Cysteine protease cathepsin X modulates immune response via activation of β₂ integrins

Abstract

Cathepsin X is a lysosomal, cysteine dependent carboxypeptidase. Its expression is restricted to cells of the immune system, suggesting a function related to the processes of inflammatory and immune responses. It has been shown to stimulate macrophage antigen-1 (Mac-1) receptor-dependent adhesion and phagocytosis via interaction with integrin β₂ subunit. Here its potential role in regulating lymphocyte proliferation via Mac-1 and the other β₂ integrin receptor, lymphocyte function-associated antigen-1 (LFA-1) has been investigated. Cathepsin X has been shown to suppress proliferation of human peripheral blood mononuclear cells, by activation of Mac-1, known as a suppressive factor for lymphocyte proliferation. On the other hand, co-localization of cathepsin X and LFA-1 supports the role of cathepsin X in regulating LFA-1 activity, which enhances lymphocyte proliferation. As shown by fluorescence resonance energy transfer, using U-937 and Jurkat cells transfected with α_L-mCFP and β₂-mYFP, recombinant cathepsin X directly activates LFA-1. The activation was confirmed by increased binding of monoclonal antibody 24, recognizing active LFA-1. We demonstrate that cathepsin X is involved in the regulation of two β₂ integrin receptors, LFA-1 and Mac-1, which exhibit opposing roles in lymphocyte activation.

Introduction

Cysteine proteases contribute to a number of processes of innate and adaptive immune responses.¹ Their active role has been reported in processing of antigenic peptides and stepwise degradation of invariant chain, Ii in major histocompatibility complex (MHC) II-dependent antigen presentation,² processing of pro-granzymes into proteolytically active forms,³ cytotoxic lymphocyte self-protection, cytokine and growth factor degradation,⁴ and modulation of integrin function.⁵ Cysteine proteases may affect the immune response also through regulation of lymphocyte proliferation.⁴ The cysteine protease cathepsin B significantly inhibits the proliferative response to phytohaemagglutinin (PHA) and its inhibition with E-64 completely restores the proliferative response.⁴ Furthermore, Ep-1, another specific inhibitor of cysteine proteases, enhances the lymphocyte proliferation induced by various concentrations of T- and B-cell mitogens such as concanavalin A and lipopolysaccharides. In vivo administration of Ep-1 enhances the delayed type hypersensitivity to bovine serum albumin.⁶ As shown in C57BL/6 mice, another cysteine protease, chymopapain C, completely inhibits the primary immune response and alters the secondary response to antibody formation.^7,8 The mechanisms by which cysteine proteases affect lymphocyte proliferation are not understood.

Several molecules on human immune cells have been reported as targets for proteolytic cleavage by cysteine proteases, including CD14, MHC I, CD54, CD59, and CD18.⁹ CD18 comprises a family of β₂ integrin molecules, including lymphocyte function-associated antigen-1 (LFA-1) (CD11a/CD18), macrophage antigen-1 (Mac-1) (CD11b/CD18), p150,95 (CD11c/CD18) and CD11d/CD18. LFA-1 is expressed on all leucocytes and is involved in a broad range of immunological processes, such as leucocyte adhesion, extravasation, migration, antigen presentation, T-lymphocyte alloantigen-induced proliferation,¹⁰ apoptosis, cytokine expression and cytotoxicity.¹¹ Mac-1 is expressed mainly on cells of myeloid lineage and plays an important role in phagocytosis, transendothelial migration of phagocytes, and the activation of neutrophils and monocytes. p150,95 is expressed by human monocytes, granulocytes, natural killer (NK) cells, and some lymphocytes, and can function as an adhesion molecule and interact with a counter-receptor on stimulated endothelium.¹² It has been shown that the functional activity of LFA-1 is involved in the activation of T-cell proliferation and consequently the immune response, whereas the activity of Mac-1 suppresses it^13–15 (see Fig. 1).

Figure 1.

Proposed mechanism of the regulation of immune response by cathepsin X. Two β₂ integrin receptors, LFA-1 and Mac-1 exhibit opposing roles in lymphocyte activation. Whereas the activation of LFA-1 leads to increased lymphocyte proliferation, activation of Mac-1 receptor suppresses it. In addition, functional interference exists between the two integrin receptors. Increased attachment and spreading via Mac-1 impairs the formation of the LFA-1 mediated homotypic aggregates of PBMC.³⁰

In our previous study we have shown that the function of β₂ integrin receptor Mac-1 during cell adhesion is mediated by cathepsin X activity.¹⁶ We suggested that this protease can be involved in other processes associated with β₂ integrin receptors, such as phagocytosis and T-cell activation, as a functional interference appears to exist between the adhesive capabilities of the three leucocyte β₂ integrins. Cathepsin X has been shown to be expressed mostly in the immune cells such as monocytes, macrophages and dendritic cells and its role in phagocytosis and the regulation of antigen presentation has been proposed.¹⁷ Whereas in U-937 pro-monocytes its localization was intracellular vesicular, in phorbol 12-myristate 13-acetate (PMA)-differentiated U-937 cells, cathepsin X was restricted predominantly to the cell membrane.¹⁶ The binding of cathepsin X to cell surface heparan sulfate proteoglycans, known as partners of integrins in focal adhesion formations, and the integrin-binding motifs, present in pro- (RGD) and in mature forms (ECD) of cathepsin X further suggest a function this enzyme might have in cell signaling and adhesion.

