This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Proteomic Analysis of HumanNeutrophil Granules*□S
George Lominadze‡, David W. Powell§, Greg C. Luerman‡, Andrew J. Link§,Richard A. Ward¶, and Kenneth R. McLeish‡¶�**
Stimulated exocytosis of intracellular granules plays acritical role in conversion of inactive, circulating neutro-phils to fully activated cells capable of chemotaxis, phag-ocytosis, and bacterial killing. The functional changes in-duced by exocytosis of each of the granule subsets,gelatinase (tertiary) granules, specific (secondary) gran-ules, and azurophil (primary) granules, are poorly defined.To improve the understanding of the role of exocytosis ofthese granule subsets, a proteomic analysis of the azuro-phil, specific, and gelatinase granules from human neutro-phils was performed. Two different methods for granuleprotein identification were applied. First, two-dimensional(2D) gel electrophoresis followed by MALDI-TOF MS anal-ysis of peptides obtained by in-gel trypsin digestion ofproteins was performed. Second, peptides from trypticdigests of granule membrane proteins were separated bytwo-dimensional microcapillary chromatography usingstrong cation exchange and reverse phase microcapillaryhigh pressure liquid chromatography and analyzed withelectrospray ionization tandem mass spectrometry (2DHLPC ESI-MS/MS). Our analysis identified 286 proteins onthe three granule subsets, 87 of which were identified byMALDI MS and 247 were identified by 2D HPLC ESI-MS/MS. The increased sensitivity of 2D HPLC ESI-MS/MS,however, resulted in identification of over 500 proteinsfrom subcellular organelles contaminating isolated gran-ules. Defining the proteome of neutrophil granule subsetsprovides a basis for understanding the role of exocytosisin neutrophil biology. Additionally, the described methodsmay be applied to mobilizable compartments of othersecretory cells. Molecular & Cellular Proteomics 4:1503–1521, 2005.
Stimulated exocytosis of membrane-bound compartmentsplays a critical role in converting inactive, circulating neutro-phils to fully activated cells capable of directed migration,phagocytosis, and killing of microbes (1). Four subsets ofmembrane-bound compartments exist in neutrophils: secre-
tory vesicles, gelatinase (tertiary) granules, specific (second-ary) granules, and azurophil (primary) granules (1, 2). Secre-tory vesicles are generated by endocytosis, resulting inmembranes rich in receptors, signaling proteins, and adhe-sion molecules, whereas their lumen contains plasma (2).Azurophil, specific, and gelatinase granules contain variousenzymes and host defense proteins in their luminal spaces,whereas their membranes contain receptors, signaling pro-teins, adhesion molecules, and enzymes (2, 3). The targetingof proteins to individual granule compartments is determinedby the timing of protein synthesis during myeloid progenitorcell differentiation, not by granule-specific sorting (4). Thus,granule proteins that are synthesized at a given stage ofcellular differentiation will be localized to the same type ofgranule (5, 6). The overlap between the timing of synthesis ofdifferent granular proteins gives rise to some overlap of con-tents among different granule subsets (6, 7).
The extent of mobilization of the three types of granuledepends on the stimulus intensity, whereas the order of mo-bilization is fixed, a process termed graded exocytosis (8, 9).Gelatinase granules are most easily mobilized followed byspecific granules and then azurophil granules. Graded exocy-tosis results in stepwise addition of granule membrane pro-teins to the plasma membrane and stepwise release of gran-ule luminal contents into the extracellular space. Thecontrolled exocytosis of neutrophil granules allows sequentialacquisition of functional responses and targeted delivery oftoxic granule proteins, thereby reducing damage to normalhost tissue.
In addition to graded granule release, neutrophil exocytosisis distinguished by two other attributes. The first is a lack ofspatial organization of granules (10–12). The random distribu-tion of granules in the cytosol of circulating neutrophils sug-gests that graded exocytosis requires a mechanism to dis-criminate among granule subsets. The second attribute iscompound exocytosis, the fusion of two or more granulesprior to their fusion with the plasma membrane (13). Com-pound exocytosis requires recognition of granule subsets astarget membranes to allow for homotypic or heterotypic fu-sion. Homotypic fusion enhances the localized delivery ofgranule contents. Heterotypic fusion allows processing ofcertain granule constituents into active forms by proteolyticcleavage (14).
The molecular mechanisms that control exocytosis of neu-
From the Departments of ¶Medicine and ‡Biochemistry and Mo-lecular Biology, University of Louisville and the �Veterans Affairs Med-ical Center, Louisville, Kentucky 40202 and the §Department of Mi-crobiology and Immunology, Vanderbilt University School ofMedicine, Nashville, Tennessee 37232
Received, May 17, 2005Published, MCP Papers in Press, June 28, 2005, DOI 10.1074/
trophil granules are poorly defined. Additionally the full com-plement of functional changes resulting from exocytosis ofdifferent granule subsets remains to be identified. A majorreason for this limited understanding of neutrophil exocytosisis an incomplete identification of membrane and luminal pro-teins of each granule subset. To address this problem, weperformed proteomic profiling of the components of azuro-phil, specific, and gelatinase granules from human neutro-phils. Two different methods for granule protein identificationwere applied. One used two-dimensional gel electrophoresis(2DE)1 followed by MALDI-TOF MS analysis of peptides ob-tained by in-gel trypsin digestion of proteins. In the other,peptides from tryptic digests of granule membrane proteinswere separated by two-dimensional microcapillary chroma-tography using strong cation exchange and reverse phasemicrocapillary high pressure liquid chromatography and ana-lyzed with electrospray ionization tandem mass spectrometry(2D HLPC ESI-MS/MS). Our analysis identified 286 proteinson the three granule subsets. Additionally optimal methods forprotein identification differed among granule subsets basedon their physical properties and luminal components.
MATERIALS AND METHODS
Neutrophil Isolation
Neutrophils (8 � 108 cells) were isolated from healthy donors usingplasma-Percoll gradients as described by Haslett et al. (15). Trypanblue staining revealed that at least 97% of cells were neutrophils with�95% viability. After isolation neutrophils were suspended in Krebs-Ringer phosphate buffer (pH 7.2) at 4 � 107 cells/ml and treated with10 �M diisopropyl fluorophosphate for 10 min on ice to inhibit pro-teases (16). The Human Studies Committee of the University of Lou-isville approved the use of human donors.
