Ultrastructural alteration of cytolytic T lymphocytes following their interaction with target cells. II. Morphogenesis of secretory granules and intracellular vacuoles.

“Secretory” granules, crystal-like structures, lysosema-like granules, ergastoplasmatic lamellae, electron-opaque ribosomes, and lipid vacuoles were detected in the cytoplasm of cytolytic T lymphocytes. After 30 to 60 min interaction with a target cell the fusion of “secretory” granules and crystal-like structures with lipids, mitochondrion degeneration, hypertrophy and reorientation of the Golgi apparatus toward the contact zone with a target cell were observed.     

Quantitative ultrastructure of cytolytic lymphocytes mediating allograft rejection in the mouse. I. Cellular alterations in T lymphocytes during specific target cell lysis

A quantitative ultrastructural analysis of cytolytic T lymphocytes (CTL) is presented which allows both the distinction of these cells from normal T lymphocytes and permits the demonstration of ultrastructural alterations of putative CTL following interaction with target cells (TC). Alloreactive CTL were generated in C57BL/10 mice receiving intraperitoneal fibroblastic allografts and target-binding splenic lymphocytes (TBSL) were concentrated by specific immunoadsorption on fibroblast monolayers. TBSL were subjected to ultrastructural quantification either at the onset of TC interaction or following 30 or 60 min incubation at 37 degrees C. By means of simple stereological relationships it was shown that, in comparison with normal, non-cytolytic splenic T lymphocytes, TBSL were slightly larger cells, displaying around 60% more cytoplasm, a similarly-sized nucleus and approximately triple the volume of Golgi apparatus. During the first 30 min of interaction with TC, the target binding surface of the TBSL plasma membrane decreased in area. This change was accompanied by a polarization of the TBSL towards the target. Incubation of lymphocytes with TC for a further 30 min resulted in a general polarization of lymphocytic cellular constituents away from the TC. These results were only attainable by objective quantitative analysis and are discussed in relation to possible mechanisms of CTL-mediated lysis.     

Reorientation of the microtubule-organizing center and the Golgi apparatus in cloned cytotoxic lymphocytes triggered by binding to lysable target cells

By immunofluorescence observations with cell couples of cloned murine cytotoxic T lymphocytes (CTL) and target cells, evidence is presented for a rapid reorientation of the microtubule-organizing center (MTOC) and the Golgi apparatus (GA) in the effector cell (but not in the target cell) toward the contact area with the target. The reorientation of the MTOC/GA and the cytotoxic activity of the CTL were inhibited reversibly by nocodazole, a microtubule-disrupting agent. In lectin-formed cell couples of CTL and neuraminidase-treated target cells, the MTOC in essentially all of the CTL was oriented toward the effector-target contact area of a lysable target cell, but was left randomly oriented with a nonlysable target cell. A similar random orientation of the effector-MTOC was also observed in cell couples of cloned natural killer cells and nonlysable targets. These findings indicate that the repositioning of the MTOC and the GA, which is shared by CTL and natural killer cells, is an essential and early event in the onset of the cytolytic mechanism. It is suggested that this reorientation serves the purpose of directing to the bound target cell secretory vesicles derived from the GA that contain cytotoxic substances.     

Exocytosis of cytolytic T-lymphocytes studied by cryofracture and stereophotogrammetry

Cryofracture, stereophotogrammetry and three-dimensional reconstruction were used to study the membranes of cytolytic T lymphocytes (CTL) and target cells (TC) conjugates. Circular bulges free of intramembrane granules and ring-shaped structures, analogous to the sites of vacuole release in secretory cells, appear on the surface of the lymphocyte plasmalemma after 30 to 60 minutes of interaction with TC. Polymorphic vacuoles 70 to 140 nm in diameter are observable in the contact space between lymphocytes and TC. Part of the vacuoles are found on the surface or near the CTL plasmalemma. The data confirm an assumption about the secretory mechanism of the cytolytic effect of T killers.     

