FD‐891, a structural analogue of concanamycin A that does not affect vacuolar acidification or perforin activity, yet potently prevents cytotoxic T lymphocyte‐mediated cytotoxicity through the blockage of conjugate formation

Cytotoxic T lymphocytes (CTL) have a myriad of lethal weapons for killing target cells, such as virus-infected and transformed cells, and utilize two distinct killing pathways, one of which depends on perforin and the other which depends on Fas ligand (FasL). These two cytotoxic pathways play an essential role in the maintenance of tissue homeostasis. CTL-mediated cytotoxicity, however, gives rise to undesirable tissue destruction, particularly in graft-versus-host disease and fulminant hepatitis. Therefore, low-molecular-weight compounds that modulate CTL effector function are desirable as potential clinical drugs, and are also useful tools for studying biochemical reactions in CTL-mediated cytotoxicity. In the course of our extensive screening, we have identified several agents that markedly inhibit perforin and/or FasL-dependent pathways, and have further clarified the molecular mechanisms of their actions in CTL-mediated cytotoxicity. 1–4 Concanamycin A (CMA, Fig. 1) belongs to the group of 18-membered macrolides, and has been shown to be a specific inhibitor of the vacuolar type H+-ATPase. 5,6 CMA neutralizes the pH of acidic organelles such as lysosomes and Golgi apparatus, which results in the perturbation of various functions of these organelles. 5,7 Lytic granules are acidic compartments present in CTL and natural killer (NK) cells, and contain various effector molecules such as perforin and granzymes. CMA raises the pH of lytic granules towards neutral pH, 8 and eventually induces the degradation and inactivation of perforin. 9,10 CMA thereby completely blocks the perforin-dependent killing pathway in CTL-mediated cytotoxicity. 2 However, the FasL-dependent killing pathway is not affected by CMA. 2 Hence, these findings demonstrate that CMA is a powerful tool for use in clarifying the contribution of these two distinct cytolytic pathways. Structures of FD-891 and concanamycin A (CMA). FD-891 ( Fig. 1) was originally isolated from the fermentation broth of Streptomyces graminofaciens A-8890, and was shown to have antitumor activity in vitro. 11,12 Like CMA, FD-891 belongs to the group of 18-membered macrolides, and is structurally related to CMA. The structural similarity between CMA and FD-891 prompted us to investigate the effects of FD-891 on CTL-mediated cytotoxicity. Unlike CMA, FD-891 profoundly blocked both perforin- and FasL-dependent cytotoxicity. In this work, we have studied the mechanism of action of FD-891 on CTL-mediated killing pathways. The H-2d-specific CD8+ CTL clone OE4 13 and the keyhole limpet haemocyanin (KLH)-specific I-Ed-restricted CD4+ CTL clone BK-1 14 were maintained in RPMI-1640 medium supplemented with 10% (vol/vol) fetal calf serum (FCS; CSL Ltd, Victoria, Australia), 50 µm 2-mercaptoethanol, 50 µg/ml of kanamycin, 8 µg/ml of tylosin tartrate and 5% (vol/vol) rat spleen cell-conditioned medium (culture supernatant of rat spleen cells stimulated with 5 µg/ml of concanavalin A for 24 hr). OE4 cells were stimulated with mitomycin C-treated spleen cells from BALB/c mice every 3 weeks. BK-1 cells were stimulated with 10 µg/ml of KLH + mitomycin C-treated spleen cells from BALB/c mice every 2 weeks. Mastocytoma P815 and B lymphoma (A20.2J and A20.HL) were maintained in RPMI-1640 medium containing 10% (vol/vol) FCS. For use as a target for BK-1 cells, A20.HL cells were pulsed with 10 µg/ml of KLH for 16 hr prior to labelling. FD-891 and CMA were produced by the same organism and prepared as described previously. 11 The experiments were performed in triplicate. 3-(2,4-Dinitroanilino)-3′-amino-N-methyldipropylamine (DAMP; Oxford Biomedical Research Inc., Oxford, MI) was used for the visualization of acidic organelles. DAMP staining was performed as described previously. 