Subcellular Fractionation for Granule Enrichment
Neutrophil granules were enriched by centrifugation on a three-layer Percoll density gradient as described by Kjeldsen et al. (17).Briefly isolated neutrophils (4 � 107/ml) were resuspended in disrup-tion buffer containing 100 mM KCl, 1 mM NaCl, 1 mM ATPNa2, 3.5 mM
MgCl2, 10 mM PIPES, and 0.5 mM PMSF and disrupted by nitrogencavitation at 380 p.s.i. and 4 °C. The cavitate was collected andsupplemented with 1.5 mM EGTA, and nuclei and intact cells wereremoved by centrifugation at 500 � g for 5 min. The postnuclearsupernatant was layered onto a discontinuous Percoll gradientformed from three 9-ml layers of Percoll prepared in a buffer contain-ing 100 mM KCl, 3 mM NaCl, 1 mM ATPNa2, 3.5 mM MgCl2, 1.25 mM
EGTA, 10 mM PIPES, and 0.5 mM PMSF to achieve final densities of
1.050, 1.090, and 1.120 g/ml. The gradient was centrifuged at37,000 � g for 30 min in an SS-34 fixed angle rotor in a Sorvall RC-5Bcentrifuge. The separated granule fractions were recovered from thegradient interfaces by aspiration, and Percoll was removed by ultra-centrifugation of each granule subset at 100,000 � g for 90 min.
Sample Preparation for 2DE
Preparation of Whole Granules—Whole granule fractions were re-suspended in 4 ml of disruption buffer and centrifuged at 100,000 �g for 20 min to obtain a solid pellet. Buffer was removed by aspiration,and the pellets were washed briefly with deionized water to removeresidual salt.
Fractionation of Granule Proteins—To separate granule proteinsinto membrane and luminal fractions, granules were resuspended in10 volumes of 0.1 M sodium carbonate, sonicated at high setting for5 s, and subjected to three freeze-thaw cycles each followed bysonication (18). Disrupted granules were incubated on ice for 30 min,and carbonate-washed membranes were pelleted at 100,000 � g.The supernatant from the carbonate wash was concentrated by ul-trafiltration through 1-kDa-cutoff centrifugal devices (MicrosepOmega, Pall, East Hills, NY), and the retentate proteins were precip-itated using chloroform-methanol desalting-precipitation (19). Theprecipitate was dried on room air and subjected to 2DE.
To separate granule proteins based on differential solubility inammonium sulfate solution, whole granules were solubilized in 3volumes of lysis buffer containing 50 mM Tris, pH 7.2, and 2% TritonX-100 and centrifuged at 20,000 � g to remove insoluble proteins.The Triton X-100-insoluble pellet was subjected to 4–12% acrylamidegradient SDS-PAGE and MALDI MS analysis, while the cleared lysatewas used for fractionation by ammonium sulfate precipitation ofproteins. One volume of 100% saturated ammonium sulfate solutionwas added to the lysate. The resulting precipitate was pelleted (sub-fraction one), and supernatant was removed. Solid ammonium sulfatewas added to the supernatant until saturation, and the protein pre-cipitate was pelleted (subfraction two). Precipitated protein pelletswere redissolved in the lysis buffer and subjected to chloroform-methanol desalting-precipitation (19).
Protein Separation by 2DE and Identification by MALDI-TOF MS
Whole granules and granule subfractions were dissolved in 160 �lof 7 M urea, 2 M thiourea rehydration buffer (Genomic Solutions, AnnArbor, MI). Proteins in all samples were separated by 2DE usingnon-linear pH 3–10 IPG strips for the first dimension and 4–12%gradient acrylamide gels for the second dimension (Invitrogen). Thegels were visualized by colloidal Coomassie staining. Proteins wereexcised and in-gel trypsin-digested, and the resulting peptides wereanalyzed by MALDI-TOF MS using the thin film sample preparationmethod as described previously (20, 21). Protein identification wascarried out by searching peptide spectra against the National Centerfor Biotechnology Information (NCBI) database using the Mascotweb-based search engine. The search parameters used were: taxon-omy, Homo sapiens; allowed error, 150 ppm; fixed modification,carbamidomethylation; variable modification, methionine oxidation;mass values, MH�; and maximum allowed missed cleavage, 1.
Protein Identification by 2D HPLC ESI-MS/MS
Granule membranes obtained following treatment with 0.1 M so-dium carbonate, as described above, were washed in 50 mM ammo-nium bicarbonate and then resuspended by sonication in 300 �l of 50mM ammonium bicarbonate. Trypsin digestion was performed byaddition of 20 ng/ml trypsin to the suspension, and samples wereincubated on a rotator overnight at 37 °C. Residual particulate mate-
rial was removed by centrifugation at 100,000 � g, and the trypsin-generated mixture was analyzed using an approach that combinedtwo-dimensional microcapillary HPLC with ESI-MS/MS (22). All tan-dem spectra were searched against H. sapiens open reading framedatabase (human.nci) using the SEQUEST algorithm (23). For singlycharged peptides, spectra with a cross-correlation score of greaterthan 1.5 were retained, whereas for multiply charged peptides, spec-tra with a cross-correlation score of greater than 2 were retained (24).The analysis was repeated three times for each granule subset, andonly proteins identified in at least two of three experiments wereassumed to be present on granules.
Quantitation of Granule Membranes and Western BlotAnalysis of Granules
Whole granule fractions were resuspended in 2 ml of disruptionbuffer and centrifuged at 100,000 � g for 20 min to obtain a solidpellet. Buffer was removed by aspiration, and the pellets were resus-pended in 100 �l of disruption buffer. Sample volume was brought upto 1 ml by water and preincubated at 37 °C. To quantitate phospho-lipid bilayers, TMA-DPH (Molecular Probes, Eugene, OR) was addedat a final concentration of 1 � 10�7 M, and fluorescence intensity wasmonitored continuously at excitation of 350 nm and emission of 430nm for 20 s on a Hitachi 4500 fluorescence spectrometer. Proteinsassociated with equal amounts of membrane from each granulesubset were loaded onto a gel for SDS-PAGE. Proteins were sepa-rated, transferred to nitrocellulose membrane, and probed with ananti-actin antibody.
RESULTS
Prior to subjecting granules to proteomic analysis, the pu-rity of each of the three granule subsets was determined.Granule subsets were analyzed by Western blotting forCD66b (specific granule marker) and by ELISA for gelatinaseand myeloperoxidase (MPO), markers for gelatinase andazurophil granules, respectively (Fig. 1). The Western blotshowed that CD66b was present in the specific granule frac-tion but essentially absent in gelatinase granule and azurophilgranule fractions. ELISA for gelatinase showed that 75% oftotal gelatinase was present in gelatinase granules, 20% wasin specific granules, and 5% was in azurophil granules,whereas 73% of MPO was in azurophil granules, 20% was inspecific granules, and 7% was in gelatinase granules. Thisdistribution of granule markers was similar to that reported byKjeldsen et al. (17, 25). Thus, optimal levels of granule subsetenrichment were obtained.
