Chemotaxis Role in Neutrophil Polarization and Leukocyte-Specific

Modulation of Mac-1 (CD11b/CD18)-Mediated Adhesion by the Leukocyte-Specific Protein 1 Is Key to Its Role in Neutrophil Polarization and This information is current as Chemotaxis of September 26, 2021. Chunjie Wang, Hisayoshi Hayashi, Rene Harrison, Basil Chiu, Jason R. Chan, Hanne L. Ostergaard, Robert D. Inman, Jan Jongstra, Myron I. Cybulsky and Jenny Jongstra-Bilen

J Immunol 2002; 169:415-423; ; Downloaded from doi: 10.4049/jimmunol.169.1.415 http://www.jimmunol.org/content/169/1/415 http://www.jimmunol.org/ References This article cites 74 articles, 37 of which you can access for free at: http://www.jimmunol.org/content/169/1/415.full#ref-list-1

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists by guest on September 26, 2021

• 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 © 2002 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Modulation of Mac-1 (CD11b/CD18)-Mediated Adhesion by the Leukocyte-Speciﬁc Protein 1 Is Key to Its Role in Neutrophil Polarization and Chemotaxis1

Chunjie Wang,*‡†† Hisayoshi Hayashi,¤** Rene Harrison,¤** Basil Chiu,*‡ Jason R. Chan,†¶ Hanne L. Ostergaard,‡‡ Robert D. Inman,*‡ʈ Jan Jongstra,*‡ Myron I. Cybulsky,†¶ and Jenny Jongstra-Bilen2*‡

Leukocyte-specific protein 1 (LSP1) is an intracellular filamentous-actin binding protein which modulates cell motility. The cellular process in which LSP1 functions to regulate motility is not yet identified. In this study, we show that LSP1 negatively regulates fMLP-induced polarization and chemotaxis of neutrophils through its function on adhesion via specific integrins. Using LSP1-deficient (Lsp1؊/؊) mice, we show increased neutrophil migration into mouse knee joints during zymosan-induced acute Downloaded from inflammation, an inflammatory model in which the number of resident synoviocytes are not affected by LSP1-deficiency. In vitro chemotaxis experiments performed by time-lapse videomicroscopy showed that purified Lsp1؊/؊ bone-marrow neutrophils exhibit an increased migration rate toward a gradient of fMLP as compared with wild-type neutrophils. This difference was observed when cells migrated on fibrinogen, but not fibronectin, suggesting a role for LSP1 in modulating neutrophil adhesion by specific integrins. LSP1 is also a negative regulator of fMLP-induced adhesion to fibrinogen or ICAM-1, but not to ICAM-2, VCAM-1, http://www.jimmunol.org/ or fibronectin. These results suggest that LSP1 regulates the function of Mac-1 (CD11b/CD18), which binds only to fibrinogen and ICAM-1 among the substrates we tested. fMLP-induced filamentous actin polarization is also increased in the absence of LSP1 when cells were layered on fibrinogen, but not on fibronectin. Our findings suggest that the increased neutrophil recruitment in ,Lsp1؊/؊ mice during acute inflammation derives from the negative regulatory role of LSP1 on neutrophil adhesion, polarization and migration via specific integrins, such as Mac-1, which mediate neutrophil responses to chemotactic stimuli. The Journal of Immunology, 2002, 169: 415Ð423.

eutrophils are among the ﬁrst leukocytes to reach an toid arthritis or inﬂammatory bowel disease, uncontrolled accumu-

inflammatory site and initiate host-defense against lations of neutrophils yield aberrantly elevated levels of inflam- by guest on September 26, 2021 N pathogens. Activated neutrophils are able to fight infec- matory compounds which become major contributors to tissue tion effectively, together with monocytes and resident tissue mac- damage (2, 3). Thus, regulation of neutrophil recruitment into in- rophages, by phagocytosis of microorganisms, or by secreting sev- flammatory sites and their clearance are critical processes assuring eral potent inflammatory compounds such as radical oxygen effective host defense without tissue injury. intermediates, proteases, or peroxidases. The importance of neu- The emigration of circulating neutrophils from blood to in- trophil function in innate immunity is best evidenced by several flamed tissue is a complex process dependent upon the coordina- human neutrophil dysfunctions whereby defects in neutrophil mi- tion of many cellular functions that culminate in directed cell gration or anti-inflammatory activity is the cause for severe sus- movement (4Ð6). Adhesion receptors are key to several of these ceptibility to bacterial infections and defective wound healing (1). processes. Transendothelial migration is initiated by neutrophil In contrast, during chronic inflammatory diseases such as rheuma- rolling mediated by the selectin family of adhesion receptors. Roll- ing neutrophils arrest on endothelium and undergo firm adhesion, *Cellular and Molecular Biology Division, Toronto Western Research Institute, and characterized by spreading and the formation of focal adhesion- †Toronto General Research Institute, University Health Network, Toronto, Canada; like structures mediated by integrin receptors. Integrins are het- Departments of ‡Immunology, ¤Biochemistry, ¶Laboratory Medicine and Pathobiol- ʈ erodimeric transmembrane proteins consisting of noncovalently ogy, and Medicine, University of Toronto, Toronto, Canada; **Cell Biology Pro- †† ␣ ␤ ␤ gram, Hospital for Sick Children, Toronto, Canada; Department of Pathology, The bound - and -chains (7). In neutrophils, the 1 integrins com- Fourth Military Medical University, Xi’an, China; and ‡‡Department of Immunology, prise a common ␤-chain (CD29) and distinct ␣-chains (CD49) (8, University of Alberta, Edmonton, Canada ␤ ␤ 9). The 2 family of integrins share their -chain (CD18) with Received for publication October 23, 2001. Accepted for publication April 29, 2002. ␣ ␤ distinct -chains (CD11). The importance of the 2 integrins in The costs of publication of this article were defrayed in part by the payment of page adhesion and transmigration is evident by the analysis of patients charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. with leukocyte adhesion deficiency type I syndrome which results 1 This work was supported with funds from the University of Toronto Arthritis Center from mutations in CD18 that lead to severe deficiency of the of Excellence (to J.J.-B., J.J., and R.I.), National Cancer Institute of Canada (to J.J. CD11/CD18 integrins and profound reduction in neutrophil re- and J.J.-B.), and Canadian Institutes of Health Research Grant MT-14151 (to M.C.); ␤ by Postdoctoral Fellowship Awards from the University of Toronto Arthritis Center cruitment during inflammation (10). The 2 integrins are critical of Excellence (to C.W.), Canadian Arthritis Network (to H.H.), Sick Children Re- for firm adhesion of rolling neutrophils on vascular surfaces and search Training Competition (to R.H.); and by a Heart and Stroke Foundation of ␣ ␤ transendothelial migration (11, 12). LFA-1 ( L 2, CD11a/CD18) Canada Research Traineeship (to J.C.). ␣ ␤ ␤ and Mac-1 ( M 2, CD11b/CD18) are the most abundant 2 inte- 2 Address correspondence and reprint requests to Dr. Jenny Jongstra-Bilen, Toronto Western Hospital, Main Pavilion, Room 13-312A, 399 Bathurst Street, Toronto, On- grins on neutrophils. Although these integrins share ICAM-1 as a tario M5T 2S8, Canada. E-mail address: [email protected] ligand, the regulation of their surface expression and their function

Copyright © 2002 by The American Association of Immunologists, Inc. 0022-1767/02/$02.00 416 LSP1 REGULATES NEUTROPHIL ADHESION AND CHEMOTAXIS during neutrophil migration appear distinct. Neutrophil activation F-actin bundles branching into hair-like cell surface projections results in significant up-regulation of surface Mac-1 levels, but not and impaired motility (41, 42). LFA-1 (13). Although the importance of both integrin receptors in In this paper, we evaluated the in vivo role for LSP1 in regu- firm adhesion of neutrophils to ICAM-1 or to endothelial surfaces lating neutrophil influx to acute inflammation in the joint and in- was demonstrated, the role of Mac-1 in recruitment of neutrophils vestigated the requirement for LSP1 in various processes which into inflammatory sites is not completely understood. Although a culminate in fMLP-induced neutrophil motility with a long-term number of studies showed inhibition of neutrophil migration by goal of elucidating the mechanism(s) by which LSP1 functions as Mac-1 blocking Abs, LFA-1 appeared to play a more important a negative regulator of neutrophil migration during inflammation. role than Mac-1 in recruitment to inflammatory sites (13Ð15). In- terestingly, increased accumulation of neutrophils was found at Materials and Methods inflammatory sites in Mac-1-deficient mice which correlated with Mice their increased survival (16). Thus, distinct or additional roles for Ϫ Ϫ 129/SvJ mice (wt and Lsp1 / ) were bred and housed in microisolators in Mac-1 have been demonstrated as compared with LFA-1 in in- the Toronto Western Research Institute Animal facility (Toronto, Ontario, flammatory response. Canada). Eight- to 12-wk-old female or male mice were used. All protocols The tight adhesion of neutrophils to extracellular matrix proteins were approved by the Animal Care Committee of the University Health is accompanied by polarization toward a chemotactic gradient that Network. mediates migration to the inflamed site. Polarization is achieved by Reagents and Abs the formations of lamellipodia (flat sheet-like structures) and filop-