15 Cells were lysed with 1% Triton-X-100, 50 m m Tris-HCl (pH 7·5), 1 m m dithiothreitol (DTT), 0·5 m m EDTA and protease inhibitor mixture (1 m m phenylmethylsulphonylfluoride, 5 m m benzamidine, 100 µm antipain, 100 µm leupeptin), and then centrifuged (for 5 min at 10 000 g) to remove insoluble materials. Postnuclear lysates (20 µg of protein) were separated by 10% sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS–PAGE) and transferred onto nitrocellulose membrane filters. The membrane filters were blotted with rat anti-mouse perforin (P1-8), 16 followed by horseradish peroxidase-conjugated anti-rat immunoglobulin G (IgG) (Amersham) and then developed using the enhanced chemiluminescence (ECL) reagent (Amersham). OE4 cells (1 × 107) were washed twice with phosphate-buffered saline (PBS) and resuspended in 250 m m sucrose, 1 m m HEPES-NaOH (pH 7·4), 1 m m EGTA (pH 7·4). Cells were disrupted through a nitrogen cavitation (450 p.s.i., 4°, 20 min) and centrifuged (for 5 min at 800 g). Resultant postnuclear lysates were layered onto 8 ml of 48% (vol/vol) Percoll gradient containing 250 m m sucrose, 1 m m HEPES-NaOH (pH 7·4), 1 m m EGTA (pH 7·4), and then centrifuged at 40 000 g at 4° for 25 min. Four-hundred microlitres of the fractions were collected from the top of the gradient. Granzyme A (N-α-benzyloxycarbonyl- l-lysine thiobenzylester [BLT] esterase) activity was used to identify the fractions containing lytic granules. Aliquots of the fractions were incubated with 200 µm of BLT (Calbiochem, San Diego, CA) and 200 µm of 5,5′-dithio-bis-(2-nitrobenzoic acid) in PBS at room temperature, and absorbance (A) at 415 nm was measured. Aliquots of the fractions were incubated with 200 µl of sheep red blood cells (8 × 107 cells/ml) in Hanks’ balanced salt solution, containing 1% bovine serum albumin and 4 m m calcium chloride, at 37° for 20 min in round-bottomed microtitre plates. After centrifugation (for 5 min at 700 g), supernatants were removed and the A415 value measured. Microtitre plates were coated with 10 µg/ml of anti-mouse CD3 (145-2C11) for 1 hr and then washed twice with PBS. OE4 cells (1 × 106/ml) were preincubated with FD-891 for 2 hr, and then transferred into anti-CD3-coated plates (100 µl/well). The plates were centrifuged (for 3 min at 300 g) and then the cells were incubated for the time-periods indicated. Aliquots of culture supernatants were removed and then measured for BLT esterase activity. For cell attachment, culture supernatants were removed and then 100 µl of 0·2% crystal violet in methanol was carefully added to each well and stained for 20 min. The plates were washed extensively with water and the dye was extracted using methanol. The A595 value was measured. Conjugate formation was performed essentially according to the method described previously. 17 OE4 and A20.2J cells were preincubated with or without FD-891 for 2 hr, and further incubated with 0·25 µm of calcein-AM (Molecular Probes, Eugene, OR) or 62·5 µg/ml of hydroethidine (Polysciences, Warrington, PA), respectively, for 30 min on ice. The stained cells were washed twice with medium to remove unincorporated fluorescent probes, and resuspended in medium. An equal number of cells (5 × 105 cells, 0·25 ml) were mixed in a single Eppendorf tube, centrifuged (for 3 min at 300 g) and incubated at 25° for 30 min. At the end of the culture period, the tubes were placed on ice. The resultant cell conjugates were resuspended carefully by pipetting, and immediately analyzed using a fluorescence-activated cell sorter (FACScalibur flow cytometer; Becton-Dickinson, Mountain View, CA). OE4 cells (5 × 105) were stained with the primary antibodies for 60 min and then stained with fluorescein isothiocyanate (FITC)-labelled secondary antibodies for 60 min. Stained cells were analysed using the FACScalibur. The primary antibodies used were: 145-2C11 (hamster anti-mouse CD3ε), M17/4.2 (rat anti-mouse lymphocyte function-associated antigen-1α[LFA-1α][CD11a]), 21-12 (mouse anti-Thy-1.2), H57-597 (hamster anti-mouse T-cell receptor [TCR]; Cedarlane, Ontario, Canada) and E23.29 (rat anti-mouse CD8α; Seikagaku Co., Tokyo, Japan). FITC-labelled secondary antibodies were purchased from commercial suppliers. The inhibitory effects of FD-891 on CTL-mediated cytotoxicity were investigated in comparison with CMA. Three different effector/target combinations were employed to examine the effect of FD-891 on either the perforin- or the FasL-dependent killing pathway. As our clone of P815 expressed only a marginal level of Fas, 2 P815 cell lysis by the CD8+ CTL clone OE4 was mediated mainly through the perforin-dependent pathway, which was almost completely inhibited by CMA ( Fig. 2a). In contrast, A20.2J cells are Fas positive and sensitive to