In this study we have assessed the role of cathepsin X in lymphocyte proliferation and studied its functional interactions with β₂ integrin receptors LFA-1 and Mac-1. By activating these two receptors cathepsin X may either augment or suppress lymphocyte proliferation and consequently regulate the extent of immune response.

Materials and methods

Cell cultures

KG-1 (CCL-246), U-937 (CRL-1593.2), Mo-T (CRL-8066) and Jurkat cell lines (TIB-152) (all ATCC, Manassas, VA) were maintained in RPMI-1640 medium, Sigma (St. Louis, MO), supplemented with 2 mm glutamine (Sigma), 2 g/l sodium bicarbonate, Riedel de Haën (St Louis, MO) and 10% fetal calf serum (FCS), HyClone (Logan, UT). Human peripheral blood mononuclear cells (PBMC) were obtained from peripheral blood from healthy donors by density gradient centrifugation (specific density, 1·077 g/ml, Sigma). After three washes in phosphate-buffered saline (PBS), pH 7·4, cells were resuspended in RPMI-1640 medium (Sigma), supplemented with 2 mm glutamine (Sigma), 2 g/l sodium bicarbonate (Riedel de Haën), and 10% FCS (HyClone).

Cysteine protease inhibitors

E-64 was from Sigma. CA-074 was from Bachem (Bubendorf, Switzerland). Neutralizing mouse monoclonal antibodies (mAb) to cathepsins X (2F12) and B (2A2) were prepared from mouse hybridoma cell lines as reported.^17,18

PHA activation of PBMC

Isolated PBMC cells were washed once with PBS and labelled with carboxy-fluorescein-diacetate succinimidyl ester (CFSE) according to the manufacturer's protocol (Human CFSE) Flow Kit, Renovar Inc., WI¹⁹). 2 × 10⁶ PBMC were resuspended in 2·5 ml of complete RPMI-1640 medium and 25 μl of PHA was added according to the manufacturer's protocol.¹⁹ Cathepsin X inhibitor 2F12 mAb (0·5 μm) was added. Its effect was compared to that of the control experiment in which 1 × 10⁶ PBMC were resuspended in 2·5 ml of medium in the absence of PHA. 2F12 mAb was tested for its effect on cell viability and proliferation of PBMC, also in the absence of PHA.

Activated T cells were immunophenotyped by flow cytometry on day 4 of PHA stimulation.

Mixed lymphocyte reaction (MLR)

KG-1 cells (4 × 10⁶ stimulator cells) were resuspended in 10 ml of complete RPMI-1640 medium containing 50 nm PMA in 75 cm² culture flasks. Cells were allowed to differentiate to monocytes/macrophages at 37° for 24 hr in a 5% CO₂ humidified atmosphere. The flasks were washed gently with PBS, pH 7·4, to remove non-adherent cells. Adherent cells were detached with PBS containing ethylenediaminetetra-acetic acid (EDTA), washed once with PBS, and resuspended in complete RPMI-1640 medium. To arrest the proliferation, they were treated with mitomycin C (25 μg/ml) at 37° for 30 min, then washed twice with PBS. Differentiated and mitomycin C-treated KG-1 cells (1 × 10⁶) were resuspended in 2·5 ml of complete RPMI-1640 medium and set up in culture flasks.

Isolated PBMC (responder cells) were washed once with PBS and labelled with CFSE cell tracking reagent¹⁹ according to the manufacturer's protocol. PBMC (3 × 10⁶) were resuspended in 2·5 ml of complete RPMI-1640 medium and added to KG-1 cells. As a control, 1 × 10⁶ PBMC were resuspended in 2·5 ml of medium and set up in a separate culture flask. The threshold for proliferating cells was set according to control PBMC, which did not proliferate. Subsets of proliferating PBMC were characterized using specific antibodies.

In a similar experiment CFSE-labelled PBMC were resuspended in 2·5 ml of conditioned medium of differentiated KG-1 cells. Conditioned medium was obtained by cultivating adhered and differentiated KG-1 cells (24 hr, 50 nm PMA, 4 × 10⁶ cells) in 5 ml of complete RPMI medium. Afterwards, the cells were removed by centrifugation (3000 g, 15 min) and medium was used as a stimulating agent.

The cysteine protease inhibitors used in the MLR were E-64 (2 μm), CA-074 (2·6 μm), 2F12 mAb (0·5 μm) and 2A2 mAb (0·5 μm). Prior to use, the inhibitors were tested for their impact on the viability of PBMC. Optimal concentrations of inhibitors were determined in a preliminary CellTiter 96® AQueous One Solution Cell Proliferation Assay (MTS colorimetric assay; Promega, Madison, WI).

