Effector granules in human T lymphocytes: the luminal proteome of secretory lysosomes from human T cells

Background Cytotoxic cells of the immune system have evolved a lysosomal compartment to store and mobilize effector molecules. In T lymphocytes and NK cells, the death factor FasL is one of the characteristic marker proteins of these so-called secretory lysosomes, which combine properties of conventional lysosomes and exocytotic vesicles. Although these vesicles are crucial for immune effector function, their protein content in T cells has so far not been investigated in detail. Results In the present study, intact membranous vesicles were enriched from homogenates of polyclonally activated T cells and initially characterized by Western blotting and electron microscopic inspection. The vesicular fraction that contained the marker proteins of secretory lysosomes was subsequently analyzed by 2D electrophoresis and mass spectrometry. The proteome analysis and data evaluation revealed that 70% of the 397 annotated proteins had been associated with different lysosome-related organelles in previous proteome studies. Conclusion We provide the first comprehensive proteome map of T cell-derived secretory lysosomes with only minor contaminations by cytosolic, nuclear or other proteins. This information will be useful to more precisely address the activation-dependent maturation and the specific distribution of effector organelles and proteins in individual T or NK cell populations in future studies.


Background
Cytotoxic T lymphocytes (CTL) and Natural Killer (NK) cells are the main cytotoxic effector cells of the immune system. In order to effectively eliminate virus-infected and tumorigenic cells, they rapidly mobilize effector molecules including granzymes, perforin, granulysin and the death factor FasL (CD178) that are presumably stored in preformed organelles termed secretory lysosomes (SL) [1]. Secretory lysosomes combine degradative properties of conventional lysosomes with characteristics of exocytotic vesicles. At the level of morphology, conventional and secretory lysosomes are hardly distinguishable and both appear to represent endpoints of an endocytotic pathway and are formed by fusion and fission of endosomes and lysosomes [2]. Similar to conventional lysosomes, large membrane areas are covered by lysosome-associated membrane-proteins (LAMPs) including LAMP-1 (CD107a), LAMP-2 (CD107b) and LAMP-3 (CD63) [3][4][5]. However, secretory effector lysosomes are characterized by a specific set of membrane and luminal marker proteins [6,7]. The current consensus is that SL of CTLs and NK cells carry the aforementioned effector proteins either in the lysosomal lumen (granzymes, perforin and granulysin) or as characteristic transmembrane compounds (FasL) [8][9][10].
Recently, we provided a protocol that allows a substantial enrichment of intact SL from in vitro expanded lymphocyte populations [11]. Employing this procedure for subcellular fractionation of a crude organelle preparation, we obtained a fraction of intact vesicles that is significantly enriched in SL marker proteins. We were thus able to report the first comprehensive analysis of the luminal proteome of secretory lysosomes from NK cells [12]. At that time, 234 different proteins were identified by mass spectrometry, 77% of which had been associated with SL or other lysosomal compartments before. Applying 2D difference gel electrophoresis, we also described a cell line-specific distribution of functionally relevant proteins in SL from human NK cell lines and primary NK cells [12].
Based on this study, it appears likely that different T cell populations utilize the SL organelles to store and mobilize lineage-specific cargo proteins. However, the proteome of secretory lysosomes in T cells has not been deciphered. To provide the first proteome map for T cell-derived SL, we enriched organelles from activated T lymphoblasts. Organelle extracts were subjected to SDS-PAGE and Western blotting to identify the FasLcontaining SL fraction. This fraction was analyzed by electron microscopy to demonstrate the enrichment of a homogeneous population of intact vesicles. In order to define the luminal proteome of the respective SL compartment, the organelles were lysed and proteins were separated by 2D gel electrophoresis. Mass spectrometry was applied to identify individual spots. We annotated 397 proteins, 70% of which had been associated with lysosome-related organelles before. With the present report, we thus provide the first comprehensive description of the content of FasL-carrying effector vesicles isolated from activated human T lymphocytes.

Results and Discussion
In our preceding analysis of the SL compartment of NK cell lines and primary NK cells, we annotated 234 individual proteins and demonstrated a cell line-specific distribution of several functionally relevant molecules including cytotoxic effector proteins, lysosomal proteases and MHC molecules [12]. As a basis to address unsolved issues regarding the maturation, function and cell type-specific composition of the cytotoxic effector compartment in T cell populations, we now analyzed the proteome of enriched secretory lysosomes from in vitro activated human T cell blasts.

