Polarized granzyme release is required for antigen-driven transendothelial migration of human effector memory CD4 T cells

Abstract

Human effector memory (EM) CD4 T cells may transmigrate across endothelial cell (EC) monolayers either in response to inflammatory chemokines or in response to TCR recognition of antigen presented on the surface of the EC. The kinetics, morphologic manifestations, and molecular requirements of chemokine- and TCR-driven transendothelial migration (TEM) differ significantly. Here we report that while the MTOC and cytosolic granules follow the nucleus across the endothelium in a uropod during chemokine-driven TEM, MTOC reorientation to the contact region between the T cell and the EC, accompanied by dynein-driven transport of granzyme-containing granules to and exocytosis at the contact region, are early events in TCR-driven but not chemokine-driven TEM. Inhibitors of either granule function or of granzyme proteolytic activity can arrest TCR-driven TEM, implying a requirement for granule discharge in the process. In the final stages of TCR-driven TEM, the MTOC precedes, rather than follows, the nucleus across the endothelium. Thus TCR-driven TEM of EM CD4 T cells appears to be a novel process that more closely resembles immune synapse formation than it does conventional chemotaxis.

Introduction

The “pioneer” T cell model of recall immunological responses proposes that rare circulating antigen-specific effector memory (EM) CD4 T cells initiate recall inflammatory response in a peripheral tissue by being the first lymphocytes to extravasate at the site where a foreign antigen has been reintroduced, e.g. by recurrent infection, and then priming the site, via production of inflammatory cytokines such as tumor necrosis factor (TNF) and interferon-γ, to activate the local microvascular endothelium, enabling recruitment of a subsequent influx of antigen-nonspecific cells through enhanced luminal expression of endothelial adhesion molecules and inflammatory chemokines (1) (2). In most mammals, with the exception of mice and rats, endothelial cells (EC) lining the blood microvessels basally express MHC class II molecules, the only well established function of which is presentation of antigenic peptides to the TCR of CD4 T cells (3–6). We and others have hypothesized that the recruitment of antigen-specific pioneer CD4 T cells may involve recognition of cognate antigen on the luminal surface of the microvascular EC. In support of this idea, we have shown, using cultured human microvascular EC, that EM CD4 T cells, freshly isolated from the circulation of adult humans, but not circulating naïve or central memory CD4 T cells, will transmigrate across EC in response to TCR engagement (7). EM CD4 T cells can also be induced to undergo TEM in response to inflammatory chemokines (8), but there are significant differences between chemokine-driven and TCR-driven TEM.

Chemokine-driven TEM of EM CD4 T cells follows the general paradigm established for TEM of other types of leukocytes, where adhesion molecules (e.g., ICAM-1 and VCAM) and inflammatory chemokines (e.g., IP-10/CXCL10) expressed on the apical EC surface engage T cell integrins (e.g., LFA-1 and VLA-4) and chemokine receptors (e.g., CXCR3), respectively, and these signals appear to be sufficient to support both capture and subsequent TEM within 15 minutes if venular levels of shear stress are provided to the adherent T cells (8, 9). In this setting, T cells recruited by chemokine signals initially roll but then spread out, firmly adhere and then crawl to the inter-EC junctions where transmigration occurs. When an EM CD4 T cell binds to EC and encounters a signal that engages the TCR, T cell spreading, uropod formation and rapid TEM are all inhibited. Instead, the T cell rounds up, extends a long (up to 20 µm), nuclear-free cytosolic extension, which we have designated as a “transendothelial protrusion” (TEP) that pushes through a region of the monolayer at or near the inter-EC junctions and then continues to tunnel underneath the EC monolayer. Over the next 30–60 minutes, the TEP is followed by the cell body, including the nucleus (7). Consistent with the disparate morphological transformations, we showed that chemokine and TCR-driven TEM depend on different regulators of the actin cytoskeleton, namely Cdc42 and Rac1, respectively (10). Moreover, only TCR-driven TEM is blocked by inhibitors of myosin IIA (10).

The molecular route taken by the T cell through the EC monolayer also differs depending on whether it is responding to chemokine or antigen. TCR-driven TEM requires interactions of specific counter-receptors on the T cell with several EC junctional molecules not required for chemokine-driven TEM, including CD31 (PECAM-1), CD99, CD155 (PVR) and CD112 (nectin-2) (11) (9), most of which have been identified as components of the lateral border recycling compartment utilized by monocytes and neutrophils for transendothelial migration (12–15). Unlike monocytes and neutrophils, but like chemokine-driven TEM of EM CD4 T cells, TCR-driven TEM also requires the application of venular levels of shear stress (7, 8).