The initial analysis of granule subset proteomes was carriedout by subjecting whole granules to 2DE. Fig. 2A shows theseparation of gelatinase granule proteins. A total of 38 pro-teins were identified by peptide mass fingerprint analysis ofMALDI MS spectral data. Thirty of these proteins were cy-toskeletal and luminal proteins. The 2D gels of specific gran-ule proteins were dominated by large amounts of luminalproteins, such as lactoferrin and lipocalin (NGAL), which werepresent in such high abundance that proteins focused poorly(Fig. 2B). A total of 26 proteins were identified by peptidemass fingerprinting. The only cytoskeletal protein identifiedwas actin (Fig. 2B). Separation of azurophil granule proteinsby 2DE revealed large amounts of highly basic proteins, such
as myeloperoxidase, that focused poorly, leading to extensivesmearing on the basic end of the gel (Fig. 2C). Only eightproteins were identified from the azurophil granule gel.
To address the problem of overabundance of luminal pro-teins, granule proteins were fractionated by one of two meth-ods prior to their separation by 2DE. First, the luminal con-tents of granules were separated from membranes by lysingthe granules in 0.1 M sodium carbonate, pH 11. At high pHluminal cationic proteins lose their positive charge and disso-ciate from negatively charged matrix (18, 26, 27). Additionallyalkaline sodium carbonate treatment has been reported toremove the actin cytoskeleton from neutrophil membranes(28). To determine what proteins were removed from granulemembranes, carbonate-soluble proteins were also subjectedto 2DE. Second, differential solubility of proteins in ammo-nium sulfate solutions was applied to granule proteins prior to2DE. Whole granules were solubilized in a buffer containingTriton X-100, and proteins were separated based on precip-itation at different concentrations of ammonium sulfate. Thefirst subfraction contained proteins precipitated in a 20%ammonium sulfate solution, whereas the second subfractioncontained the proteins soluble in 20% ammonium sulfate butinsoluble in 100% ammonium sulfate.
Fractionation of gelatinase granules with sodium carbonateprior to 2DE resulted in identification of 26 proteins of whicheight proteins were not detected on whole granule gels (Fig. 3,A and B). On the other hand, 20 proteins identified by 2DE of
FIG. 1. Assessment of purity of granule fractions. Granule frac-tions separated by centrifugation on Percoll gradients were analyzedfor the content of CD66b (specific granule marker), gelatinase (gela-tinase granule marker), and myeloperoxidase (azurophil granulemarker). A, Western blotting analysis of CD66b content in granulefractions. B, ELISA for gelatinase. C, ELISA for myeloperoxidase.
Proteomic Analysis of Human Neutrophil Granules
Molecular & Cellular Proteomics 4.10 1505
whole granules were not observed on either carbonate-washed membrane or carbonate-soluble protein gels. Most ofthese proteins were below 30 kDa, suggesting that low mo-lecular weight proteins were lost during fractionation. Fig. 3, Cand D, shows 2DE separation of ammonium sulfate-precipi-tated proteins from gelatinase granules. This fractionationallowed identification of 38 proteins, 22 of which were notpreviously identified. Use of ammonium sulfate precipitationalso identified all but three proteins detected following frac-tionation with sodium carbonate. Ten proteins seen in wholegelatinase granule gels were not detected after either frac-tionation method, indicating that protein fractionation is com-plementary to analysis of intact gelatinase granules. A total of62 proteins were identified from gelatinase granules by 2DEprotein separation and MALDI MS (Table I).
Fractionation of specific granule proteins with sodium car-bonate resulted in identification of more proteins than extrac-tion by ammonium sulfate precipitation. Sodium carbonatetreatment prior to 2DE resulted in identification of 51 proteinsfrom membranes and 27 proteins in the supernatant of which42 proteins were not detected on gels from whole specificgranules (Fig. 4, A and B). Ammonium sulfate precipitationallowed identification of 45 proteins, only 10 of which were notseen on gels from whole granules. Ammonium sulfate precip-itation did not lead to identification of any protein not identi-fied after fractionation with sodium carbonate (Fig. 4, C andD). Collagenase and cathepsin X were identified on the wholespecific granule gels but not after extraction by sodium car-bonate. A total of 70 proteins were identified from specificgranules (Table I).
Fractionation of azurophil granule proteins by sodium car-bonate or by ammonium sulfate precipitation failed to improvevisualization and identification of proteins. The 2D gels ofcarbonate-washed membranes were largely devoid of pro-teins (only four proteins were identified), whereas 2DE ofcarbonate-soluble proteins failed to reveal proteins notfound on gels of whole granule proteins (data not shown).Likewise fractionation of azurophil granule proteins by am-monium sulfate precipitation failed to improve the numberof proteins visualized by 2DE (data not shown). Thus, onlyeight proteins from azurophil granules were identified by2DE (Table I).
For all granule subsets, the protein fraction precipitated by100% ammonium sulfate contained primarily luminal proteins.Only two proteins remained in solution after precipitation ofspecific and gelatinase granule proteins with 100% ammo-nium sulfate, MRP-14 and MRP-8, whereas the correspond-ing sample from azurophil granules was devoid of protein(data not shown). To identify any proteins that failed todissolve in the TX-100-containing buffer and thus could notbe detected in ammonium sulfate precipitated fractions,TX-100-insoluble pellets were also subjected to SDS-PAGE.Only actin, vimentin, CD11b, lactoferrin, and gelatinasewere detected in pellets from specific and gelatinase gran-
FIG. 2. 2D gels of whole granules. Neutrophil granules were sub-jected to 2DE, gels were stained with colloidal Coomassie Blue dye,the spots were excised and in-gel digested with trypsin, and theresultant peptides were analyzed by MALDI-TOF MS. A, gelatinasegranules. B, specific granules. C, azurophil granules.
Proteomic Analysis of Human Neutrophil Granules
1506 Molecular & Cellular Proteomics 4.10
ules (data not shown), whereas myeloperoxidase, cathepsinG, and lysozyme were found in the TX-100-insoluble frac-tion of azurophil granules. A total of 87 proteins were iden-tified from granule subsets by the three 2DE approaches ofwhich 56 were cytoskeletal or luminal proteins. Membraneproteins, except for CD11b and CD18, were not visualizedby 2DE.
