrapure LPS. (A and B) Luminometric measurement of reactive oxygen production. (C and D) IL-1β (C) and MIP-2 (D) levels determined from cell-free supernatants after incubation for 24 h. Kinetic curves (A) show mean and SD of a representative experiment. Bar graph shows mean and SEM from three (B) or four to five (C and D) independent experiments. RLU, relative luminescence unit. Two-way ANOVA genotype × stimulus interaction analysis; n.s., not significant; **, P < 0.01; ***, P < 0.001. See the text for actual P values.

Figure 2.

Myeloid Src-family kinases mediate MSU crystal–induced arthritis in vivo. (A–G) WT or Hck^−/−Fgr^−/−Lyn^−/− mice or various single and double mutants (D and E) were subjected to gouty arthritis by injecting the right hind paws of the mice with 1 mg MSU crystal suspended in 50 μl PBS, along with injecting left hind paws with PBS alone as controls. MSU crystal–induced inflammation was followed by photographing on day 1 (A; scale bar: 10 mm), measuring the paw volume by plethysmometry (B–E), and examining hypernociception (F and G). Panel A shows a representative experiment. Quantitative curves and bar graphs show mean and SEM from 11 to 13 (B), 32–33 (C), 5–13 (D and E; except for a single Hck^−/−Lyn^−/− mouse), and 11–16 (F and G) mice per genotype from 3 (B), 5 (C), 2–3 (D and E), and 2 (F and G) independent experiments. Two-way ANOVA genotype × stimulus (A–E) or genotype × time (F and G) interaction analysis; *, P < 0.05; **, P < 0.01; ***, P < 0.001. See the text for actual P values.

Figure 3.

Analysis of the inflammatory microenvironment. (A–L) WT or Hck^−/−Fgr^−/−Lyn^−/− mice were subjected to gouty arthritis as described in Fig. 2. (A–D) In vivo MPO (A and B) and NADPH-oxidase (C and D) activity was determined by chemiluminescence imaging after i.p. injection of luminol or lucigenin, respectively. Color-coded photon flux intensity is superposed on the grayscale photo of the mice after 24 h of MSU injection (A and C) and quantified in defined regions of interest at the indicated time points (B and D). (E–G) The hind paws were flushed after 24 h of MSU injection and the number of neutrophils (E) or monocytes/macrophages (F and G) was determined by flow cytometry. (H–L) The cell-free supernatants of the tissue infiltrates were probed using a commercial cytokine array (H and I) or analyzed by ELISA for the indicated pro-inflammatory mediators (J–L). Panels A, C, and H show representative images. Bar graphs show mean and SEM from 6–12 (B and D), 8–20 (E), 3 (F and G), 2 (I), and 13–27 (J–L) mice per genotype from 2 (B and D), 8 (E), 1 (F and G), 2 (I), and 4–7 (J–L) independent experiments. Two-way ANOVA genotype × stimulus interaction analysis; *, P < 0.05; **, P < 0.01; ***, P < 0.001. See the text for actual P values. Source data are available for this figure: SourceData F3.

Figure 4.

Impaired phagocytic activity and intracellular signaling but normal migratory ability of Hck^−/−Fgr^−/−Lyn^−/− neutrophils. (A–C) WT or Hck^−/−Fgr^−/−Lyn^−/− neutrophils were stimulated with 1 mg/ml MSU crystals followed by assessment of phagocytosis of the MSU crystals (A and B), intracellular phosphorylation (C). Where indicated, the cells were pretreated with 10 μM cytochalasin D. (D and E) Mixed bone marrow chimeras carrying CD45.1-expressing WT and CD45.2-expressing WT, Itgb2^−/−, or Hck^−/−Fgr^−/−Lyn^−/− hematopoietic cells were subjected to MSU crystal–induced arthritis as described above. 24 h later, the hind paws were flushed, the ratio of CD45.1- and CD45.2-expressing neutrophils in the paw infiltrate was determined by flow cytometry and compared to the ratio in the peripheral blood. Panel D shows the results for each individual mouse, whereas panel E shows the accumulation of the CD45.2-expressing WT, Itgb2^−/−, and Hck^−/−Fgr^−/−Lyn^−/− neutrophils relative to the CD45.1-expressing WT cells. Panels A and C show representative images from two to three independent experiments. Panel B shows mean and SEM from two to six independent experiments. Panel E shows data mean and SEM of five to six chimeras per group from three independent experiments. Two-way ANOVA inhibitor × stimulus or genotype × stimulus interaction analysis (B) and one-way ANOVA (E); n.s., not significant; **, P < 0.01; ***, P < 0.001. See the text for actual P values. Source data are available for this figure: SourceData F4.

Figure S3.