Fas-mediated apoptosis, and killed by OE4 cells through both perforin- and FasL-dependent pathways. As shown in Fig. 2(b), CMA inhibited the cytolysis of A20.2J cells by ≈ 50%, indicating that the contributions of perforin- and FasL-dependent pathways are comparable. In both cases, FD-891 completely blocked the cytolysis of P815 and A20.2J cells by OE4 cells. The CD4+ CTL clone BK-1 kills target cells exclusively through the FasL-dependent pathway, which was totally resistant to CMA ( Fig. 2c). FD-891 blocked the FasL-dependent lytic pathway of BK-1 cells ( Fig. 2c). In contrast to CMA, these data demonstrate that FD-891 inhibits both perforin- and FasL-dependent CTL-mediated cytotoxicity. FD-891 blocks perforin- and Fas ligand (FasL)-dependent cytolytic activity mediated by CD8+ and CD4+ cytotoxic T lymphocyte (CTL) clones. (a) (b) CD8+ CTL clone OE4 was preincubated with different concentrations of concanamycin A (CMA) or FD-891 for 2 hr, and then incubated with 51Cr-labelled P815 (a) or A20.2J cells (b) for 4 hr in the presence of the inhibitors (open squares). Target cells alone were incubated with the inhibitors for 4 hr (filled squares). The effector : target (E : T) ratio was 2. (c) CD4+ CTL clone BK-1 was preincubated with different concentrations of CMA or FD-891 for 2 hr, and then incubated with 51Cr-labelled keyhole limpet haemocyanin (KLH)-pulsed (open squares) or unpulsed (filled squares) A20.HL cells for 6 hr in the presence of the inhibitors. The E : T ratio was 5. To determine whether FD-891 affects effector or target cells, both cell types were pretreated with or without 100 n m of FD-891 for 2 hr, and then incubated for 4 hr in the absence of FD-891 ( Fig. 3). Although A20.2J pretreatment partially decreased cytolysis, OE4 pretreatment completely abrogated the killing activity. Thus, the direct target of FD-891 seems to be effector cells rather than target cells. However, this is not caused by the induction of apoptosis or necrosis by FD-891, as FD-891 caused neither a significant decrease in 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT)-reducing activity nor release of the cytoplasmic enzyme lactate dehydrogenase (data not shown). Effector cytotoxic T lymphocytes (CTLs) are the main target of FD-891. OE4 or 51Cr-labelled A20.2J cells were pretreated with (+) or without (–) 100 n m of FD-891 for 2 hr. Cells were washed twice with medium and then co-cultured for 4 hr. The effector : target (E : T) ratio was 2. As CMA specifically inhibits the acidification of vacuolar organelles, we tested whether FD-891 is able to neutralize the acidic pH in the CTL clone ( Fig. 4). OE4 cells were incubated with 1 µm of FD-891 or 100 n m of CMA, and subjected to DAMP staining. Non-treated OE4 cells exhibited a number of fluorescent dots, which correspond to acidic granules ( Fig. 4a). CMA treatment markedly decreased the number of fluorescent dots ( Fig. 4b), whereas DAMP-positive granules were still detectable in cells treated with an excess of FD-891 ( Fig. 4c). These data indicate that FD-891 is incapable of neutralizing acidic pH. FD-891 does not block acidification of cytoplasmic granules. OE4 cells were preincubated without (a) or with 100 n m of concanamycin A (CMA) (b) or 1 µm of FD-891 (c) for 90 min, and further incubated with 30 µm of 3-(2,4-dinitroanilino)-3′-amino-N-methyldipropylamine (DAMP) for 30 min in the presence of the inhibitors. The DAMP-positive granules were stained and observed under fluorescence microscopy. Upon neutralization of acidic pH in the lytic granules, CMA eventually induces the degradation and inactivation of perforin. 9,10 Although FD-891 exhibited no obvious effects on vacuolar acidification, we addressed the possibility that FD-891 directly affects perforin activity in the lytic granules. FD-891 showed no significant effects on the perforin content at concentrations up to 1 µm, although 100 n m of CMA drastically decreased the perforin content ( Fig. 5a). Lytic granules were separated by using Percoll density-gradient centrifugation ( Fig. 5b). Granzyme A activity was detected in two distinct fractions (i.e. light granule and dense granule fractions) and perforin activity was detected only in the dense granule fractions. CMA completely blocked the perforin activity, accompanied by a significant conversion of granzyme A activity from dense granule fractions to light granule fractions. This might be caused by the vacuolation of lytic granules. 