The proximal TCR signaling machinery of EM CD4 T cells interacting with antigen presented by EC during TCR-driven TEM resembles that of TCR-mediated T cell activation in response to antigen presentation by a professional antigen-presenting cell (APC). For example, in both processes, ZAP70 is activated at the region of interface between the T cell and the cell displaying the antigen. Activated ZAP70 then phosphorylates adapter proteins at the region of cell-cell contact and eventually leads to the activation of Vav, the GTP exchange protein that activates Rac (10). However, while zones of molecules may be clearly defined in the contact area made between conventional APC and T cells, termed the immunological synapse (IS), such discrete organization is not readily apparent in the contact region between EM CD4 T cells and EC.

The organization of intracellular organelles is controlled and reflects the function of the cell. During migration, T cells develop a cellular polarization in which most organelles and the microtubular organizing center (MTOC) move to the rear of the cell, trailing behind the nucleus, where they recruit and provide energy for molecular motors that provide the contraction for forward propulsion (16, 17). The positioning of the MTOC in migrating leukocytes is unique; in all other non-hematopoietic cells, the MTOC is located between the leading edge and the nucleus. When T cells encounter cognate antigen on the surface of a conventional APC, the MTOC is repositioned to the IS, a process that is essential for polarized secretion of cytokines and secretory lysosomes (lytic granules) by T helper cells and CTL, respectively (18) (19). In the present study, we have analyzed MTOC positioning that occurs within EM CD4 T cells during TEM in response to TCR signals. We find that the MTOC moves to the contact region between the T cell and EC, where granzyme-containing granules are recruited by dynein to permit TEP formation and TEM, during which the MTOC precedes, rather than follows, the nucleus.

Materials and Methods

Cells and reagents

All human materials were obtained from de-identified blood or tissue donors under protocols approved by the Yale Human Investigation Committee. CIITA-transduced human dermal microvascular EC (CIITA HDMEC) and FcR-transduced (FcR) HDMEC were generated using retroviral vectors and characterized as described (7). Prior to flow experiments, CIITA HDMEC were incubated in the presence of 10 ng/ml recombinant human TNF (rhTNFα, R&D Systems) for 18–28 hours to upregulate adhesion molecules, and with 100 ng/ml recombinant TSST-1 (Toxin Technology, Inc.) for 30 minutes on CIITA HDMEC to allow TCR-mediated activation of all T cells utilizing Vβ2 gene segment to form their TCR, approximately 5–20% of the circulating T cells in most donors, or 10 µg/ml OKT3 or control IgG for 30 min on FcR HDMEC to activate all or none of the T cells, respectively.

Leukapheresis was performed on healthy volunteer adult donors and PBMCs were enriched by Ficoll-Hypaque density gradient centrifugation prior to cryopreservation of aliquotted cells. Total peripheral blood CD4 T cells were isolated from the cryopreserved samples by positive selection with CD4 Dynabeads magnetic beads and released with Detachabead (Dynal). Memory (CD4⁺CD45RA⁻) T cells were enriched by depletion of CD45RA⁺ cells from CD4 T cells using anti-CD45RA mAb (eBiosciences) and pan-mouse IgG beads (Dynal) and EM cells were further enriched by depleting CD4 central memory cells with anti-CCR7 mAb (BioLegend) and pan-mouse IgG beads. Approximately 90% of the initial T cell population as well as essentially all other leukocyte types were removed by these manipulations. The remaining 10% of CD4 T cells, highly enriched for CD4 EM T cells, were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum, 2 mM glutamine, and penicillin/streptomycin overnight prior to assays. To inhibit dynein, myosin IIA, and actin depolymerization, CD4 EM T cells were treated with 25 µM ciliobrevin D (Calbiochem), 50 µM blebbistatin (−) (Calbiochem), and 10 nM jasplakinolide (Calbiochem), respectively, or vehicle (DMSO) 30 min and washed prior to flow TEM assays. For granule inhibition, CD4 EM T cells were treated for 2 h with 0.1 µM concanamycin A (Sigma), 30 nM Bafilomycin A1 (Tocris), or vehicle (DMSO) and washed before flow assay. For serine protease inhibition, cells were treated with 10 µM DCI (Calbiochem), 10 µM TLCK (Sigma) or vehicle (DMSO) for 30 min and washed prior to flow assay.