Activated T cells were immunophenotyped by flow cytometry on day 5 of MLR co-culture.

Mo-T-cell proliferation

Proliferation of Mo-T lymphocytes (T lymphocyte cell line²⁰) was determined with a MTS colorimetric assay on a Tecan microplate reader. Mo T cells were resuspended in conditioned medium of differentiated U-937 cells to yield a final concentration of 4 × 10⁵ cells/ml. One hundred μl of this suspension was added to each well of 96-well microtitre plate and different concentrations of cysteine protease inhibitors were added and cells further incubated at 37° for 24 hr. The control sample contained only the medium of differentiated U-937 cells and appropriate volume of solvent in which inhibitor was diluted (PBS or medium). After incubation, proliferation of Mo-T cells was assessed by MTS colorimetric assay. Twenty μl of MTS reagent was added to the wells and incubated for additional 2 hr at 37°. Cell proliferation was determined by eqn 1, where A_{test cells} and A_{control cells} are the absorbance of formazan determined for cells treated with different concentrations of protease inhibitors and for non-treated control cells, respectively:

(1)

Cathepsin X in lymphocyte aggregation

PBMC (2 × 10⁶) were resuspended in 2·5 ml of complete RPMI-1640 medium and 25 μl of PHA (Renovar Inc) was added according to the manufacturer's protocol. Mo T cells were resuspended in conditioned medium of differentiated U-937 cells to yield a final concentration of 4 × 10⁵ cells/ml. Both, PBMC and Mo T cells were cultured in the presence or absence of 2F12 mAb (0·5 μm) or CA-074 (2 μm). After 4 days transmission microscopy was performed to observe the cell cluster formation.

Flow cytometry

Different cell lineages, i.e. KG-1, differentiated KG-1 (24 hr, 50 nm PMA), U-937, differentiated U-937 (24 hr, 50 nm PMA), Mo-T and Jurkat cells (4 × 10⁵ cells) were grown in 24-well culture plates (Costar Corning, New York, NY), either stimulated with PHA for 24 hr or not (Sigma). 2F12 mAb was added prior to the activation at a concentration of 0·5 μm. Cells were then washed with PBS and analysed by flow cytometry for expression of LFA-1 or Mac-1.

MAbs labelled with phycoerythrin (PE) or Alexa-488 (Beckton Dickinson Immunocytometry System; Beckton Dickinson, Inc., San Jose, CA) in conjunction with propidium iodide (PI), were used for three-colour flow cytometry. The following antibodies were used for labelling: anti-CD3–PE, anti-CD4–PE, anti-CD8–PE, anti-CD19–PE, anti-CD11a–PE, anti-cathepsin X2F12 Alexa- and anti-CD11b–Alexa-488. A Simultest control (mouse immunoglobulin G1 (IgG1)–PE and mouse IgG1–Alexa-488) was used as background control. From forward scatter (FSC) versus PI plot, the population of live cells was gated for further analysis. From CFSE versus PE plot proliferated subpopulations were distinguished from specific antibody staining. Three-colour flow cytometry was performed on a FACSCalibur (Becton Dickinson, Inc.).

Immunofluorescence microscopy

Visualization of LFA-1 activation by fluorescence resonance energy transfer (FRET)

Plasmid DNA constructs for α_L-m-cyan fluorescent protein (CFP) and β₂-m-yellow fluorescent protein (YFP) transformed in Escherichia coli were provided by Dr Tim Springer (Adggene). The plasmids were isolated and purified from each colony using the Eppendorf midiprep kit. Isolated plasmids were transiently transfected into U-937 and Jurkat cells using Lipofectamine (Invitrogen, San Diego, CA). A 12-well plate was used to produce transfectants expressing α_L-mCFP alone, β₂-mYFP alone, and both constructs together. The DNA was incubated for 20 min with Lipofectamine in serum-free medium and added to 1 × 10⁶ U-937 cells in a 12-well plate. After 4 hr incubation at 37°, the medium was replaced. After 24 hr, cells were placed in LabTek chambered coverglass system (Nalge Nunc International, Rochester, NY). Cells were imaged first without any stimulating factors, and time lapse images were collected during treatment with 5 μm recombinant cathepsin X.²¹ A Zeiss LSM 510 META spectral imaging system was used to image cells expressing each of the two vectors alone and those expressing both. CFP laser excitation (458 nm) or YFP laser excitation (514 nm) was performed for determining the FRET effect or transfection efficiency, respectively.

Co-localization studies

Mo-T cells were grown for 24 hr in 24-well culture plates (Costar Corning) on poly-L-lysine precoated coverslips. KG-1 cells (4 × 10⁵/ml) were differentiated with 50 nm PMA as described above under MLR, resuspended in growth medium and seeded (1 × 10⁵) in 24-well culture plates (Costar Corning) on poly-L-lysine precoated coverslips. PBMC (3·5 × 10⁵) were added and allowed to proliferate for 5 days in a co-culture, in the presence or absence of CA-074 (2·6 μm). Afterwards, the cells were centrifuged for 2 min at 1000 g to achieve optimal attachment. Before labelling, cells were fixed with methanol (−20°) for 5 min and permeabilized by 0·01% Triton-X-100 in PBS, pH 7·4, for 10 min. Non-specific staining was blocked with 3% bovine serum albumin in PBS, pH 7·4, for 1 hr.