Ultrastructure of the conjugates of cytolytic T-lymphocytes and target cells]

Conjugates of target cells and of cytological T-lymphocytes obtained on the 11th day after alloimmunization were investigated. The conjugates formed small and medium lymphocytes; mature secretory granules, crystal-like structures and lipids were revealed in their cytoplasm. The lymphocyte is spherical, the area of contact with the target cell does not exceed 5 to 15%. Cytolysis of target cells is observed after 30 to 60 minutes of incubation. The lymphocyte becomes flattened, its nucleus acquires an oval form, and the area of contact with the target cell increases considerably. At the same time hypertrophy and change of orientation of Golgi's complex to the area of contact with the target cell, coalescence of the secretory granules with lipids and crystal-like structures, the appearance of immature secretory granules and vacuolar degeneration of mitochondria are demonstrated. The lymphocyte membrane becomes "desquamated"; structures connected with it, named "membranosomes" are described. It is suggested that the secretory processes in the cytoplasm of cytolytic T-lymphocytes are activated in their interaction with target cells.     

Motility and ultrastructure of large granular lymphocytes on lipid bilayers reconstituted with adhesion receptors LFA-1, ICAM-1, and two isoforms of LFA-3

Large granular lymphocytes, mediators of NK activity, bind to other cells using both the LFA (lymphocyte function-associated)-1-ICAM and the CD2-LFA-3 adhesion pathways. Here we have studied the motility and ultrastructure of large granule lymphocyte (LGL) on lipid bilayers containing purified LFA-1, ICAM-1, and the transmembrane and glycophosphatidylinositol isoforms of LFA-3. LGLs moved at 8 microns/min on ICAM-1 but poorly (less than 1 microns/min) on its receptor pair LFA-1. TM-LFA-3 promoted locomotion at a rate close to ICAM-1, whereas the cells were less motile on GPI-LFA-3. The difference in the rates of locomotion on the two isoforms of LFA-3 is presumably attributable to their difference in anchoring and lateral mobility in the bilayer. In spite of the variation in motility the ultrastructure of the adhering cells was similar on all four ligands. LGLs contacted the membrane variably, i.e., cells adhering only in a few small areas or in larger areas were detected on each ligand. The relative percentage of the plasma membrane facing the lipid bilayer was greatest on ICAM-1 and least on the transmembrane isoform of LFA-3, demonstrating no correlation with motility. The ratio of adjacent plasma membrane to lipid bilayer was virtually constant for all four ligands. Activation of the LGLs with a combination of CD2 mAb T11(2) and T11(3) (T11(2/3) mAb) reduced the movement on ICAM-1 and virtually immobilized the cells on the other bilayers. In the presence of T11(2/3) mAb, the area of cell membrane attaching to bilayers containing ICAM-1 and GPI-LFA-3 was decreased and the percentage of plasma membrane facing other cells was increased. No preferential orientation of the Golgi apparatus or degranulation was detected in the absence or presence of T11(2/3) mAb, but a significantly lower percentage of LGLs on ICAM-1 contained a profile of the Golgi apparatus after exposure to T11(2/3) mAb. The results demonstrate that the motility of LGLs depends on the type of receptor in the opposing bilayer, the receptor mobility in the bilayer, and the activation of the cells. The ultrastructure of LGLs binding to any of the adhesion molecules does not have the characteristics of LGLs in cytolytic contact with target cells, suggesting that the mediation of an attack on a target requires more complex stimulus than any one of the single adhesion proteins tested here.     

Retractile processes in T lymphocyte orientation on a stimulatory substrate: morphology and dynamics

T cells of the immune system target infected and tumor cells in crowded tissues with high precision by coming into direct contact with the intended target and orienting the intracellular Golgi apparatus and the associated organelles to the area of the cell-cell contact. The mechanism of this orientation remains largely unknown. To further elucidate it we used three-dimensional microscopy of living T cells presented with an artificial substrate mimicking the target cell surface. The data indicate that long, finger-like processes emanate from the T cell surface next to the intracellular Golgi apparatus. These processes come in contact with the substrate and retract. The retraction accompanies the reorientation of the T cell body which brings the Golgi apparatus closer to the stimulatory substrate. Numerical modeling indicates that considering the forces involved the retraction of a process attached with one end to the cell body near the Golgi apparatus and with the other end to the substrate can bring the Golgi apparatus to the substrate by moving the entire cell body. The dynamic scenarios that are predicted by the quantitative model explain features of the reorientation movements that we measured but could not explain previously. We propose that retraction of the surface processes is a force-generating mechanism contributing to the functional orientation of T lymphocytes.     