Zymosan, fMLP, PMA, fibrinogen, fibronectin (human plasma), poly-L- Downloaded from odia (thin needle-like projections) at the leading edge and uropod lysine, BSA (IgG-free, low endotoxin), and HBSS with or without Ca2ϩ/ at the rear of the cell (17). Cycles of attachment to the extracellular Mg2ϩ were purchased from Sigma-Aldrich (St. Louis, MO); Immulon 2HB substrate at the front and detachment from the rear are mediated by microtiter plates, used in adhesion assays, were obtained from Dynex temporal and spatial changes in the localization and strength of Technologies (Chantilly, VA). Mouse ICAM-1 was purified from NS-1 B adhesive forces through integrin receptors (18, 19). Lamellipodia lymphoma cell membranes by affinity chromatography (43). Human chi- meric VCAM-1/Fc and ICAM-2/Fc were purchased from R&D Systems and filopodia are devoid of cytoplasm but contain mainly micro- (Minneapolis, MN). FITC-conjugated fMLP, Calcein-AM, and Alexa 3 filaments that are formed by local filamentous actin (F-actin) po- Fluor 488 Phalloidin were obtained from Molecular Probes (Eugene, OR). http://www.jimmunol.org/ lymerization. F-actin filaments are bundled in filopodia and cross- For flow cytometry, mAbs against Mac-1 (FITC-anti-CD11b, M1/70; linked in lamellipodia (20). Thus, proteins which promote F-actin Serotec, Oxford, U.K.), Gr-1 (PE-conjugated), corresponding isotype controls (Cedarlane Laboratories, Hornby, Ontario, Canada), or LFA-1 polymerization, or bundle or cross-link actin filaments, are key to (FD441.8; a kind gift from Dr. F. Takei, Terry Fox Laboratory, Vancouver, provide the rigidity required for these structures to extend, and ␣ ␣ British Columbia, Canada), 4 or 5 (PS/2 or BMA-5, respectively; a kind modulations in their expression have major impact on motile prop- gift from Dr. B. Chan, University of London, Ontario, Canada) in con- erties of cells (21Ð25). junction with biotinylated anti-rat Abs and streptavidin-PE (Cedarlane Laboratories) were used. F(abЈ) of goat anti-rabbit IgG (Fc-specific) and The mouse leukocyte-specific protein 1 (LSP1; formerly lym- 2 anti-FITC Abs were obtained from Caltag Laboratories (Burlingame, CA), phocyte-specific protein 1 but renamed to reflect its expression in

unconjugated blocking Abs against mouse Mac-1 (M1/70), LFA-1 (M17/ by guest on September 26, 2021 macrophages and neutrophils) (26, 27) is a 52-kDa intracellular 5.2), and CD18 (2E6) were purified from corresponding hybridoma cul- F-actin binding protein which accumulates on the cortical cy- tures (American Type Culture Collection, Manassas, VA), isotype control toskeleton (28Ð30). LSP1 mRNA and protein is expressed in ma- rat IgG2b (low endotoxin) was obtained from BD PharMingen (San Diego, CA). ture and immature B and T cells, macrophages, and neutrophils (31Ð33). Human and mouse LSP1 share the same expression pat- Zymosan-induced acute inflammation in mouse knee joints terns (26, 34Ð37) and the amino acid identity between the two Intra-articular injection of 180 ␮g zymosan was performed along the su- species is 85% for the basic C-terminal half containing multiple prapatellar ligament into the joint cavity of anesthetized mice (44). The F-actin binding domains and 53% for the acidic N-terminal half. contralateral knee was injected similarly with an equal volume of nonpy- LSP1 has two putative EF-hand motifs near the N terminus and rogenic saline. Both knee injections were completed by injection of the dye ϩ ␮ ␮ rLSP1 binds Ca2 . LSP1 is a substrate for mitogen-activated pro- fast green (1.5 l) separated from zymosan by a small air bubble (0.1 l) to assess the success of the injection. The intra-articular injections were tein kinase activated protein kinase 2 (37) and for protein kinase C performed using a microinjector equipped with an operating microscope (38). (Zeiss, Toronto, Ontario, Canada) attached to a MedVet anesthetic system. Ϫ Ϫ Adult Lsp1 / mice generated in our laboratory show no sig- After 24 h, swelling of the knees was determined by using a microcaliper nificant differences from wild-type (wt) mice in the development (Fisher Scientific, Whitby, Ontario, Canada), knee joints were dissected, of the lymphoid and myeloid lineage (27). They display a 3-fold fixed in 10% buffered formalin and decalcified in a 25% formic acid, 0.4 M sodium citrate solution. H&E staining was performed by Cell Path Tech- increase in resident peritoneal macrophages but normal propor- nologies (Toronto, Canada). Inflammation was evaluated by two indepen- tions of B and T cells within the lymphoid population. Thiogly- dent scientists in a blinded fashion by scoring the influx of inflammatory collate-induced influx of monocytes and neutrophils into the peri- cells into the subsynovium and joint cavity (synovial fluid). Ϫ/Ϫ toneum is increased in Lsp1 mice. In vitro transmigration Isolation of bone marrow (BM) neutrophils assays using elicited neutrophils demonstrated increased chemoat- Ϫ Ϫ Ϫ/Ϫ Mouse BM cells were flushed from the femur and tibia of wt and Lsp1 / tractant-induced motility of Lsp1 neutrophils. These results 2ϩ 2ϩ suggest that LSP1 is a negative regulator of neutrophil motility in male mice with Ca /Mg -free HBSS. Cells were washed and resus- pended in 45% Percoll (Amersham Pharmacia Biotech, Uppsala, Sweden) vivo and in vitro. Studies by Howard and colleagues (39, 40) sup- and neutrophils were purified from Percoll density gradients (45). RBC port this conclusion. LSP1 is over-expressed and a 89-kDa protein were lysed with E-lyse solution (Cardinal Associates, Santa Fe, NM). Cells ϩ ϩ is under-expressed in patients with neutrophil actin dysfunction were washed with ice-cold Ca2 /Mg2 -free HBSS (for Calcein-AM label- 2ϩ 2ϩ (NAD) disorder 47/89. Neutrophils from these patients and trans- ing) or HBSS with Ca /Mg and 0.1% BSA (HB, for all other studies). Flow cytometry revealed high expression of the murine granulocytic fectants with high expression of LSP1 display formations of large marker Gr-1 typically on 75Ð80% of the isolated cells. Chemotaxis and chemokinesis assays 3 Abbreviations used in this paper: F-actin, filamentous actin; LSP1, leukocyte-specific protein 1; wt, wild type; BM, bone marrow; RT, room temperature; RFI, relative Chemokinesis was performed by adding fMLP directly to BM neutrophils fluorescent intensity. and chemotaxis was performed by using a point source of fMLP derived The Journal of Immunology 417