Cathepsin X was labelled with Alexa Fluor 488 (Molecular Probes, Carlsbad, CA)-labelled mouse 2F12 mAb that recognizes the mature active form. For CD11a labelling, primary antibodies were goat anti-human integrin α_L N-18 from Santa Cruz Biotechnology (Santa Cruz, CA). For active LFA-1 labelling, primary antibody was mAb 24 provided by Dr Nancy Hogg (Imperial Cancer Research Fund, London, UK). After 2 hr incubation with primary antibodies, cells were washed with PBS and treated with Alexa Fluor 633-labelled donkey anti-goat or Alexa Fluor 488-labelled rabbit anti-mouse secondary antibodies (Molecular Probes) for 1 hr. After the final wash with PBS, ProLong Antifade kit (Molecular Probes) was used to mount coverslips on glass slides. Controls were run in the absence of primary antibodies.

Fluorescence microscopy was performed using Carl Zeiss LSM 510 confocal microscope. Alexa Fluor 488 or Alexa Fluor 633 were excited with an argon (488 nm) or He/Ne (633 nm) laser and emission was filtered using narrow band 505–530 nm, LP 560 nm and LP 650 nm filters, respectively. Images were analysed using Carl Zeiss LSM Image software 3.0.

Statistical analysis

spss PC software package (Release 13.0) was used for statistical analysis. The difference between the groups was evaluated using the non-parametric Mann–Whitney test. P values of <0·05 were considered to be statistically significant.

Results

PHA activation of PBMC

PHA, which exerts its action via β₂ integrin receptors, was used to induce the proliferation of PBMC in order to show any possible effect of cathepsin X activity on β₂ integrin dependent PBMC proliferation. In the presence of 2F12 mAb, a specific inhibitor of cathepsin X, the proliferation of PBMC under PHA stimulation was increased and there were more proliferating lymphocyte blast cells (Fig. 2). The percentage of all PBMC proliferated was 23·2%, as against 7·3% in the control. In the absence of PHA, however, 2F12 mAb did not have any effect on PBMC proliferation (data not shown).

Figure 2.

Inhibition of cathepsin X increases the proliferation of PBMC induced with PHA. In order to show any possible effect of cathepsin X activity on β₂ integrin dependent PBMC proliferation, PHA, which exerts its action via β₂ integrin receptors, was used to induce the proliferation of PBMC. The FSC–SSC (side scatter) density plot shows a marked increase of proliferating blast cells in the presence of the cathepsin X neutralizing 2F12 mAb relative to control cells stimulated in the absence of inhibitor. R1 indicates the percentage of the proliferating cell population. One representative experiment out of three is presented.

Mixed lymphocyte reaction (MLR)

In another β₂ integrin receptor dependent proliferation assay, MLR with differentiated KG-1 cells as stimulator cells, several cysteine protease inhibitors were tested for their ability to promote the proliferation of PBMC (Fig. 3a). The proliferation of both, CD4⁺ and CD8⁺ T lymphocytes was increased by these inhibitors. Proliferation increased by 24·1% with CA-074 (inhibitor of cathepsins B and X)²² and by 11·7% with 2A2 mAb (specific inhibitor of cathepsin B) (Fig. 3b). Inhibition of cathepsin X with 2F12 mAb in MLR resulted in significant increase (P < 0·005) in proliferation of PBMC at concentrations 0·5, 1 and 2 μm (Fig. 3c). The concentration 0·5 μm, causing 43·3% increase in proliferation, was used in further flow cytometry proliferation analysis (Fig. 4). A marked increase was also evident when allogeneic PBMC were used as stimulator cells, although the overall proliferation was smaller (data not shown). When conditioned medium was used as a stimulating agent, PBMC again proliferated more in the presence of 2F12 mAb (Figs 3c and 4). To exclude any mitogenic effect of monoclonal antibodies or cysteine protease inhibitors used, we show that, lymphocytes did not proliferate in control flasks devoid of stimulating cells or conditioned medium (Figs 3 and 4 controls), regardless of the presence of 2F12 or 2A2 mAb or cysteine protease inhibitors.

Figure 3.

