SOS1, ARHGEF1, and rho-GEFs Mediate JAK-Dependent LFA-1 Activation by

This information is current as Lara Toffali, Alessio Montresor, Michela Mirenda, Giorgio of September 29, 2021. Scita and Carlo Laudanna J Immunol 2017; 198:708-717; Prepublished online 16 December 2016; doi: 10.4049/jimmunol.1600933

http://www.jimmunol.org/content/198/2/708 Downloaded from

Supplementary http://www.jimmunol.org/content/suppl/2016/12/15/jimmunol.160093 Material 3.DCSupplemental http://www.jimmunol.org/ References This article cites 36 articles, 19 of which you can access for free at: http://www.jimmunol.org/content/198/2/708.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision by guest on September 29, 2021 • No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2017 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

SOS1, ARHGEF1, and DOCK2 rho-GEFs Mediate JAK-Dependent LFA-1 Activation by Chemokines

Lara Toffali,*,† Alessio Montresor,*,† Michela Mirenda,* Giorgio Scita,‡,x and Carlo Laudanna*,†

JAK-dependent activation of the rho module of integrin affinity triggering mediates -induced leukocyte adhesion. However, the signaling events linking JAKs to rho small GTPase activation by chemokines is still incompletely described. In this study, we show that 1 (SOS1), rho guanine nucleotide exchange factor (GEF)1 (ARHGEF1), and dedicator of cytokinesis (DOCK)2 GEFs mediate CXCL12-induced LFA-1 activation in human primary T . Downregulated expression of SOS1, ARHGEF1, and DOCK2 impairs LFA-1–mediated rapid T adhesion as well as underflow arrest on ICAM-1 induced by CXCL12. Moreover, LFA-1 affinity triggering by CXCL12 is impaired by SOS1, ARHGEF1, and DOCK2 downregulation. Notably, the three GEFs are all critically involved in chemokine-induced RhoA and Rac1 activation, thus Downloaded from suggesting the occurrence of a SOS1 specificity shift in the context of chemokine signaling. Accordingly, SOS1, ARHGEF1, and DOCK2 are tyrosine phosphorylated upon chemokine signaling with timing coherent with rapid LFA-1 affinity activation. Importantly, chemokine-induced tyrosine phosphorylation of these GEFs is fully mediated by JAK tyrosine kinases. Unexpectedly, and differently from VAV1, tyrosine phosphorylation of SOS1, ARHGEF1, and DOCK2 is completely inhibited by pertussis toxin pretreatment, thus suggesting different routes of rho-GEF triggering upon CXCR4 engagement. Taken together, these findings reveal a deeper level of complexity in the rho-signaling module, with at least four different rho-GEFs cooperating in http://www.jimmunol.org/ the regulation of chemokine-induced integrin activation, possibly suggesting the emergence of stochastic concurrency in signaling mechanisms controlling leukocyte trafficking. The Journal of Immunology, 2017, 198: 708–717.

eukocyte recruitment is a central homeostatic process in cooperate to increase cell adhesiveness, integrin affinity upregu- the . Although consisting of an ensemble lation plays a prominent role in rapid leukocyte arrest underflow L of continuous, mainly overlapping, cellular and bio- (2–5). Following –coupled receptor engagement by chemical events, involving many different molecules both at the chemoattractants, a complex, still incompletely defined, inside-out qualitative and quantitative level, leukocyte recruitment is never- signaling network is assembled controlling the induction of by guest on September 29, 2021 theless modeled as a sequence of discrete steps, including cap- integrin conformational changes leading to increased affinity for turing, rolling, activation, arrest, crawling to sites of exit, and endothelial ligands. Among many investigated signaling events, transmigration through the barriers of endothelial cells and the Rho (3, 5) and Rap (6, 7) small GTPases are the most studied basement membrane. The transition from rolling to arrest, which is signaling mechanisms linking chemokine-triggered signal trans- the step when leukocytes become fully resistant to the pushing duction to rapid integrin affinity regulation. More recently, we force of the blood flow, is critically controlled by apically presented have demonstrated the critical role of protein tyrosine kinases chemoattractants, such as chemokines, leading to rapid integrin (PTKs) of JAKs, in particular JAK2 and JAK3, as upstream ac- activation (1). Although different modalities of integrin activation tivators of the rho module of LFA-1 affinity triggering in human T lymphocytes (8). In this context, we also started the character- ization of the biochemical link between JAKs and rho small *Department of Medicine, University of Verona, Verona 37134, Italy; †Center for Biomedical Computing, University of Verona, Verona 37134, Italy; GTPases demonstrating the involvement of VAV1 rho guanine ‡Italian Foundation for Cancer Research Institute for Molecular Oncology, Milan nucleotide exchange factor (GEF) in JAK-mediated rho module 20139, Italy; and xDepartment of Oncology and Hemato-Oncology, Italian Database of Operational Losses, University of Milan, Milan 20142, Italy activation by CXCL12. Small GTPase activity is tightly regulated by the rapid turnover Received for publication May 31, 2016. Accepted for publication November 16, 2016. between GDP/GTP-bound states, promoted by GTPase-specific This work was supported by Italian Association for Cancer Research Grant IG16797, GEFs. GEF activity is controlled by different mechanisms, in- Ministero dell’Istruzione, dell’Universita` e della Ricerca Grant PRIN 2009, Nano- cluding protein–protein interactions, protein–lipid interaction, medicine project funding from the University of Verona and the Fondazione Cariverona (to C.L.), Italian Association for Cancer Research Grant IG10168, European Research binding to second messengers, and posttranslational modifications Council Advanced Grant ERC 268836, and by Worldwide Cancer Research Grants (9). Tyrosine phosphorylation is one of the best characterized AICR-09-0582 and 14-0335 (to G.S.). mechanisms of GEF activation (10). For instance, phosphorylation Address correspondence and reprint requests to Prof. Carlo Laudanna, Department of of tyrosine 174 is crucial to VAV1 GEF activity (11). Moreover, Medicine, University of Verona, Strada le Grazie 8, Verona 37134, Italy. E-mail address: [email protected] son of sevenless 1 (SOS1), a ras-specific GEF, manifests a spec- The online version of this article contains supplemental material. ificity switch upon tyrosine phosphorylation, becoming capable of promoting GDP/GTP exchange on Rac1 (12, 13). Notably, VAV1 Abbreviations used in this article: ARHGEF1, rho guanine nucleotide exchange factor 1; DOCK, dedicator of cytokinesis; GEF, guanine nucleotide exchange factor; gene silencing causes only a partial inhibition of LFA-1 affinity P1, Penetratin 1; PTK, protein tyrosine kinase; PTx, pertussis toxin; siRNA, small triggering, and of RhoA and Rac1 activation, suggesting that other interfering RNA; SOS1, son of sevenless 1; TKIP, tyrosine kinase inhibitor peptide. rho-GEFs are at play in CXCL12-induced LFA-1 affinity upreg- Copyright Ó 2017 by The American Association of Immunologists, Inc. 0022-1767/17/$30.00 ulation. Consistent with the latter notion, dedicator of cytokinesis www.jimmunol.org/cgi/doi/10.4049/jimmunol.1600933 The Journal of Immunology 709