Contact regions of cytotoxic T lymphocyte-target cell conjugates.

The binding of cytotoxic T lymphocytes to target cells was studied by ultrastructural and tracer techniques. It was found that binding was achieved through interaction of the microvilli of both cells and that only a relatively small proportion of the cell surface was involved. Short points of contact, averaging 1500 Å in length, were the main form of junction. Periodic substructures were observed in some of the contact points. The transfer of cytoplasmic content from effector to target cell and vice versa was investigated, but no fluorescein or ⁵¹Cr-labeled components were transferred during the interaction. Examination of cell organelle localization during the interaction revealed that microfilaments were the only cellular components which localized at the contact area; the well-developed Golgi apparatus of the cytotoxic lymphocytes was randomly distributed.     

The role of T cells in anti-herpes simplex virus immunity. I. Induction of antigen-specific cytotoxic T lymphocytes.

Mice infected with herpes simplex virus develop little or no cytotoxic T lymphocyte (CTL) response. However, in lymph nodes (LN's) draining a local site infected with HSV, antigen-specific CTL precursors are sensitized, which upon transfer to in vitro culture conditions develop within 72 hr into effective CTL. The in vivo blockade of CTL differentiation can be overcome by cyclophosphamide, suggesting that a cyclophosphamide-sensitive mechanism blocks the in vivo generation of HSV-immune CTL. The cytolytic activity of HSV-immune CTL is H-2 restricted and antigen specific. Thus CTL sensitized toward HSV type 1 discriminate between syngeneic targets infected with either the immunologic HSV variant type 1 or type 2 (and vice versa). H-2-matched target cells exposed for 30 min to infectious HSV are lysed within 60 min of contact with CTL. Since HSV replication is believed to require more than 4 to 5 hr, the data suggest that either the expression of HSV-dependent "early proteins" takes place within 30 to 90 min or cell membrane-integrated HSV virion represents the target antigen of CTL.     

Staphylococcal enterotoxins direct and trigger CTL killing of autologous HLA-DR+ mononuclear leukocytes and freshly prepared leukemia cells

Attempts have been made to induce cytolytic T cells to kill target cells that do not express the appropriate target molecules by crosslinking the T cells and the target cells in various ways. One successful strategy has been to use heteroconjugates or bispecific monoclonal antibodies reacting with T cell molecules with activating properties (e.g., mab directed to CD3/TCR) and target cell surface antigens. In this report we show that Staphylococcal enterotoxins (SE) direct human T lymphocytes to execute cytotoxicity toward MHC class II-expressing Raji cells, but not against MHC class II-deficient Raji mutant RJ 2.2.5. Both HLA-DR+ and HLA-DR- effector T lymphocytes are effective in the killing of Raji cells coated with SE. The Staphylococcal enterotoxin-dependent cell-mediated cytotoxicity (SDCC) is a rapid T lymphocyte-mediated cytolytic mechanism killing the targets within an hour of incubation. HLA-DR+ target cells are sensitized to be killed within minutes of incubation with picomolar concentrations of SE. SE-sensitized Raji cells remain targets for SDCC after overnight culture at 37 degrees C, demonstrating that the sensitive state is relatively stable. SEA- and SEB-selective cytolytic T cell lines were established to illustrate the clonal variability of SDCC effectors with respect to SE specificity. We also demonstrate that autologous monocytes and activated T lymphocytes as well as B lymphocytes and freshly prepared HLA-DR+ leukemic cells are excellent targets in SDCC.     

Ultrastructural changes during lysis of L929 target cells by class II-restricted influenza virus-specific murine cytotoxic T-lymphocyte clones

Lysis of virus-infected L929 target cells transfected with the H-2 class II IAk gene by class II-restricted influenza virus-specific murine cytotoxic T lymphocyte (CTL) clones was studied by electron microscopy and compared with lysis of L929 cells by class I-restricted CTL clones. T lymphocytes predominantly approached the basal surface of target cells grown on a plastic dish and also approached uninfected L929 target cells, although virus maturation exhibited no polarity with respect to the cell surface site. After incubation for 30 min, the target cell nuclei began to change: chromatin became irregularly redistributed and aggregated, and the nuclei appeared swollen. Later, electron-dense and -light areas of nuclei became segregated, and the cytoplasm became disorganized with many vacuoles. The ultrastructural changes of target cells during lysis by class I- and class II-restricted CTL clones appeared to be similar. These findings and other cytotoxicity data of class I and class II CTLs are discussed.     

