The Linda Jacobs Chair in Immune and Stem Cell Research
E-mail: ronen.alon@weizmann.ac.il
Phone: +972 8 934 2482
Fax: +972 8 934 2724
Location: Wolfson Bldg., Room 330
|
Prof. Ronen Alon
The Linda Jacobs Chair in Immune and Stem Cell Research E-mail: ronen.alon@weizmann.ac.il Phone: +972 8 934 2482 Fax: +972 8 934 2724 Location: Wolfson Bldg., Room 330 |
| Research Interests | Findings & Objectives | Acknowledgements | Members | Publications |
Research Interests:
- Chemokine signaling to leukocyte integrins at endothelial and extravascular contacts
Circulating immune cells and hematopoietic progenitors must exit blood vessels near specific target sites of injury, inflammation or tissue repair. The vessel wall at these sites displays specific combinations of traffic signals in the form of adhesion molecules (selectins, integrins) and chemotactic cytokines (chemokines) which operate in sequence to recruit only specific circulating subsets with proper receptors to these signals (Scheme 1).
As these processes take place under shear stress, these traffic molecules have evolved to operate under specialized kinetic and mechanical contexts. Using special flow chambers which simulate blood flow in the circulation (Scheme 2) we attempt to dissect how these molecules and their cytoplasmic associations with the cell cytoskeleton mediate cell adhesion and exit through blood vessels. Videomicroscopy of immune cells interacting with vessel cells (Movie 1 and Movie 2) and subcellular staining of both adhesion receptors and their specific cytoplasmic regulators allow us to follow spatially and temporally how these molecules mediate leukocyte exit across the blood vessel walls and how antigenic signals are integrated into stoppage of extravascular lymphocytes on dendritic cells and macrophages. This information is key for the development and implementation of therapeutic tools to fight autoimmunity, allergy, heart injury, atherosclerosis, organ rejection and metastasis.
Scheme 1: A multistep model for leukocyte and stem cell recruitment at target vessel walls. Selectins, chemoattractants (chemokines) and integrins coordinate in capturing a circulating leukocyte to the vessel wall at specific sites and in generating resistance to detachment by the blood shear forces. Further leukocyte exposure to apical endothelial chemokines and shear flow trigger leukocyte transmigration through the endothelial barrier.
Scheme 2: The in vitro flow chamber apparatus and attached videomicroscopy setup used in the lab for recording real time adhesion and migration of immune cells under conditions simulating physiological blood flow.
- Leukocyte trafficking to lungs of heavy smokers: implications for susceptibility to chronic obstructive pulmonary diseases (COPD) and progression of epithelial lung cancers Smoking is the major cause of obstructive airway diseases commonly termed COPD (chronic obstructive pulmonary diseases) and a major risk for lung cancers. COPD is a chronic inflammatory process in the small airways. In spite of increasing amount of data on the role of specific adhesion receptors and chemotactic cytokines in leukocyte trafficking to the lung in health as well as in asthmatic processes, there is still little information on the key trafficking routes taken by particular subsets of leukocytes to enter and function at sites of obstructive chronic pulmonary diseases. Well controlled attenuation of this trafficking is likely to interfere with COPD progression. Interference with the function of effector leukocytes within smoke associated sites of inflammation is expected to also modulate smoke inducted lung tumorigenesis. The major aim of our studies is to identify key trafficking routes for inflammatory monocytes and effector lymphocytes that can be promising targets for anti-inflammation therapy for heavy smokers with high risk to develop COPD. A second aim is to investigate if blocking the trafficking of specific inflammatory cell subsets to the lung can attenuate COPD progression and at the same time reduce the progression of early transformed tumor cells into aggressive lung cancers. To reach these goals, we plan to first establish controlled murine experimental models for smoke induced COPD, define if and how endothelial selectins, specific integrin receptors, and different chemokine receptors control leukocyte entry into smoke triggered inflamed lung and contribute to COPD pathogenesis, and finally manipulate any of these specific trafficking molecules and test how attenuated trafficking and altered functions of specific immune regulatory cells affect COPD pathogenesis.
- Chemokine activation of leukocyte integrins at endothelial contacts under shear stress
The ability of rolling leukocytes to arrest on target endothelial sites depends on rapid activation of their integrins. Studying LFA-1 activation as a paradigm for leukocyte integrins, we find that rapid LFA-1 activation by endothelial-displayed chemokines (Movie 3 and Movie 4) involves local GPCR signals which trigger conformational rearrangements of the integrin within millisecond contacts (Scheme3). These spatially confined events involve the cytoskeletal regulatory GTPases RhoA and Rap-1 and require proper integrin activation and anchorage to the cortical actin cytoskeleton via cytoskeletal linkers such as talin and Kindlin-3. An inherited integrin activation deficiency identified by us, termed LAD-III, is linked to deficiency in Kindlin-3, a hematopoietic member of the Kindlin adaptor family. A possible crosstalk between chemokine stimulated Rap-1 and RhoA GTPases and these integrin cytoskeletal linkers and their associated effectors is currently under investigation in both normal and malignant lymphocytes.