9 FD-891 affected the distribution of granzyme A activity, but exhibited only a slight reduction in perforin activity recovered in the dense granule fractions. These observations indicate that the inhibitory mechanism of FD-891 is quite different from that of CMA, despite their structural similarity. FD-891 does not affect perforin activity in lytic granules. (a) OE4 cells were treated for 2 hr with different concentrations of concanamycin A (CMA) or FD-891. Perforin content was measured by Western blotting. (b) OE4 cells were treated for 2 hr with 100 n m of cytotoxic T lymphocytes (CTLs) or with 1 µm of FD-891. The cells were disrupted through a nitrogen cavitation, and the postnuclear lysates were separated by using Percoll density-gradient centrifugation. Fractions 1–24 correspond to top to bottom fractions. Each fraction was tested for granzyme A activity (filled squares) and perforin activity (open squares). Lytic granules are released upon activation, such as target cell recognition. Immobilized anti-CD3 can stimulate effector CTL to exocytose the lytic granules. To examine the effect of FD-891 on granule exocytosis, we measured the release of granzyme A activity into the culture medium. As shown in Fig. 6(a), FD-891 prevented the granule exocytosis in a dose-dependent manner. The inhibitory concentrations of FD-891 were very similar to those required for CTL-mediated cytotoxicity. To investigate whether FD-891-treated OE4 cells had reduced binding capacity to immobilized anti-CD3 or had a defect in signal transduction downstream of receptor engagement, OE4 cells were cultured in anti-CD3-coated plates for a short time-period and the number of plate-bound cells were measured using crystal violet staining ( Fig. 6b). Although OE4 cells were weakly adherent to cell culture plates, immobilized anti-CD3 significantly augmented the cell attachment ( Fig. 6b, inset). FD-891 prevented the attachment of OE4 cells at concentrations greater than 10 n m, and demonstrated a strong correlation with the inhibition of granule exocytosis. FD-891 inhibits granule exocytosis mainly as a result of the blockage of cytotoxic T lymphocyte (CTL) binding to immobilized anti-CD3. (a) OE4 cells were pretreated with different concentrations of FD-891 for 2 hr, and then stimulated with plate-coated anti-CD3 for 4 hr. Granzyme A (N-α-benzyloxycarbonyl- l-lysine thiobenzyl [BLT] esterase) activity in culture supernatants was measured. (b) OE4 cells were pretreated for 2 hr with different concentrations of FD-891 and then incubated for 20 min with plate-coated anti-CD3. The plate-bound cells were stained with crystal violet. Inset represents OE4 cells that bound to either non-treated or anti-CD3-treated plates. As the TCR/CD3 complex on effector cells plays a major role in the recognition of target cells, the observation that FD-891-treated OE4 cells did not bind to immobilized anti-CD3 implicates that FD-891 might prevent conjugate formation between effector and target cells. OE4 and A20.2J cells were pretreated with different concentrations of FD-891 for 2 hr, stained with fluorochromes and co-cultured at 25° for 30 min ( Fig. 7). A brief centrifugation induced substantial conjugate formation. The conjugate formation was largely prevented when OE4 cells were exposed to > 30 n m of FD-891. The inhibitory concentrations correlated with those required for the killing activity. Treatment of A20.2J cells with excess FD-891 resulted in the moderate blockage of the conjugate formation. FD-891 blocks conjugate formation in a dose-dependent manner. OE4 and A20.2J cells were preincubated for 2 hr with different concentrations of FD-891, and then stained for 30 min on ice with Calcein-AM or hydroethidine, respectively. The cells were mixed in a single tube, centrifuged, then incubated at 25° for 30 min. Conjugate formation was analysed by using a fluorescence-activated cell sorter (FACScalibur). The lower right, upper left and upper right quadrants contain OE4, A20.2J and conjugates, respectively. The left panel represents a mixture of OE4 and A20.2J cells, not centrifuged, which contains only 1·6% conjugates. The percentages of the conjugates are: OE4 treatment 46·1% (none), 27·5% (10 n m), 9·5% (32 n m), 6·1% (100 n m), 3·5% (320 n m), 2·2% (1000 n m); A20.2J