Cysteine protease inhibitors have different effects in MLR, which depends on the β₂ integrin receptors. (a) KG-1 cells were differentiated with PMA for 24 hr and pretreated with mitomycin C prior to MLR. PBMC were labelled with CFSE and co-incubated with differentiated KG-1 cells for 5 days in the presence or absence of cysteine protease inhibitors: 2A2 mAb (specific inhibitor of cathepsin B) and CA-074 (inhibitor of cathepsins B and X). Afterwards, cells were labelled with anti-CD3, anti-CD4, anti-CD8 or anti-CD19 and proliferation was measured by flow cytometry. (b) Proliferation of PBMC was only slightly increased with CA-074, a synthetic inhibitor of cathepsin B and cathepsin X. The effect of 2A2 mAb and 2F12 mAb, the inhibitors of cathepsins B and X, respectively, reveals that the increase in proliferation is attributable to cathepsin X activity. The bars represent mean ± SD of three experiments. (c) U-937 cells were differentiated with PMA for 24 hr in a microtitre plate and co-cultured with PBMC as responder cells at different concentrations of 2F12 mAb. Inhibition of cathepsin X activity caused a concentration dependent increase of PBMC proliferation. Proliferation was significantly increased at concentrations marked with asterisk (**for P < 0·005 and *for P = 0·106). Proliferation was measured with an MTS colorimetric assay on a Tecan microplate reader. The bars represent mean ± SD of three experiments.

Figure 4.

Inhibition of cathepsin X increases PBMC proliferation in MLR. KG-1 cells were differentiated with PMA for 24 hr and pretreated with mitomycin C prior to MLR. PBMC were labelled with CFSE and co-incubated with differentiated KG-1 cells for 5 days in the presence or absence of cathepsin X inhibitor 2F12 mAb. Alternatively, PBMC were stimulated with conditioned medium of differentiated KG-1 cells only. Afterwards, cells were labelled with anti-CD3, anti-CD4, anti-CD8 mAbs, respectively and proliferation was measured by flow cytometry. Inhibition of cathepsin X profoundly increased proliferation of PBMC. In the absence of differentiated KG-1 cells or conditioned medium as stimulus, 2F12 mAb did not affect cell proliferation. One representative experiment out of three is present.

Mo-T-cell proliferation

To distinguish the effect of cathepsin X activity on LFA-1 from that on Mac-1, we examined the proliferation of Mo-T lymphocytes, which do not express Mac-1; therefore, the influence of cathepsin X activity via Mac-1 receptor can be excluded. In contrast to PHA activation and MLR, the proliferation of Mo-T lymphocytes stimulated with the conditioned medium, obtained from U-937 cells was reduced in the presence of 2F12 mAb or CA-074 inhibitor (Fig. 5). E-64, a non-specific inhibitor of cysteine proteases, did not show a similar effect (Fig. 6). None of the inhibitors had any effect on the proliferation of Mo-T cells in the absence of the conditioned medium.

Figure 5.

Reduced proliferation of Mo-T lymphocytes with inhibition of cathepsin X. Mo-T lymphocytes, which do not express Mac-1 receptor, were cultured in conditioned medium of differentiated U-937 cells in the presence of different inhibitors of cysteine proteases. Inhibition of cathepsin X by 2F12 mAb caused reduced proliferation of Mo-T lymphocytes (P = 0·02); synthetic inhibitor CA-074 showed a similar trend, although the reduction was not statistically significant. E-64, however, did not effect proliferation of Mo-T lymphocytes. The bars represent mean ± SD of three experiments.

Figure 6.

Formation of cell clusters is altered by cathepsin X activity. Similar to lymphocyte proliferation, cathepsin X differentially affects cell cluster formation via LFA-1 or Mac-1 integrin receptors. PBMC were stimulated with PHA for 4 days in the presence or absence of 2F12 mAb. Inhibition of cathepsin X resulted in formation of markedly larger cell clusters, that arise via LFA-1–ICAM-1 homotypic interactions. On the other hand, inhibition of cathepsin X with 2F12 mAb in Mo-T activation with conditioned medium diminished intercellular Mo-T aggregation.

Cathepsin X in lymphocytes aggregation

To demonstrate that cathepsin X is involved in increased spheroid formation via LFA-1, we determined the aggregation of lymphocytes in the presence or absence of cathepsin X inhibitors, 2F12 mAb and CA-074. PBMC spheroids were much larger when PBMC were cultured in the presence of either 2F12 mAb (Fig. 6) or CA-074 (data not shown) during PHA stimulation. On the other hand, when Mo-T lymphocytes were stimulated with conditioned medium, the formation of cell clusters was diminished in the presence of the same inhibitors (Fig. 6).

Expression of LFA-1 and Mac-1

In order to elucidate whether cathepsin X activity affects the expression of integrin receptors immunophenotype analysis was performed. It is evident that PHA treatment barely modified the expression of LFA-1 in U-937, (Fig. 7a) KG-1 cells and Mo-T cells (data not shown). Expression of LFA-1 in U-937 cells was, however, increased when the cells were differentiated with 50 nm PMA for 24 hr. On the other hand, LFA-1 expression was not changed when the cells were treated with 2F12 mAb during stimulation, nor when 2F12 mAb was added to non-stimulated cells (data not shown).

Figure 7.