FasL-associated secretory lysosomes in activated lymphocytes
We and others have shown that in CTLs, preformed FasL accumulates in the limiting membrane of secretory lysosomes with late endosome or multi-vesicular-body structure and there co-localizes with characteristic lysosomal marker proteins including CD63 or lysosomal hydrolases and cytoskeletal adapter proteins [7][8][9][13][14][15]. Confocal laser-scanning microscopy (CLSM) was applied to confirm that FasL also might serve as a marker for secretory lysosomes in in vitro expanded PHA-stimulated T lymphocytes used in the present study. As depicted in Figure 1, we detected an apparent co-localization of CD63 with FasL, granzyme A and the lysosomal protease cathepsin B. It should be mentioned that a common or distinct localization of LAMP-3 (CD63) and FasL is still controversially discussed. Several reports suggest a co-localization of FasL with granule proteins, such as cathepsin D, CD63, granzyme B, perforin and LAMP-1 in a single granular entity [8,9] whereas other studies indicate that CD63 and FasL are located in distinct subcellular compartments [16].
Our protocol for the enrichment of secretory lysosomes yielded six separate fractions that were subjected to further analysis by Western blotting or 2D gel electrophoresis. To demonstrate an effective enrichment of the SL fraction, we first separated the proteins of individual fractions by SDS-PAGE and stained for characteristic organelle marker proteins after Western blotting. As shown in Figure 2, indicated by the high abundance of FasL, CD63 and cathepsin D, SL were enriched in fraction 2. Although LAMP-1 was also enriched in this fraction, the presence of this lysosomal membrane protein in other fractions might indicate the complex composition of the lysosomal compartment in general and that other lysosome-related vesicles might exist with distinct Figure 1 In T cell blasts, FasL associates with lysosomal vesicles. PHA blasts (d14) were fixed and stained for FasL with NOK1 and Alexa Fluor488-conjugated donkey anti-mouse IgG or for granzyme A with GrA-11 FITC-conjugated mAbs or for cathepsin B with polyclonal goat anti-cathepsin B (N-19) antibodies and Alexa Fluor488-conjugated donkey anti-goat IgG. After extensive washing, all samples were stained for CD63 with Alexa Fluor555-conjugated mAb MEM-259. Nuclei were visualized by DAPI (bar: 10 μm). biophysical properties that separate at different media densities. As further indicators for the effective organelle enrichment and separation, we used cytochrome oxidase subunit IV (CoxIV) as a marker for mitochondria (see enriched organelles and fraction 5 in Figure 2) and pancadherin as a marker for the plasma membrane (only present in whole cell lysates). Of note, all proteins that were enriched in separate fractions were of course also present in the enriched organelle (EO) fraction placed on the gradient. However, due to the the relatively low abundance of individual proteins in the EO fraction, Western blot detection at the displayed exposure time did only reveal very faint bands. This is in agreement with our previous report [11] in which we showed a massive enrichment of FasL in fraction 2 while in the starting EO material from different T cell populations, FasL was almost not detectable at the same exposure time.
Regarding the "purity" of the obtained fraction, it should be stressed that most if not all enrichment protocols published so far do not allow a "purification" rather than an "enrichment" of a given organelle population. This is presumably based on the fact that lysosome formation and protein loading is a highly dynamic process that implies fusion and fission of several membraneous compartments and a complex protein sorting and transport machinery. For the initial characterization of enriched SL [11], we already pointed to potential "contaminations" in fraction 2, using antibodies against EEA1, a putative marker for endosomes, or Bip/Grp78, a marker for ER, respectively. Interestingly, during these analyses, golgin, a marker for the golgi apparatus/cisternae was only detected in fractions 3-6, but not in fractions 1 and 2 [11]. For the present study, we thus restricted ourselves to routinely check for the marker proteins depicted in Figure 2.

The enriched SL fraction consists of homogeneous intact vesicles
In addition to the biochemical analysis of the individual fractions, we visualized the obtained lysosomal fraction 2 by electron microscopy in comparison to the putative mitochondrial fraction 5. Figure 3 provides characteristic overview pictures of the two fractions. In both cases, the organelles within one fraction display a high degree of homogeneity with respect to their morphology ( Figure  3A,C). At higher magnification, the characteristics of the organelles in fraction 2 become apparent. These membranous vesicles are round-shaped with a maximum size of about 700 nm and display a characteristic electron density. In contrast, organelles of fraction 5 are characterized by irregular internal membranous structures ( Figure 3B,D) as expected for mitochondria.