Scheme 3: A proposed model for rapid integrin activation by endothelial-immobilized chemokines. Flow chamber analysis is shown in movies 3 and 4. Bidirectional activation switches the integrin (LFA-1 in this demonstration) from an inactive bent state to an extended state within 0.1 sec (step 1). This critical event primes the integrin to bind its endothelial ligand and undergo a further conformational shift (step 2, depicted by the red star). This activation of the integrin headpiece causes further separation of the integrin subunit tails (step 3). Associations of the integrin subunit tails with talin and additional adaptors provide critical mechanical stabilization of the nascent adhesive bond.The GTPase RhoA was previously implicated in CCL21-triggered LFA-1 affinity triggering in T lymphocytes and in LFA-1 dependent adhesion strengthening to ICAM-1. In a follow up study, in collaboration with C. Laudanna (Verona) we showed that a specific RhoA 23/40 effector region is vital for the earliest LFA-1 dependent arrests of lymphocytes on HEVs. Blocking the RhoA 23/40 region in human T lymphocytes in vitro impaired the subsecond CXCL12-triggered LFA-1 mediated T cell arrest on ICAM-1 by preventing the induction of an extended LFA-1 conformational state. In contrast to the two prototypic chemokines, CXCL12 and CCL21, the inflammatory chemokine CXCL9 triggered robust LFA-1 mediated T lymphocyte adhesion to ICAM-1 at subsecond contacts independently of the RhoA 23/40 region. CXCL9 also did not induce inside-out conformational changes in the LFA-1 ectodomain, suggesting that weak endothelial displayed chemokines can activate LFA-1 through outside-in stabilization following ICAM-1 binding, rather than by conformational activation prior to ligand binding. Our results collectively suggest that the 23/40 region of RhoA regulates chemokine induced inside-out LFA-1 extension prior to ligand binding, but is not required for a variety of non chemokine signals that strengthen LFA-1-ICAM-1 bonds via outside in conformational changes, such as TCR signals (see below).
- The role of talin1 and Kindlin-3 in integrin activation and lymphocyte adhesiveness to inflamed vessels under shear stress
Our recent work has shed light on the earliest signals transmitted by chemokines through their G-protein coupled receptors (GPCRs) to distinct leukocyte integrins (Scheme 4). In lymphocytes, the integrin adaptor talin1 was found to control major conformational changes of both LFA-1 and VLA-4 integrins implicated in lymphocyte arrest on endothelial ligands. We also found that VLA-4 anchorage to the cytoskeleton is enhanced by the cytoskeletal adaptor paxillin, which cooperates with talin1 to promote the activation of this integrin under external strain but is not essential for chemokine stimulated integrin activation. Experiments with T cells in which talin or paxillin transcription has been silenced suggest that these two adaptors regulate distinct and complementary roles in VLA-4 adhesiveness to the major endothelial ligand VCAM-1.
In addition, we found a major role for the coactivating adaptor, Kindlin-3 (absent in T cells, neutrophils, and platelets from LAD-III patients) in chemokine mediated integrin activation in human leukocytes. A subgroup of these patients also have defective expression of a PLC-regulated GEF (guanine exchange factor) for Rap-1, CalDAG-GEFI (CDGI), but the loss of Kindlin-3 masks the effect of this deficiency in this subgroup of patients.
Scheme 4: A postulated scheme for rapid chemokine signaling to lymphocyte LFA-1 under shear flow.(Top) A rolling leukocyte tethered to an integrin ligand must encounter juxtaposed chemokine at the site of integrin activation, possibly a single microvillus. A quaternary complex between integrin, ligand, chemokine and G-protein coupled receptor (GPCR) must form within milliseconds to locally activate the integrin ligand complex via a bi-directional signaling event (bottom). Only a fully activated integrin can arrest the rolling leukocyte on the vessel wall while partially activated integrins can participate in rolling adhesions (see Figure 1).
Upon initial encounter, endothelial-bound chemokine transduces leukocyte GPCRs signals which convert the inactive (folded) integrin to its extended conformation (step 1).