2DE-based approaches are biased against membrane pro-teins, low abundance proteins, and proteins at the extremesof isoelectric point and molecular mass. To address theseissues, we used an approach that couples 2D HPLC withESI-MS/MS analysis (22, 24). This high sensitivity mass spec-trometry-based approach that allows direct analysis of com-plex protein mixtures was applied to granule membranesfollowing removal of luminal and cytoskeletal proteins with
sodium carbonate. Only proteins present in at least two ofthree experiments or proteins also identified by 2DE-MALDIMS were considered as valid granule proteins. By this criteriaa total of 247 proteins were identified from all granule subsetsof which 48 were also identified by 2DE. Table II lists identifiedproteins by granule subset, method of identification, andfunctional classification of the protein. A total of 86 proteinswere identified only from gelatinase granules, 28 proteinswere identified from only specific granules, and 26 proteinswere identified from only azurophil granules. A number ofproteins were identified on multiple granule subsets, including79 proteins from gelatinase and specific granules, five pro-teins from specific and azurophil granules, and 62 proteinsfrom all three granule subsets. The peptide sequences de-tected by 2D HPLC ESI-MS/MS can be found in Supplemen-
FIG. 3. 2D gels of gelatinase granule proteins fractionated by carbonate lysis and by ammonium sulfate precipitation of TritonX-100-solubilized proteins. Shown is fractionation of gelatinase granule proteins by carbonate lysis (A and B) and by ammonium sulfateprecipitation (C and D). A, a 2D gel of proteins that remained on the membrane after granule lysis in 0.1 M sodium carbonate buffer. B, proteinsthat were solubilized in 0.1 M sodium carbonate. C, proteins precipitated in 20% ammonium sulfate solution. D, proteins that remained insolution in 20% ammonium sulfate but were precipitated in 100% ammonium sulfate solution.
Proteomic Analysis of Human Neutrophil Granules
Molecular & Cellular Proteomics 4.10 1507
TABLE IProteins identified from 2D gels by MALDI-TOF MS
Eighty-seven proteins were identified from all 2D gels of neutrophil granules. The table lists the gene and protein designation, granule type,number of peptides matched, the percent protein coverage by these peptides, and the calculated molecular weight search (MOWSE) score foreach protein. SOD, superoxide dismutase; GE, gelatinase; SP, specific; AZ, azurophil; VDAC, voltage-dependent anion channel; ARP,actin-related protein; LSP1, lymphocyte-specific protein 1; BP, binding protein; TOLLIP, Toll-interacting protein; Arc, ARP-related complex;Grp, glucose regulated protein; Hsp, heat shock protein.
Entrez gene Entrez ID Protein name GranulePeptides
tal Tables 1 and 2. Supplemental Table 1 lists the proteinsidentified at least twice in three experiments and the se-quences of corresponding peptides. Proteins identified oncein a given granule subset by 2D HPLC ESI-MS/MS and iden-tified by either 2DE or at least twice by 2D HPLC ESI-MS/MSin another granule subset are listed in Supplemental Table 2.Sequences of corresponding peptides and the percent cov-erage of identified proteins are also shown.
To eliminate proteins present as contaminants of granulesubsets, proteins that were detected by only one of the three2D HPLC ESI-MS/MS experiments were classified as rejectedproteins. Supplemental Table 3 lists the rejected proteins foreach granule subset. A total of 344 proteins were rejectedfrom gelatinase granules, 211 were rejected from specificgranules, and 230 were rejected from azurophil granules.Most rejected proteins were of ribosomal, nuclear, and mito-chondrial origin or were hypothetical proteins, suggesting thatthe granule isolation method resulted in contamination with
small amounts of other subcellular organelles. That somecontamination did occur is supported by identification of sev-eral proteins listed in Table II. The benzodiazepine receptorwas described previously to be present on mitochondria (29).Lamin B receptor and Sad1 are of nuclear origin (30, 31),suggesting that membranes from these structures contami-nated granule preparations. The presence of signaling pro-teins (G protein-coupled receptor kinase 6; p21-activated ki-nase 5; small GTPases of the Ras, Rab, and Rho family; andheterotrimeric G protein subunits) and membrane traffickingproteins (syntaxin 2, syntaxin 3A, syntaxin 8, syntaxin 10, andsyntaxin 11) in the list of rejected proteins suggests the ex-clusion criteria eliminated some proteins that are likely com-ponents of the granule proteome. Functional classification ofproteins was performed by a literature search for each proteinin the PubMed database.
Receptors and Cytoskeletal Membrane Anchors—Thisgroup included transmembrane proteins involved in adhesion,
TABLE I—continued
Entrez gene Entrez ID Protein name GranulePeptides
transmigration, cellular activation, and cytoskeletal anchoringto the membrane, including components of lipid rafts. Of the40 proteins, 12 were identified only from gelatinase granules,19 were found on both specific and gelatinase granules, sixwere found on all three granule types, and three were foundonly on azurophil granules. Proteins unique to specific gran-ules were not found in this group. Formyl peptide receptors,complement component receptor 1 (CR1), CD11b, CD18,SCAMPs, Stomatin, CD63, and LAMPs have been identifiedpreviously as components of neutrophil granules (1, 2, 32).
Channels and Transporters—Proteins that function in facil-itating the movement of solutes across lipid bilayers wereincluded in this category. Thirteen proteins were identifiedonly on gelatinase granules, 10 were from gelatinase and spe-cific granules, two were from all three granule subtypes, and
two were only from azurophil granules. These proteins includedproton pumps and transporters for metal ions, glucose, adeninenucleotides, amines, and steroids. Eleven of these proteinswere described previously to be of mitochondrial origin.
GTPases—This group included both monomeric and het-erotrimeric GTPases and two GTPase-activating proteins. Ofthe total of 21 identified proteins, 10 were found on gelatinasegranules only, one was only from specific granules, twowere from specific and gelatinase granules, and eight werefrom all three granule subsets. Monomeric GTPases of theRab family and Cdc42 have been shown to be involved ingranule exocytosis in other cell types (33–35). Of note,Rab27A was shown previously to associate with many typesof granules and play a crucial role in exocytosis by modu-lating granule binding with cytoskeleton and with proteins
FIG. 4. 2D gels of specific granule proteins fractionated by carbonate lysis and by ammonium sulfate precipitation of TX-100-solubilized proteins. Shown is fractionation of specific granule proteins by carbonate lysis (A and B) and by ammonium sulfate precipitation(C and D). A, a 2D gel of proteins that remained on the membrane after granule lysis in 0.1 M sodium carbonate buffer. B, proteins that weresolubilized in 0.1 M sodium carbonate. C, proteins precipitated in 20% ammonium sulfate solution. D, proteins that remained in solution in 20%ammonium sulfate but were precipitated in 100% ammonium sulfate solution.