The luminal proteome of enriched SL as analyzed by 2D-PAGE and mass spectrometry
In order to obtain a comprehensive list of putative luminal proteins of secretory lysosomes, enriched fraction 2 vesicles of PHA-stimulated T lymphoblasts were subjected to 2D-PAGE. More than 1600 spots from 6 replicate gels were subsequently subjected to proteolytic  cleavage and lead to the mass spectrometric identification of 1335 spots. Due to repetitive identifications at respective spot locations in different gels, the actual number of identified individual spots decreased to 742. The resulting proteome map is shown as an overview in Figure 4. Additional information on identified proteins and images of individual quadrants to match proteins to respective spots are given as additional files 1, 2 and 3 (Table S1, FigureS1, Dataset S1). Multiple (up to six) identifications in separate gels from individual secretory lysosomes preparations from T cells of different donors also underscore the reproducibility of the isolation protocol [11]. Overall, the identified spots represent a total of 397 separate protein entries in the NCBI database that are listed according to their protein names, the predicted subcellular distribution and function in Table 1. Importantly, based on database annotations combining proteome analyses of different organelles [17], 70% of the 397 proteins were assigned to lysosomal or secretory vesicles (including cytolytic granules (CG), lysosomes (LY), exosomes (EX), endosomes (EN), melanosomes (ME), platelet granules (PL) and synaptosomes (SY)) (Table 1, Figure 5). The majority of the remaining 30% was classified as proteins of unknown (11%) or cytosolic (11%) localization, and as cytosolic or nuclear proteins (CY/NU, 3.5%). The low percentage of mitochondrial (MT, 1.5%), nuclear (NU, 0.8%), plasma membrane (PM, 0.3%) or endoplasmic strictly reticulum-associated          (ER, 1.5%) and peroxisomal proteins (PE, 0.3%) again underscores the selective enrichment of lysosomal organelles in the present study. In terms of function, the classification revealed a large heterogeneity and a broad spectrum of potential activities. However, as expected, proteins associated with degradation, signal transduction, trafficking and immunity formed about 35% of the total proteome of enriched SL ( Figure 5B). The important role of these organelles in cytotoxicity is also supported by the identified effector molecules perforin (#280) and granzyme A (#707, 717, 720, 724). Interestingly, and in contrast to the published SL proteome of NK cells [12], we did not detect significant amounts of granzyme B at the respective position in 2D gels from T cell blasts. However, this is in line with our previous observation that granzyme B might be stored in a separate compartment formed by electron dense granules that do not contain transmembrane FasL and that sediment as fraction 6 in our enrichment gradient [11]. To prove this result and address this issue in more detail, we started to analyze fraction 2 and fraction 6 vesicles (granzyme B granules). The direct comparison of the two granule populations by 2D DIGE and Western blotting clearly verified the result of the present analysis and provided first biochemical and proteomic evidence for two distinct species of cytotoxic effector vesicles in T cell blasts [18].
Surprisingly, it is still unknown to date whether functionally distinct TCRαβ and TCRγδ T cells, CD4 + and CD8 + T cells, vδ1 + and vδ2 + T cells, or normal and leukemic T cells also differ in terms of protein content and function of their lysosomal compartment(s). Based on the present description of the luminal proteome of FasL-containing secretory lysosomes in fully differentiated T cells, it will be possible to directly compare the content of cytotoxic effector organelles in different T cell subpopulations, e.g. by 2D difference gel electrophoresis. In addition, based on a larger set of marker proteins, the maturation of effector vesicles in the course of T cell activation can now be addressed in detail. Of note, using the applied protocol, we identified the luminal rather than the membrane proteome of this vesicular population. In addition, one has to consider that due to methodological limitations, the applied 2D technique might cover only about 20-30% of the total proteome and thus might be complemented in future studies employing LC-coupled mass spectrometric approaches.

Conclusion
We provide the first comprehensive proteome map of T cell-derived secretory lysosomes with only minor contaminations by cytosolic, nuclear or other proteins. This information will be useful to more precisely address the activation-dependent maturation and the specific distribution of effector organelles and proteins in individual T or NK cell populations in future studies.

Cells
Human peripheral blood mononuclear cells (PBMC) were isolated from buffy coat preparations by Ficoll density gradient centrifugation. For the generation of PHA-stimulated lymphoblasts, T cells were purified by magnetic cell sorting (MACS) using cell isolation kits from Miltenyi Biotech (Bergisch Gladbach, Germany). The cells were stimulated with phytohemagglutinin A (PHA, 0.5 μg/ml, Remel, Lenexa, KS, USA) and expanded in the presence of irradiated EBV-transformed B cells and allogenic PBMC and subsequently with recombinant interleukin 2 (rIL-2, 100 U/ml, Chiron GmbH, Marburg, Germany). Before the cells were analyzed on day 12-14, dead cells were removed by Ficoll-gradient centrifugation resulting in a > 98% pure T cell population as judged by CD3 FACS analysis.

Subcellular fractionation
For subcellular fractionation and enrichment of secretory lysosomes, at least 4x10 8 T cells were used. The fractionation procedure has been recently described in detail [11]. Briefly, the cells were mechanically disrupted and organelles were enriched by differential centrifugation steps. The enriched organelles were then loaded on a discontinuous density gradient (4.4 ml volume) with 27%, 22.5%, 19%, 16%, 12%, 8% Optiprep ® which is a 60% Iodixanol solution (Sigma, Deisenhofen, Germany) and subjected to ultracentrifugation. Interphases were collected from the top of the gradient resulting in six 400 μl fractions named and numbered 1 to 6. The protein content in each fraction was determined using a Coomassie Protein Assay Reagent (Thermo, Rockford, IL, USA).