from a micropipette. Micropipettes with tips 30Ð100 ␮m in diameter were immunofluorescence microscope to determine the proportion of cells with filled with 10Ϫ3 M fMLP in 1% low-melt agarose (FMC Bioproducts, head-to-tail morphology. The overall F-actin levels in adhered cells was de- Rockland, ME). Neutrophils (0.4 ϫ 106 in HB) were plated on fibrinogen termined from merged stacked confocal images of cells, acquired under non- or fibronectin-coated coverslips and were incubated for 10 min in a 37¡C saturating laser intensities, using NIH Image 1.62 software. Cell outlines were heated chamber to allow for adhesion. Nonadherent cells were removed by manually traced on images and pixel intensities in the Alexa Fluor 488 phal- several washes with HB. Time-lapse video microscopy was performed by loidin-stained areas were expressed as relative fluorescent intensity (RFI). capturing differential interference contrast images every 15 s under a Leica DM IRB microscope using MetaFluor software (Universal Imaging, West Statistical Analysis Chester, PA). Image acquisition was initiated 1 min before the placement Ϯ of the micropipette or addition of fMLP into the chamber to obtain resting All data are expressed as mean values 1 SEM. Unpaired Student’s t tests neutrophil images. The average speed was calculated from the total mi- or Welch’s t tests were performed using InStat software (GraphPad Soft- gration distance determined by MetaMorph software (Universal Imaging) ware, San Diego). and the elapsed time. In chemokinesis experiments the response criterion was cell movement at least over a distance corresponding to one cell body. Results Increased influx of Lsp1Ϫ/Ϫ neutrophils into zymosan-injected Coating with extracellular matrix proteins knee joints ␮ ␮ Fibrinogen (1 mg/ml), fibronectin (20 g/ml), or ICAM-1 (1 g/ml) in We previously reported increased accumulation of Lsp1Ϫ/Ϫ neu- PBS were added to microtiter wells or to acid-washed coverslips and incubated for2hat37¡C. Unbound proteins were removed, plates coated trophils as compared with wt during thioglycollate-induced peri- Ϫ Ϫ with ICAM-1 were blocked with 2% BSA (IgG free, low endotoxin; Sig- tonitis. Lsp1 / mice exhibit increased resident macrophage levels ma-Aldrich) in PBS for 30 min at room temperature (RT). For coating with in the peritoneum but not in the lung, indicating that LSP1 defi- Ј VCAM-1/Fc and ICAM-2/Fc, goat anti-human IgG (Fc-specific F(ab )2, ciency does not affect all resident macrophage populations. Be- Downloaded from 100 ␮g/ml) were added to each well for 60 min at RT, followed by blocking with 2% BSA before the addition of 20 ␮g/ml VCAM-1/Fc or ICAM- cause, in our former study, we did not eliminate the possibility that 2/Fc overnight at 4¡C. All coating steps were followed by three washes the increased level of resident macrophages contribute to the in- with PBS. Poly-L-lysine was coated (10 ␮g/ml) as described by the creased influx of neutrophils to the peritoneum, we chose to induce supplier. acute inflammation in another tissue where resident cell levels For adhesion assays under shear flow, 20 ␮l goat anti-rabbit IgG (Fc- Ј ␮ ␮ would be unaffected by LSP1-deficiency. specific F(ab )2,50 g/ml) and fibrinogen (100 g/ml) were adsorbed onto the center of a 35-mm dish for 60 min in a humidified atmosphere at RT; The thin cell layer of resident synoviocytes along the synovial http://www.jimmunol.org/ 10 ␮l rabbit anti-FITC Abs (20 ␮g/ml) were subsequently added for 60 membrane are composed of macrophages and fibroblasts (48, 49). min. After blocking with 0.5% BSA overnight at 4¡C, 10 ␮l fMLP-FITC To verify whether the unstimulated synovial lining displays a sim- ␮ Ϫ Ϫ (20 g/ml) were added at RT. ilar number of resident cells in the knee joints from wt vs Lsp1 / Neutrophil adhesion assays mice, we counted 200Ð500 total cells/mouse synovial lining and determined the average cell number per unit length. We found no For static adhesion assays, Calcein-AM-labeled (27) BM neutrophils in HB (106/100 ␮l) were plated into precoated microtiter wells (in triplicate) and differences between five mice from each genotype (not shown), incubated at 37¡C with 7% CO for 60 min following the addition of 100 confirming that the resident cells of unstimulated knee joints are 2 Ϫ Ϫ ␮l agonists. Plates were centrifuged upside-down as described (45) and a equivalent. To determine the accumulation rate of Lsp1 / vs wt CytoFluor 2300 fluorometric plate reader (Millipore, Nepean, Ontario, neutrophils in this system, we injected zymosan into the knee by guest on September 26, 2021 Canada) was used to calculate the percentage of adhered cells. A standard joints. Within 24 h, zymosan induced a typical acute inflammation curve was constructed with increasing amounts of labeled cells added into a separate plate. characterized by swelling of the joints and the influx of inflam- For assays under shear flow, BM neutrophils (106/ml) settled for 2 min matory cells (50, 51). Swelling which is a manifestation of edema in a parallel flow chamber as previously described (46). Shear stress was due to increased vascular permeability was not affected by LSP1- 2 applied at 4 dynes/cm for 30 s. Cells were observed under an inverted deficiency (Table I). Neutrophils, identified by their typical mul- phase contrast microscope, videotaped, and analyzed (46). tilobed morphology by H&E staining, were the predominant cells fMLP-induced Mac-1 up-regulation and F-actin polymerization accumulating in the joint at this early time point of inflammation. Fig. 1 illustrates an example for each genotype analyzed and Table For Mac-1 up-regulation, BM neutrophils in HB (0.5 ϫ 106 cells) were prewarmed for 4 min at 37¡C and stimulated with different concentrations I summarizes the average scores for inflammation. Fig. 1, A and B, of fMLP for 10 min, and stained with FITC-conjugated anti-Mac-1, PE- shows control joints injected with saline, which exhibit similar ϩ conjugated anti-Gr-1, or isotype controls (47). Mac-1 live cells were an- morphology in both genotypes. Zymosan-induced accumulation of high alyzed within the gated Gr-1 neutrophils by flow cytometry using FACS- neutrophils was significantly increased in the synovial cavity of Calibur and CellQuest Software (BD Biosciences, San Jose, CA). Ϫ/Ϫ For F-actin measures, BM neutrophils in HB (2.5 ϫ 106/ml) were pre- Lsp1 mice (Fig. 1, D and F) as compared with wt (Fig. 1, C and warmed at 25¡C and incubated with or without 10Ϫ6 M fMLP for different times, fixed with 3.7% paraformaldehyde in PBS for 20 min at RT, washed, and permeabilized with 0.1% Triton X-100 in PBS for 5 min. After washing, cells were stained with Alexa Fluor 488 phalloidin for 30 min and Table I. Histological evaluation of acute inflammation in zymosan- a analyzed by flow cytometry as above. injected mouse knees

Fluorescence microscopy Joint Inflammationc b BM neutrophils (0.4 ϫ 106 cells) in HB were preincubated on fibrinogen- or Genotype Swelling (L:R) Subsynovium Cavity fibronectin-coated coverslips for 10 min before incubation with or without ϩ/ϩ Ϯ Ϯ Ϯ Ϫ5 Lsp1 1.17 0.03 (8) 1.50 0.20 (14) 2.07 0.27 (14) ء(M fMLP for 30 min at 37¡C with 5% CO2. After washing briefly with Lsp1Ϫ/Ϫ 1.19 Ϯ 0.03 (7) 1.92 Ϯ 0.14 (13) 3.08 Ϯ 0.26 (13 10 prewarmed HB, the adhered cells were fixed with 4% paraformaldehyde solution (Canemco, Quebec, Canada) for 30 min at RT. Cells were stained for a Values are mean Ϯ 1 SEM. Numbers in parentheses correspond to the number F-actin as described above. Coverslips were mounted with DAKO fluorescent of mice analyzed. mounting medium (Carpinteria, CA). Slides were examined with a Zeiss LSM b Swelling of the joints was measured after removal of the surrounding muscles 510 scanning laser confocal mounted on a Zeiss Axiovert 100 microscope with and was expressed as a ratio of zymosan over saline-injected knee joint (L:R, left to a 100ϫ oil immersion objective. Alexa Fluor 488 phalloidin-stained neutro- right). c The extent of inflammation in the femorotibial joints was scored as the exudation phils were visualized with a fluorescein filter set (515 nm dichroic mirror, 530 of neutrophils into the joint cavity and infiltration into the subsynovium using a linear nm long-pass barrier filter) using Argon laser (488 nm) illumination. Digital scale 0Ð4 according to cell number and area occupied; 0 ϭ normal histology, 4 ϭ images were prepared using Adobe Photoshop 6.0 software (Adobe Systems, extensive coverage with densely packed neutrophils. .p Ͻ 0.02 ,ء Mountain View, CA). Two hundred cells were counted under a Leica DMRB 418 LSP1 REGULATES NEUTROPHIL ADHESION AND CHEMOTAXIS

FIGURE 1. Increased influx of LSP1-deficient neutrophils to the zymosan-injected knee joints. Histology sections of saline- (A and B) or zymosan- (C and F) injected femorotibial joints from wt (ϩ/ϩ, A, C, and E) and Lsp1Ϫ/Ϫ (Ϫ/Ϫ, B, D, and F) mice were stained with H&E. The regions indicated with arrowheads in C and D are illustrated in higher magnifica- tion in E and F, respectively. Note that control joints (A and B) display similar morphology of subsynovium (s) and the synovial lining (ar- Downloaded from rows) in both genotypes. The exudate neutrophils (e) are denser and occupy a larger area in Lsp1Ϫ/Ϫ mice (D and F) than in wt mice (C and E). jc, joint cavity; s, subsynovium; i, in- filtrate; e, exudate; c, cartilage. http://www.jimmunol.org/ by guest on September 26, 2021