TEM assays

CIITA HDMEC and FcR HDMEC were grown to confluence on 20 µg/ml human plasma fibronectin-coated 35 mm coverglasses, treated with TNF and loaded with TSST-1 and OKT3, respectively, as described (7), washed twice with RPMI/10% FBS, and assembled with a parallel plate flow chamber apparatus (Glycotech) using the 0.01 inch height, 5 mm wide slit gasket provided by the manufacturer. On a 37°C heating surface, CD4⁺CD45RA⁻CCR7^lowCD62L^low (EM) human T cells (10⁶ cells/500 µl) suspended in the same medium were loaded onto the EC monolayer at 0.75 dyne/cm² for 2 minutes, followed by washing with medium only at 1 dyne/cm² for 15 or 50 minutes. Samples were then fixed with 3.7% formaldehyde in PBS, stained with anti-Vβ2TCR mAb (Beckman Coulter), followed by Alexafluor 488 or 546-conjugated goat or donkey anti-mouse IgG, mounted on slides using mounting medium containing DAPI (Prolong Gold, Invitrogen), and examined by microscopy with a Zeiss Axiovert 200M microscope. A FITC filter was used to detect FITC or Alexafluor 488-stained cells, a TRITC filter was used to detect Alexafluor 546-stained cells, and a DAPI filter used to detect DAPI-stained nuclei. Using a 40X/0.60 korr Ph2 objective, phase contrast optics were used to determine whether CD4 T cells were either on top or underneath the HDMEC monolayer²⁶ The percentage of transmigrated CD4 T cells were calculated for 100–200 cells per sample by analyzing five to ten groups of 20 cells each, calculating the percentage for each group, and calculating the mean and s.e.m. for the groups.

For whole cell quantification of granzyme A in transmigrated and non-transmigrated EM CD4 T cells, fluorescence of individual cells in images taken with 40X objective were quantified using Image J.

Confocal Microscopy

To visualize MTOC of EM CD4 T cells transmigrating on CIITA HDMEC plus TSST-1, samples were stained with anti-Vβ2TCR mAb (Beckman Coulter), Alexa Fluor 488-conjugated goat anti-mouse IgG (Invitrogen), permeabilized with 0.5% Triton/PBS, blocked with mouse IgG, stained with Zenon-Alexa Fluor 555 (Invitrogen)-labeled anti-α-tubulin mAb (Sigma), re-fixed with methanol/acetone (50/50), stained with rabbit anti-γ-tubulin (Sigma), Alexa Fluor 647-conjugated chicken anti-rabbit IgG, and mounted on slides using mounting medium containing DAPI (Prolong Gold, Invitrogen). To visualize lytic granules, samples were permeabilized, stained with anti-granzyme A mAb (BioLegend), Alexa Fluor 546-conjugated donkey anti-mouse IgG, blocked with mouse IgG, stained with Zenon-fluorescein (Invitrogen)-labeled anti-Vβ2TCR mAb (Beckman Coulter), Alexa Fluor 488-conjugated rabbit anti-FITC (Invitrogen), Alexa Fluor 488-conjugated goat anti-rabbit, Zenon-Alexa Fluor 647 (Invitrogen)-labeled anti-α-tubulin mAb (Sigma) and mounted on slides using mounting medium containing DAPI (Prolong Gold, Invitrogen). Alternatively, samples were stained with anti-Vβ2TCR mAb (Beckman Coulter), Alexa Fluor 488-conjugated goat anti-mouse IgG (Invitrogen), permeabilized with 0.5% Triton/PBS, blocked with mouse IgG, stained with Zenon-Alexa Fluor 555 (Invitrogen)-labeled anti-granzyme A mAb (BioLegend), re-fixed with methanol/acetone (50/50), stained with rabbit anti-γ-tubulin (Sigma), Alexa Fluor 647-conjugated chicken anti-rabbit IgG, and mounted on slides using mounting medium containing DAPI (Prolong Gold, Invitrogen). To visualize extracellular and intracellular VE-cadherin at sites of TCR-driven TEM, samples were permeabilized, stained with goat polyclonal Ab to intracellular carboxy terminus VE-cadherin (Santa Cruz Biotechnology C-19), rabbit polyclonal Ab to extracellular VE-cadherin aa 1–258 (LifeSpan BioSciences), and anti-Vβ2TCR mAb (Beckman Coulter) followed by Alexa Fluor 647-conjugated chicken anti-goat IgG, Alexa Fluor 546-conjugated donkey anti-rabbit IgG, and Alexa Fluor 488-conjugated donkey anti-mouse IgG (Invitrogen) and mounted on slides using mounting medium containing DAPI. Alternatively, samples were stained with anti-Vβ2TCR mAb, Alexa Fluor 488-conjugated donkey anti-mouse IgG, permeabilized, followed by goat polyclonal Ab to intracellular carboxy terminus VE-cadherin (C-19, Santa Cruz Biotechnology), Alexa Fluor 647-conjugated chicken anti-goat IgG, blocked with mouse IgG, then Zenon Alexa Fluor 555 (Invitrogen)-labeled anti-VE-cadherin mAb (clone 16B1, eBiosciences) and mounted on slides using mounting medium containing DAPI. To visualize MTOC and lytic granulesof EM CD4 T cells transmigrating on FcR HDMEC plus IgG and OKT3 or HUVEC plus chemokines, samples were fixed with methanol/acetone, stained with rabbit anti--γ-tubulin, anti-granzyme A mAb, Alexa Fluor 647-conjugated chicken anti-goat IgG, Alexa Fluor 546-conjugated donkey anti-mouse IgG, FITC-conjugated anti CD45 mAb, Alexa Fluor 488-conjugated rabbit anti-FITC and mounted on slides using mounting medium containing DAPI.