(DOCK)2 has been proposed to participate in VLA-4–mediated Human primary cells integrin activation (14). Moreover, JAK-mediated phosphorylation Human primary CD3+ lymphocytes were isolated from whole blood of of rho-GEF 1 (ARHGEF1) has been reported in an angiotensin II– healthy donors by Ficoll and Percoll gradients. Purity of T lymphocyte dependent murine vascular model of hypertension (15). preparation was evaluated by flow cytometry after staining with FITC- conjugated anti-CD3 Ab and was .95%. Isolated T lymphocytes were In this study, we pursued a deeper characterization of the upstream kept at 37˚C in standard adhesion buffer (PBS, 1 mM CaCl2, 1 mM MgCl2, signaling mechanisms linking JAK PTKs to integrin activation and 10% FBS [pH 7.2]) and used within 1 h. The University of Verona Ethics investigated the role of SOS1, ARHGEF1, and DOCK2 as media- Committee approved experimentation involving human primary cells. tors of chemokine-triggered JAK-dependent rho module activa- Static adhesion assay tion leading to LFA-1 affinity upregulation in human primary T lymphocytes. We show that SOS1, ARHGEF1, and DOCK2 are Human primary T lymphocytes were suspended at 3 3 106 cells/ml in phosphorylated on tyrosine by the CXCR4 ligand CXCL12 in a standard adhesion buffer. Adhesion assays were done on 18-well glass slides coated with human 1 mg/ml ICAM-1 in PBS; 20 ml of cell sus- time frame coherent with rapid integrin triggering. By using JAK- pension was added to the wells and stimulated for 120 s at 37˚C with 5 ml specific inhibitors, we further demonstrate that CXCL12-induced of CXCL12, 0.05 mM final concentration or Mn2+, 1 mM final concen- tyrosine phosphorylation of SOS1, ARHGEF1, and DOCK2 is tration. After rapid washing, adherent cells were fixed in ice-cold 1.5% mediated by JAK activity. Moreover, small interfering RNA glutaraldehyde in PBS and counted by computer-assisted enumeration. (siRNA)–mediated downregulation of SOS1, ARHGEF1, and Underflow adhesion assay DOCK2 expression impaired T lymphocyte ability to adhere to 6 ICAM-1 in static and underflow conditions and blocked LFA-1 Human primary T lymphocytes were suspended at 10 cells/ml in standard

adhesion buffer. Cell behavior in underflow conditions were studied with Downloaded from affinity upmodulation by CXCL12. Individual gene silencing of the BioFlux 200 system (Fluxion Biosciences). Forty-eight–well plate these rho-GEFs strongly reduced chemokine-triggered RhoA microfluidics were first cocoated overnight at room temperature with and Rac1 activation. Surprisingly, and differently from what was 2.5 mg/ml human E-selectin and 5 mg/ml human ICAM-1 in PBS. Before previously observed for VAV1 activation, pertussis toxin (PTx) use, microfluidic channels were washed with PBS and then coated with pretreatment of T lymphocytes completely prevented chemokine- 4 mM CXCL12 in PBS for 3 h at room temperature and the assay was done at wall shear stress of 2 dyne/cm2. After extensive washing of microfluidics induced tyrosine phosphorylation of SOS1, ARHGEF1, and with adhesion buffer, the behavior of interacting lymphocytes was recor- DOCK2. Collectively, these results identify SOS1, ARHGEF1, ded on digital drive with a fast CCD video camera (25 frames/s, capable of http://www.jimmunol.org/ and DOCK2 as novel upstream mediators of the rho module of 1/2 subframe 20-ms recording) and analyzed subframe by subframe. Single 2 integrin activation by chemokines, regulated by both JAK PTKs as areas of 0.2 mm were recorded for at least 60 s. Interactions of $20 ms were considered significant and scored. Lymphocytes that remained firmly well as by Gai signaling, and cooperating with VAV1 to mediate adherent for at least 1 s were considered fully arrested. Cells arrested for at LFA-1 affinity triggering by chemokines. The data further point to least 1 s and then detached, or for 10 s and then remaining adherent, thus the possible emergence of stochasticity in concurrent signaling including also events of adhesion stabilization, were scored separately and networks controlling leukocyte trafficking. plotted as independent groups. Rolling interacting and arrested cell be- haviors were automatically detected and quantified with BeQuanti.

Materials and Methods Measurement of LFA-1 affinity states by guest on September 29, 2021 Reagents Lymphocytes suspended in standard adhesion buffer at 2 3 106 cells/ml FBS was from Lonza; human CXCL12, ICAM-1/Fc, and human E-selectin/Fc were stimulated for the indicated time points with 0.05 mM CXCL12 under were from R&D Systems; FITC goat secondary Ab to mouse was from Sigma- stirring at 37˚C in the presence of KIM127 (reporter for extended con- Aldrich; KIM127 mAb recognizing LFA-1 in low-intermediate affinity state formation epitope possibly corresponding to an intermediate affinity state) and IB4 CD18 blocking mAb were from American Type Culture Collection; (17–19) or 327C (reporter for extended conformation epitope related to 327C and 327A mAbs recognizing LFA-1 in the high affinity state were provided a high affinity state) (20). After rapid washing, the cells were stained by Dr. K. Kikly (Eli Lilly); mouse monoclonal anti-ARHGEF1 (ab56462) was with FITC-conjugated secondary Ab and analyzed by cytofluorimetric from Abcam; mouse monoclonal anti-DOCK2 (sc-365242) and rabbit poly- quantification. clonal anti-SOS1 (sc-10803) were from Santa Cruz Biotechnology; mouse siRNA technique anti–phospho-SOS1, specifically detecting phosphorylation on Tyr1196 leading to Rac1-GEF specificity shift, was developed by Prof. G. Scita (S. Gerboth Lymphocytes were plated at 5 3 106 cells/ml in RPMI, 2 mM glutamine, et al., submitted for publication); 4G10 anti–phospho-tyrosine mouse mAb was and 10% FBS for 2 h before silencing. Cells, suspended in nucleofector from Millipore; rabbit anti-actin Ab was from Sigma-Aldrich; tyrphostin buffer at 5 3 107 cells/ml, were electroporated using the Amaxa Nucle- AG490 was from Sigma-Aldrich; WHI-P154 was from Calbiochem; siRNAs ofector (Amaxa Biosystems) in presence of specific siRNAs according to for ARHGEF1 and SOS1 were designed and chemically synthesized by the manufacturer’s instructions. Dharmacon and provided as premixed pools (ON-TARGETplus SMARTpool). The efficiency of siRNA nucleoporation was evaluated with FITC- SOS1 siRNA sequences are: 59-UAGUAGCAGUCUUAGAAUA-39,59- conjugated siRNA, and efficacy of gene silencing was evaluated by spe- GGAUAUGUUUCAUCAAGAU-39,59-GGCAGAAAUUCGACAA UAU-39, cific immunoblotting at 24, 48, and 72 h after transfection. From rho-GEF 59-dAACAGAAGCUGAUCGCAUA-39. ARHGEF1 siRNA sequences are: siRNAs, scrambled sequences were randomly generated by using Vector 59-UGACGUGGCGGGUGACUAA-39,59-AAACUGGUGUGCUCUCAUC- NTI software package and used as negative controls. Intensities of band 39,59-CCACGGCCCUUCGGAAAGU-39,59-UAUACGAGCU GGUGG- signals were quantified by densitometric analysis (Quantity One; Bio-Rad) CACA-39. DOCK2 (bases 4489–4509) siRNA duplexes were designed as by using ImageQuant LAS 4000 (GE Healthcare Life Sciences). reported before (14) and purchased from Dharmacon (custom-designed ON-TARGETplus): 59-UGCCAUAGAAACCAUGUCCdTdT-39. Cell viability Nanovector technology Primary T lymphocytes were labeled for 20 min at 4˚C with Vybrant DyeCycle Violet/SYTOX AADvanced apoptosis kit (Thermo Fisher Sci- The Penetratin 1 (P1)–tyrosine kinase inhibitor peptide (TKIP) peptide entific) following the manufacturer’s instruction. Cell viability was (RQIKIWFQNRRMKWKKGWLVFFVIFYFFR) encompassed the com- assessed by FACS analysis. plete P1 sequence (16 aa), an inserted glycine to allow flexibility of the fusion peptide, and the TKIP sequence (12 aa) blocking specifically JAK2 Immunoprecipitation and Western blot autophosphorylation (16). The P1-TKIP peptide displayed the following properties: 29 aa, m.w. 3969.89, Pi 12.02. Stock solutions (10 mM) in Cells were lysed in ice-cold 1% Nonidet P-40 buffer, containing phos- DMSO were kept at 220˚C and diluted in adhesion buffer immediately phatase inhibitors and complete protease inhibitor mixture (Roche), and before the experiments. The control P1 and the JAK2-blocking (P1-TKIP) lysates were quantified by Bradford assay (Bio-Rad). Protein expression peptides were synthesized by GenScript. Standard treatment of cells with levels were analyzed by Western blot with 7.5% SDS-PAGE. In particular, peptides was for 1 h at 37˚C in 24- or 6-well plates. SOS1, ARHGEF1, and DOCK2 were revealed by immunoblotting with 710 SOS1, ARHGEF1, AND DOCK2 IN LYMPHOCYTE ADHESION anti-SOS1 (Santa Cruz Biotechnology), anti-ARHGEF1 (Abnova), anti-DOCK2 chemical inhibitor), or with a P1-coupled peptide carrying the (Santa Cruz Biotechnology), and anti–b-actin (Sigma-Aldrich) for loading previously described TKIP sequence (P1-TKIP: Trojan nanovector control, followed by appropriate HRP-linked secondary Ab (Amersham Bio- blocking selectively JAK2 autophosphorylation and subsequent sciences) incubation, ECL revelation (Millipore), and quantification by using ImageQuant LAS 4000 (GE Healthcare). For immunoprecipitation assays, 20 3 activation). Pretreatment with these inhibitors led to a complete 106 lymphocytes, treated as described, were stimulated with 0.1 mMCXCL12 inhibition of CXCL12-induced SOS1, ARHGEF1, and DOCK2 for indicated times and then lysed in ice-cold 1% Nonidet P-40 buffer, con- tyrosine phosphorylation (Fig. 1C, 1D), thus confirming that SOS1, taining phosphatase inhibitors and complete protease inhibitor mixture. After ARHGEF1, and DOCK2 are direct targets of tyrosine phosphor- quantification by Bradford assay, lysates were incubated at 4˚C for 4 h under rotary agitation with protein G–Sepharose beads (GE Healthcare) previously ylation by chemokine-triggered JAK PTKs. conjugated overnight with specific Abs depending on specific protein enrich- Collectively, these data show that SOS1, ARHGEF1, and ment. After elution with sample buffer, samples were analyzed by Western blot. DOCK2, together with VAV1, are involved in a JAK-mediated After incubation with specific primary Abs and appropriate HRP-linked sec- signaling pathway induced by chemokines, and, possibly, in the ondary Abs (Amersham Biosciences), immunoreactive bands were visualized control of integrin activation in human primary T lymphocytes. by ECL detection (Millipore) and were quantified by using ImageQuant LAS 4000 (GE Healthcare). In pull-down assays, specific combinations of Abs were SOS1, ARHGEF1, and DOCK2 mediate static and underflow used depending on specific rho-GEF analyzed. In particular, for SOS1, immu- noprecipitation with anti-SOS1 (Santa Cruz Biotechnology)-conjugated beads adhesion on ICAM-1 and Western blot with anti–phospho-SOS1 primary Ab were used. For ARH- To study the involvement of SOS1, ARHGEF1, and DOCK2 in GEF1, immunoprecipitation with 4G10 anti–phospho-tyrosine (Millipore)- conjugated beads and Western blot with anti-ARHGEF1 primary Ab (Abnova) CXCL12-induced LFA-1–mediated adhesion in T lymphocytes, were used. For DOCK2, immunoprecipitation with anti-DOCK2 (Santa Cruz we took advantage of siRNA technique by using GEF-specific