Transmission electron microscopy
Enriched organelles of fractions 2 and 5 were fixed with a mixture of 3% paraformaldehyde and 0.05% glutaraldehyde in PBS at 4°C overnight, washed in PBS, postfixed in 2% OsO 4 , dehydrated in ethanol, and embedded in araldite (Sigma, Deisenhofen, Germany). Ultrathin sections were mounted on formvar-coated grids and double-stained with a saturated solution of uranyl acetate in 70% methanol and lead citrate. The grids were examined with a Zeiss EM 900 transmission electron microscope equipped with a digital camera system.

2D electrophoresis, image analysis and spot picking
The 2D electrophoresis was performed as described before [11]. Briefly, SL pellets of fraction 2 were lysed on ice for 30 min with 30 μl lysis buffer (pH 8.5) containing 7 M urea, 2 M thiourea, 30 mM Tris, 4% CHAPS. The supernatant was recovered after centrifugation for 20 min at 20.000 × g at 4°C. A total amount of 250 μg of protein was mixed with rehydration buffer (7 M urea, 2 M thiourea, 4% CHAPS, 2% (v/v), IPG buffer pH 3-11 and 2% (w/v) DTT) and applied by cuploading onto 24 cm non-linear pH 3-11 IPG gel strips for isoelectric focusing (IEF). The second dimension was performed on 26 × 20 cm large 12.5% polyacrylamide gels after reduction and alkylation using the Ettan DALTsix large vertical electrophoresis system from GE Healthcare. The gels were removed from the glass plates, stained with Flamingo Pink (Bio Rad), mounted on a non-backed gel frame, scanned on a Typhoon Trio imager (GE Healthcare) and analyzed using Image Master 6.0 (GE Healthcare). Selected spots were picked with a 2 mm picking head. The picked gels were again scanned to verify the correct location of the punched spots.
In-gel tryptic digestion and mass spectrometry Gel plugs were washed with water and 12.5 mM ammonium bicarbonate (ABC) in 50% acetonitrile (ACN) and dehydrated in pure ACN. The dry gel pieces were rehydrated with 100 ng sequencing-grade trypsin (Serva, Heidelberg, Germany) in 5 mM ABC and tryptic in-gel digestion was performed at 37°C overnight. For peptide extraction, 0.3% trifluoroacetic acid (TFA) in ACN was added and the samples were sonicated for 15 min. The liquid phases were collected, lyophilized, redissolved in 0.5 to 1 μl MALDI matrix solution (3.2 mg/ml α-cyanohydroxycinnamic acid (Sigma) in 65% ACN/0.1% TFA), spotted onto 192-well stainless steel MALDI plates and air-dried. The samples were analyzed by peptide mass finger printing in positive reflectron mode followed by MSMS analyses of the most apparent five peptides using the 4700 Proteomics Analyzer mass spectrometer (Applied Biosystems, Framingham, MA, USA) as described elsewhere [12]. Peptide mass spectra were processed by internal calibration with autolytic fragments of porcine trypsin with 25 ppm mass tolerance. MSMS spectra were acquired using default calibration updated prior to the run. Spectral data were searched against human proteins in the NCBI database (Homo sapiens, 192,176 entries) using MASCOT V2.0 (Matrix Sciences, London, UK).

Database analysis
Database searches with MASCOT were performed using the following parameters: the modification on cysteine residues by carbamidomethylation was set as obligate, methionine oxidation was considered as a potential modification; the maximum number of missed tryptic cleavages was one; the monoisotopic masses were considered and the mass tolerance was set to ± 50 ppm, and the fragment-ion mass tolerance was set to 0.2 Da (MS/MS). A protein was accepted to be identified when the total protein score reached or exceeded the MASCOT score threshold (≥ 65 with a probability of identification greater 95%). A repeated search against a randomized decoy database (http://www.matrixscience. com/help/decoy_help.html) using the decoy.pl script and identical search parameters let to a false-positive rate of 1.2%.
The classification according the localization and function of individual proteins was based on the Uni-Prot knowledge base, the iHOP database [19] and the iProXpress database [17] available through the Protein Information Resource (PIR) (GUMC, Washington DC, USA). Identified proteins were searched in this organelle-proteome reference dataset according to their Uni-Prot numbers.

Additional material
Additional file 1: Table S1. List of identified spots in enriched SL preparations from activated T cells. 742 spots representing 397 proteins were identified and annotated according to Figure S1 A-D. Proteins (3) are listed with spot numbers (1), the number of iterant identifications (2), respective NCBI accession (4) and Uni-Prot (5) numbers, theoretical molecular weights (MW) (6) and isoelectric points