Images were captured with a Leica TCS SP5 Spectral Confocal Microscope, 405UV using a 63X oil immersion objective and sequential scanning with 405 Diode, argon and He/Ne laser excitation lines of 405 nm, 488 nm, 543 nm, and 633 nm. Typically, 6–12 Z slices spanning portions of or the entire T cell were captured.

Staining was quantified with Image J.

FACS analysis

Flow Cytometry Standard (FCS) files of CD4 T cells fixed, permeabilized and stained with anti-granzyme A mAb (BioLegend) or IgG control, Alexa Fluor 647-conjugated goat anti-mouse IgG, blocked with mouse IgG, stained with Alexa Fluor 488-conjugated anti-CCR7 (BioLegend) or IgG control were acquired using an LSRII flow cytometer (Becton Dickenson) with FACSDiva software (Becton Dickenson) and analyzed using Flowjo software (Tree Star).

Statistics

For experiments in which more than two groups were compared, statistical significance was determined by one-way ANOVA using a 95% confidence interval and the Tukey post-test (Prism 4.0 for Macintosh). Statistical error is expressed as s.e.m. For experiments in which two groups were compared, a t-test was used.

Results

To simultaneously analyze chemokine- and TCR-driven TEM, both of which are shear-stress-dependent, we use HDMEC monolayers overlaid with a superantigen, namely toxic shock syndrome toxin 1 (TSST-1), and analyze the morphology and location of isolated human peripheral blood EM CD4 T cells allowed to interact with the EC monolayer in a flow chamber apparatus that applies venular levels of shear stress. In this system, those EM CD4 T cells that express a TCR containing a Vβ2 gene segment (Vβ2TCR+, that can be identified by a mAb) will interact with and become activated by TSST-1 bound to MHC Class II molecules expressed on the EC. As reported previously, T cells that do not express a Vβ2-containing TCR (Vβ2TCR-) are not activated by TSST-1 and begin to transmigrate within minutes in response to TNF-induced chemokines displayed on the surface of HDMEC. (Vβ2TCR+ T cells respond similarly to Vβ2TCR-T cells in the absence of TSST-1.) These cells display polarity common to all other mammalian leukocytes, characterized by the MTOC in the trailing uropod both while crawling on the EC surface as well as during diapedesis (Figure 1A). While TNF-treated HDMEC display sufficient levels of endogenous chemokines to stimulate transmigration, TNF-treated HUVEC do not and require exogenously added chemokine. This difference provides an opportunity to test if the observed morphology is dependent upon specific chemokines. In point of fact, the same morphological patterns are iduced by SDF-1 and IP-10. (Supplemental Figure 1). However, essentially all of the Vβ2TCR+ cells activated by TSST-1 adopt a radially symmetric morphology and round up. During this process, the MTOC moves to a position in between the T cell nucleus and EC surface; this also occurs in a small percentage of Vβ2TCR-cells, as would be expected as a result of alloantigen recognition by a small subset (<1%) of EM CD4 T cells (Figure 1B,C). A proportion of the Vβ2TCR+ cells then form TEPs that penetrate and tunnel beneath the EC monolayer. The point of TEP insertion through the EC monolayer is at or very near the inter-EC junctions as shown by IF using antibodies to the intracellular or extracellular domains of VE-cadherin (Supplemental Figure 2). The MTOC in these cells is positioned at the base of, and sometimes within, the TEP (Figure 1D). Subsequently, the nucleus follows the TEP under the EC monolayer to complete transmigration. In contrast to all other migrating mammalian leukocytes, the MTOC precedes the nucleus into the TEP during the process of TEM (Figure 1E). To confirm that these morphological changes were triggered by a TCR signal rather than by superantigen independent of the TCR, we repeated these observations using FcR HDMEC displaying the anti-CD3 mAb OKT3, a widely used surrogate for polyclonal TCR signaling. T cells induced to undergo TEM in response to OKT3 bound to FcR HDMEC displayed the same pattern of MTOC dynamics as T cells responding to TSST-1 (Supplemental Figure 3).