Proteomic Analysis of Human Neutrophil Granules
1510 Molecular & Cellular Proteomics 4.10
that control membrane fusion (36). Rab27A was a compo-nent of all three granule types. To our knowledge this is thefirst report of the presence of Rab27A in neutrophils. Cdc42was also found on all three granule subsets. This Rho familyGTPase stimulates exocytosis by activating the inositol1,4,5-trisphosphate and calcium second messenger signal-ing pathway (35).
Structural Proteins and Adaptors—We included cytoskel-etal proteins and actin-binding proteins in this group. Elevenproteins were identified only on gelatinase granules, 13 werefrom gelatinase and specific granules, four were from specificgranules only, 10 were from all three granule types, and twowere from only azurophil granules. The presence of actin,tubulin, and vimentin on all three granule subtypes is consist-ent with a role for these cytoskeletal structures in neutrophilexocytosis (1, 2, 37, 38). Visual inspection of 2D gels sug-gested that greater amounts of cytoskeletal proteins wereassociated with gelatinase granules than specific granules,which in turn had more than azurophil granules. To confirmthis observation, Western blotting of whole granules for actinwas performed following normalization for membrane contentusing a membrane-binding fluorescent dye (Fig. 5). Theamount of actin associated with gelatinase granules was dra-matically greater than that associated with specific granules.A minimal amount of actin was associated with azurophilgranules.
Kinases and Phosphatases—This group included tyrosinekinases, tyrosine phosphatases, and one kinase scaffoldingprotein, MEK1-binding protein. Tyrosine kinases were identi-fied previously on neutrophil granules, and inhibition of tyro-sine kinase activity prevents neutrophil degranulation (39–42).To our knowledge, this is the first report of a tyrosine phospha-tase association with neutrophil granules. Tyrosine phosphata-ses are localized to granules in other cells, and they have beenshown to participate in signaling leading to exocytosis (43–46).
Luminal and Host Defense Proteins—This group includedproteins localized to granule lumens and luminal and mem-brane proteins that participate in host defense, such as themembrane proteases neprilysin and leukolysin. Four proteinswere identified on gelatinase granules only, 13 proteins werefrom gelatinase and specific granules, 11 were proteins onlyfrom specific granules, 18 were from all three granule subsets,three were from specific and azurophil granules, and 13 werefrom azurophil granules only. All of the luminal proteins in thislist are annotated on the NCBI database as proteins destinedeither to lysosomal or to secretory compartments. Proteins oflysosomal origin were found previously in azurophil granulesin agreement with the hypothesis that azurophil granules arelysosome-related organelles (1, 2). Among the identified pro-teins in this group was calreticulin, a component of the endo-plasmic reticulum (47). Calreticulin may also be a bona fidegranule protein, however, as it acts as a chaperone for neu-trophil granule luminal proteins (48) and is present on theextracellular surface of the neutrophil plasma membrane (49).
Calreticulin was identified as a cytolytic component of T lym-phocyte granules (50).
Membrane Traffic and Fusion—In this group four proteinswere found on gelatinase granules only, one was from gela-tinase and specific granules, two were from specific granulesalone, four were from all three granule types, and two werefrom azurophil granules only. Hunc18b (Munc18-2) andUnc-13 homolog 3 (Munc-13) have not been previously iden-tified in neutrophils. These proteins are members of a family ofproteins that bind to and modulate the activity of membranefusion proteins (51, 52). A hypothetical protein with a C2domain, which is known to be involved in membrane fusionevents (53), was identified on gelatinase granules. Syntaxin 7and VAMP 8 were expressed on all three granule subsets.These two membrane fusion proteins have not been previ-ously identified from neutrophil granules.
Redox Proteins—Proteins involved in redox reactions, in-cluding a transmembrane protein related to thioredoxin andthe components of the neutrophil NADPH oxidase, p22-phoxand gp91-phox, were included in this group. Ten proteinswere found only on gelatinase granules, one was from gela-tinase and specific granules, two were from specific granulesalone, and three were from all three granules. Components ofthe NADPH oxidase were found on all three granule subsets.
Miscellaneous Proteins—This group included proteins notclassified into the other groups or for which no function hasbeen described. Twenty-one proteins were identified on gel-atinase granules, 16 were from gelatinase and specific gran-ules, six were from specific granules only, 10 were from allthree granule types, two were from specific and azurophilgranules, and four were from azurophil granules only. Amongthe proteins found were chaperones such as PDI and Hsp70and 11 hypothetical proteins. Hemoglobin was also found ongelatinase and specific granules, suggesting that hemoglobinfrom lysed red blood cells binds to phospholipid membranes.Hemoglobin was completely removed from granule mem-branes by sodium carbonate treatment.
DISCUSSION
Using two approaches, 2DE followed by MALDI MS and 2DHPLC ESI-MS/MS, 286 granule and granule-associated pro-teins from human neutrophils were identified. Only 87 proteinswere identified by 2DE and MALDI MS of which 56 wereabundant structural or luminal proteins. The only transmem-brane-spanning proteins identified after 2DE were the inte-grins CD11b and CD18, both of which have a single trans-membrane domain. The limited protein identification usingintact granules was marginally enhanced by fractionation ofgranule proteins using sodium carbonate treatment or ammo-nium sulfate precipitation. These results emphasize the rec-ognized limitations to 2DE-based proteomics, namely, thebias against low abundance proteins, transmembrane pro-teins, and proteins at the extremes of isoelectric point andmolecular mass (54, 55). Despite these limitations, 39 proteins
Proteomic Analysis of Human Neutrophil Granules
Molecular & Cellular Proteomics 4.10 1511
TABLE IIAll proteins identified on neutrophil granules by 2DE MALDI-TOF MS and SCX-RP ESI-MS/MS
A total of 286 proteins were identified, and proteins were catalogued by function based on a literature search for each protein in the PubMeddatabase. The gene and protein designation, method of identification, and granule subset on which the protein was found are listed. To belisted as a component of the granule subset, a protein was either identified from a granule subset by MALDI-MS analysis of 2D gels, detectedby at least two of three 2D HPLC ESI-MS/MS experiments, or detected by 2D HPLC ESI-MS/MS experiment once in one granule subset andadditionally detected by 2DE MALDI-MS or at least twice by 2D HPLC ESI-MS/MS in a different granule subset (indicated by an asterisk nearthe method designation in “Granule typeMETHOD” column). In the method designation, LCMS is an abbreviation for 2D HPLC ESI-MS/MS, and2D is an abbreviation for 2DE MALDI-TOF MS. SOD, superoxide dismutase; hnRNP, heterogeneous nuclear ribonucleoprotein; GE, gelatinase;SG, specific granule; AG, azurophil granule; ICAM, intracellular adhesion molecule; CEACAM, carcinoembryonic antigen-related cell adhesionmolecule; ADAM, a disintegrin and metalloprotease; LSP1, lymphocyte-specific protein 1; SNAP, synaptosomal-associated protein; ARP,actin-related protein; VAMP, vesicle associated membrane protein; Grp, glucose-regulated protein; Hsp, heat shock protein.