GTP bound RhoA and Rac1 are involved in this step in the case of lymphocyte LFA-1. This critical chemokine-driven inside-out event primes the integrin to transiently bind endothelial ligands on the counter endothelial surface. The various I-domains undergo further conformational shifts upon extracellular ligand binding (step 2), resulting in further integrin activation (outside-in). This ligand-driven step is predicted to result in further separation of the integrin subunit tails, conditional on the presence of talin nearby the ligand occupied integrin, and force transduction (step 3). Clustering of ligand-occupied integrins (not shown) can rapidly follow. Force transduction may lead to talin activation, recruitment of vinculin and crosslinking of integrin-talin complexes to the cortical actin cytoskeleton, all within seconds after initial leukocyte arrest.
The additional involvement of Rap1 and its effectors RIAM and RAPL (not shown) in GPCR transduced inside-out activation is suggested for LFA-1 activation by chemokines, whereas PKC is necessary for full VLA-4 activation.
- Invasive filopodia, key functional units in lymphocyte transendothelial migration (TEM)
Endothelial chemokines are instrumental for integrin mediated lymphocyte adhesion and TEM through vascular barriers. Addressing how chemokine signals promote these two processes using flow chamber assays, we found that following arrest, both T and B lymphocytes rapidly crawl on inflamed endothelium in an LFA-1 and chemokine dependent manner. We also found that endothelial presented chemokines trigger lymphocyte crawling by stimulating high affinity (HA) LFA-1 at numerous focal submicron dots underneath the entire adhesive zone reversibly generated by the rapidly crawling lymphocyte (Figure 1).
Surface bound chemokines and ICAM-1 are both necessary and sufficient for these quantal LFA-1 dots to form and rapidly dissociate under physiological shear flow. Interestingly, chemokine stimulated VLA-4 remains clustered at the rear of crawling lymphocytes, shortly after microvillar collapse and in proximity to other microvillar surface receptors like PSGL-1 and CD43. Notably, shear forces applied on crawling lymphocytes increased the density of both LFA-1 dots and of adhesive filopodia extended from the base of the crawling lymphocyte. Strikingly, a fraction of these filopodia became highly invasive into the endothelial cell body even prior to transendothelial crossing. These filopodia are not classical podosomes recently observed in T blasts crossing activated endothelium. The density of invasive filopodia increased with the magnitude of chemokine and ICAM-1 signals and closely correlated with the ability of T cells to successfully transmigrate upon reaching paracellular junctions. During TEM, T cells extend large subluminal lamellopodia, enriched with HA-LFA-1 and subluminal ICAM-1, suggesting that basal endothelial ICAM-1 guides T cell TEM (Figure 1).
Similar LFA-1 dots have not yet been observed in vivo due to poor resolution of integrin imaging, but several EM reports detected invasive protrusions underneath lymphocytes associated with multiple endothelial beds in vivo. We propose that HA-LFA-1 dots are key shear resistant adhesive units which give rise to adhesive and invasive filopodia that scan the endothelial surface for apical and junctional guidance signals for diapedesis. Our recent results also suggest the involvement of the CDC42 GTPase in this chemokine triggered, shear force-facilitated LFA-1-dependent process. In addition, we were among the first groups to identify novel transcellular routes taken by neutrophils and small subsets of T effector cells to cross endothelial barriers (Movie 5 and Figure 2).
Figure 1: Distribution of LFA-1 subsets in focal dots generated by peripheral blood T cells crawling and sending out leading lamelliopodia into sub-endothelial compartments. A. LFA-1 (red, right) clusters in numerous focal dots underneath T cells crawling over chemokine presenting cytokine activated HUVEC whereas α4 integrins (green. left) remain in the uropod. Post fixation staining. B. A transmission EM image of a T cell sending out invasive filopodia via a paracellular EC junction (red arrowhead). C. Top view of high affinity (HA) LFA-1 (green) engaged with sub-endothelial ICAM-1 (red) in an actively transmigrating T cell. Note the tight engagement between sub-endothelial ICAM-1 (red dots, right panel) and the HA LFA-1 (green dots, left panel). HA LFA-1 enrichment at the apical side of the T cell’s leading edge nearby ICAM-1 (bottom image) suggests that sub-endothelial ICAM-1 guides the transmigrating T cell via serial engagements with HA LFA-1.
