Hax1 depletion impairs uropod retraction and reduces RhoA activation. (A) Lentiviral knockdown of Hax1 (Hax1 shRNA A and B) in PLB-985 cells display significant depletion of Hax1 compared with control shRNA. Quantification was normalized to vinculin loading control. ***, P < 0.01; ANOVA with Tukey posttest. (B) Still image from a time-lapse video of control shRNA and Hax1 shRNA A and B PLB-985 cells during chemotaxis on 10 µg/ml fibrinogen (Videos 4 and 5). Arrowheads indicate elongated uropods of Hax1-deficient PLB-985 cells during chemotaxis. Bar, 100 µm. (C) Quantification of uropod length in Hax1-deficient cells compared with control. Control shRNA (n = 279) cells, Hax1 shRNA A (n = 275) cells, and Hax1 shRNA B (n = 241) cells were from three independent experiments. The graph represents means ± 95% confidence interval. ***, P < 0.01; one-way ANOVA with Tukey posttest. (D) GST-RBD pull-down assays show a 10-fold decrease in RhoA activity in Hax1-deficient cells compared with control. ***, P < 0.01; t test. (E) Control and Hax1-deficient PLB-985 cells expressing utrophin-GFP during chemotaxis to 1 µM fMLP on fibrinogen (Videos 6 and 7). For Rho-associated protein kinase inhibition, cells were incubated with 30 µM Y-27632 for 30 min before imaging. Fluorescence intensity from the front to the back of the cell was determined by line scan analysis (yellow lines). Bar, 10 µm. (F) Quantification of line scan analysis of utrophin-GFP localization from E. P-values were obtained by ANOVA with Tukey posttest. Red lines indicate the mean fluorescence ratio from three independent experiments. Error bars are means ± SEM from three independent experiments. A.U., arbitrary unit.

Hax1- and RhoA-deficient PLB-985 cells display increased integrin-mediated adhesion. (A) Adhesion assay of differentiated control and Hax1-deficient cells on fibronectin (Fn), fibrinogen (Fbg), and vitronectin (Vn). (B) Adhesion assay of control and Hax1-deficient cells on 10 µg/ml fibrinogen. For antibody blocking, cells were incubated with β2 or αvβ3 integrin function-blocking antibodies as described in Materials and methods. (C) Lentiviral knockdown of RhoA in PLB-985 cells. (D) Adhesion assay of differentiated control and RhoA-deficient PLB-985 cells on 10 µg/ml fibrinogen. (E) Migration velocities of differentiated control and RhoA-deficient cells during chemotaxis in microfluidic gradient generators. P-values were obtained by ANOVA with Tukey posttest (control shRNA, n = 23 cells and RhoA shRNA 1 and RhoA shRNA 2, n = 22 cells; horizontal lines indicate the mean velocity from three independent experiments). *, P < 0.05. Error bars indicate means ± SEM from three independent experiments.

Hax1 expression in PLB-985 cells and localization during chemotaxis. (A) Schematic of Hax1 modified from (Jeyaraju et al., 2009). Binding sites for HS1 (Suzuki et al., 1997), Gα13 (Radhika et al., 2004), and β6 integrin (Ramsay et al., 2007) are shown. Hax1-GFP constructs are used in C. (B) Hax1 expression is decreased upon differentiation. Vinculin, HS1, and actin were included as controls and show increased expression with neutrophil differentiation. Quantification of expression represents the ratio of differentiated to undifferentiated cells as mean intensity ± SDs from three independent experiments as described in Materials and methods. (C) Still image from time-lapse videos of utrophin-RFP and Hax1-GFP localization in PLB-985 cells. Utrophin-RFP was coexpressed with full-length Hax1-GFP (Video 1), Hax1 1–113–GFP (Video 2), and Hax1Δ113-GFP (Video 3) in differentiated PLB-985 cells and then imaged during chemotaxis on 10 µg/ml fibrinogen in response to a needle containing 1 µM fMLP. Bar, 10 µm.

Directed migration is impaired in Hax1-deficient PLB-985 cells using microfluidic gradient generators. (A) Schematic of the microfluidic device showing stable gradient formation with Alexa Fluor 488–conjugated dextran quantified by line scan analysis (yellow line). C/Csource, concentration of the line scan over concentration of the source. (B) Image taken from a time-lapse video of control and Hax1 knockdown cells during chemotaxis on fibrinogen (Videos 8 and 9). Hax1-deficient cells display impaired migration and develop elongated tails (arrowheads). Insets are zoomed regions indicated by the white squares. Bar, 100 µm. (C) Representative velocity and angle (insets) histogram of control and Hax1-deficient cells during chemotaxis on fibrinogen in the absence and presence of β2 or αvβ3 integrin function-blocking antibodies (from one of three independent experiments; n = 10 cells tracked over 30 min during chemotaxis). (D) Mean velocity heat map from three independent experiments in triplicate of control and Hax1 knockdown cells during chemotaxis on fibronectin (Fn), fibrinogen (Fbg), and vitronectin (Vn) and in the presence of integrin function-blocking antibodies on fibrinogen. n = 30 cells per replicate. ***, P < 0.01; t test.

Constitutively active RhoA and the N-terminal region of Hax1 rescue neutrophil adhesion. (A) Adhesion assays of control cells stably expressing GFP and Hax1 shRNA A cells expressing GFP, or an shRNA-resistant GFP-Hax1, GFP–Hax1 1–113, and GFP-Hax1Δ113 were plated on 10 µg/ml fibrinogen in the presence or absence of 200 nM fMLP. (B) Adhesion assay of control and Hax1-deficient cells plated on 10 µg/ml fibrinogen in the presence or absence of 200 nM fMLP and 10 µM colchicine for 30 min. (C) Adhesion assay of control and Hax1-deficient cells stably expressing GFP or GFP-RhoA (Q63L). (B and C) ***, P < 0.01; one-way ANOVA with Tukey posttest. (D) Migration velocities of control and Hax1-deficient cells stably expressing GFP, GFP-RhoA (Q63L), GFP-Hax1, GFP–Hax1 1–113, and GFP-Hax1Δ113 during chemotaxis to fMLP in microfluidic chemotaxis devices. Horizontal lines indicate the mean velocity. (A and D) Means with common letters indicate P > 0.05 by ANOVA with Tukey posttest.