Entrez gene Entrez protein Protein name Granule typeMETHOD
Receptors and transmembranecytoskeletal anchors
BZRP NP_000705 Benzodiazepine receptor GELCMS
FCGR3B NP_000561 CD16, Fc�RIIIb GELCMS
CR1 NP_000564 CD35, CR1 GELCMS
NP_000642SPN NP_003114 CD43, sialophorin GELCMS
CD9 NP_001760 CD9, motility-related protein GELCMS
FLOT2 NP_004466 Flotillin 2 GELCMS
LBR NP_919424 Lamin B receptor GELCMS
NP_002287GP9 NP_000165 Platelet glycoprotein IX GELCMS
LOC388015 XP_370776 Similar to RTI1 GELCMS, SPLCMS
HRNR XP_373868 Similar to hornerin GELCMS, SPLCMS*
NP_001009931
Proteomic Analysis of Human Neutrophil Granules
Molecular & Cellular Proteomics 4.10 1517
were identified only by the 2DE-based approach.The 2DE analysis of whole granules revealed that the quality
of protein separation paralleled the density of the lumen ma-trix for each granule subset (1, 2, 17, 25). This is likely due tothe presence of large amounts of basic proteins in the granulelumens that interfere with IEF and render the less abundantmembrane-associated proteins unable to focus. This obser-vation led to attempts to fractionate granule proteins withalkaline sodium carbonate or precipitation with various con-centrations of ammonium sulfate. The sodium carbonatemethod fractionated granule proteins largely into luminal andmembrane fractions. This method resulted in effective 2DEseparation of specific granule proteins possibly becausethese granules contain large amounts of moderately cationic
luminal proteins that can be dissociated from the luminalmatrix and granule membranes. The less dense granule ma-trix of gelatinase granules was associated with adequate pro-tein resolution by 2D gels, resulting in minimal improvementafter sodium carbonate extraction of luminal proteins. Sodiumcarbonate extraction did not improve separation and identifi-cation of proteins from azurophil granules, which contain ahighly packed matrix of acid mucopolysaccharide and highlycationic myeloperoxidase (56). Sodium carbonate treatmentof gelatinase, but not specific granules, also resulted in thesignificant loss of low molecular weight proteins comparedwith 2DE of whole granules (compare Fig. 2A with Fig. 3, Aand B).
The failure of sodium carbonate extraction to improve pro-tein identification of gelatinase or azurophil granulesprompted us to seek an alternative method of protein frac-tionation based on a different physical characteristic. Ammo-nium sulfate precipitation fractionates proteins according totheir solubility in concentrated ammonium sulfate solutions.For neutrophil granules, this method separated cytoskeletaland cytoskeleton-binding proteins from more hydrophilic lu-minal proteins. This fractionation method enhanced proteinseparation and identification of gelatinase granule proteins(Fig. 3, C and D). Fractionation of specific and azurophil
FIG. 5. Assessment of actin content on neutrophil granules byWestern blotting. Membrane lipid content was quantitated in neu-trophil granules by TMA-DPH lipophilic dye, and granule proteinswere loaded such that amounts of granule membranes loaded wereequal among different granule subsets. Proteins were transferred tonitrocellulose and probed for actin with an anti-actin antibody.
TABLE II—continued
Entrez gene Entrez protein Protein name Granule typeMETHOD
SEMG1 NP_002998 Semenogelin I GELCMS*, SP2D
NP_937782SEMG2 NP_002999 Semenogelin II GELCMS*, SP2D
TMEM30A NP_060717 Hypothetical protein FLJ10856 SPLCMS, AZLCMS
DDX4 NP_061912 DEAD (Asp-Glu-Ala-Asp) box polypeptide 4 AZLCMS
ARL10B NP_620150 Hypothetical protein BC015408 AZLCMS
FLJ22662 NP_079105 Hypothetical protein FLJ22662 AZLCMS
SQSTM1 NP_003891 Sequestosome 1 AZLCMS
Proteomic Analysis of Human Neutrophil Granules
1518 Molecular & Cellular Proteomics 4.10
granule proteins by ammonium sulfate precipitation did notimprove protein separation or identification. This was likelydue to the fact that most of the protein content in thesegranules was of luminal origin and, therefore, hydrophilic.Such proteins precipitate at high concentrations of ammo-nium sulfate, and therefore they precipitated in the 100%ammonium sulfate solution. More hydrophobic proteins,which were mostly cytoskeletal and were less abundant, pre-cipitated in the 20% ammonium sulfate solution. The use ofother concentrations of ammonium sulfate did not result inimproved fractionation of granule proteins (data not shown).Thus, sodium carbonate treatment was more effective foridentification of specific granule proteins, whereas ammoniumsulfate precipitation was more effective for identification ofgelatinase granule proteins. Neither approach improved pro-tein identification from azurophil granules. These results indi-cate that optimal protein extraction and separation of granuleproteins by 2DE is highly dependent on the properties of thegranule matrix and luminal proteins.
To overcome the limitations of 2DE-based protein identifi-cation, granule proteins were also identified by strong cationexchange reverse phase two-dimensional chromatographycombined with ESI-MS/MS, commonly referred to as directanalysis of large protein complexes (DALPC) or multidimen-sional protein identification technology -spanning and mem-brane-associated proteins, sodium carbonate extraction wasused to remove luminal and cytoskeletal proteins (18, 26–28,58). As shown by the comparison of Fig. 2 with Figs. 3 and 4,the use of sodium carbonate significantly reduced high abun-dance luminal proteins, whereas the reduction in cytoskeletonwas modest. 2D HPLC ESI-MS/MS identified 247 proteinsfrom the three granule subsets.