Biotechnology)-conjugated beads and Western blot with 4G10 anti–phospho- siRNAs to selectively and individually downregulate the expres- Downloaded from tyrosine primary Ab (Millipore) were used. sion of SOS1, ARHGEF1, and DOCK2. To silence SOS1 or RhoA and Rac1 activation assays ARHGEF1, we nucleoporated human T lymphocytes with two distinct GEF-specific pools of siRNAs targeting four distinct re- RhoA and Rac1 activations were determined using G-LISA activation assay kits gions of each protein. The maximal efficiency of silencing for (Cytoskeleton), by adaptation of the protocols as previously reported (5). these two was observed after 48 h from nucleo- Lymphocytes, nucleoporated in the presence of specific siRNAs as indi-

∼ ∼ http://www.jimmunol.org/ cated, were stimulated or not with 0.1 mM CXCL12 for 120 s and were lysed poration with a mean decrease of 50% in SOS1 and 60% in for 15 min in the lysis buffer indicated in the manufacturer’s protocol. ARHGEF1 expression (Fig. 2A, 2B). Lymphocyte nucleoporation Protein concentrations were quantified by Bradford assay. Equal amounts of with a pool of scrambled siRNAs, used as control, was ineffective. proteins were added in triplicate to a 96-well plate coated with the RhoA- To silence DOCK2, we designed a specific single siRNA duplex binding domain of rhotekin or anti–Rac1-GTP protein and incubated at 4˚C for 30 min. Wells containing only lysis buffer were used as blank samples. targeting 4489–4509 of DOCK2 human gene sequence as described After washing, the amount of RhoA-GTP bound or Rac1-GTP bound to each previously (14). Nucleoporation of human primary T lymphocytes well was revealed by an anti-RhoA Ab or anti-Rac1 Ab, followed by a with this specific single duplex siRNA efficiently diminished the secondary HRP-labeled Ab and detection of HRP. Signals were measured intracellular content of DOCK2 in a time-dependent manner, with a microplate spectrophotometer (Victor X5 multilabel plate reader, with maximal decrease of ∼60% at 24 h after transfection PerkinElmer) by quantifying absorbance at 490 nm. (Fig. 2A, 2B). Importantly, nucleoporation with individual specific by guest on September 29, 2021 Statistical analysis siRNAs did not affect protein expression of the other GEFs Statistical analysis was performed by calculating mean and SD from different (Supplemental Fig. 1). experiments; significance was calculated by ANOVA followed by a Dunnett Using this approach, we tested the involvement of the three GEFs post hoc test versus control condition indicated in each graph. All statistical in integrin activation by chemokines by first checking the adhesive analyses were performed using GraphPad Prism 6 (GraphPad Software). capability of silenced T lymphocytes in static adhesion assays. SOS1 downregulation resulted in a consistent reduction (nearly a Results 40% decrease) of CXCL12-induced adhesion to ICAM-1, wheras CXCL12 triggers tyrosine phosphorylation of SOS1, ARHGEF1 and DOCK2 silencing had a lower (∼30% decrease), ARHGEF1, and DOCK2 in a JAK PTK-dependent manner but statistically significant impact on static adhesion (Fig. 2C). Phosphorylation on tyrosine residues was shown to modify the GEF Notably, nucleoporation with GEF-specific siRNAs did not glob- activity of SOS1 (12) and ARHGEF1 (15). To further characterize ally affect integrin availability to ligand binding, as assessed by the functional link between CXCL12-triggered JAK PTKs and the Mn2+-induced adhesion on ICAM-1 (Supplemental Fig. 2A), with activation of rho small GTPases, we first tested the ability of specificity of integrin binding to ICAM-1 confirmed by the in- CXCL12 to change the status of tyrosine phosphorylation of hibitory activity of the CD18 blocking Ab IB4 (Supplemental SOS1, ARHGEF1, and DOCK2 with timing consistent with rapid Fig. 2B). Finally, nucleoporation did not impair cell viability cell adhesion and integrin affinity triggering. (Supplemental Fig. 2C). In pull-down tyrosine phosphorylation assays in CXCL12- Next, we evaluated the adhesion of silenced T lymphocytes in stimulated human primary T lymphocytes, we found that SOS1, underflow conditions. Underflow adhesion was assessed by means ARHGEF1, and DOCK2 are rapidly phosphorylated on tyrosine of the BioFlux 200 microfluidic system in microchannels cocoated residues by CXCL12 in a time-dependent manner (Fig. 1A, 1B). with E-selectin, ICAM-1, and CXCL12. Cell behavior was clas- Notably, these data show, to our knowledge for the first time, the sified and scored as rolling cells or cells adherent for 1 s or for 10 s, capability of chemokines to trigger rapid tyrosine phosphorylation conditions that enable the analysis of effects indicating both rapid of SOS1 and ARHGEF1. Our previous data on the role of JAK2 arrest (1 s arrested cells) and, possibly, adhesion stabilization (10 s and JAK3 in integrin activation and in VAV1 phosphorylation and arrested cells). Downregulation of SOS1, ARHGEF1, or DOCK2 activation by CXCL12 may suggest that JAK PTKs are involved reduced CXCL12-triggered rapid arrest of silenced T lymphocytes, in the activation of other rho-specific GEFs in the context of both at 1 s or 10 s of arrest time (Fig. 2D). As expected, the in- signaling mechanisms regulating LFA-1 activation by chemokines. hibition of arrest was accompanied by a concomitant increase of To prove the functional link between JAK PTKs and SOS1, ARHGEF1, rolling cells. Thus, SOS1, ARHGEF1, and DOCK2 rho-GEFs and DOCK2, human T lymphocytes were treated with AG490 mediate CXCL12-triggered rapid arrest with the most prominent (JAKs-specific chemical inhibitor), WHI-P154 (JAK3-selective effect exerted by SOS1, as also observed in static adhesion assays. The Journal of Immunology 711 Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 1. CXCL12 rapidly triggers SOS1, ARHGEF1, and DOCK2 activation in a JAK PTK-mediated signaling pathway in human primary T lymphocytes. (A and B) Time course of SOS1 (left), ARHGEF1 (middle), and DOCK2 (right) tyrosine phosphorylation by CXCL12. Primary T lymphocytes were treated with buffer (NT) or with 0.1 mM CXCL12 for the indicated times; specific protein phosphorylation was analyzed by immunoprecipitation followed by Western blot analysis. Immunoblot (A) and densitometric quantifications (B) of immunoreactive bands of SOS1, ARHGEF1, and DOCK2 protein tyrosine phosphorylation are shown. Immunoblots (A) correspond to one representative experiment of three independent experiments; upper panels show bands of protein phosphorylated in tyrosine; lower panels show total proteins. (B) The densitometric quantifications of band intensity of p-SOS1, p-ARHGEF1, and p-DOCK2, normalized to the total amount of each protein, correspond to mean values of three independent experiments. Error bars represent SD. *p , 0.05, **p , 0.01 versus NT. (C and D)SOS1, ARHGEF1, and DOCK2 activation is mediated by JAK PTK activity. T lymphocytes, pretreated with AG490, P1-TKIP, and WHI-P154 for 1 h, were stimulated with buffer (NT and Control) or with 0.1 mM CXCL12 for 120 s and specific protein tyrosine phosphorylation was analyzed by immunoprecipitation followed by Western blot analysis. Immunoblot (C) and densitometric quantifications (D) of immunoreactive bands of SOS1 (left), ARHGEF1 (middle), and DOCK2 (right) protein tyrosine phosphorylation are shown; upper panels show bands of protein phosphorylated in tyrosine; lower panels show total proteins. Immunoblots (A) correspond to one representative experiment of three independent experiments. (D) The densitometric quantifications of band intensity of phosphorylated p-SOS1, p-ARHGEF1, and p-DOCK2, normalized by the total amount of each protein, correspond to mean values of three independent experiments. Error bars represent SD. *p , 0.01, **p , 0.001 versus control.