Figure 1.

MTOC localization during chemokine- and TCR-driven TEM of EM CD4 T cells. Representative T cells are stained for the MTOC marker γ-tubulin (white) in addition to α-tubulin (red), Vβ2TCR (green) and nuclei (DAPI, blue) and analyzed by confocal microscopy. The confocal images are taken in a plane parallel to the slide surface. A. Chemokine-driven TEM (note the absence of Vβ2TCR staining) at 5 min flow. The lower panels are at a plane under the EC, and the upper panels are 2.52 µm above. The white dot in the γ-tubulin column identifies the MTOC. B. TCR-driven TEM at 5 min. The lower panels are at the apical surface of the EC, and the upper panels are 2.68 µm above. C. Graph shows % of MTOC localized between the T cell nucleus (sub-nuclear) and EC for Vβ2TCR- (VB2-) and Vβ2TCR+ (VB2+) cells. N=106 and 113 for VB2- and VB2+, respectively. D,E. 3D projection images of 12 confocal slices of merged images of cells stained as in B, viewed from the top (top panels) and side (bottom panels); the white dot indicated by the arrow is the MTOC. D is a cell after 20 min, showing the MTOC at the base of the TEP and E after 50 min flow, showing MTOC in front of the nucleus under the EC. The arrows point to the MTOC. Scale bars are 2.5 µm (A,B) and 5 µm (D,E).

Given the strikingly different MTOC dynamics between chemokine- and TCR-driven TEM, we wondered if perturbing MTOC movement would affect TEM. Recently it was shown that dynein and myosin IIA functioned together to control MTOC movement to the IS in T cells (20), potentially resolving the disparate and contradictory results obtained in many studies investigating myosin IIA alone (21). We have shown previously that inhibition of myosin IIA by blebbistatin had no effect on chemokine-driven TEM, but specifically reduced TCR-driven TEM (10). Dynein inhibition by ciliobrevin D also specifically reduced TCR-driven TEM (Figure 2A). However, despite the effects of these agents on TCR-driven TEM, early MTOC movement to the region of contact with the EC surface in our system was not affected by inhibition of myosin IIA and/or dynein (Figure 2B).

Figure 2.

Dynein and myosin IIa control TCR-driven TEM but not MTOC orientation to the T cell/EC contact region. A. TEM assays of EM CD4 T cells treated with vehicle (veh) or ciliobrevin D (CB). Graphs show %TEM of cells lacking the Vβ2TCR (VB2-, which do not interact with TSST-1 and transmigrate in response to chemokines on the EC; chemokine-driven TEM) and % TEM of cells expressing Vβ2TCR (VB2+, those that interact with TSST-1 presented by the EC; TCR-driven TEM) after 50 min flow. Graphs display combined data from 3 separate experiments, representing 300 (VB2-) and 600 (VB2+) cells each condition. *, p<0.0001. B. TEM assays of EM CD4 T cells treated with vehicle (veh), ciliobrevin D (CB), blebbistatin (bleb) or both CB and bleb after 5 minutes flow were stained for Vβ2TCR, γ-tubulin and DAPI and analyzed by confocal microscopy. Graph shows percentage of cells with the MTOC located between the T cell nucleus (sub-nuclear) and EC. N=50 from 2 experiments.