Figure 2: Shear stress induces massive apicali invaginations in neutrophils and lymphocytes. Left: TEM through PAF-presenting, TNAFα-activated HUVEC. A. Ultrastructural analysis of neutrophils migrating under shear flow across PAF-presenting HUVEC prestimulated with TNFα. Shown are serial cross sections of two neutrophils adhering to and migrating across neighboring ECs. B. A representative neutrophil left under shear on IL-8 presenting HUVEC prestimulated with TNFα. Right: Two T lymphocytes sending invasive filopodia into the body of a cytokine stimulated endothelial cell in the presence of shear forces. The density of these filopodia is increased by up to 5 fold in the presence of maximal stimulatory conditions (shear stress, high levels of apical chemokines).
The Rho/Rac GTPases are in situ triggered by signals from endothelial chemokines signals, and function as key regulators of leukocyte integrin activation and motility. In lymphocytes, a major portion of chemokine-mediated Rac activation depends on the CDM adaptor DOCK2. Using a novel ex vivo model for real time analysis of murine leukocyte subsets, we previously identified a specialized role for this adaptor in chemokine-triggered integrin-independent lymphocyte motility on endothelial and extracellular matrix barriers but not in chemokine-triggered integrin-mediated adhesiveness or the transmigration of lymphocytes through a chemokine-presenting endothelial barrier. We have found that DOCK2 is also crucial for activation of a subset of Rac mediated activities which control the ability of lymphocytes to crawl away from the arrest site to endothelial junctions, a process mediated primarily by their LFA-1 integrin. Although implicated in integrin activation, Rap-1 is not required for this integrin-independent chemokine triggered lymphocyte motility.
Scheme 5: Chemokine stimulated lymphocyte crawling via high affinity LFA-1 microclusters and millipede-like contacts. Endothelial bound chemokines trigger Rap-1 and Rho family GTPases that stimulate LFA-1. The high affinity LFA-1 is arranged in ICAM-1 engaged focal dots which are critical for shear resistant adhesion of the newly arrested lymphocyte. Adhesive filopodia are subsequently generated by additional chemokine signals. The conversion of adhesive into invasive filopodia allow the crawling lymphocyte to probe the endothelium for sites of TEM and possibly integrate additional signals when it reaches the paracellular endothelial junction. Serial engagements with sublumenal endothelial ICAM-1 guide the leading edge of the transmigrating lymphocyte. This image was included in a preview article by S. G. Ward (Immunity, March 20, 2009) that discusses our paper (Shulman, Immunity 30, 384-396).Current studies are aimed at elucidating how integrins occupied by ICAM-1 participate in this and other modes of leukocyte transendothelial migration. A current follow up study suggests that effector Th1 lymphocytes use to a lesser extent shear stress signals for their transendothelial migration. These lymphocytes also do not need signals from endothelial chemokines to generate integrin dependent invasive filopodia. Notably, these effector T cells use their preexistent high affinity LFA-1 to trigger outside in machineries that mediate much faster TEM than resting T cells. It appears that key residues on LFA-1 expressed by these effector cells are modified with unique serine/threonine phosphorylation patterns, recently shown to alter LFA-1 conformation and mobility on the plasma membrane. Currently, we are comparing the roles of Src kinases, PI3Ks, DOCK2 and Rap-1 effectors in the TEM of Th1 vs. resting lymphocyte across distinct inflamed endothelial barriers. The possibility that effector/inflammatory lymphocytes, but not resting memory T cells, use unique Src programs for integrin activation is pharmacologically attractive for introduction of selective integrin inhibitory drugs for these lymphocyte subsets.
- Role of lymph node chemokines in lymphocyte scanning of dendritic cells
Lymphocytes entering the T cell zone in peripheral lymph nodes are exposed to high levels of CCR7 and CXCR4 binding chemokines. Our recent work suggests that these extravascular chemokines preferentially operate to trigger T cell motility and encounter of antigen presenting dendritic cells (DCs) in stromal-presented (two-dimensional) rather than soluble states. We have developed an in vitro live imaging set-up to follow how human T cells locomoting on immobilized chemokines encounter DCs and how specific antigenic signals stop lymphocyte subsets on differently stimulated DCs (Figure 3 and Movies 6, 7).
Figure 3: In vitro imaging of T cells (labeled with the inert dye BCECF) locomoting over immobilized CCL21, the key chemokine found in lymph node T zones. T cells are rapidly locomoting around monocyte-differentiated dendritic cells (grey), derived from a matching donor, but fail to generate any stable contacts in the absence of an antigen. The image is taken from Movie 6. The predominant T zone chemokine is CCL21. Most lymphocytes rapidly locomote on networks of fibroblastic reticular stromal cells (FRCs) presenting this chemokine in an immobilized state as they touch and scan FRC associated dendritic cells for antigenic signals. Assessing the mechanisms by which this chemokine and ICAM-1 can support random motility, we found that a lawn of immobilized chemokines (CCL21, CCL19 or CXCL12) is necessary and sufficient for rapid T cell motility, whereas soluble counterparts of these chemokines fail to promote motility even in the presence of integrin ligands. Notably, the ability of weak GPCR-ligand bonds to support contractile motility did not involve homologous GPCR desensitization, allowing the lymphocyte to retain persistent motility for hours without noticeable decay in responsiveness to these surface-presented chemokines.