SOS1, ARHGEF1, and DOCK2 mediate CXCL12-triggered recognize conformers of extended LFA-1 corresponding to low- LFA-1 affinity triggering intermediate and high affinity states, respectively. Decreased individual To further support the regulatory role of SOS1, ARHGEF1, and expression of each of the three rho-GEFs resulted in significant DOCK2 in integrin activation by chemokines, we investigated blockade of LFA-1 intermediate and high affinity states induced by SOS1, ARHGEF1, and DOCK2 involvement in CXCL12-induced CXCL12, with a stronger reduction observed for SOS1 (Fig. 3A, LFA-1 affinity upregulation by using the reporter mAbs KIM127 Supplemental Figs. 3A, 4A) as compared with ARHGEF1 (Fig. 3B, and 327C (Fig. 3, Supplemental Figs. 3, 4), which specifically Supplemental Figs. 3B, 4B) and DOCK2 (Fig. 3C, Supplemental Figs. 712 SOS1, ARHGEF1, AND DOCK2 IN LYMPHOCYTE ADHESION Downloaded from http://www.jimmunol.org/ by guest on September 29, 2021

FIGURE 2. SOS1, ARHGEF1, and DOCK2 control LFA-1–mediated static and underflow adhesion of human primary T lymphocytes. (A and B) Specific protein downregulation. Lymphocytes were nucleoporated with a pool of four scrambled or SOS1-specific (left) or ARHGEF1-specific (middle) siRNAs and kept in culture for 48 h. Lymphocytes were nucleoporated with single duplex of scrambled or DOCK2-specific (right) siRNA and kept in culture for 24 h. Immunoblot evaluation (A) and densitometric quantification (B) of immunoreactive bands of SOS1, ARHGEF1, and DOCK2 expression are shown. Each immunoblot (A) shows the specific protein content compared with the total amount of actin and represents one of five independent experiments. In each densitometric quantification (B) of immunoreactive bands the relative ratio of the band intensity of total SOS1, ARHGEF1, and DOCK2 was nor- malized by the level of NT intensity. Mean values from four independent experiments are shown; error bars represent SD. *p , 0.01 versus scrambled siRNAs. (C) In static adhesion assay, T lymphocytes, downregulated in specific rho-GEFs expression as in (A), were spotted on 18-well glass slides coated with human ICAM-1 and were stimulated for 120 s with buffer (No) or 0.05 mM CXCL12 (Agonist). Adherent cells number was reported as mean values of four independent experiments in triplicate; error bars represent SD. *p , 0.01 versus scrambled siRNAs. (D) In underflow adhesion assays, T lymphocytes, downregulated in specific rho-GEFs expression as in (A), were fluxed at 2 dyne/cm2 in microcapillary tubes (BioFlux 200 system) coated with E-selectin, ICAM-1, and CXCL12. Cell behavior was expressed as percentage of rolling or arrested cells for 1 s or for 10 s over the total interacting cells. Reported are the mean values of three independent experiments. Error bars represent SD. *p , 0.05, **p , 0.01 versus scrambled siRNAs. The Journal of Immunology 713

(Fig. 4C) in human primary T cells stimulated with CXCL12. The individual ablation of the three GEFs led to a significant and ro- bust reduction in CXCL12-induced activation of both RhoA (Fig. 4, left) and Rac1 (Fig. 4, right). Collectively, these data demonstrate that in human primary T lymphocytes, SOS1, ARHGEF1, DOCK2 (the current study), and VAV1 (8) are all required for optimal JAK-dependent acti- vation of RhoA and Rac1, thus activating the signaling cascade leading to LFA-1 affinity modulation and dependent adhesion induced by chemokines. SOS1, ARHGEF1, and DOCK2 tyrosine phosphorylation by CXCL12 depends on the activity of heterotrimeric Gai proteins