The failure of dynein and myosin IIA inhibition to block MTOC orientation indicated that other mechanisms were involved. Centripetal actin flow is driven by actin polymerization at the outer ring and depolymerization in the middle of the IS, and low concentrations of jasplakinolide that selectively inhibit actin depolymerization disrupt IS structure, whereas inhibition of myosin IIA has no effect (22). In our system, treatment of EM CD4 T cells with low concentrations of jasplakinolide reduced MTOC movement to the area between the T cell nucleus and EC, strongly inhibited TCR-driven TEM, but only modestly affected chemokine-driven TEM (Figure 3). These results suggest that the centripetal actin flow driven by depolymerization of actin, rather than dynein and myosin IIA activities, supports MTOC orientation to the contact region in EM CD4 T cells interacting with antigen presented by EC under flow and that MTOC orientation to the T cell EC contact region may be a prerequisite for TCR-driven TEM.

Figure 3.

Actin depolymerization is necessary for MTOC orientation to the T cell/EC contact region and TEM. TEM assays of EM CD4 cells treated with vehicle (veh) or jasplakinolide (jasplak). Left graph shows % MTOC in Vβ2TCR+ cells located between the T cell nucleus (sub-nuclear) and EC at 5 min flow, N=41 from 2 experiments. Middle and right graphs show %TEM of cells lacking the Vβ2TCR (VB2-, which do not interact with TSST-1 and transmigrate in response to chemokines on the EC; chemokine-driven TEM) and % TEM of cells expressing Vβ2TCR (VB2+, those that interact with TSST-1 presented by the EC; TCR-driven TEM), respectively, after 50 min flow. Data combined from 3 experiments, representing 300 cells each condition. *, p<0.0001.

We next considered what role dynein may play in the later stage of TCR-driven TEM. Dynein is a minus-end directed motor protein that transports cargo along microtubules toward the MTOC. In cytotoxic lymphocytes (CTL), perforin and granzymes are stored in specialized secretory lysosomes termed lytic granules and are transported in a minus-end directed movement along microtubules toward the MTOC upon TCR stimulation (23, 24). Once at the IS, lytic granules fuse with the cell membrane and deliver granzymes into the target cell in a perforin-dependent process. CD4 T cells also form granules, although these appear to preferentially express granzyme A rather than granzyme B or perforin (Figure 4A, (25) and data not shown). Granzyme A-containing granules colocalize with the MTOC in the uropod of T cells undergoing chemokine-driven TEM (Figure 4B) but disperse radially within minutes after TCR activation (Figure 4C) and subsequently concentrate in the vicinity of the MTOC at the base of the TEP (Figure 4D, Supplemental Figure 3). Ciliobrevin D prevents granzyme A accumulation at the MTOC (Figure 4E,F), suggesting that dynein functions to translocate granules toward the MTOC and near the point of T cell/EC contact in EM CD4 T cells during TCR-driven TEM.

Figure 4.

Dynein controls lytic granule accumulation to the MTOC. A. Lytic granules are expressed in EM CD4 cells. FACS plots showing staining of total CD4 cells with IgG controls (left) and granzyme A and CCR7 (right). Note that the granzyme A positive cells are expressed only in the CCR7-low EM population. B–E. Localization of granzyme A-containing granules during chemokine- and TCR-driven TEM. The numbers in the right hand margin indicate the number of micrometers above the plane of the lowest panel, which is underneath the monolayer in B and D, and at the apical surface of the EC in C and E. Montage in B shows Granzyme A concentrated in the uropod during chemokine-driven TEM at 5 min flow, and montages C and D show the redistribution of Granzyme A in TCR-driven TEM at 5 and 30 min flow, respectively. Note the diffuse distribution at 5 min and the accumulation around the MTOC at the stalk of the TEP at 30 min. Montage in E shows Granzyme A localization in ciliobrevin-treated cells at 30 min flow. Note the similarity to the diffuse staining of the untreated cells at 5 min (C). Scale bars are: B. 2.5 µm; C. 2.5 µm; D. 5 µm; E. 2.5 µm. F. Dynein controls accumulation of lytic granules to the MTOC in TCR-driven TEM. Confocal pictures were analyzed for the accumulation of Granzyme A near the MTOC in TCR-driven TEM assays of cells treated with vehicle (veh) or ciliobrevin D (CB) at 30 min flow. Graph shows percentage of cells with concentrated granzyme A in the vicinity of the MTOC at the stalk of the TEP, as in D, of 13 (veh) and 19 (CB) cells analyzed.