E) (Table I). Although increased levels of infiltration into the sub- is related to its role in adhesion and whether such a role would be synovium were also found in Lsp1Ϫ/Ϫ mice, the differences were specific for a particular class of integrins. not statistically significant. These results demonstrate that LSP1 ␤ negatively regulates neutrophil recruitment into the joint during LSP1 is a negative regulator of 2 integrin-mediated adhesion ␤ ␤ acute inflammation independently from changes in resident cell Neutrophils require an optimal level of adhesion via 1 and 2 numbers. integrins on activated endothelial surfaces to mediate spreading ␤ ␤ Ϫ/Ϫ and chemotaxis (19, 52). The 1 or 2 family of integrins on neu- Increased chemotaxis of Lsp1 neutrophils on fibrinogen trophils mediate binding to extracellular matrix proteins or their ␣ ␤ ␣ ␤ We previously demonstrated that in the absence of LSP1, there is counterreceptors such as fibronectin ( 4 1, 5 1), VCAM-1 ␣ ␤ ␣ ␤ ␣ ␤ ␣ ␤ ␣ ␤ ␣ ␤ increased in vitro transmigration of thioglycollate-elicited neutro- ( 4 1, 9 1)orfibrinogen ( M 2, X 2), ICAM-1 ( L 2, M 2) ␣ ␤ phils in response to chemoattractants, fMLP, or KC. These find- or ICAM-2 ( L 2). ings raised the question of whether the increased chemotaxis of We performed in vitro adhesion assays with BM neutrophils Ϫ/Ϫ ␤ Lsp1 neutrophils depends on their activation state since we using various substrates which serve as ligands for 2 (Fig. 2)or ␤ used elicited neutrophils. Furthermore, the transmigration assays 1 (Fig. 3) integrins. Fig. 2, A, B, and D, illustrates fMLP and do not distinguish between increased speed per cell or an increased PMA-induced tight adhesion to ICAM-1, fibrinogen, or ICAM-2, ␤ number of cells responding. To better elucidate the role of LSP1 in respectively, which are mediated by 2 integrins. Adhesion of un- neutrophil motility, we performed time-lapse in vitro fMLP-in- stimulated wt and Lsp1Ϫ/Ϫ neutrophils is similar with each sub- duced chemokinesis and chemotaxis assays on fibrinogen or fi- strate tested. Upon fMLP stimulation, we observed a significant bronectin using BM neutrophils purified from Lsp1Ϫ/Ϫ and wt mice. Chemokinesis assays, in which a uniform concentration of a fMLP (10Ϫ4 or 10Ϫ5 M) induced undirectional movement, re- Table II. Chemotaxis of BM neutrophils vealed similar number of cells responding (75Ð80 of 100 cells/ Speed (micrometers genotype, not shown) on either substrate. Chemotaxis assays were Genotype Substratum per minute) n used to determine the average migration rate toward a gradient of ϩ ϩ fMLP per cell. Table II shows that in the absence of LSP1, neu- Lsp1 / Fibrinogen 5.16 Ϯ 0.72 13 27 ءLsp1Ϫ/Ϫ 7.02 Ϯ 0.51 trophils gained a significant increase in their average speed during ϩ ϩ Lsp1 / Fibronectin 8.85 Ϯ 0.61 16 chemotaxis on fibrinogen-coated coverslips. Interestingly, the av- Lsp1Ϫ/Ϫ 8.58 Ϯ 0.54 21 erage speed of wt and Lsp1Ϫ/Ϫ neutrophils was similar on fi- a A point source of fMLP was used for chemotaxis as described in Materials and bronectin-coated coverslips. These data raised the question of Methods. Values are mean Ϯ 1 SEM; n, the number of cells tracked. .p Ͻ 0.05 ,ء whether the negative regulatory role of LSP1 in neutrophil motility The Journal of Immunology 419

FIGURE 3. LSP1 is not required for adhesion to VCAM-1/Fc or ﬁ- bronectin. Assays (four to six per genotype) of tight adhesion were performed for 60 min with wt ()orLsp1Ϫ/Ϫ (f) neutrophils to VCAM-1/Fc (A)orﬁbronectin (B) in response to 10Ϫ4 M fMLP, 25 ng/ml PMA, or the

carrier, 0.1% DMSO. Downloaded from

sion were also observed with ICAM-1 (Fig. 2A) while adhesion of Lsp1Ϫ/Ϫ neutrophils to ﬁbrinogen was signiﬁcantly increased as

compared with wt in the presence of PMA. The lack of difference http://www.jimmunol.org/ in PMA-induced adhesion to ICAM-1 of wt and Lsp1Ϫ/Ϫ neutro- FIGURE 2. LSP1 is a negative regulator of neutrophil adhesion to phils may be due to their similar binding via LFA-1 as was shown ICAM-1 and fibrinogen, but not to ICAM-2. Assays (four to eight per for ICAM-2 (Fig. 2D). genotype) of tight adhesion (A, B, and D) or adhesion under shear force, 4 ␤ dyn/cm2 (C), were performed for 60 min (A, B, and D) with wt ()or To investigate a role for LSP1 in 1 integrin-mediated firm ad- Lsp1Ϫ/Ϫ (f) neutrophils to ICAM-1 (A), fibrinogen (B and C), or ICAM- hesion, we performed adhesion assays with extracellular matrix 2/Fc (D) in response to 10Ϫ4 M fMLP, 25 ng/ml PMA, or the carrier, 0.1% proteins VCAM-1 or fibronectin. Fig. 3, A and B, shows that un- -stimulated neutrophils of both genotypes bind similarly at back ,ء .DMSO (A, B, and D), and as described in Materials and Methods for C -p Ͻ 0.0001. ground levels to VCAM-1 or fibronectin. Interestingly, the in ,ءء ;p Ͻ 0.04 crease in adhesion by fMLP or PMA to these substrates is by guest on September 26, 2021 indistinguishable between wt and Lsp1Ϫ/Ϫ neutrophils. Neutrophil increase in tight adhesion of wt neutrophils to fibrinogen, but not adhesion was similar between the two genotypes when lower (5 to ICAM-1, while fMLP-stimulated Lsp1Ϫ/Ϫ neutrophils adhered ␮g/ml) or higher (100 ␮g/ml) concentrations of fibronectin were more tightly (1.5- to 2-fold) to both extracellular proteins (Fig. 2, used to coat the wells. The fMLP-induced adhesion on 100 ␮g/ml A and B). We obtained a similar increase in fMLP-induced adhe- fibronectin was: wt 19.93% Ϯ 1.60; Lsp1Ϫ/Ϫ 18.67% Ϯ 2.86 (n ϭ sion of Lsp1Ϫ/Ϫ neutrophils when cells were layered on wells 3), indicating that the 20 ␮g/ml fibronectin used throughout this coated with five times less fibrinogen (0.2 mg/ml): wt 8.90 Ϯ paper is saturating. Ϫ/Ϫ Ϯ ϭ Ͻ ␤ 0.83%; Lsp1 12.56 1.05% (n 5, p 0.03), indicating that Our attempts to confirm the identity of the 2 integrin for which saturating amounts of fibrinogen have been used throughout this LSP1 is a negative regulator of adhesion by use of blocking Abs study. Because under physiological conditions neutrophils sense against murine integrin receptors have not been successful. We and respond to stimulus under shear flow, we repeated the fMLP- have used anti-CD11b (M1/70), anti-LFA-1 (M17/5.2, FD441.8), induced adhesion experiments to fibrinogen under conditions and anti-CD18 (2E6) Abs which we purified from hybridoma su- where neutrophils were subjected to a shear force of 4 dyn/cm2 pernatants to block adhesion (13). However, these and the control (Fig. 2C). Under these conditions, 58.2 Ϯ 1.6% of wt neutrophils IgG2b Abs (low endotoxin) stimulated, rather than blocked, adhe- Ϯ Ϫ/Ϫ Ј and 86.9 1.4% of Lsp1 neutrophils remained adhered to sion of BM neutrophils from both genotypes even when F(ab )2 fibrinogen. A shear force of 10 dyn/cm2 yielded similar results (not were used. As a test for the involvement of integrins in the in- shown). We have thus confirmed by two different assays that LSP1 creased adhesion of LSP1-deficient neutrophils, we repeated the deficiency results in enhanced fMLP-induced adhesion to fibrino- adhesion assays in the absence of Ca2ϩ and Mg2ϩ which are re- gen. These results, together with significantly increased adhesion quired for integrin function (7, 53, 54). Table III illustrates that ␤ to ICAM-1, suggest strongly that LSP1 negatively regulates 2 unstimulated, fMLP-, or PMA-induced neutrophils of either geno- integrin-mediated adhesion. Although fibrinogen is a ligand for type were unable to bind to ICAM-1 in the absence of the divalent Mac-1, the increased binding of Lsp1Ϫ/Ϫ neutrophils to ICAM-1 ions and consequently, the increase in binding of Lsp1Ϫ/Ϫ neutro- may be the result of binding to both Mac-1 and LFA-1. To verify phils to ICAM-1 as compared with their wt counterparts was abol- the capacity of neutrophil adhesion to LFA-1 in the presence or ished. Similarly, PMA-induced adhesion to ICAM-2 was also absence of LSP1, we conducted adhesion assays to the specific abolished in the absence of Ca2ϩ and Mg2ϩ (not shown). In con- LFA-1 ligand ICAM-2. Although, fMLP did not induce adhesion trast, adhesion of unstimulated or stimulated neutrophils of either to ICAM-2 in wt or in LSP1-deficient neutrophils (Fig. 2D), PMA genotype to poly-L-lysine, which does not involve integrins, was induced similar adhesion to ICAM-2 in neutrophils from both ge- not significantly affected by the absence of Ca2ϩ and Mg2ϩ.Itis notypes showing that murine neutrophils are capable of adhering notable that LSP1 deficiency did not affect adhesion to poly-L- to this extracellular protein. Similar levels of PMA-induced adhe- lysine under any of the conditions tested. 420 LSP1 REGULATES NEUTROPHIL ADHESION AND CHEMOTAXIS