Human monocyte-mediated lysis of antibody-coated tumor cells: the role of the cytoskeleton in monocytes

Human monocytes (M phi) show high cytolytic activity towards antibody-coated tumor cells (AbK562). In this report, the relationship between the cytoskeleton in the M phi and the M phi cytolytic activity has been investigated. The actin filament inhibitors cytochalasin B and dihydrocytochalasin B (H2CB) both reduced M phi-mediated lysis of AbK562 cells by approximately 50% at a concentration of 1 microM. This concentration of H2CB did not inhibit the number of target cells bound to M phi. Dihydrocytochalasin B did not inhibit the M phi ability to release cytotoxic protein factors, suggesting that H2CB does not inhibit lysis by inhibiting release of cytotoxic protein factors. Immunofluorescence microscopy showed a rapid accumulation of actin filaments towards the contact area in more than 80% of the examined M phi-AbK562 conjugates. Exposure to H2CB did not prevent this accumulation, but caused aggregation of the accumulated actin filaments in the contact area with the target cell. Accumulation of actin filaments did not occur toward tumor cells not coated with antibodies. Scanning and thin section electron microscopy demonstrated large M phi pseudopodia directed toward the AbK562 cells, with close apposition of the effector and target cell membranes with interdigitations. The formation of the M phi pseudopodia was inhibited by exposure to H2CB. These observations indicate that M phi membrane motility toward AbK562 cells is closely related to M phi-mediated lysis of AbK562 cells. Immunofluorescence microscopy of the microtubule-organizing center (MTOC) and the Golgi apparatus revealed that both the MTOC and the Golgi apparatus in M phi reoriented towards the bound AbK562 cells in approximately 45% of the examined M phi-AbK562 conjugates. The microtubule-depolymerizing drugs colchicine and vinblastine did not inhibit M phi-mediated lysis of AbK562 cells at concentrations which disrupted the microtubule arrays in the M phi. The carboxylic ionophore monensin, which blocks Golgi-derived secretion, inhibited M phi-mediated lysis of AbK562 to a lesser extent as compared to H2CB. These results suggest that microtubule functions are of less importance in M phi-mediated lysis of AbK562 cells as compared to actin filament functions. However, the MTOC and the Golgi apparatus could participate in M phi-mediated lysis of AbK562 cells by mechanisms related to secretion of cytotoxic molecules.     

Endocytosis of an antibody ricin A-chain conjugate (immuno-A-toxin) adsorbed on colloidal gold. Effects of ammonium chloride and monensin

An immunotoxin (IT) formed by a specific antibody coupled to the ricin A chain was adsorbed on colloidal gold particles (IT-Au). Binding and internalization of IT-Au in human lymphoblastic CEM cells were studied using electron microscopy. IT-Au showed specific cytotoxic activity toward the target cells. After 1 h at 4 degrees C, IT-Au were linked diffusely to the plasma membrane with 45% of the particles regrouped in clusters. Upon transfer to 37 degrees C, the particles carrying the ligand were regrouped more frequently and internalized into the cell by endocytosis through smooth microinvaginations or coated pits of the plasma membrane. After 15 min, IT-Au was observed in endocytic vacuoles, or receptosomes, in tubular structure near the Golgi apparatus and in lysosomes. Entry of IT-Au into lysosomes was rapid (around 50% of intracellular IT-Au particles after 30 min). NH4Cl or monensin, well-known potentiators of immunotoxin activity, when present in incubation medium, altered neither the processes nor the rate of IT-Au endocytosis. In the presence of either of these substances, IT-Au accumulated in the normal or often enlarged endocytic vacuoles, and entry into the lysosomes was slowed down (50% of particles after 2 h 15 min). We conclude that this intense slowing-down in the speed of IT-Au transportation into lysosomes and the functional modifications of these organelles help to explain the increased efficacy of immunotoxins in the presence of potentiators.     