The distribution of proteins among the granule subsetsindicates that approximately half were present on more thanone granule subset. The presence of proteins in more thanone granule subset could be due to neutrophil granule bio-genesis or to cross-contamination among the granule subsetsfollowing separation by Percoll gradient centrifugation. Gran-ule biogenesis occurs during myeloid cell maturation withdifferent granules forming at different stages of differentiation.The protein content of these granules is determined by thetiming of protein synthesis relative to formation of differentgranule subsets, not by selective targeting of proteins todifferent granules (1–7). The relative purity of granule subsetpreparations was addressed to determine the quality of theproteome for each granule. The distribution of markers foreach granule subset showed granule separation similar to thatreported by other groups (17, 25). Western blot analysis forCD66b, a marker for specific granules, showed that gelatinaseand azurophil granules were not contaminated by specificgranules. The presence of gelatinase in the specific granulepreparation is consistent with a previous study showing that60% of myeloperoxidase-negative granules contain gelatin-ase in addition to a marker of specific granules, lactoferrin (1,
2). The presence of myeloperoxidase in specific and gelatin-ase granules may be due to the extremely basic nature of thisprotein, which allows myeloperoxidase released from azuro-phil granules to bind to phospholipid membranes of othergranules. Based on current separation techniques and knowngranule markers, however, minimal cross-contaminationamong the granule subsets cannot be excluded.
We attempted to establish a valid list of proteins for eachgranule type by rejecting proteins that were identified in onlyone of three 2D HPLC ESI-MS/MS experiments. A total ofover 500 proteins, including nuclear, mitochondrial, riboso-mal, and cytosolic proteins, were rejected using this exclusioncriterion. A few contaminating proteins may have still beenpresent in granule proteomes as several mitochondrial mem-brane transporters were identified in more than one granulepreparation, especially in the lighter gelatinase granule frac-tion. This is likely due to the fact that gelatinase granulessediment at a density of 1.08 g/ml in Percoll (17, 25), whereasmitochondria sediment at a density of 1.05 g/ml (59). Histoneswere also identified in all three granule subsets. Their pres-ence may reflect nuclear contamination from the reportedbreakage of 16% of nuclei during nitrogen cavitation of neu-trophils (17, 25) and subsequent nonspecific binding to gran-ule membranes. Histones recently were described to co-lo-calize with granule enzymes in neutrophil extracellular traps(60). Subcellular fractionation showed that murine macro-phage granules contained histones (61). Histones were alsosecreted by amnion cells (62). Thus, histones may be com-ponents of neutrophil granules. The list of rejected proteinscontained several proteins, including syntaxins and synapto-brevin-like proteins, that regulate granule membrane fusion.Although none of these proteins have been identified previ-ously on neutrophil granules, these results suggest the pos-sibility that some granule proteins may have been excluded bythe criteria used in this study. The presence of false positiveand false negative results indicates that defining subcellularorganelle proteomes using highly sensitive mass spectrome-try-based techniques is limited by the ability to purify theseorganelles. Ultimately confirmation that a specific protein ispresent in an organelle will require histologic studies.
Analysis of the distribution of proteins among the granulesubsets revealed differences among the subsets that suggestfunctional heterogeneity. The total number of proteins and thenumber of proteins in each functional classification, except forluminal proteins, increased from azurophil to specific to gel-atinase granules. The number of luminal proteins identifiedwas greatest in specific granules (45 proteins), whereasazurophil (34) and gelatinase (35) granules contained similarnumber of proteins. Gelatinase granules are the most easilymobilized granules in neutrophils, and, therefore, these gran-ules will undergo exocytosis at an earlier stage of neutrophilactivation than specific or azurophil granules. Gelatinasegranule membranes contain a large number of membranereceptors and adhesion molecules and are associated with
Proteomic Analysis of Human Neutrophil Granules
Molecular & Cellular Proteomics 4.10 1519
the largest amount of actin cytoskeleton and cytoskeletalregulatory proteins. These results support a role for gelatinasegranules in enhancing the plasma membrane expression ofmolecules necessary for neutrophil adherence to and migra-tion through inflamed vascular endothelium and for subse-quent chemotaxis. On the other hand, azurophil granulesprimarily fuse with phagosomes, while exocytosis is negligi-ble. These granules contain the largest number of luminalbactericidal proteins, while there is a paucity of membrane,cytoskeletal, and GTP-binding proteins. The greater complex-ity of the specific granule lumen and the presence of a signif-icant number of transmembrane and membrane-associatedproteins suggest that these granules represent a transitionalphase that can contribute to neutrophil activation throughexocytosis or provide bactericidal proteins to phagosomes.Our results also suggest a differential role of the actin cy-toskeleton in regulation of granule exocytosis. Gelatinasegranules, the most easily mobilizable granule subset, are as-sociated with the largest amount of actin cytoskeleton. On theother hand, azurophil granules are associated with very littleactin, and they fail to exhibit exocytosis unless the subplasmamembrane actin cytoskeleton is disrupted (1, 2). Once again,specific granules represent a transition between these twoextremes. Previous reports have emphasized the role of thesubplasma membrane actin cytoskeleton as a barrier to gran-ule exocytosis (63). The present study suggests that the actincytoskeleton associated with granules also plays an activerole in mediating exocytosis.
To our knowledge, this is the first proteomic study of neu-trophil granules. The importance of regulated exocytosis indetermining the activation state of neutrophils and the use ofregulated exocytosis by a large number of other cells indicatethe value of defining granule proteomes. The methodologydescribed herein provides an approach to defining granuleproteomes. In our experience, 2DE identified and revealedrelative distribution of more abundant proteins in granules,whereas 2D HPLC ESI-MS/MS identified less abundant pro-teins and hydrophobic proteins not amenable to detection by2DE. The sensitivity of this latter approach demands thatgranules be isolated to a high degree of purity and/or exper-imental approaches be used to eliminate false positive proteinidentification. Our data show that neutrophil granules arecomplex organelles. Analysis of the more than 200 proteinsassociated with neutrophil granules will allow a better under-standing of the contribution of each of the granule subsets toneutrophil biology.
* The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be herebymarked “advertisement” in accordance with 18 U.S.C. Section 1734solely to indicate this fact.
□S The on-line version of this article (available at http://www.mcponline.org) contains supplemental material.
** To whom correspondence should be addressed: Molecular Sig-naling Group, Kidney Disease Program, Donald E. Baxter Research
Bldg., 570 S. Preston St., University of Louisville, Louisville, KY40202. Tel.: 502-852-0014; Fax: 502-852-4384; E-mail: [email protected].