Table III. Ca2ϩ/Mg2ϩ-dependent adhesion to ICAM-1a

% Adhesion with Ca2ϩ,Mg2ϩ % Adhesion Ca2ϩ,Mg2ϩ-Free

cnt fMLP PMA cnt fMLP PMA

ICAM-1 wt 5.1 Ϯ 1.1 (7) 6.7 Ϯ 0.5 (7) 20.7 Ϯ 1.1 (7) 1.5 Ϯ 1.0 (2) 2.0 Ϯ 1.7 (2)† 1.9 Ϯ 1.0 (2)† †(Ϯ 2.1 (7) 1.0 Ϯ 0.5 (2) 0.7 Ϯ 0.5 (2)† 1.2 Ϯ 0.6 (2 22.7 ء(KO 5.1 Ϯ 1.0 (7) 15.5 Ϯ 2.5 (7 Plysine wt 27.4 Ϯ 4.3 (3) 16.8 Ϯ 3.7 (3) 13.6 Ϯ 2.5 (3) 22.5 Ϯ 1.6 (2) 13.9 Ϯ 5.5 (2) 21.8 Ϯ 2.5 (2) KO 25.4 Ϯ 5.6 (3) 19.6 Ϯ 2.0 (3) 17.7 Ϯ 2.2 (3) 20.9 Ϯ 3.9 (2) 10.7 Ϯ 2.0 (2) 24.4 Ϯ 6.8 (2)

a Tight adhesion of BM neutrophils to ICAM-1 and to poly-L-lysine (Plysine) in response to 10Ϫ4 M fMLP, 25 ng/ml PMA, or the carrier 0.1% DMSO (cnt) was performed in the presence or absence of Ca2ϩ/Mg2ϩ for 60 min. For adhesion in the absence of Ca2ϩ/Mg2ϩ, 0.1 mM EDTA was added to Ca2ϩ/Mg2ϩ-free HBSS. Values are mean Ϯ 1 SEM. Numbers in parentheses correspond to the number of experiments. .p Ͻ 0.01 when compared to wt ,ء †, p Ͻ 0.02 when compared to the corresponding values in the presence of Ca2ϩ/Mg2ϩ.

Taken together, these results demonstrate that LSP1 is not a neutrophils from both genotypes (Fig. 4B). Neutrophils stimulated ubiquitous negative regulator of adhesion but it is involved in in- with 10Ϫ6 M fMLP displayed a maximal F-actin polymerization at Downloaded from tegrin-mediated adhesion. In addition, our results also demonstrate 30 s which declined and reached a plateau level after 1 min. LSP1 that the requirement for LSP1 depends on the class of integrin deficiency did not substantially change this kinetics except at 2 ␤ being activated. In this study, we have identified 2 integrins, and min where a small (14.3%) but statistically significant decrease in among them most likely Mac-1, rather than LFA-1, as one of the F-actin content in Lsp1Ϫ/Ϫ neutrophils was observed. No differ- integrins being regulated by LSP1. ences in F-actin content were observed with 10Ϫ5 M fMLP in any of the time points tested (not shown). Taken together, these results LSP1 is not required for fMLP-induced Mac-1 up-regulation http://www.jimmunol.org/ and F-actin polymerization demonstrate a lack of LSP1 involvement in F-actin polymerization in fMLP-induced neutrophils. Neutrophils up-regulate their surface Mac-1 levels within minutes of agonist stimulation (55, 56). We asked whether the negative regulatory role of LSP1 in Mac-1-mediated adhesion could be at the level of Mac-1 expression in fMLP-stimulated neutrophils. Fig. 4A shows a robust increase in surface Mac-1 levels (3- to 4-fold) 10 min after fMLP stimulation which was similar for wt and Ϫ/Ϫ

Lsp1 neutrophils. This demonstrates that LSP1 is not involved by guest on September 26, 2021 in the up-regulation of Mac-1 surface expression in stimulated ␣ ␤ neutrophils. Surface expression of other integrins, LFA-1, 4 1 or ␣ ␤ 5 1, which do not signiﬁcantly change after fMLP stimulation, were similar in neutrophils from both genotypes (not shown). We had previously shown that rLSP1 does not modify the kinetics of F-actin polymerization in vitro (29). We investigated whether in BM neutrophils LSP1 plays a role in fMLP-induced F-actin polymerization. F-actin levels are similar in unstimulated

FIGURE 5. LSP1 negatively regulates neutrophil polarization on fibrinogen. Confocal images of F-actin stained wt (A and B) and Lsp1Ϫ/Ϫ (C and FIGURE 4. LSP1 is not required for fMLP-induced Mac-1 up-regula- D) neutrophils, stimulated with 10Ϫ5 M fMLP for 30 min on fibrinogen- tion or F-actin polymerization. A, wt () and Lsp1Ϫ/Ϫ (f) neutrophils coated coverslips. A1 and C1 correspond to the optical sections at the most were incubated without (cnt) or with carrier (0.1% DMSO), or with indi- basal region of the cell adjacent to the coverslips. Each subsequent frame cated concentrations of fMLP for 10 min and surface Mac-1 levels were (2Ð5) represents a 0.4-␮m increment from the level of the previous section. determined by flow cytometry in two experiments. B, wt (-▫-) or Lsp1Ϫ/Ϫ B and D are merged images from other fields of view. Insets are merged (--) neutrophils were incubated with 10Ϫ6 M fMLP at 25¡C for indicated images from additional cells. White arrows indicate lamellipodia, open times and F-actin levels were measured by flow cytometry in three arrowheads indicate trailing filopodia. Note the more extensive F-actin experiments. stained polarizations in Lsp1Ϫ/Ϫneutrophils. (Bars ϭ 10 ␮m). The Journal of Immunology 421