Blockade of heterotrimeric Gai proteins by PTx was originally shown to prevent lymphocyte homing in vivo (21) and to inhibit human T lymphocyte adhesion by CXCL12 in vitro (8). We also showed that the activation of Rap1A and rho small GTPases by CXCL12 is differentially affected by PTx treatment, which

completely abrogates Rap1 activity, whereas it only partially Downloaded from reduces RhoA and Rac1 activation. Notably, PTx treatment had no effects on JAK autophosphorylation or on JAK-dependent tyrosine phosphorylation of VAV1, pointing to the existence of heterotrimeric Gai protein–independent mechanism of JAK, and VAV1, activation by chemokines (8). These results further suggest the

existence of a complex interplay between different upstream http://www.jimmunol.org/ chemokine receptor signal transducers, concurrently involving both heterotrimeric Gai proteins and JAK PTKs. Hence, we next tested whether SOS1, ARHGEF1, and DOCK2 tyrosine phosphoryla- tion might also be independent of the transducing activity of FIGURE 3. SOS1, ARHGEF1, and DOCK2 control the CXCL12- heterotrimeric Ga proteins. Unexpectedly, however, we found induced LFA-1 intermediate and high affinity state. LFA-1 affinity up- i that PTx pretreatment of human T lymphocytes completely regulation detected by specific reporter mAb is shown. Lymphocytes were nucleoporated with a pool of four scrambled (Control) or SOS1-specific inhibited CXCL12-induced tyrosine phosphorylation of SOS1 (A) or ARHGEF1-specific (B) siRNAs and kept in culture for 48 h, or (Fig. 5A), ARHGEF1, (Fig. 5B) and DOCK2 (Fig. 5C). These latter lymphocytes were nucleoporated with single duplex of scrambled (Control) data show that, at variance with respect to VAV1, JAK PTKs and by guest on September 29, 2021 or DOCK2-specific (C) siRNA and kept in culture for 24 h. Lymphocytes Gai-dependent signaling events converge on SOS1, ARHGEF1, were stimulated for 120 s with buffer or 0.05 mM CXCL12 under stirring and DOCK2 to concurrently regulate integrin activation by in the presence of KIM127 (left) or in presence of 327C (right). Induction chemokines (Fig. 6). of intermediate or high state was quantified by cytofluorimetric analysis and reported as fold increase between buffer and CXCL12-stimulated cells. The values were then normalized on the expression level of control Discussion cells (lymphocytes nucleoporated with scrambled siRNAs). Reported are Leukocyte recruitment is tightly regulated by several intracellular the mean values of four different experiments. Error bars represent SD. signal transduction mechanisms triggered, in leukocytes, by ad- *p , 0.01 versus control. hesion molecules, such as selectins and integrins, which generate outside-in signaling events, and by chemotactic factors, such as 3C, 4C). Notably, LFA-1 total expression was completely unaffected chemokines, which trigger inside-out signaling pathways. Che- motactic factors act at different levels because they may regulate by GEF silencing (data not shown). Thus, SOS1, ARHGEF1, and cell tethering and rolling by activating, for instance, the b integrin DOCK2 are involved in CXCL12-induced LFA-1 affinity modula- 1 VLA-4 (22), cell arrest by increasing integrin affinity and clus- tion and dependent adhesion of human normal T lymphocytes. tering, and cell and microenvironmental positioning These findings, together with our previous data on VAV1 (8), indi- within the tissues. Furthermore, chemoattractants may also induce cate that at least four distinct rho-GEFs play nonredundant, regu- gene expression (23). It is not surprising that such a phenome- latory roles in integrin-mediated adhesion induced by chemokines nological variety is regulated by extremely complex signaling in human primary T lymphocytes, possibly highlighting a differ- machinery, and a complete description of chemoattractant- ential quantitative contribution of distinct rho-GEFs to the signaling triggered signaling mechanisms is mandatory to fully character- pathway triggered by CXCL12 and controlling LFA-1 activation. ize the different steps of leukocyte migration and the consequent behavior within the tissues. In the context of leukocyte interaction SOS1, ARHGEF1, and DOCK2 mediate CXCL12- triggered with the endothelial cells, rapid upmodulation of integrin affinity activation of RhoA and Rac1 by chemokines is the main modality of integrin activation sup- We have previously shown that the small GTPases RhoA, Rac1 porting immediate arrest underflow (2, 3). A subset of the small GTPases, and their effectors Rap1A, PLD1, and PIP5K1C are chemoattractant-triggered signaling network is devoted to the all implicated in regulating CXCL12-induced LFA-1 affinity regulation of integrin affinity activation, with the involvement of triggering (5, 8). However, the possibility that SOS1, ARHGEF1, at least 67 signaling molecules (24), although many gaps in the and DOCK2 may mediate JAK-dependent RhoA and Rac1 acti- network are still present and we are far from a complete charac- vation by chemokines was never investigated. To this end, we terization of the pathway intermediates linking very upstream silenced SOS1 (Fig. 4A), ARHGEF1 (Fig. 4B), and DOCK2 transducers, such as heterotrimeric Gai proteins or JAK PTKs, to 714 SOS1, ARHGEF1, AND DOCK2 IN LYMPHOCYTE ADHESION

FIGURE 4. SOS1, ARHGEF1, and DOCK2 mediate RhoA and Rac1 activation by CXCL12. RhoA (left) and Rac1 (right) activation (GTP-bound state) analysis by G-LISA assay. Lymphocytes were nucleoporated with a pool of four scrambled (NT and Control) or SOS1-specific (A) or ARHGEF1-specific (B) siRNAs and kept in culture for 48 h or lymphocytes were nucleoporated with single Downloaded from duplex of scrambled (Control) or DOCK2-specific (C) siRNA and kept in culture for 24 h. Lymphocytes were stimulated with 0.1 mM CXCL12 for 120 s and then lysed following the manufacturer’s protocol. The levels of GTP-bound RhoA and Rac1 were measured (absor- bance at 490 nm) by Victor X5 multilabel plate reader.

The values are reported as fold increase over non- http://www.jimmunol.org/ stimulated cells (NT) and represent the mean values of three independent experiments in duplicate. Error bars represent SD. *p , 0.01 versus control. by guest on September 29, 2021

downstream effectors, such rap and rho small GTPases and more recent data identified biallelic mutations in DOCK2 gene in distal effectors, such as Talin1 and Kindlin3. five patients characterized by early-onset invasive infections. GEFs specific for small GTPases are obvious candidates to fill T lymphocytes isolated from these patients are characterized by an the functional gaps between JAK PTKs and the rho module of impaired chemokine-induced migration coupled to a reduced Rac1 integrin activation. The encodes for ∼80 rho-GEFs, activation in vitro (26). These data strongly suggested the possible divided into two unrelated gene families: the DBL family, with 69 role of DOCK2 in chemokine-induced, rho-mediated, LFA-1 affinity members, and the DOCK family, with 11 members. DOCK pro- modulation. Furthermore, among other rho-GEFs, we also focused teins, containing the conserved catalytic domain DOCK homology the analysis on SOS1 and ARHGEF1 GEFs. These are interesting region 2, are structurally and mechanistically unrelated to the targets of investigation because SOS1 is an atypical small GTPase DBL family, which is characterized by the presence of a Dbl GEF capable of switching its target specificity from Ras to Rac1 upon homology domain. Available data suggest that DOCK GEFs may tyrosine phosphorylation (27), whereas ARHGEF1, a specific mem- trigger GDP/GTP exchange of Rac and/or Cdc42, but not of ber of RGS-GEFs family, has been demonstrated to be directly ac- RhoA, whereas the Dbl family can activate RhoA, Rac1, and/or tivated by JAK2 PTK, through phosphorylation of Tyr738,ina Cdc42. Collectively, the two families of rho-specific GEFs exceed vascular model of hypertension (15). Importantly, a possible regula- by at least 4-fold the number of small GTPases, implying that a tory role for SOS1 and ARHGEF1 in integrin activation by chemo- single rho small GTPase can be activated by multiple GEFs and attractants was never addressed. that, at the same time, a single GEF can activate many different In this study, by applying a siRNA approach to downmodulate rho GTPases (25). Our previous data demonstrated that CXCL12- rho-GEFs in primary human T lymphocytes, we demonstrate, in induced LFA-1 activation is only partially regulated by VAV1 agreement with the previously formalized four criteria defining the GEF activity, thus suggesting the possible involvement of other role of signaling molecules in integrin affinity activation by che- rho-GEFs in this process. The involvement of multiple rho-GEFs, mokines (24), the concurrent involvement of SOS1, ARHGEF1, other than VAV1, in LFA-1 activation is also supported by the and DOCK2 in the activation of LFA-1 affinity triggering and involvement of DOCK2 in VLA-4 regulation (14). Interestingly, mediated adhesion by chemokines. Specifically, our data show The Journal of Immunology 715