Inhibition of cathepsin X does not change expression of LFA-1 or Mac-1. (a) U-937 cells were stimulated with PMA or PHA for 24 hr or 4 days, respectively and (b) PBMC were stimulated with PHA or in a MLR experiment with differentiated KG-1 cells as stimulator cells for 4 or 5 days, respectively. Afterwards, cells were stained for specific cell surface markers with the indicated conjugated mAbs and flow cytometry was performed. Results shown are representative of triplicate experiments. Cathepsin X and Mac-1 receptor show similar pattern of expression, both being only slightly increased in stimulation with PHA, but more with PMA. LFA-1 is expressed abundantly on control cells and expression is increased with different stimulation. Treatment of U-937 cells or PBMC with 2F12 mAb (bold line) did not affect expression of Mac-1 or LFA-1 relative to untreated cells (grey area).

Both, Mo-T lymphocytes (Fig. 7a) and PBMC (Fig. 7b) expressed LFA-1 to a great extent.

The expression of Mac-1 after PHA treatment was only slightly changed in U-937 cells, but was markedly increased in differentiated U-937 cells. The addition of 2F12 mAb did not influence the expression of Mac-1, in either non-stimulated or PHA treated cells.

Expression of cathepsin X

Expression of cathepsin X in stimulator cells, as well as in PBMC during PHA activation and MLR was determined using Alexa-488 labelled 2F12 mAb. In U-937 cells the expression of cathepsin X was increased slightly when the cells were stimulated with PHA and to a greater extent in differentiated U-937 cells (U-937 cells stimulated with PMA) (Fig. 7a). Thus, from this perspective, differentiated cells are more convenient as stimulator cells in MLR. Similar expression can be observed also in Mo-T cells (Fig. 7a) and KG-1 cell line (data not shown).

Visualization of LFA-1 activation by FRET

As there was no obvious effect of cathepsin X activity on LFA-1 and Mac-1 expression, we evaluated a possible effect on LFA-1 activation. To detect CFP and YFP emission of LFA-1 integrin constructs and activation (affinity change) of LFA-1²³ in transfected cells we used a Zeiss LSM 510 META spectral imaging system.²⁴ The cells being observed for FRET were first visualized with excitation at 514 nm to determine YFP transfection efficiency. At 514 nm only YFP is excited and fluoresces. Next, the cells expressing β₂-mYFP were excited at 458 nm and their emission spectrum recorded. At 458 nm CFP is excited and fluoresces. However, if there is energy transfer, the fluorescence emission spectrum of YFP is seen, as the fluorescence from CFP excites YFP. The emission spectrum of non-activated LFA-1 shows peak emission at 530 nm, indicating FRET from CFP to YFP. Before addition of recombinant cathepsin X to the medium, U-937 and Jurkat cells expressed non-activated LFA-1, as can be observed from the emission spectra (Fig. 8). However, after the addition of recombinant cathepsin X, CFP emission gradually increased, indicating a decrease in FRET and an increase in the amount of activated LFA-1 (Fig. 8). Mean ratio between YFP (483 nm) and CFP (526 nm) emission changed from 3·97 ± 0·13 at time 0 min to 2·30 ± 0·06 at time 40 min, which is significantly different (P < 0·0001). CFP emission can be seen in cells transfected only with α_L-mCFP and in cells with active LFA-1 receptor. Because we observed only the cells transfected with β₂-mYFP, the CFP emission is caused only by active LFA-1 receptor. As a control, emission spectra of cells expressing α_L-mCFP or β₂-mYFP were recorded with excitation at 458 or 514 nm respectively (data not shown).

Figure 8.

FRET experiment showing activation of LFA-1 with cathepsin X. U-937 cells were transfected with plasmid DNA constructs for α_L-mCFP and β₂-mYFP. After 24 hr, cells were imaged first without any stimulating factors, and time lapse images were collected during treatment with 5 μm recombinant cathepsin X. A Zeiss LSM 510 META spectral imaging system was used to image cells. CFP laser excitation (458 nm) or YFP laser excitation (514 nm) was performed for determining FRET effect or transfection effectiveness, respectively. The emission spectrum of non-activated LFA-1 shows peak emission at 530 nm, indicating FRET from CFP to YFP. Before addition of recombinant cathepsin X to the medium, U-937 expressed non-activated LFA-1, as observed from the emission spectrum. However, after the addition of recombinant cathepsin X, the spectrum gradually showed a CFP emission signal (marked with black arrows), indicating a decrease in FRET and an increase in the amount of activated LFA-1. The first time point, marked as 0 min, was taken just before the addition of recombinant cathepsin X. We selected 16 circular regions of interest (diameter 15 pixels) as marked on the first panel. Scale bar: 10 μm. Right panels show mean emission spectra for each image. Error bars represent SEM. As a control, emission spectra of cells at 514 nm were recorded to confirm YFP expression in cells. The red line on the graph represents the emission spectra of cells, transfected only with CFP.

Co-localization of LFA-1 and cathepsin X by immunofluorescence microscopy

To demonstrate co-localization of cathepsin X with LFA-1 on lymphocytes, immunofluorescence microscopy was performed. LFA-1 is extensively expressed on Mo-T lymphocytes (Fig. 9a) and enables formation of homotypic intercellular spheroids. As can be seen from the contour plot and the mask of the pixels above the threshold in both channels (blue colour), LFA-1 and cathepsin X co-localize predominantly in cell–cell junctions, that is in cell contact areas (Fig. 9b and c). In a co-culture of KG-1 and PBMC, co-localization of LFA-1 and cathepsin X was most pronounced in PBMC spheroids (Fig. 9b). In the presence of higher levels of cathepsin X the activation of LFA-1 was more pronounced (Fig. 9d and e).