In conventional antigen presentation, cytosolic granules are exocytosed at the immune synapse where their cargo may exert their functions. Therefore, if exocytosis of granules were important for TCR-driven TEM, transmigrated cells should retain reduced levels of granzyme A. Also, since EM CD4 cells, as fractionated by CCR7 expression (CCR7 low), also contain a subpopulation of granzyme A low cells (Figure 4A), we would expect to see those cells that had not transmigrated to be granzyme A low if granzyme A were important for TCR-driven TEM. Consistent with these predictions, quantification of granzyme A at 5 min and 50 min flow show that cells at 50 min that had failed and those that had transmigrated displayed reduced numbers of granzyme A positive granules compared to the starting population (Figure 5A). To test whether granule exocytosis plays a role in TCR-driven TEP formation and TEM, we treated T cells with concanamycin A (CMA) and Bafilomycin A1 (Baf A), which disrupt granule function by inhibiting vesicular ATPases. CMA and Baf A decreased TCR-driven TEP formation and TEM, but not chemokine-driven TEM (Figure 5B). Analysis of granzyme A abundance and localization show that CMA-treated cells contained more granzyme A that accumulated at the T cell-EC interface (Figure 5C), confirming that CMA treatment prevents granule fusion and loss of granule contents that would normally occur during TCR-driven TEM. Treatment of T cells with the cell-permeable serine protease inhibitors 3,4 dichloroisocoumarin (DCI) and N-tosyl-L-lysyl-chloromethylketone (TLCK), both of which inhibit granzyme A, also reduce TCR-driven but not chemokine-driven TEM (Figure 6).

Figure 5.

Inhibition of granule exocytosis impairs TCR-driven TEP formation and TEM. A. Granzyme A expression was measured by immunofluorescence in Vβ2TCR+ cells at 5 min and 50 min flow. Cells at 50 min were divided into cells that were on the apical surface (A) or had transmigrated (TM). Graph shows individual values from 45 cells per group as well as average and SEM horizontal bars. B. TEM assays of EM CD4 cells treated with vehicle (veh), concanamycin A (CMA, upper graphs) and Bafilomycin A1 (Baf A, lower graphs) at 30 min. Graphs show % TEM and % TEP (cells that have formed a TEP but have not transmigrated) of data combined from 3 experiments, representing at least 300 cells per condition. *, p<0.0001. C. CMA blocks lytic granule release at the area of T cell-EC contact. Left graph shows total granzyme A per cell (*, p < 0.05) and right graph shows distribution of granzyme A from the top (away from EC surface) to bottom in cells treated with vehicle (veh) and CMA. Data from 17 (veh) and 18 (CMA) cells analyzed.

Figure 6.

Inhibition of granzyme activity impairs TCR-driven TEP formation and TEM. A. TEM assays of EM CD4 cells pre-treated with vehicle (veh) or DCI at 50 min flow. Left graph represents chemokine-driven TEM, middle graph TCR-driven TEM, and right graph TEP formation of the Vβ2TCR+ cells. Graphs show % TEM or % TEP of data combined from 3 experiments, representing at least 300 cells per condition. *, p<0.0001. B. TEM assays of EM CD4 cells pre-treated with vehicle (veh) or TLCK at 50 min flow. Left and right graphs represent chemokine-driven and TCR-driven TEM, respectively. Graphs show % TEM of data combined from 3 experiments, representing at least 300 cells per condition. *, p<0.0001. C. TEM assays of EM CD4 T cells treated with DCI and TLCK on FcR HDMEC loaded with OKT3. Graphs show % TEM of data combined from 2 experiments , 200 cells analyzed for each condition. *, p<0.0001.

Discussion

Confocal analysis of transmigrating EM CD4 T cells reveal that the MTOC and granules are differentially localized depending upon whether the cells are transmigrating in response to chemokine or antigen presented by the EC. Included in these observations is that, in contrast to chemokine-driven TEM (or lymphocytes moving in response to chemokines in general), the MTOC precedes the nucleus across the endothelium during TCR-driven TEM. Like the MTOC, Granzyme A-containing granules remain sequestered in the trailing uropod in cells responding to chemokine. In response to antigen, the granules disperse, then accumulate near the MTOC at the T cell/EC interface, and are reduced in number during TEM, consistent with granule exocytosis. Our study has been performed with EM CD4 T cells directly isolated from human peripheral blood. This approach minimizes alterations that develop in cell lines and our results are likely to be more applicable to human biology then rodent models, but the trade off is that genetic approaches to manipulate the T cells are not feasible and we must rely on pharmacological inhibitors to probe the role of granules and their contents on EM CD4 T cell TEM. With the caveat that all pharmacological inhibitors have non-specific side effects, we interpret the effects of CB, CMA, Baf A, DCI and TLCK to indicate that polarization and exocytosis of serine protease activity stored in granules is necessary for TCR-driven TEP formation and TEM, but not chemokine-driven TEM of EM CD4 T cells.