treatment 28·1% (none), 20·3% (100 n m), 10·3% (1000 n m). Finally, we tested whether FD-891 affects the expression of cell surface receptors such as TCR/CD3 ( Fig. 8). OE4 cells were treated with 100 n m of FD-891 for 2 hr, incubated with primary antibodies reactive to CD3, CD8, LFA-1, Thy-1.2 and the TCR, and then were incubated with relevant FITC-conjugated secondary antibodies. The expression of LFA-1 and Thy-1.2 was only marginally affected, whereas the same treatment significantly decreased the expression of TCR, CD3 and CD8. Therefore, the inhibitory effect of FD-891 on CTL-mediated cytotoxicity might be partly caused by a decreased expression of the TCR/CD3/CD8 complex. FD-891 decreases preferentially the expression of the T-cell receptor (TCR)/CD3 complex. OE4 cells were treated with or without 100 n m of FD-891 for 2 hr. The cells were stained with anti-CD3, anti-TCR, anti-CD8, anti-lymphocyte function-associated antigen-1 (LFA-1) and anti-Thy-1.2 primary antibodies, followed by fluorescein isothiocyanate (FITC)-labelled secondary antibodies (unbroken lines). Dotted lines indicate cells that were stained with secondary antibodies alone. In this work we have investigated the inhibitory mechanism of FD-891 on CTL-mediated killing pathways. In contrast to CMA, FD-891 was found to block both perforin- and FasL-dependent killing pathways. FD-891 did not affect vacuolar acidification and only slightly decreased perforin activity in lytic granules. However, FD-891 markedly blocked CTL binding to target cells or immobilized anti-CD3. Moreover, FD-891 induced a reduction in the cell-surface expression of TCR, CD3 and CD8. These results suggest that FD-891-induced TCR down-regulation is responsible, at least partly, for the blockage of CTL–target conjugate formation and subsequent cytolysis of target cells. Concanamycins possess an 18-membered lactone ring and a 6-membered hemiketal ring in which 23-OH is glycosylated. It has been reported that are specific inhibitors of vacuolar type with effects on the other two types of and 5,6 The that the 18-membered lactone ring with the 6-membered hemiketal ring is for the inhibitory effects on vacuolar type activity 6 as well as In contrast, the is for this activity. These findings are with our observation that FD-891 was to inhibit vacuolar acidification in our that to are also specific for inhibition of vacuolar type to Although a number of structural of and 18-membered have been no relevant to our have been to compounds that to of might also be For which is a of the group of macrolides, is to be a specific inhibitor of but does not inhibit vacuolar type H+-ATPase. The that belongs to the group of blocks the and which is essential the of T-cell However, these inhibitors exhibited quite different in CTL-mediated cytotoxicity. As one of the inhibitors that showed inhibitory similar to FD-891 and was also reported to block conjugate the CTL–target conjugate to an effect on the binding of the TCR to 20 In was also shown that that to T-cell a and that the CD3 with the These findings demonstrate the role of the in TCR binding and However, FD-891 did not obvious effects on in a (data not which was used to demonstrate that the can the 24 FD-891 induced a in CTL (data not to 25 Thus, FD-891 does not to prevent but rather might modulate the A single major complex complex can up to ≈ 200 TCR, and TCR is to TCR The capacity to the is by a reduction in the number of that the TCR is a that CTL attachment to target cells, FD-891-induced reduction of the TCR/CD3/CD8 complex might be at least partly for the blockage of CTL binding to target cells or immobilized anti-CD3. However, as the between the TCR and is and has a other molecules might play a role in CTL–target conjugates. this is the LFA-1 is a as binding is enhanced by TCR An effect of FD-891 was to the in cell lines It was reported that such as CD3 which to of the T-cell 30 Therefore, is that FD-891 such as the CD3 that are essential for T-cell FD-891-induced reduction of might eventually block the of binding of molecules in CTL–target conjugate formation. the primary target of FD-891 might to identify molecules that are for the of CTL-mediated cytotoxicity. are to for of the This work was in by a for the Research from the of and Research to and a from the of and of

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.