Competitive in vivo migration assays. (A–C) Representative flow cytometric histograms of competitive in vivo migration experiments in mixed bone marrow chimeras with CD45.1-expressing WT and CD45.2-expressing WT (A), Itgb2^−/− (B), or Hck^−/−Fgr^−/−Lyn^−/− (C) neutrophils, shown in Fig. 4, D and E (see Fig. 4 for more detail).

Figure S4.

Analysis of MSU crystal–induced DNA externalization. (A–D) Nuclei of WT and Hck^−/−Fgr^−/−Lyn^−/− mouse neutrophils (A and B), or of WT neutrophils treated with vehicle or 100 nM dasatinib (C and D) were stained with 0.5 μg/ml Hoechst dye followed by addition of 0.2 mg/ml MSU crystals along with 50 nM Sytox-Green DNA to visualize extracellular DNA. Panels A and C are representative images (scale bar: 200 µm), and panels B and D show the quantification of the externalized DNA area as mean and SEM from four (A and B) or three (C and D) independent experiments. Two-way ANOVA genotype × stimulus or inhibitor × stimulus interaction analysis; n.s., not significant. See the text for actual P values.

Figure 5.

Dasatinib inhibits MSU crystal–induced proinflammatory functions of human neutrophils. Human neutrophils were pretreated with 100 nM or the indicated concentrations of dasatinib or 10 μM cytochalasin D, followed by stimulation with 1 mg/ml MSU crystals. (A) Luminometric measurement of reactive oxygen production. (B and C) LTB₄ release determined from cell-free supernatants after 18 h incubation with the crystals of unprimed (B) or TNF-primed (10 ng/ml TNF; C) neutrophils. (D) Analysis of intracellular p38/ERK phosphorylation. (E–H) Phagocytosis of the MSU crystals assessed by flow cytometry (E and G) or phase contrast microscopy (F and H; scale bar: 20 µm). Panels D, E, and F show representative images. Bar graph shows mean and SEM from four to six (A), two to three (B and C), and three (G and H) independent experiments. RLU, relative luminescence unit; AUC, area under the curve. Two-way ANOVA inhibitor × stimulus interaction analysis (A–C and G) and one-way ANOVA (H); n.s., not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001. See the text for actual P values. Source data are available for this figure: SourceData F5.

Figure 6.

Dasatinib inhibits MSU crystal–induced arthritis in vivo. (A–I) Mice were treated with or without 50 mg/kg dasatinib every 12 h, starting either 1 h prior (A–H) or 4 h after (I) injection of MSU crystals into the hind paws of the experimental animals as described in Figs. 2 and 3. (A, B, and I) Arthritis was followed by measuring paw swelling using a plethysmometer. (C and D) In vivo MPO activity was determined by chemiluminescence imaging. Representative color-coded photon flux intensity superposed on the grayscale photo of the mice (C) and quantified in defined regions of interest (D). (E–H) Analysis of the inflammatory microenvironment by flow cytometry (E) or ELISA (F–H). Panels B–H show results obtained 24 h after MSU crystal injection. Panels C and D show representative color-coded photon flux intensity superposed on the grayscale photo of the mice (C) and quantified in defined regions of interest (D). Kinetic curves and bar graphs show mean and SEM from 9 to 14 (A), 14–24 (B), 7 (D), 4–15 (E–H), and 14–24 (I) mice per group from 3 (A), 2 (D), 3–5 (E–H), and 3–5 (I) independent experiments. Two-way ANOVA inhibitor × stimulus interaction analysis; n.s., not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001. See the text for actual P values.

Figure 7.

The Hck^−/−Fgr^−/−Lyn^−/− mutation attenuates inflammation in the motheaten viable model. (A) Representative images of the hind paws of WT, Ptpn6^me-v/me-v, and Ptpn6^me-v/me-vHck^−/−Fgr^−/−Lyn^−/− mice. (B and C) H&E-stained sections (B; upper panels: 10× magnification, scale bar, 100 µm; lower panels: 40× magnification, scale bar, 25 µm) and weight (C) of lungs from Ptpn6^me-v/me-v and Ptpn6^me-v/me-vHck^−/−Fgr^−/−Lyn^−/− mice. (D) Survival of the mice of the indicated genotypes. (E–G) neutrophils isolated from mice of the indicated genotypes were stimulated on fibrinogen surface (E and G) or by using 1 nM PMA (F), and superoxide release (E and F) or the percentage of cellular adhesion (G) was determined. Panels A and B show representative images from four independent experiments. Bar graphs show mean and SEM of 15–19 mice per genotype (C) and mean and SEM from three independent experiments (G). Survival curves show the data of 10–29 animals per genotype. Kinetic curves on panels E and F show representative experiments performed in triplicate (mean and SEM) from four independent experiments. Cells were prepared from three to four mice per genotype and pooled for each experiment. One-way ANOVA (C and G), log-rank test (D); n.s., not significant; ***, P < 0.001. See the text for actual P values.