LSP1-deficiency leads to increased polarization of neutrophils proximity of lamellipodia which often had more extensive trailing on fibrinogen filopodia formations than in wt (Fig. 5D, arrowheads). Quantita- tively, on fibrinogen, there was a 2-fold increase (2.06 Ϯ 0.19, n ϭ Because LSP1 exhibits F-actin bundling capacity (29, 41), it was Ϫ/Ϫ logical to determine whether the specific negative regulatory role 3) in the proportion of Lsp1 neutrophils which exhibit highly for LSP1 in fMLP-induced neutrophil adhesion to fibrinogen, but polarized F-actin with head-to-tail morphology as compared with wt. On fibronectin, the extent of head-to-tail polarization of F-actin not to fibronectin, would be accompanied by similar differences in Ϫ/Ϫ F-actin polarization on these substrates. Adhesion was induced by in wt neutrophils was similar to that of Lsp1 neutrophils (Fig. fMLP on fibrinogen- or fibronectin-coated coverslips and neutro- 6, A and B, and inset B). Lamellipodia could be observed through- phils were stained for F-actin. When analyzed by confocal micros- out the different optical sections of both genotypes (Fig. 6, A or B, Ϫ/Ϫ Ϯ copy, a small number of unstimulated neutrophils with typical un- 1Ð6). The ratio of polarized wt vs Lsp1 neutrophils was 1.11 ϭ polarized spherical appearance were found on coverslips coated 0.16, n 3. with either substrate (see insets in Fig. 6, A1 and B1). With fMLP, RFI of F-actin is indistinguishable between fMLP-treated neu- Ϯ both wt and Lsp1Ϫ/Ϫ neutrophils underwent polarization on either trophils from the two genotypes plated on fibrinogen (wt, 40.2 Ϫ/Ϫ Ϯ Ϯ Ϫ/Ϫ substrate (Figs. 5 and 6). On fibrinogen, these changes were much 3.5; Lsp1 , 38.7 3.1) or fibronectin (wt, 30.7 5.5; Lsp1 , Ϯ more pronounced in Lsp1Ϫ/Ϫ neutrophils than in wt, while they 26.1 6.7). Thus, consistent with the lack of LSP1 involvement were similar between the two genotypes on fibronectin. As shown in F-actin polymerization in fMLP-induced neutrophils in suspen- in Fig. 5, A1, A2, and B, wt neutrophils on fibrinogen displayed sion (Fig. 4), these data show that LSP1-deficiency does not affect polarized F-actin patches at the anterior end (lamellipodia, white the overall F-actin levels stimulated by fMLP even when neutro- Downloaded from arrows) which were lost in regions further away from the substra- phils exhibit increased adhesion and F-actin polarization on tum (Fig. 5A, 3Ð5). Polarized wt neutrophils displayed one or two fibrinogen. trailing filopodia at the posterior end (Fig. 5A, open arrowheads; Fig. 5B, inset). Lsp1Ϫ/Ϫ neutrophils on fibrinogen (Fig. 5C) dis- Discussion played larger areas of intense F-actin caps as compared with wt Although it has previously been reported that resident macro- (white arrows). These caps were thicker, as they could be seen in phages disappear from the inflamed site within the first hour of optical sections away from the substratum (see Fig. 5C, 1Ð5). thioglycollate injection to the peritoneum, resident macrophages http://www.jimmunol.org/ Some neutrophils from both genotypes had more than one lamel- do participate in the production of proinflammatory agonists (such ␣ lipodia (insets), but again the Lsp1Ϫ/Ϫ neutrophils displayed more as TNF- , GM-CSF, IL-8, and other chemokines) that activate polarized structures. We also found a significant number of neutrophils (57, 58). We reported previously that neutrophil influx Lsp1Ϫ/Ϫ neutrophils with bright actin-stained focal contacts at the into the peritoneum is increased in response to thioglycollate in Lsp1Ϫ/Ϫ mice which harbor increased levels of resident peritoneal macrophages. Therefore, it was important to study the role of LSP1 in neutrophil migration in vivo, independently from changes in resident cell levels to distinguish between intrinsic vs environ- mental factors influencing the migration of activated neutrophils. by guest on September 26, 2021 In this study, we found that in unstimulated knee joints the number of the synoviocytes was indistinguishable between the two genotypes while the accumulation in the synovial fluid of Lsp1Ϫ/Ϫ neutrophils was significantly increased in response to zymosan. These findings demonstrate that the increased neutrophil recruitment in Lsp1Ϫ/Ϫ mice is independent from the resident cell numbers and is likely due to the changes in the chemotactic characteristics of neutrophils. Our in vitro studies using single cell analysis by time-lapse video microscopy of BM neutrophils further confirmed the notion that LSP1 deficiency affects the locomotive machinery of neutrophils because we showed that the average migration rate per cell toward a gradient of fMLP is greater in the absence of LSP1 and this does not require previous activation of neutrophils. Interest- ingly, we also found that this was dependent on the substrate on which the neutrophils migrate. Accelerated migration of LSP1- deficient neutrophils occurred on fibrinogen, but not fibronectin, raising the possibility that the role of LSP1 on neutrophil migration stems from a specific role for this protein in adhesion via certain integrins. In vitro adhesion assays using different extracellular substrate proteins indeed showed that fMLP-stimulated Lsp1Ϫ/Ϫ neu- ␤ trophils exhibit increased adhesion to the 2 integrin ligands, fibrinogen (Mac-1), or to ICAM-1 (LFA-1 and Mac-1) but not to FIGURE 6. LSP1 does not modulate the polarization of neutrophils on ICAM-2 (LFA-1). Adhesion to ␤ integrin ligands VCAM-1 or fibronectin. Confocal images of F-actin-stained wt (A) and Lsp1Ϫ/Ϫ (B) 1 Ϫ fibronectin were similar. These results suggest strongly for the first neutrophils, stimulated with 10 5 M fMLP for 30 min on fibronectin- coated coverslips. Panels 1Ð6 illustrate similar image acquisition as de- time that the negative regulatory role of LSP1 in cell migration is scribed in Fig. 5. Insets in A1 and B1 illustrate typical merged images of dependent on its role in adhesion via specific integrins. Although unstimulated neutrophils and the inset in B6 illustrates a merged image of adhesion to fibrinogen may not exclusively occur via Mac-1, the Ϫ/Ϫ another polarized Lsp1Ϫ/Ϫ neutrophil. Note a similar extent of F-actin po- fact that increased adhesion of Lsp1 neutrophils occurs both to larization in neutrophils of both genotypes. (Bars ϭ 10 ␮m). ICAM-1 and fibrinogen and not to ICAM-2, fibronectin, and 422 LSP1 REGULATES NEUTROPHIL ADHESION AND CHEMOTAXIS

VCAM-1 strongly suggests that at least one integrin modulated by that integrins need to be relieved from the cytoskeletal constraints ␤ LSP1 is Mac-1, while LFA-1 and 1 integrins are not negatively to move along the plane of the membrane for clustering (70Ð72). regulated by this protein. The tight adhesion of wt neutrophils to Thus, lack of LSP1 and the resulting flexibility of the actin fila- fibronectin or VCAM-1 was stronger than to fibrinogen or ICAM-1 ments may promote the movement of Mac-1 integrins within the and similarly the extent of polarization and average migration rate of different polarized compartments of adhered neutrophils or along wt neutrophils were higher on fibronectin than fibrinogen. Thus, at the plane of the plasma membrane to induce high avidity binding ␤ least in the case of 1 integrins, the strength of adhesion may com- to the ligand. pensate for the negative regulatory role of LSP1. LSP1 is overexpressed in neutrophils obtained from patients During chemotaxis, neutrophils undergo transient shape with NAD 47/89 who suffer from recurrent infections (39, 40). In changes by sending forward lamellipodia in the direction of mi- U937 or melanoma cell transfectants, overexpression of LSP1 in- gration. These attach to the extracellular matrix and allow the cell hibits cell migration (42). These findings are in agreement with our body to move toward this new attachment. Cells also release at- results that LSP1 is a negative regulator of neutrophil chemotaxis. tachments of their posterior ends (6, 59). In this study, we observed LSP1 overexpression in neutrophils may lead to an abnormally striking differences in fMLP-induced F-actin polarization between rigid cytoskeleton that impedes dynamic actin rearrangement dur- wt and LSP1-deficient neutrophils when they were induced to ad- ing adhesion and migration. Indeed, aberrant hair-like protrusions here on fibrinogen, while adhesion to fibronectin did not result in with extensive F-actin bundles were found in NAD 47/89 neutro- such differences. It is notable that in suspended neutrophils, in the phils, which may account for their inability to spread in vitro and absence of any attachment, fMLP-stimulated F-actin polarization, emigrate into sites of infection in vivo (39). as measured by percent-polarized cells or the area occupied by In conclusion, we showed that LSP1 negatively regulates neu- Downloaded from F-actin caps was similar between wt and Lsp1Ϫ/Ϫ neutrophils (not trophil chemotaxis and polarization through its regulatory function shown). These findings support the notion that LSP1 is not in- on adhesion via specific integrins such as Mac-1. We suggest that volved directly in fMLP-induced polarization of F-actin but rather these findings account for our in vivo observations that LSP1 mod- emphasized the function of LSP1 in regulating the gradient of ulates neutrophil accumulation during acute inflammation. Future adhesive strength which develops during chemotaxis. studies will focus on the mechanisms by which LSP1 regulates