Delta 9-tetrahydrocannabinol decreases cytotoxic T lymphocyte activity to herpes simplex virus type 1-infected cells.

The purpose of this study was to examine the effect of delta 9-tetrahydrocannabinol (delta 9-THC), the major psychoactive component of marijuana, on T lymphocyte functional competence against herpes simplex virus Type 1 (HSV1) infection. Spleen cells from C3H/HeJ (H-2k) mice primed with HSV1 and exposed to delta 9-THC were examined for anti-HSV1 cytolytic T lymphocyte (CTL) activity. Flow cytometry was used to determine whether delta 9-THC altered T cytotoxic (Lyt-2+) and T helper (L3T4+) lymphocyte numbers or cell ratios. Nomarski optics microscopy was used to determine whether effector lymphocytes from drug-treated mice were able to bind to virally infected L929 (H-2k) target cells. Cytotoxicity assays demonstrated that CTL from mice exposed to delta 9-THC were deficient in anti-HSV1 cytolytic activity. delta 9-THC in vivo treatment had little effect on the number of T lymphocytes expressing the Lyt-2 or L3T4 antigens. Nomarski optics microscopy revealed that the CTL from the drug-treated mice were able to bind specifically to the HSV1-infected targets. However, delta 9-THC in vivo exposure affected CTL cytoplasmic polarization toward the virus-infected target cell. CTL granule reorientation toward the effector cell-target cell interface following cell conjugation occurred at a lower frequency in co-cultures containing CTL from drug-treated mice. These results suggest that delta 9-THC elicits dysfunction in CTL by altering effector cell-target cell postconjugation events.     

Characteristics of the secretory apparatus of memory T-cells

It has been shown that by day 7--8 of cultivation, large lymphocytes and lymphoblasts disappear, the DNA synthesis and cytolytic activity decrease in lymphocyte suspension enriched with a fraction of lymphoblasts obtained on the 5th day of mixed lymphocyte culture. The cytoplasm of medium-size and small lymphocytes adsorbed on the surface of target-cells manifests no signs of secretion: there are no tubular structures or rough reticulum, the Golgi complex is underdeveloped. It is suggested that there is a relationship between the secretory apparatus and cytolytic activity of T-killers.     

Lectin-dependent cytotoxicity of human peripheral blood lymphocytes

The evaluative technique of lymphocyte cytolytic activity in human peripheral blood has been designed. The target cells, lectin (Con A) concentration and incubation time for measuring cytolytic activity of lymphocytes pre-purified from adherent cells have been selected. The mean values of lectin-dependent cytotoxicity in peripheral blood of 50 healthy donors are presented.     

Concanavalin A-Induced Morphological Changes and Its Binding Sites in Starfish Follicle Cells as Revealed by Electron Microscopy

Concanavalin A (Con A) stimulates the production in starfish follicle cells of 1-methyladenine, a hormone which induces oocyte maturation. We have therefore investigated Con A-induced morphological changes and Con A-binding sites in the follicle cell using native Con A and horseradish peroxidase- or ferritin-labeled Con A (HRP-Con A, Fer-Con A). After isolated follicle cells were incubated with Con A (1 mg/ml), vacuoles, the Golgi complex and multivesicular body-like organelles (MVBs) became prominent in most of the cells. After follicle cells were prefixed and then incubated with Fer-Con A for 60 min, tagged ferritin was diffusely and randomly distributed as single or small clustered particles on the cell surface. The incubation of isolated follicle cells with Fer-Con A for 10 min before fixation resulted in numerous ferritin particles localized along the internalized membrane, and also in vacuoles, MVBs and small lysosome-like structures. After 60 min incubation with Fer-Con A, ferritin was further located in large lysosome-like structures and in vesicles near and in the Golgi area as well as in the organelles described above. HRP-Con A binding sites were also observed in vacuoles and MVBs of the intact cells.
These results suggest that Con A binds at first to the cell surface and causes rapid internalization and that membrane-bound Con A is easily endocytosed into vacuoles, MVBs and lysosome-like structures, and is later incorporated in some vesicles in the Golgi area.     