Most surprisingly, when the lymph node chemokines were encountered by motile T cells together with high levels of FRC or DC expressed integrin ligands, like ICAM-1 and VCAM-1, the integrin receptors to these ligands remained largely silent unless lymphocytes were exposed to external shear forces. Furthermore, neither the extent of motility nor the ability of motile cells to occasionally stop on chemokine-CAM coated surfaces was affected by integrin blockage. In vivo analysis of murine T cells lacking functional LFA-1 confirmed these results, although the speed of motile T cells was slightly reduced. Thus, LFA-1 does not contribute to any detectable T cell-T cell or T cell-DC sticking in vivo or in vitro, consistent with effective silencing of chemokine clustered LFA-1 in the absence of antigenic signals.
In conclusion, while LFA-1 and VLA-4 adhesiveness is robustly activated when lymphocytes encounter chemokines and juxtaposed integrin ligands on endothelial cells in the presence of shear forces, these integrins are generally silenced in shear-free extravascular environments.
Figure 4: Surface-presented CCL21 triggers robust motility and polarized redistribution of LFA-1 and VLA-4 in migrating T cells. Live images of LFA-1 (top) or of the α4 integrins VLA-4 and α4b7 (bottom) in T cells migrating in vitro. Representative images were taken at the indicated time intervals. Cells were labeled with either Alexa568-tagged TS2/4 (non-blocking anti-αL) or Alexa488-tagged B5G10 (non-blocking anti-α4). Upper rows depict consecutive merged DIC and fluorescence images. Lower rows depict spectrum analysis of the relative fluorescence intensity.
- Active role of ICAM-1 in TCR stimulated LFA-1 activation and T cell stoppage on DCs
Lymphocyte stoppage on dendritic cells (DC) requires activation of LFA-1 by T cell receptor (TCR) signals, but the molecular basis of this activation is elusive. Using antibodies for specific LFA-1 conformations, we found that TCR activation in resting lymphocytes was insufficient to trigger LFA-1 extension or headpiece opening. However, TCR signals facilitate rapid conformational activation of LFA-1 by immobile ICAM-1, and promote lymphocyte spreading without requirement for cytosolic Ca2+.
The T cell-DC synapse is mediated by numerous scattered focal dots of ICAM-1-rearranged high affinity LFA-1. The CCL21 chemokine dramatically accelerates LFA-1 responsiveness to TCR stop signals. Interestingly, T cells locomoting on immobilized CCL21 and CXCL12 cluster their LFA-1 in the leading edge (Fig. 4) and thereby prime it to rapid activation by TCR signals. Thus, promotile chemokines not only increase T cell scanning of DCs but sensitize LFA-1 to arrest these motile lymphocytes upon encounter of correct antigen. The integrin adaptor, Kindlin-3 is critical for LFA-1 affinity triggering by TCR signals but dispensable for lymphocyte motility on chemokines.
Our results suggest that TCR signals activate LFA-1 conditional to its occupancy by ICAM-1 within submicron focal dots. We suggest that these dots are the quantal adhesive units of the lymphocyte-DC synapse and that the characteristic ring of LFA-1 observed in many in vitro studies of T-B synapses is probably a terminating supramolecular structure rather than the initial prerequisitory adhesive unit required for T cell stoppage on antigen presenting cells.
Figure 5: The quantal adhesive units of TCR stimulated T cells spread on DCs are scattered focal dots enriched with high affinity LFA-1. T cells were incubated with OKT3 (10 µg/ml) and labeled with a trace of a non blocking (Alexa 568 labeled) anti LFA-1 mAb. Lymphocytes were allowed to spread on prespread DCs for 5 min and fixed. A merge DIC/fluorescence image of LFA-1 in a representative T cell spread on a DC is shown. Bar, 3 µm. Our current research is supported by ISF, GIF, BSF, Minerva and the FAMRI Foundation.
Staff Scientist: Dr. Sara Feigelson Consultant: Dr. Valentin Grabovsky Post-doctoral Fellow: Dr. Sigal Nakav Ph.D. Students: Eujenia Manevich-Mendelson Ronit Pasvolsky Ziv Shulman