FIGURE 5. SOS1, ARHGEF1, and DOCK2 activa- tion by CXCL12 is dependent on heterotrimeric Gai protein signaling. Lymphocytes were treated with buffer (NT and Control) or 2 mg/ml PTx for 3 h and stimulated with 0.1 mM CXCL12 for 120 s, and specific protein phosphorylation was analyzed by immunoprecipitation followed by Western blot analysis. Immunoblot (left) and densitometric quantifications (right) of immunore- Downloaded from active bands of SOS1 (A), ARHGEF1 (B), and DOCK2 (C) protein tyrosine phosphorylation are shown. Im- munoblots (left) correspond to one representative experiment of three independent experiments. The densitometric quantifications (right) of band intensity of p-SOS1, p-ARHGEF1, and p-DOCK2, normalized by

the total amount of each protein, correspond to mean http://www.jimmunol.org/ values of three independent experiments. Error bars represent SD. *p , 0.05, **p , 0.01 versus control. by guest on September 29, 2021

that: 1) SOS1, ARHGEF1, and DOCK2 are tyrosine phosphorylated well as heterotrimeric Gai protein signaling. Thus, in the rho module of by CXCL12 in a time frame consistent with LFA-1 affinity triggering; integrin activation, we observe a variety of dependencies on the activity 2) tyrosine phosphorylation of SOS1, ARHGEF1 and DOCK2 is of the two very upstream traducers (Fig. 6). Notably, such functional mediated by JAK PTKs; 3) SOS1, ARHGEF1, and DOCK2 mediate cooperation between different pathways downstream of the same re- LFA-1 affinity activation and dependent lymphocyte rapid ad- ceptor was already reported for fMLP-stimulated cooperation of VAV1 hesion; 4) SOS1, ARHGEF1, and DOCK2 mediate chemokine- andP-Rex1inmurineneutrophils(28). Mechanistically, our data may induced RhoA and Rac1 activation by CXCL12; and 5) SOS1, be accounted for by a role exerted by lipid metabolism in the coor- ARHGEF1, and DOCK2 tyrosine phosphorylation is prevented by dinated regulation of GEF activity by JAKs and heterotrimeric Gai PTx, thus implying a concurrent role for Gai-mediated signaling. proteins. Indeed, GEF activity is regulated by the presence of specific The emerging view is that of a complex interrelationship between domains able to bind , such as the pleckstrin homology JAK- and heterotrimeric Gai protein–mediated signaling. Indeed, (PH) domain in SOS1 and ARHGEF1 and atypical DOCK homology JAK activation is insensitive to heterotrimeric Gai protein inhibition region 1 (DHR1) domain in DOCK2. These domains may regulate by PTx, and heterotrimeric Gai protein function was not blocked by protein targeting to membrane, thus bringing these GEFs close to JAK JAK inhibition (JAK inhibitors do not inhibit intracellular calcium PTKs and making them available to subsequent tyrosine phosphory- spikes induced by chemokines) (8). Thus, JAK and heterotrimeric lation and full activation. Moreover, phospholipids binding to Dbl- Gai protein act as two independent upstream transducers of chemo- related pleckstrin homology domains may directly modulate kine receptors. Moreover, VAV1 activation is fully dependent on JAK the exchange activity of the adjacent Dbl homology domains activity but insensitive to PTx inhibition, whereas RhoA and Rac1 through allosteric processes (29, 30). For instance, PIP3 has been activation is fully dependent on JAK activity but only partially de- reported to increase the Rac GEF activity of VAV1 (31) and P-Rex1 pendent on heterotrimeric Gai protein signaling (8). In contrast, (32)andalsoofSOS1inaSOS1/ABI1/EPS8complex(13). Rap1A activation is fully dependent on both JAK and heterotrimeric Our present and past (8) data show that at least four rho-GEFs Gai protein activity. Finally, and similarly to Rap1A, our latter data cooperate to mediate JAK-dependent activation of the rho mod- show that SOS1, ARHGEF1 and DOCK2 chemokine-induced tyro- ule of integrin affinity triggering by chemokines in human pri- sine phosphorylation and activation are either dependent on JAK as mary lymphocytes, with an apparent prominent role for SOS1 and 716 SOS1, ARHGEF1, AND DOCK2 IN LYMPHOCYTE ADHESION Downloaded from

FIGURE 6. Model of concurrent involvement of VAV1, SOS1, DOCK2, and ARHGEF1 GEFs in JAKs and Gai-dependent rho module activation

controlling LFA-1 affinity triggering by chemokines. CXCR4 engagement by CXCL12 induces LFA-1 affinity triggering through the concurrent upstream http://www.jimmunol.org/ activity of JAK PTKs and heterotrimeric Gai protein leading to activation of multiple rho-GEFs. Whereas VAV1 activation is dependent only on JAK kinase activity, SOS1, DOCK2, and ARHGEF1 are equivalently controlled by Gai protein and JAKs. In turn, VAV1, SOS1, ARHGEF1, and DOCK2 cooperate to activate RhoA and Rac1. Additionally, Rap1A activation is dependent on both JAK-activated RhoA and PLC regulation by heterotrimeric Gai protein. Together, the Rho and Rap modules coordinately regulate the activity of several actin/integrin-binding proteins (FERMT3, Kindlin3; TLN1, Talin1; CYTH1, Cytohesin-1; RIAM and RASSF5, RAPL), thus leading to the regulation of LFA-1–mediated rapid arrest.