Figure 9.

Cathepsin X and LFA-1 co-localization and activation. (a) Mo-T cells were grown in 24-well culture plates (Costar Corning) on poly-L-lysine precoated coverslips for 24 hr. (b and c) PBMC were added to differentiated KG-1 cells and co-incubated for 5 days in MLR. In co-localization experiments fluorescent dyes (cathepsin X-green fluorescence, LFA-1-red fluorescence) were imaged sequentially in a line-interlace mode to eliminate cross talk between the channels. Threshold value was determined arbitrarily, based on negative control co-localization experiments of cathepsin B and cathepsin X. The threshold level for this display was set to 40% of the maximal brightness level. The mask of the pixels above the threshold in both channels (blue colour) and the contour plot are shown for images demonstrating co-localization. (d) and (e) Cathepsin X upregulated Jurkat cells (Jevnikar et al., submitted) express significantly more active LFA-1 (green fluorescence) detected with mAb 24 (e) as the wild type Jurkat cells (d), confirming the activation of LFA-1 by cathepsin X. Scale bars represent 50 μm.

Discussion

In our recent study we showed that the cysteine dependent carboxypeptidase cathepsin X interacts with β₂ integrin receptors in differentiated macrophages and promotes cell adhesion by activation of Mac-1.¹⁶ Because of the extended functions of β₂ integrin receptors in cell signalling and activation, other roles of cathepsin X are to be expected. More than 30 years ago it was suggested that a macrophage ‘catheptic’ carboxypeptidase activity plays a role in T cell activation.^25,26 We have also observed that Mo-T lymphocytes can be stimulated by the conditioned medium from differentiated U-937 cells and preliminary results showed that the inhibition of cathepsin X activity in conditioned medium reduced the proliferation of T lymphocytes. We therefore decided to elucidate the role of cathepsin X in β₂ integrin receptor dependent activation of T-lymphocytes. Therefore, human PBMC were selected as an experimental cell model and were activated either with PHA or in a MLR, the methods, known to be Mac-1 receptor dependent. Our results show that in both, PBMC proliferation is regulated by cathepsin X activity.

The activation of T lymphocytes with PHA requires macrophage–lymphocyte interaction^27,28 and it has been shown that PHA exerts its action via interaction with β₂ subunit (CD11b) of Mac-1 receptor.^27,28 In recent studies, it has been shown that the constitutive activity of Mac-1 on antigen-presenting cells inhibits antigen presentation and down-regulates T-cell activation,^15,29 presumably because of suppression of inflammatory cytokine release.²⁹ The antibodies directed against CD11/CD18 inactivate Mac-1 activity,²⁸ resulting in higher T-cell proliferation and reduced monocyte adhesion. By analogy, the activation of Mac-1 by cathepsin X¹⁶ results in diminished activation and proliferation of T lymphocytes and vice versa, the inhibition of cathepsin X by 2F12 mAb diminishes adhesion of monocytes through the Mac-1 receptor¹⁶ and enhances T lymphocyte proliferation.

Similarly, in a MLR, alloantigen-induced T-cell proliferation could be increased by Mac-1 inhibition.¹⁵ In the present study, the inhibition of cathepsin X by 2F12 mAb also markedly enhanced proliferation of PBMC. The proliferation was increased by CA-074, an inhibitor of cathepsins B and X, showing that the effect is attributable to the inhibition of cathepsin X activity and not to non-specific action of the monoclonal antibody. 2A2 mAb, a specific inhibitor of cathepsin B, induced a smaller increase in proliferation than CA-074. Our results are in accord with those of Daneri-Nevarro et al.,⁴ which showed restored proliferation of lymphocytes by the general cysteine protease inhibitor E-64 and this effect is shown here to be attributable to cathepsin X, and only to a lesser extent to other cysteine proteases.

The enhanced proliferation of PBMC with 2F12 mAb was also observed when PBMC were treated only with the conditioned medium from differentiated KG-1 cells. This indicates that the presence of cathepsin X in the conditioned medium diminishes T-cell proliferation. Using enzyme-linked immunosorbent assay we confirmed that cathepsin X is indeed secreted into the medium (data not shown). Again, cathepsin X inhibition enhanced T lymphocyte proliferation because of activation of Mac-1 receptor, as noted above.