Hannigan et al. (60) recently reported that LSP1-deficient blood Mac-1 function in adhesion. It also remains to be investigated http://www.jimmunol.org/ neutrophils exhibit impaired chemotaxis toward a gradient of the whether this negative regulatory role of LSP1 affects other neu- chemokine KC on glass coverslips, which they attributed to undi- trophil functions involving Mac-1, i.e., phagocytosis and cytotox- rectional protrusions produced in the absence of LSP1. Our data icity (72Ð74). show the importance of the substrate used for understanding the role of LSP1 in chemotaxis. They also reported that LSP1 is not Acknowledgments required for adhesion to fibrinogen in response to KC or PMA. We are indebted to Dr. Sergio Grinstein for his generous support and ad- Besides the differences in the experimental conditions, these authors vice. We thank Drs. Gregory Downey, Hedy Ginzberg, and Amos Cohen used total bone marrow cells as opposed to purified neutrophils, for their advice. We thank Mian Chen and Haiyan Xiao for their help with which may have masked the increased adhesion of neutrophils. the mouse colony. by guest on September 26, 2021 A specific role of LSP1 in the functioning of Mac-1 rather than LFA-1 can be considered if LSP1 interacts with Mac-1 directly or References through an adapter protein and provides a linkage to F-actin in 1. Kuijpers, T. W., R. S. Weening, and D. Roos. 1999. Clinical and laboratory work-up of patients with neutrophil shortage or dysfunction. J. Immunol. Meth- unstimulated neutrophils. Given that LSP1 is an abundant protein ods 232:211. with F-actin bundling capacity, we suggest that LSP1 deficiency 2. Pillinger, M. H., and S. B. Abramson. 1995. The neutrophil in rheumatoid ar- results in a more flexible actin-cytoskeleton architecture which un- thritis. Rheum. Dis. Clin. N. Am. 21:691. 3. Dallegri, F., and L. Ottonello. 1997. Tissue injury in neutrophilic inflammation. dergoes more efficient rearrangements during adhesion. We envis- Inflamm. Res. 46:382. age that in wt cells, LSP1 undergoes modification by phosphory- 4. Downey, G. P. 1994. Mechanisms of leukocyte motility and chemotaxis. Curr. lation and dissociates from F-actin to allow for cytoskeletal Opin. Immunol. 6:113. 5. Granger, D. N., and P. Kubes. 1994. The microcirculation and inflammation: rearrangement as has been described for the macrophage-enriched modulation of leukocyte-endothelial cell adhesion. J. Leukocyte Biol. 55:662. myristoylated alanine-rich C kinase substrate or L-plastin, two 6. Gumbiner, B. M. 1996. Cell adhesion: the molecular basis of tissue architecture other F-actin bundling proteins (61, 62). LSP1 could be phosphor- and morphogenesis. Cell 84:345. 7. Hynes, R. O. 1992. Integrins: versatility, modulation, and signaling in cell ad- ylated by mitogen-activated protein kinase activated protein kinase hesion. Cell 69:11. ␤ 2 of the p38 pathway (37) which has been implicated in fMLP- 8. Werr, J., X. Xie, P. Hedqvist, E. Ruoslahti, and L. Lindbom. 1998. 1 integrins induced chemotaxis and adhesion of neutrophils (63). The reg- are critically involved in neutrophil locomotion in extravascular tissue in vivo. J. Exp. Med. 187:2091. ␣ ␤ ulatory function of LSP1 in adhesion may also involve its recently 9. Shang, T., T. Yednock, and A. C. Issekutz. 1999. 9 1 integrin is expressed on demonstrated role in modulating extracellular signal-regulated kinase human neutrophils and contributes to neutrophil migration through human lung and synovial fibroblast barriers. J. Leukocyte Biol. 66:809. activity (64), because fMLP-induced extracellular signal-regulated 10. Fischer, A., B. Lisowska-Grospierre, D. C. Anderson, and T. A. Springer. 1988. kinase activation contributes to Mac-1-mediated adhesion of Leukocyte adhesion deficiency: molecular basis and functional consequences. neutrophils (65Ð 67). Immunodefic. Rev. 1:39. 11. Bienvenu, K., and D. N. Granger. 1993. Molecular determinants of shear rate- In stimulated neutrophils, new expression of Mac-1 receptors dependent leukocyte adhesion in postcapillary venules. Am. J. Physiol. 264: occurs on the surface by exocytosis of the secretory vesicles stor- H1504. ing these integrins. We showed that the basal and fMLP-up-regu- 12. Luscinskas, F. W., A. F. Brock, M. A. Arnaout, and M. A. Gimbrone, Jr. 1989. Endothelial-leukocyte adhesion molecule-1-dependent and leukocyte (CD11/ lated surface levels of Mac-1 are not LSP1-dependent. There is CD18)-dependent mechanisms contribute to polymorphonuclear leukocyte adhe- evidence that activation of neutrophils induces the trafficking of sion to cytokine-activated human vascular endothelium. J. Immunol. 142:2257. 13. Ding, Z. M., J. E. Babensee, S. I. Simon, H. Lu, J. L. Perrard, D. C. Bullard, existing Mac-1 to uropod while newly expressed Mac-1 receptors X. Y. Dai, S. K. Bromley, M. L. Dustin, M. L. Entman, et al. 1999. Relative are found at the anterior of the cell forming new adhesion sites (68, contribution of LFA-1 and Mac-1 to neutrophil adhesion and migration. J. Im- 69). In addition, activation of surface integrins through avidity munol. 163:5029. 14. Rosen, H., and S. Gordon. 1987. Monoclonal antibody to the murine type 3 change occurs through signals generated by chemotactic receptors complement receptor inhibits adhesion of myelomonocytic cells in vitro and in- in a cytoskeleton-dependent fashion. There is increasing evidence flammatory cell recruitment in vivo. J. Exp. Med. 166:1685. The Journal of Immunology 423

15. Jutila, M. A., L. Rott, E. L. Berg, and E. C. Butcher. 1989. Function and regu- articular cartilage of the murine knee joint by RT-PCR: kinetics of stromelysin lation of the neutrophil MEL-14 antigen in vivo: comparison with LFA-1 and and IL-1 mRNA levels during arthritis. Rheumatol. Int. 16:197. MAC-1. J. Immunol. 143:3318. 45. Lowell, C. A., L. Fumagalli, and G. Berton. 1996. Deficiency of Src family 16. Coxon, A., P. Rieu, F. J. Barkalow, S. Askari, A. H. Sharpe, U. H. von Andrian, kinases p59/61hck and p58c-fgr results in defective adhesion-dependent neutrophil ␤ M. A. Arnaout, and T. N. Mayadas. 1996. A novel role for the 2 integrin functions. J. Cell Biol. 133:895. CD11b/CD18 in neutrophil apoptosis: a homeostatic mechanism in inflammation. 46. Chan, J. R., S. J. Hyduk, and M. I. Cybulsky. 2001. Chemoattractants induce a ␣ Immunity 5:653. rapid and transient upregulation of monocyte 4 integrin affinity for vascular cell 17. Sanchez-Madrid, F., and M. A. del Pozo. 1999. Leukocyte polarization in cell adhesion molecule 1 which mediates arrest: an early step in the process of em- migration and immune interactions. EMBO J. 18:501. igration. J. Exp. Med. 193:1149. 18. Greenwood, J. A., and J. E. Murphy-Ullrich. 1998. Signaling of de-adhesion in 47. Jongstra-Bilen, J., B. Vukusic, K. Boras, and J. E. Wither. 1997. Resting B cells cellular regulation and motility. Microsc. Res. Tech. 43:420. from autoimmune lupus-prone New Zealand Black and (New Zealand Black ϫ 19. Cox, E. A., and A. Huttenlocher. 1998. Regulation of integrin-mediated adhesion New Zealand White)F1 mice are hyper-responsive to T cell-derived stimuli. during cell migration. Microsc. Res. Tech. 43:412. J. Immunol. 159:5810. 20. Mitchison, T. J., and L. P. Cramer. 1996. Actin-based cell motility and cell 48. Athanasou, N. A. 1995. Synovial macrophages. Ann. Rheum. Dis. 54:392. locomotion. Cell 84:371. 49. Edwards, J. C. 1995. Synovial intimal fibroblasts. Ann. Rheum. Dis. 54:395. 21. Lauffenburger, D. A., and A. F. Horwitz. 1996. Cell migration: a physically 50. Keystone, E. C., H. U. Schorlemmer, C. Pope, and A. C. Allison. 1977. Zymosan- integrated molecular process. Cell 84:359. induced arthritis: a model of chronic proliferative arthritis following activation of 22. Gluck, U., and A. Ben-Ze’ev. 1994. Modulation of ␣-actinin levels affects cell the alternative pathway of complement. Arthritis Rheum. 20:1396. motility and confers tumorigenicity on 3T3 cells. J. Cell Sci. 107:1773. 51. van den Berg, W. B., M. W. Kruijsen, L. B. van de Putte, H. J. van Beusekom, 23. Rodriguez Fernandez, J. L., B. Geiger, D. Salomon, and A. Ben-Ze’ev. 1993. M. van der Sluis-van der Pol, and W. A. Zwarts. 1981. Antigen-induced and Suppression of vinculin expression by antisense transfection confers changes in zymosan-induced arthritis in mice: studies on in vivo cartilage proteoglycan syn- cell morphology, motility, and anchorage-dependent growth of 3T3 cells. J. Cell thesis and chondrocyte death. Br. J. Exp. Pathol. 62:308. Biol. 122:1285. 52. Lowell, C. A., and G. Berton. 1999. Integrin signal transduction in myeloid 24. Roberts, A. W., C. Kim, L. Zhen, J. B. Lowe, R. Kapur, B. Petryniak, A. Spaetti, leukocytes. J. Leukocyte Biol. 65:313. 53. Dransfield, I., C. Cabanas, A. Craig, and N. Hogg. 1992. Divalent cation regu-