VAV1 compared with ARHGEF1 and DOCK2. Functional over- represent, de facto, averaged data derived from populations of un- lapping in chemokine-triggered signaling was already demonstrated synchronized cells, the obtained findings represent the sum of the previously, such as for RhoA and Rac1 small GTPases (5), or for possibilities in terms of participation of a specific category of signal by guest on September 29, 2021 JAK2 and JAK3 PTKs (8). Besides simply representing evolutionary transducers to the integrin regulatory network. This does not remove redundancy possibly leading to compensatory mechanisms, perhaps concurrency, but introduces a level of stochasticity in signaling suggested by the partial recovery of LFA-1 affinity inhibition during mechanisms triggered by chemotactic factors. The apparent com- longer time of CXCL12 stimulation (see Supplemental Fig. 3), plexity of the signaling mechanisms regulating leukocyte trafficking functional overlapping more likely reflects a mechanism of signaling could simply derive from a nondeterministic usage of the signaling quantitative fine tuning in response to different environmental cues, molecules most readily available, at a specific time point and sub- regulated by quantitative variability in protein copy number, affinity cellular location, to participate in the building of a localized signaling for ligands, catalytic activity, or temporal activation. Previous data module (a signalosome) dedicated to the regulation of a specific cell indicate that quantitative regulation of signaling events controlling function. In this context, the different molecules involved in the reg- leukocyte arrest is essential to generate a flexible signaling system ulation of integrin activity by chemokines could be possibly regarded able to efficiently respond to a variety of environmental conditions as interchangeable transistors in electronic circuits, where the final (33); consequently, the concept of concurrency has been included in function is mainly dependent on the architecture more than the nature the general multistep model of leukocyte recruitment (34). However, of the individual components. Thus, differently from other, more de- the growing complexity that is emerging from the ongoing scrutiny terministic, signaling mechanisms, such as cytokine or TcR signaling, of the signaling mechanisms controlling integrin activation, which, in which the repertoire of involved signaling components is more with our new data, now reach the noteworthy number of 70 involved restricted and well defined, chemokine signaling could be based on signaling proteins, may suggest an updated vision of the global stochastic assembling of membrane-localized regulatory modules, signaling machinery activated by chemokines and controlling leu- otherwise involving the most readily accessible proteins out of a wide kocyte trafficking. Indeed, note that many available data are obtained repertoire of components, thus allowing reaching a threshold of signal in primary cells, which are unlikely to be functionally synchronized. independently of the individuality of the single components. Further If we consider the global pool of signaling molecules expressed by a data need to be obtained to clarify this hypothetical framework, such leukocyte, it is possible, if not likely, that the expression levels and/ as, for instance, analyzing several GEFs simultaneously, by applying or the intracellular distributions of the individual components of a multiple rho-GEFs silencing or developing new chemical compounds regulatory pathway are not uniform throughout the experimentally targeting rho-GEF activity (35, 36) or testing the distribution of acti- analyzed cellular pool. Additionally, such uniformity is note likely to vated signaling proteins within a cell population. be maintained over time in a single cell. Thus, it is possible that, Collectively, our results provide a deeper characterization of the upon chemokine triggering, the global repertoire of 70 signaling inside-out signaling pathway controlling chemokine-induced LFA-1 proteins regulating integrin activation is not always fully accessible in affinity triggering and dependent adhesion of human primary every single cell, and a differential availability, and thus usage, may T lymphocytes. These data suggest the emergence of a very complex occur within the cell population. Because the experimental data interplay between the signaling components of the rho-module, possibly The Journal of Immunology 717 highlighting the prominent regulatory role of the circuitry versus the 17. Robinson, M. K., D. Andrew, H. Rosen, D. Brown, S. Ortlepp, P. Stephens, and individual molecular components in the signaling network controlling E. C. Butcher. 1992. Antibody against the Leu-CAM beta-chain (CD18) pro- motes both LFA-1- and CR3-dependent adhesion events. J. Immunol. 148: 1080– leukocyte trafficking. 1085. 18. Shimaoka, M., M. Kim, E. H. Cohen, W. Yang, N. Astrof, D. Peer, A. Salas, A. Ferrand, and T. A. Springer. 2006. AL-57, a ligand-mimetic antibody to Acknowledgments integrin LFA-1, reveals chemokine-induced affinity up-regulation in lymphocytes. We are grateful to Dr. Kristine Kikly (Eli Lilly and Company) for providing Proc. Natl. Acad. Sci. USA 103: 13991–13996. the anti–LFA-1 mAbs 327C and 327A. We thank the Center for Biomed- 19. Stanley, P., A. Smith, A. McDowall, A. Nicol, D. Zicha, and N. Hogg. 2008. ical Computing of the University of Verona for providing Victor X5 and Intermediate-affinity LFA-1 binds alpha-actinin-1 to control migration at the leading edge of the T cell. EMBO J. 27: 62–75. BioFlux technologies. 20. Lum, A. F., C. E. Green, G. R. Lee, D. E. Staunton, and S. I. Simon. 2002. Dynamic regulation of LFA-1 activation and arrest on intercellular adhesion molecule 1 (ICAM-1) in shear flow. J. Biol. Chem. 277: 20660–20670. Disclosures 21. Bargatze, R. F., and E. C. Butcher. 1993. Rapid G protein-regulated activation The authors have no financial conflicts of interest. event involved in lymphocyte binding to high endothelial venules. J. Exp. Med. 178: 367–372. 22. Grabovsky, V., S. Feigelson, C. Chen, D. A. Bleijs, A. Peled, G. Cinamon, References F. Baleux, F. Arenzana-Seisdedos, T. Lapidot, R. R. Lobb, et al. 2000. Subsecond induction of a integrin clustering by immobilized chemokines stimulates leu- 1. Takada, Y., X. Ye, and S. Simon. 2007. The integrins. Genome Biol. 8: 215. 4 kocyte tethering and rolling on endothelial vascular cell adhesion molecule 1 2. Constantin, G., M. Majeed, C. Giagulli, L. Piccio, J. Y. Kim, E. C. Butcher, and under flow conditions. J. Exp. Med. 192: 495–506. C. Laudanna. 2000. Chemokines trigger immediate b2 integrin affinity and 23. Cassatella, M. A., F. Bazzoni, M. Ceska, I. Ferro, M. Baggiolini, and G. Berton. mobility changes: differential regulation and roles in lymphocyte arrest under flow. Immunity 13: 759–769. 1992. IL-8 production by human polymorphonuclear leukocytes. The chemo- 3. Giagulli, C., E. Scarpini, L. Ottoboni, S. Narumiya, E. C. Butcher, G. Constantin, attractant formyl-methionyl-leucyl-phenylalanine induces the gene expression Downloaded from and C. Laudanna. 2004. RhoA and z PKC control distinct modalities of LFA-1 and release of IL-8 through a pertussis toxin-sensitive pathway. J. Immunol. 148: activation by chemokines: critical role of LFA-1 affinity triggering 3216–3220. in lymphocyte in vivo homing. Immunity 20: 25–35. 24. Montresor, A., L. Toffali, G. Constantin, and C. Laudanna. 2012. Chemokines 4. Kim, M., C. V. Carman, W. Yang, A. Salas, and T. A. Springer. 2004. The and the signaling modules regulating integrin affinity. Front. Immunol. 3: 127. primacy of affinity over clustering in regulation of adhesiveness of the integrin 25. Tybulewicz, V. L., and R. B. Henderson. 2009. Rho family GTPases and their regulators in lymphocytes. Nat. Rev. Immunol. 9: 630–644. aLb2. J. Cell Biol. 167: 1241–1253. 5. Bolomini-Vittori, M., A. Montresor, C. Giagulli, D. Staunton, B. Rossi, 26. Dobbs, K., C. Domı´nguez Conde, S. Y. Zhang, S. Parolini, M. Audry, J. Chou,