The substrate(s) of the serine protease activity that must be digested to permit TCR-driven TEM is unknown. Given that the granules in EM CD4 T cells lack perforin, and are therefore unlikely to enter the cytoplasm of the EC, the substrate is presumably extracellular. Extracellular substrates of granzymes described to date include several components of the extracellular matrix, as well pro-urokinase and proteinase-activated receptors receptors (26). However, since exocytosed proteolytic activity seems to be localized to the T cell/EC contact region, and TCR-driven TEM proceeds through a route requiring EC junctional proteins PECAM-1, CD99, PVR, and nectin-2, we hypothesize that the substrate(s) in question may have a connection to the lateral border recycling compartment, in which several of these junctional proteins reside and participate in TEM (15). VE-cadherin, which is localized to the EC junctions but is not a component of the lateral border recycling compartment, could also be a target, although the preservation of extracellular epitopes of this molecule detected by IF microscopy during TCR-driven TEM makes this seem less likely. The effect of TLCK indicates that the substrate contains Arg/Lys at the P1 position (the amino acid N-terminal from the cleaved peptide bond), acted on by the tryptase family of proteases that include Granzyme A. Whether Granzyme A is the acting protease in this process could not be definitively determined due to the unavailability of more specific inhibitors and the inability to use siRNA knockdown in our system; the phenotype of the freshly isolated EM CD4 T cells changes during the time in culture needed for knockdown. Nevertheless, by using Granzyme A as a marker of granule movement during TEM, the cumulative evidence supports a model where granzyme exocytosed at the T cell/EC contact region is necessary for TCR-driven TEM, but is not involved in chemokine-driven TEM.

These results also provide an explanation for the delayed kinetics of TCR-driven TEM compared to chemokine-driven TEM, namely that more time is required for the granules to accumulate near the MTOC and undergo exocytosis. Also, since neither central memory nor naïve CD4 T cells (both CCR7 high) form Granzyme A-containing granules, the role of these organelles may also contribute to the inability of these T cell populations to transmigrate in response to antigen presented by EC.

Granzyme A-containing granules are directed to the site of T cell/EC contact by the motor protein dynein, which moves cargo along microtubules towards the MTOC. In contrast to other T cell-APC systems, dynein and myosin IIA are not responsible for the orientation of the MTOC to the T cell/EC interface. Rather, centripetal actin flow, driven by the polymerization of actin at the periphery and depolymerization at the center, seems to play a role in MTOC orientation to the T cell/EC contact region. This, then, is similar to the movement of signaling microclusters in an IS established between motile T cells and conventional APCs (a dynamic observed in vivo), in which actin depolymerization establishes an interior F-actin-poor region toward which TCRs and cSMACs move (22).

Indeed, the orientation of the MTOC to the T cell/EC contact region is another example in a series of observations that correlates the interaction of an EM CD4 T cell with EC during TCR-driven TEM with that of an IS formed between T cells and conventional APCs. And like the IS formed between a T helper cell and B cell, or CTL and target, in which the orientation of the MTOC is necessary for polarized cytokine secretion that drives antibody production and lytic granule delivery that kills cells, respectively, the orientation of the MTOC to the T cell/EC interface has a critical functional outcome, namely traversal across the endothelium of EM CD4 T cells that have been activated by antigen presented by the EC. The cytokines produced by these activated EM CD4 T cells may then act on the local environment to facilitate the recruitment of a large number of antigen non-specific cells to that site.

In summary, antigen recognition by circulating EM CD4 T cells, increasingly appreciated as a key event in the initiation of memory responses at peripheral sites, induces TEM in a manner that closely resembles IS formation between a T cell and a conventional APC. In this case, MTOC orientation to the region of contact between the T cell and EC allows dynein-mediated delivery of granzyme-containing granules to and exocytosis at the contact region. Inhibition of this process offers new targets for therapeutic intervention.

Abbreviations List

Baf A

Bafilomycin A1

CB

ciliobrevin D

CMA

concanamycin A

DCI

3,4-dichloroisocoumarin

EC

endothelial cell

EM

effector memory

IS

immune synapse

MTOC

microtubule organizing center

TEM

transendothelial migration

TEP

transendothelial protrusion

TLCK

N-tosyl-L-lysine chloromethyl ketone