Reorientation and fusion of cytotoxic T lymphocyte granules after interaction with target cells as determined by high resolution cinemicrography

The interaction of murine cytotoxic T lymphocyte (CTL) clones with human lymphoblastoid target cells was studied in thin preparations by using high resolution cinemicrography. CTL not bound to target cells were morphologically polar, possessing a broad leading edge containing the nucleus, and a tapered tail containing a large number of granules. The CTL were observed to move by the extension of pseudopods from the leading edge. Initial contact with a target cell was made via the leading edge of the CTL. If the human target cell expressed the appropriate HLA antigen, distinct morphologic changes occurred in the CTL as early as 2 min after initial contact. The CTL rounded up, and the nucleus moved from a position adjacent to the zone of contact to be replaced by the cytoplasmic granules. Redistribution of the granules was completed as early as 10 min after initial contact was made. These morphologic changes did not occur when the CTL made contact with other CTL, or with target cells that did not express the appropriate HLA antigens. In studies that make use of Nomarski optics, an apparent fusion of CTL cytoplasmic granules with the membrane in the vicinity of the target cell contact area was observed 4 min after binding, and before granule reorientation was complete. These data provide direct evidence for the occurrence of both reorientation of the cytoplasmic contents and granule fusion in CTL with a time course similar to that of administration of the lethal hit.     

The cytotoxic activity of human natural killer cells requires an intact secretory apparatus

We have analyzed the effect of various inhibitors of cellular secretion and motility on the cytolytic activity of human natural killer (NK) cells. As effector cells we used highly purified peripheral blood lymphocytes consisting of 75–85% large granular lymphocytes (LGL) that have previously been shown to be responsible for the NK activity in man. Treatment of the effector cells with a carboxylic ionophore monensin inhibited irreversibly the NK-cell-mediated killing. This drug is known to interrupt the vesicular traffic of Golgi-derived vesicles and thus the results strongly suggested that secretory processes are required in the cytolytic activity of human NK cells. In the monensin-treated effector cells large amounts of glycoprotein accumulated in the Golgi area within 24 hr of incubation. The lytic activity did not require intact microtubules since effector cells in which vinblastine-induced tubulin-containing paracrystals were demonstrated still mediated normal NK activity. Energy was required in the human NK-cell-mediated cytolysis. The lethal hit stage of the cytolytic activity was preceded by formation of intimate contacts between effector and target cells and required active cell movement and divalent cations.     

Identification of a novel mechanism for LFA-1 organization during NK cytolytic response

The elimination of transformed and viral infected cells by natural killer (NK) cells requires a specialized junction between NK and target cells, denominated immunological synapse (IS). After initial recognition, the IS enables the directed secretion of lytic granules content into the susceptible target cell. The lymphocyte function-associated antigen (LFA)-1 regulates NK effector function by enabling NK-IS assembly and maturation. The pathways underlying LFA-1 accumulation at the IS in NK cells remained uncharacterized. A kinase anchoring protein 350 (AKAP350) is a centrosome/Golgi-associated protein, which, in T cells, participates in LFA-1 activation by mechanisms that have not been elucidated. We first evaluated AKAP350 participation in NK cytolytic activity. Our results showed that the decrease in AKAP350 levels by RNA interference (AKAP350KD) inhibited NK-YTS cytolytic activity, without affecting conjugate formation. The impairment of NK effector function in AKAP350KD cells correlated with decreased LFA-1 clustering and defective IS maturation. AKAP350KD cells that were exclusively activated via LFA-1 showed impaired LFA-1 organization and deficient lytic granule translocation as well. In NK AKAP350KD cells, activation signaling through Vav1 was preserved up to 10 min of interaction with target cells, but significantly decreased afterwards. Experiments in YTS and in ex vivo NK cells identified an intracellular pool of LFA-1, which partially associated with the Golgi apparatus and, upon NK activation, redistributed to the IS in an AKAP350-dependent manner. The analysis of Golgi organization indicated that the decrease in AKAP350 expression led to the disruption of the Golgi integrity in NK cells. Alteration of Golgi function by BFA treatment or AKAP350 delocalization from this organelle also led to impaired LFA-1 localization at the IS. Therefore, this study characterizes AKAP350 participation in the modulation of NK effector function, revealing the existence of a Golgi-dependent trafficking pathway for LFA-1, which is relevant for LFA-1 organization at NK-lytic IS.