J. D. Pollock, J. B. Borneo, G. B. Bradford, et al. 1999. Deficiency of the he- Downloaded from matopoietic cell-specific Rho family GTPase Rac2 is characterized by abnormal- lation of the function of the leukocyte integrin LFA-1. J. Cell Biol. 116:219. ities in neutrophil function and host defense. Immunity 10:183. 54. Lundgren-Akerlund, E., E. Berger, and K. E. Arfors. 1992. Effect of divalent 25. Sun, H. Q., K. Kwiatkowska, D. C. Wooten, and H. L. Yin. 1995. Effects of CapG cations on adhesion of polymorphonuclear leukocytes to matrix molecules in overexpression on agonist-induced motility and second messenger generation. vitro. J. Leukocyte Biol. 51:603. J. Cell Biol. 129:147. 55. Berger, M., J. O’Shea, A. S. Cross, T. M. Folks, T. M. Chused, E. J. Brown, and 26. Pulford, K., M. Jones, A. H. Banham, E. Haralambieva, and D. Y. Mason. 1999. M. M. Frank. 1984. Human neutrophils increase expression of C3bi as well as Lymphocyte-specific protein 1: a specific marker of human leukocytes. Immu- C3b receptors upon activation. J. Clin. Invest. 74:1566. nology 96:262. 56. Borregaard, N., L. Kjeldsen, K. Lollike, and H. Sengelov. 1993. Granules and 27. Jongstra-Bilen, J., V. L. Misener, C. Wang, H. Ginzberg, A. Auerbach, vesicles of human neutrophils: the role of endomembranes as source of plasma http://www.jimmunol.org/ A. L. Joyner, G. P. Downey, and J. Jongstra. 2000. LSP1 modulates leukocyte membrane proteins. Eur. J. Haematol. 51:318. populations in resting and inflamed peritoneum. Blood 96:1827. 57. Melnicoff, M. J., P. K. Horan, and P. S. Morahan. 1989. Kinetics of changes in peritoneal cell populations following acute inflammation. Cell. Immunol. 118:178. 28. Klein, D. P., S. Galea, and J. Jongstra. 1990. The lymphocyte-specific protein 58. Gordon, S., I. Fraser, D. Nath, D. Hughes, and S. Clarke. 1992. Macrophages in LSP1 is associated with the cytoskeleton and co-caps with membrane IgM. J. Im- tissues and in vitro. Curr. Opin. Immunol. 4:25. munol. 145:2967. 59. Huttenlocher, A., R. R. Sandborg, and A. F. Horwitz. 1995. Adhesion in cell 29. Jongstra-Bilen, J., P. A. Janmey, J. H. Hartwig, S. Galea, and J. Jongstra. 1992. migration. Curr. Opin. Cell Biol. 7:697. The lymphocyte-specific protein LSP1 binds to F-actin and to the cytoskeleton 60. Hannigan, M., L. Zhan, Y. Ai, and C. K. Huang. 2001. Leukocyte-specific gene through its COOH-terminal basic domain. J. Cell Biol. 118:1443. 1 protein (LSP1) is involved in chemokine KC-activated cytoskeletal reorgani- 30. Zhang, Q., Y. Li, and T. H. Howard. 2000. Human lymphocyte-specific protein zation in murine neutrophils in vitro. J. Leukocyte Biol. 69:497. 1, the protein overexpressed in neutrophil actin dysfunction with 47-kDa and 61. Zhou, X., and J. Li. 2000. Macrophage-enriched myristoylated alanine-rich C 89-kDa protein abnormalities (NAD 47/89), has multiple F-actin binding do- kinase substrate and its phosphorylation is required for the phorbol ester-stimu- by guest on September 26, 2021 mains. J. Immunol. 165:2052. lated diffusion of ␤ integrin molecules. J. Biol. Chem. 275:20217. 31. Klein, D. P., J. Jongstra-Bilen, K. Ogryzlo, R. Chong, and J. Jongstra. 1989. 2 62. Jones, S. L., J. Wang, C. W. Turck, and E. J. Brown. 1998. A role for the Lymphocyte-specificCa2ϩ-binding protein LSP1 is associated with the cytoplas- actin-bundling protein L-plastin in the regulation of leukocyte integrin function. mic face of the plasma membrane. Mol. Cell. Biol. 9:3043. Proc. Natl. Acad. Sci. USA 95:9331. 32. Jongstra, J., M. E. Ittel, N. N. Iscove, and G. Brady. 1994. The LSP1 gene is 63. Zu, Y. L., J. Qi, A. Gilchrist, G. A. Fernandez, D. Vazquez-Abad, D. L. Kreutzer, expressed in cultured normal and transformed mouse macrophages. Mol. Immu- C. K. Huang, and R. I. Sha’afi. 1998. p38 mitogen-activated protein kinase ac- nol. 31:1125. tivation is required for human neutrophil function triggered by TNF-␣ or FMLP 33. Brennan, L., and J. Jongstra. 1996. Low expression of the Lsp1 gene in early mouse stimulation. J. Immunol. 160:1982. T-lymphomas induced by N-methyl-N-nitrosourea. Carcinogenesis 17:771. 64. Cao, M. Y., F. Shinjo, S. Heinrichs, J. W. Soh, J. Jongstra-Bilen, and J. Jongstra. 34. Jongstra-Bilen, J., A. J. Young, R. Chong, and J. Jongstra. 1990. Human and 2001. Inhibition of anti-IgM-induced translocation of protein kinase C ␤I inhibits mouse LSP1 genes code for highly conserved phosphoproteins. J. Immunol. 144: ERK2 activation and increases apoptosis. J. Biol. Chem. 276:24506. 1104. 65. Pillinger, M. H., A. S. Feoktistov, C. Capodici, B. Solitar, J. Levy, T. T. Oei, and 35. Kadiyala, R. K., B. W. McIntyre, and A. M. Krensky. 1990. Molecular cloning M. R. Philips. 1996. Mitogen-activated protein kinase in neutrophils and enucle- and characterization of WP34, a phosphorylated human lymphocyte differentia- ate neutrophil cytoplasts: evidence for regulation of cell-cell adhesion. J. Biol. tion and activation antigen. Eur. J. Immunol. 20:2417. Chem. 271:12049. 36. Li, Y., A. Guerrero, and T. H. Howard. 1995. The actin-binding protein, lym- 66. Pillinger, M. H., C. Capodici, P. Rosenthal, N. Kheterpal, S. Hanft, M. R. Philips, phocyte-specific protein 1, is expressed in human leukocytes and human myeloid and G. Weissmann. 1998. Modes of action of aspirin-like drugs: salicylates in- and lymphoid cell lines. J. Immunol. 155:3563. hibit ERK activation and integrin-dependent neutrophil adhesion. Proc. Natl. 37. Huang, C. K., L. Zhan, Y. Ai, and J. Jongstra. 1997. LSP1 is the major substrate Acad. Sci. USA 95:14540. for mitogen-activated protein kinase-activated protein kinase 2 in human neutro- 67. Philips, M. R., J. P. Buyon, R. Winchester, G. Weissmann, and S. B. Abramson. phils. J. Biol. Chem. 272:17. 1988. Up-regulation of the iC3b receptor (CR3) is neither necessary nor sufficient 38. Carballo, E., D. Colomer, J. L. Vives-Corrons, P. J. Blackshear, and J. Gil. 1996. to promote neutrophil aggregation. J. Clin. Invest. 82:495. Characterization and purification of a protein kinase C substrate in human B cells: 68. Francis, J. W., R. F. Todd III, L. A. Boxer, and H. R. Petty. 1989. Sequential identification as lymphocyte-specific protein 1 (LSP1). J. Immunol. 156:1709. expression of cell surface C3bi receptors during neutrophil locomotion. J. Cell 39. Coates, T. D., J. C. Torkildson, M. Torres, J. A. Church, and T. H. Howard. 1991. Physiol. 140:519. An inherited defect of neutrophil motility and microfilamentous cytoskeleton 69. Hughes, B. J., J. C. Hollers, E. Crockett-Torabi, and C. W. Smith. 1992. Re- associated with abnormalities in 47-kD and 89-kD proteins. Blood 78:1338. cruitment of CD11b/CD18 to the neutrophil surface and adherence-dependent 40. Howard, T., Y. Li, M. Torres, A. Guerrero, and T. Coates. 1994. The 47-kD cell locomotion. J. Clin. Invest. 90:1687. protein increased in neutrophil actin dysfunction with 47- and 89-kD protein 70. van Kooyk, Y., and C. G. Figdor. 2000. Avidity regulation of integrins: the abnormalities is lymphocyte-specific protein. Blood 83:231. driving force in leukocyte adhesion. Curr. Opin. Cell Biol. 12:542. 41. Howard, T. H., J. Hartwig, and C. Cunningham. 1998. Lymphocyte-specific pro- 71. Stewart, M. P., A. McDowall, and N. Hogg. 1998. LFA-1-mediated adhesion is tein 1 expression in eukaryotic cells reproduces the morphologic and motile ab- regulated by cytoskeletal restraint and by a Ca2ϩ-dependent protease, calpain. normality of NAD 47/89 neutrophils. Blood 91:4786. J. Cell Biol. 140:699. 42. Li, Y., Q. Zhang, R. Aaron, L. Hilliard, and T. H. Howard. 2000. LSP1 modulates 72. Williams, M. A., and J. S. Solomkin. 1999. Integrin-mediated signaling in human the locomotion of monocyte-differentiated U937 cells. Blood 96:1100. neutrophil functioning. J. Leukocyte Biol. 65:725. 43. Berg, N. N., and H. L. Ostergaard. 1995. Characterization of intercellular adhe- 73. Ross, G. D. 2000. Regulation of the adhesion versus cytotoxic functions of the ␣ ␤ sion molecule-1 (ICAM-1)-augmented degranulation by cytotoxic T cells: Mac- 1/CR3/ M 2-integrin glycoprotein. Crit. Rev. Immunol. 20:197. ICAM-1 and anti-CD3 must be co-localized for optimal adhesion and stimula- 74. van Spriel, A. B., J. H. Leusen, M. van Egmond, H. B. Dijkman, K. J. Assmann, tion. J. Immunol. 155:1694. T. N. Mayadas, and J. G. van de Winkel. 2001. Mac-1 (CD11b/CD18) is essential 44. Van Meurs, J. B., P. L. Van Lent, L. A. Joosten, P. M. Van der Kraan, and for Fc receptor-mediated neutrophil cytotoxicity and immunologic synapse for- W. B. Van den Berg. 1997. Quantification of mRNA levels in joint capsule and mation. Blood 97:2478.