On the other hand, antibodies to LFA-1 inhibit alloantigen-induced T-cell proliferation and cytotoxic activity in MLR,¹⁴ when present at the time of initial T-cell receptor/antigen engagement. Although anti-LFA-1 antibodies impede the generation of functionally active T cells, the expression of receptors CD3, CD4, intracellular adhesion molecule-1 (ICAM-1), and LFA-1 is not altered.¹⁴ To distinguish the effect of cathepsin X activity on LFA-1 from that on Mac-1, we examined the proliferation of Mo-T lymphocytes in conditioned medium from differentiated U-937 cells. In this system inhibition of cathepsin X diminished the proliferation of Mo-T lymphocytes, which can be explained by its action on LFA-1 receptor that is expressed abundantly on Mo-T lymphocytes (data not shown), whereas Mac-1 is not. The fact that the formation of Mo-T homotypic intercellular spheroids, presumably via LFA-1–ICAM-1 binding was altered by 2F12 mAb additionally supports promotive role of cathepsin X for lymphocyte proliferation via activation of LFA-1. Furthermore, up-regulation of cathepsin X in lymphocytes significantly augments intercellular spheroids formation via LFA-1–ICAM-1 as well as increases ICAM-1 attachment and spreading of lymphocytes (Jevnikar et al., submitted).

However, Mac-1 mediated adhesion competes or interferes with LFA-1-dependent intercellular adhesion and, hence, the attachment and spreading via Mac-1 greatly impairs the formation of the LFA-1-mediated homotypic aggregates.³⁰ In agreement with our previous results,¹⁶ inhibition of cathepsin X activity during the activation of PBMC impaired Mac-1 activation and spreading, and promoted aggregation via LFA-1.^31,32

The FRET experiment confirmed that cathepsin X activates LFA-1 receptor. Addition of recombinant cathepsin X to the medium of transfected cells with α_L-mCFP and β₂-mYFP resulted in a change of emission spectra on the membrane from YFP emission to CFP emission spectrum, indicating activation of LFA-1 receptor. Co-localization of the active cathepsin X and LFA-1 on Mo-T and PBMC lymphocytes supports the role of cathepsin X in the regulation of LFA-1 activity. The co-localization was most pronounced in the cell-cell junctions in spheroids of PBMC, indicating that cathepsin X interacts with LFA-1 receptor on lymphocytes. The presence of activated LFA-1 receptor was further confirmed, by activation reporter mAb 24.

Importantly, β₂ integrins have opposing roles in lymphocyte activation. Active β₂ integrins on T lymphocytes (LFA-1) are required for optimal formation of the immunological synapse and T-cell activation,³³ but on antigen-presenting cells (Mac-1) they inhibit antigen presentation and T-cell activation.^15,34,35 As mentioned previously, Mac-1 activity interferes with LFA-1 mediated aggregation³⁰ and it has been proposed that dysregulated adhesion and formation of immunological synapse caused by Mac-1 activity might explain down-regulated T-cell activation.¹⁵ In this regard, it has been proposed that via constitutively active Mac-1 on macrophages, in contrast to functionally inactive Mac-1 on dendritic cells, antigen presentation in vivo is restricted to dendritic cells.¹⁵ As shown in our previous study, the localization of cathepsin X in macrophages differs from that in dendritic cells,¹⁷ which may have an impact on different regulation of Mac-1 receptors in the two cell types.

The switch between active and inactive forms of LFA-1 and Mac-1 is controlled by a membrane proximal motif of the cytoplasmic tail of CD18 (β₂ subunit), that exists in a state of competition with the α subunit promoting a low-affinity state and the cytoskeletal protein talin promoting a high affinity state.³⁶ Therefore, the β₂ subunit, through which the adhesion properties of LFA-1 and Mac-1 are regulated, may have different roles in the function of these two receptors.⁸ Within both, α and β subunits, there appear to be several potential proteolytic cleavage sites that have different effects on these receptors. The treatment of different cells with 2F12 mAb did not change the expression of LFA-1 receptor, suggesting therefore a conformational change or truncation of LFA-1 structure leading to a switch of affinity state. Similarly, cathepsin X activity does not change the expression of β₂ subunit in Mac-1 receptor, but rather affects its activation, enhancing cell adhesion. Nevertheless, at this step, the mechanism of β₂ integrin receptors activation by cathepsin X remains elusive.

In conclusion, we have shown that, in addition to promotion of adhesion and phagocytosis in monocytes/macrophages, cathepsin X may also modulate the proliferation of lymphocytes via LFA-1 and Mac-1 receptors. The activation of Mac-1 receptor results in suppression of lymphocyte proliferation and cluster formation whereas the activation of LFA-1 promotes proliferation of lymphocytes and consequently immune response. Catheptic carboxypeptidase activity thus indeed affects T cell activation, as proposed by Dessaint et al.²⁵ and Katz et al.,²⁶ and we have shown that this effect is attributable to activation of the β₂ integrin receptors.

Glossary

Abbreviations

PHA

phytohaemagglutinin

LFA-1

lymphocyte function-associated antigen-1

Mac-1

macrophage antigen-1

FRET

fluorescence resonance energy transfer

PBMC

peripheral blood mononuclear cells

MLR

mixed lymphocyte reaction

PMA

phorbol 12-myristate 13-acetate

CFP

cyan fluorescent protein

YFP

yellow fluorescent protein

ICAM-1

intracellular adhesion molecule-1