M. Martinello, G. Constantin, and C. Laudanna. 2009. Regulation of conformer- E. Haapaniemi, S. Keles, I. Bilic, S. Okada, et al. 2015. Inherited DOCK2 de- http://www.jimmunol.org/ specific activation of the integrin LFA-1 by a chemokine-triggered Rho signaling ficiency in patients with early-onset invasive infections. N. Engl. J. Med. 372: module. Nat. Immunol. 10: 185–194. 2409–2422. 6. Ebisuno, Y., K. Katagiri, T. Katakai, Y. Ueda, T. Nemoto, H. Inada, J. Nabekura, 27. Innocenti, M., P. Tenca, E. Frittoli, M. Faretta, A. Tocchetti, P. P. Di Fiore, and T. Okada, R. Kannagi, T. Tanaka, et al. 2010. Rap1 controls lymphocyte adhe- G. Scita. 2002. Mechanisms through which Sos-1 coordinates the activation of sion cascade and interstitial migration within lymph nodes in RAPL-dependent Ras and Rac. J. Cell Biol. 156: 125–136. and -independent manners. Blood 115: 804–814. 28. Lawson, A. J., P. Rischbieter, A. Numanoglu, N. Wieselthaler, and 7. Katagiri, K., N. Ohnishi, K. Kabashima, T. Iyoda, N. Takeda, Y. Shinkai, S. J. Beningfield. 2011. Imaging the Rex vein preoperatively using wedged K. Inaba, and T. Kinashi. 2004. Crucial functions of the Rap1 effector molecule hepatic venous portography. Pediatr. Radiol. 41: 1246–1249. RAPL in lymphocyte and trafficking. Nat. Immunol. 5: 1045–1051. 29. Baumeister, M. A., K. L. Rossman, J. Sondek, and M. A. Lemmon. 2006. The 8. Montresor, A., M. Bolomini-Vittori, L. Toffali, B. Rossi, G. Constantin, and Dbs PH domain contributes independently to membrane targeting and regulation C. Laudanna. 2013. JAK tyrosine kinases promote hierarchical activation of Rho of guanine nucleotide-exchange activity. Biochem. J. 400: 563–572. and Rap modules of integrin activation. J. Cell Biol. 203: 1003–1019. 30. Rossman, K. L., D. K. Worthylake, J. T. Snyder, D. P. Siderovski, S. L. Campbell, by guest on September 29, 2021 9. Bos, J. L., H. Rehmann, and A. Wittinghofer. 2007. GEFs and GAPs: critical and J. Sondek. 2002. A crystallographic view of interactions between Dbs and elements in the control of small G proteins. Cell 129: 865–877. Cdc42: PH domain-assisted guanine nucleotide exchange. EMBO J. 21: 1315– 10. Rossman, K. L., C. J. Der, and J. Sondek. 2005. GEF means go: turning on RHO 1326. GTPases with guanine nucleotide-exchange factors. Nat. Rev. Mol. Cell Biol. 6: 31. Das, B., X. Shu, G. J. Day, J. Han, U. M. Krishna, J. R. Falck, and D. Broek. 167–180. 2000. Control of intramolecular interactions between the pleckstrin homology 11. Lo´pez-Lago, M., H. Lee, C. Cruz, N. Movilla, and X. R. Bustelo. 2000. Tyrosine and Dbl homology domains of Vav and Sos1 regulates Rac binding. J. Biol. phosphorylation mediates both activation and downmodulation of the biological Chem. 275: 15074–15081. Mol. Cell. Biol. activity of Vav. 20: 1678–1691. 32. Welch, H. C., W. J. Coadwell, C. D. Ellson, G. J. Ferguson, S. R. Andrews, 12. Sini, P., A. Cannas, A. J. Koleske, P. P. Di Fiore, and G. Scita. 2004. Abl- H. Erdjument-Bromage, P. Tempst, P. T. Hawkins, and L. R. Stephens. 2002. dependent tyrosine phosphorylation of Sos-1 mediates growth-factor-induced P-Rex1, a PtdIns(3,4,5)P3- and Gbg-regulated guanine-nucleotide exchange Rac activation. Nat. Cell Biol. 6: 268–274. Cell 13. Innocenti, M., E. Frittoli, I. Ponzanelli, J. R. Falck, S. M. Brachmann, P. P. Di factor for Rac. 108: 809–821. Fiore, and G. Scita. 2003. Phosphoinositide 3-kinase activates Rac by entering in 33. D’Ambrosio, D., C. Albanesi, R. Lang, G. Girolomoni, F. Sinigaglia, and a complex with Eps8, Abi1, and Sos-1. J. Cell Biol. 160: 17–23. C. Laudanna. 2002. Quantitative differences in chemokine receptor engagement 14. Garcı´a-Bernal, D., E. Sotillo-Mallo, C. Nombela-Arrieta, R. Samaniego, generate diversity in integrin-dependent lymphocyte adhesion. J. Immunol. 169: Y. Fukui, J. V. Stein, and J. Teixido´. 2006. DOCK2 is required for chemokine- 2303–2312. promoted human T lymphocyte adhesion under shear stress mediated by the 34. D’Ambrosio, D., P. Lecca, G. Constantin, C. Priami, and C. Laudanna. 2004. Concurrency in leukocyte vascular recognition: developing the tools for a pre- integrin a4b1. J. Immunol. 177: 5215–5225. 15. Guilluy, C., J. Bre´geon, G. Toumaniantz, M. Rolli-Derkinderen, K. Retailleau, dictive computer model. Trends Immunol. 25: 411–416. L. Loufrani, D. Henrion, E. Scalbert, A. Bril, R. M. Torres, et al. 2010. The Rho 35. Nishikimi, A., T. Uruno, X. Duan, Q. Cao, Y. Okamura, T. Saitoh, N. Saito, exchange factor Arhgef1 mediates the effects of angiotensin II on vascular tone S. Sakaoka, Y. Du, A. Suenaga, et al. 2012. Blockade of inflammatory responses and blood pressure. Nat. Med. 16: 183–190. by a small-molecule inhibitor of the Rac activator DOCK2. Chem. Biol. 19: 488– 16. Flowers, L. O., H. M. Johnson, M. G. Mujtaba, M. R. Ellis, S. M. Haider, and 497. P. S. Subramaniam. 2004. Characterization of a peptide inhibitor of Janus kinase 36. Evelyn, C. R., X. Duan, J. Biesiada, W. L. Seibel, J. Meller, and Y. Zheng. 2014. 2 that mimics suppressor of cytokine signaling 1 function. J. Immunol. 172: Rational design of small molecule inhibitors targeting the Ras GEF, SOS1. 7510–7518. Chem. Biol. 21: 1618–1628. 1 Supplemental Figure 1. Specific SOS1, ARHGEF1 or DOCK2 siRNAs do not affect other GEF

2 proteins expression. Lymphocytes were nucleoporated with a pool of four scrambled or SOS1 (A)

3 or ARHGEF1 (B) specific siRNAs and kept in culture for 48 h. Lymphocytes were nucleoporated

4 with single duplex of scrambled or DOCK2 (C) specific siRNA and kept in culture for 24 h.

5 Immunoblot evaluation of immunoreactive bands of SOS1, ARHGEF1, DOCK2 and actin

6 expression. A representative experiment of three independent experiments is reported.

1 Supplemental Figure 2. Specific SOS1, ARHGEF1 and DOCK2 siRNAs do not affect Mn2+

2 induced adhesion and cell viability. Lymphocytes were nucleoporated with a pool of four scrambled

3 or SOS1 (left) or ARHGEF1 (middle) specific siRNAs and kept in culture for 48 h. Lymphocytes

4 were nucleoporated with single duplex of scrambled or DOCK2 (right) specific siRNA and kept in

5 culture for 24 h. In static adhesion assay (A) T lymphocytes, down-regulated in specific rho-GEFs,

6 were spotted on 18-well glass slides coated with human ICAM-1 and stimulated for 120 s with

7 Mn2+ 1 mM. Adherent cells number was reported as mean values of three independent experiments

8 in triplicate; error bars represent SD. (B) Cells down-regulated in GEFs expression pre-treated or

9 not with IB4 blocking LFA1 antibody, were spotted on 18-well glass slides coated with human

10 ICAM-1 and stimulated for 120 s with buffer (No) or 0.05 µM CXCL12. Adherent cells number

11 was reported as mean values of three independent experiments in triplicate; error bars represent SD;

12 *, P < 0.01, versus scrambled siRNAs; **, P < 0.001, versus CXCL12. (C) Cytofluorimetric

13 analysis of cells viability by double staining with Vybrant (apoptosis) and Sytox (necrosis).

14 Histograms represent percentage of double negative cells from three independent experiments.

15 Error bars are SD.

1 Supplemental Figure 3. Time course of CXCL12 induced LFA-1 high affinity state. Lymphocytes

2 were nucleoporated with a pool of four scrambled (Control) or SOS1 (A) or ARHGEF1 (B) specific

3 siRNAs and kept in culture for 48 h or lymphocytes were nucleoporated with single duplex of

4 scrambled (Control) or DOCK2 (C) specific siRNA and kept in culture for 24 h. Lymphocytes were

5 stimulated for 30, 60, 120 and 180 s with buffer or 0.05 µM CXCL12 under stirring in presence of

6 327C. Induction of high state was quantified by cytofluorimetric analysis and reported as fold

7 increase between buffer and CXCL12-stimulated cells at each time point. Values were then

8 normalized on the expression level of control cells (lymphocytes nucleoporated with scrambled

9 siRNAs). Here are reported the mean values of three independent experiments. Error bars are SD; *,

10 P < 0.01, versus Control.

Toffali et al.

1 Supplemental Figure 4. Representative flow cytometry data. Lymphocytes were nucleoporated with

2 a pool of four scrambled (single and bold lines) or SOS1 (A) or ARHGEF1 (B) specific siRNAs

3 and kept in culture for 48 h or lymphocytes were nucleoporated with single duplex of scrambled or

4 DOCK2 (C) specific siRNA and kept in culture for 24 h (dashed and dotted lines, for each specific

5 protein silencing). Lymphocytes were stimulated for 120 s with buffer (single lines for control and

6 dotted lines for specific silencing) or 0.05 µM CXCL12 (bold lines for control and dashed lines for

7 specific silencing) under stirring in the presence of KIM127 (left) or in presence of 327C (right).

8 Induction of intermediate or high state was quantified by cytofluorimetric analysis. Results are

9 presented as histograms of fluorescence intensity in log scale. Here are reported a representative

10 experiment of four independent experiments.

1