REFERENCES
1. Faurschou, M., and Borregaard, N. (2003) Neutrophil granules and secre-tory vesicles in inflammation. Microbes Infect. 5, 1317–1327
2. Borregaard, N., and Cowland, J. B. (1997) Granules of the human neutro-philic polymorphonuclear leukocyte. Blood 89, 3503–3521
3. Gullberg, U., Bengtsson, N., Bulow, E., Garvicz, D., Lindmark, A., andOlson, I. (1999) Processing and targeting of granule proteins in humanneutrophils. J. Immunol. Methods 232, 201–210
4. Borregaard, N., Theilgaard-Monch, K., Sørensen, O. E., and Cowland, J. B.(2001) Regulation of human neutrophil granule protein expression. Curr.Opin. Hematol. 8, 23–27
5. Borregaard, N., Sehested, M., Nielsen, B. S., Sengelov, H., and Kjeldsen, L.(1995) Biosynthesis of granule proteins in normal human bone marrowcells. Gelatinase is a marker for terminal neutrophil differentiation. Blood85, 812–822
6. Le Cabec, V., Cowland, J. B., Calafat, J., and Borregaard, N. (1996) Tar-geting of proteins to granule subsets is determined by timing and not bysorting: the specific granule protein NGAL is localized to azurophil gran-ules when expressed in HL-60 cells. Proc. Natl. Acad. Sci. U. S. A. 93,6454–6457
7. Cowland, J. B., and Borregaard, N. (1999) The individual regulation ofgranule protein mRNA levels during neutrophils maturation explains theheterogeneity of neutrophil granules. J. Leukoc. Biol. 66, 989–995
8. Sengelov, H., Kjeldsen, L., and Borregaard, N. (1993) Control of exocytosisin early neutrophil activation. J. Immunol. 150, 1535–1543
9. Sengelov, H., Follin, P., Kjeldsen, L., Lollike, K., Dahlgren, C., and Borre-gaard, N. (1995) Mobilization of granules and secretory vesicles during exvivo exudation of human neutrophils. J. Immunol. 154, 4157–4165
10. Bainton, D. F., and Farquhar, M. G. (1966) Origin of granules in polymor-phonuclear leukocytes. Two types derived from opposite faces of theGolgi complex in developing granulocytes. J. Cell Biol. 28, 277–301
11. Bainton, D. F. (1999) Distinct granule populations in human neutrophils andlysosomal organelles identified by immuno-electron microscopy. J. Im-munol. Meth. 232, 153–168
12. Robinson, J. M., Kobayashi, T., Seguchi, H., and Takizawa, T. (1999)Evaluation of neutrophil structure and function by electron microscopy:cytochemical studies. J. Immunol. Methods 232, 169–178
13. Lollike, K., Lindau, M., Calafat, J., and Borregaard, N. (2002) Compoundexocytosis of granules of human neutrophils. J. Leukoc. Biol. 71,973–980
14. Sørensen, O. E., Follin, P., Johnsen, A. H., Calafat, J., Tjabringa, G. S.,Hiemstra, P. S., and Borregaard, N. (2001) Human cathelicidin, hCAP18,is processed to the antimicrobial peptide LL-37 by extracellular cleavagewith proteinase 3. Blood 97, 3951–3959
15. Haslett, C., Guthrie, L., Kopaniak, M., Johnston, J., and Henson, P. (1985)Modulation of multiple neutrophil functions by preparative methods oftrace concentrations of bacterial lipopolysaccharide. Am. J. Pathol. 119,101–110
16. Amrein, P. C., and Stossel, T. P. (1980) Prevention of degradation of humanpolymorphonuclear leukocyte proteins by diisopropylfluorophosphate.Blood 56, 442–447
17. Kjeldsen, L., Sengelov, H., Lollike, K., Nielsen, M. H., and Borregaard, N.(1994) Isolation and characterization of gelatinase granules from humanneutrophils. Blood 83, 1640–1649
18. Howell, K. E., and Palade, G. E. (1982) Hepatic Golgi fractions resolved intomembrane and content subfractions. J. Cell Biol. 92, 822–832
19. Wessel, D., and Flugge, U. I. (1984) A method for the quantitative recoveryof protein in dilute solution in presence of detergents and lipids. Anal.Biochem. 138, 141–143
20. Powell, D. W., Rane, M. J., Joughin, B. A., Kalmukova, R., Hong, J. H.,Tidor, B., Dean, W. L., Pierce, W. M., Klein, J. B., Yaffe, M. B., andMcLeish, K. R. (2003) Proteomic identification of 14-3-3� as a mitogen-activated protein kinase-activated protein kinase 2 substrate: role indimer formation and ligand binding. Mol. Cell. Biol. 23, 5376–5387
21. Singh, S., Powell, D. W., Rane, M. J., Millard, T. H., Trent, J. O., Pierce,W. M., Klein, J. B., Machesky, L. M., and McLeish, K. R. (2003) Identifi-cation of the p16 Arc subunit of the Arp2/3 complex as a substrate of
MAPK-activated protein kinase 2 by proteomic analysis. J. Biol. Chem.278, 36410–36417
22. Sanders, S. L., Jennings, J., Canutescu, A., Link, A. J., and Weil, A. P.(2002) Proteomics of the eukaryotic transcription machinery: identifica-tion of proteins associated with components of yeast TFIID by multidi-mensional mass spectrometry. Mol. Cell. Biol. 22, 4723–4738
23. Eng, J. K., McCormack, A. L., and Yates, J. R. (1994) An approach tocorrelate tandem mass-spectral data of peptides with amino-acid-se-quences in a protein database. J. Am. Soc. Mass Spectrom. 5, 976–989
24. Link, A. J., Eng, J., Scheiltz, D. M., Carmack, E., Mize, G. J., Morris, D. R.,Garvik, B. M., and Yates, J. R. (1999) Direct analysis of protein com-plexes using mass spectrometry. Nat. Biotechnol. 17, 676–682
25. Kjeldsen, L., Sengelov, H., and Borregaard, N. (1999) Subcellular fraction-ation of human neutrophils on Percoll density gradients. J. Immunol.Methods 232, 131–143
26. Fujiki, Y., Hubbard, A. L., Fowler, S., and Lazarow, P. B. (1982) Isolation ofintracellular membranes by means of sodium carbonate treatment: ap-plication to endoplasmic reticulum. J. Cell Biol. 93, 97–102
27. Fujiki, Y., Fowler, S., Shio, H., Hubbard, A. L., and Lazarow, P. B. (1982)Polypeptide and phospholipid composition of the membrane of rat liverperoxisomes: comparison with endoplasmic reticulum and mitochondrialmembranes. J. Cell Biol. 93, 103–110
28. Nebl, T., Pestonjamasp, K. N., Leszyk, J. D., Crowley, J. L., Oh, S. W., andLuna, E. J. (2002) Proteomic analysis of a detergent-resistant membraneskeleton from neutrophil plasma membranes. J. Biol. Chem. 277,43399–43409
29. Lin, D., Chang, Y. J., Strauss, J. F., and Miller, W. L. (1993) The humanperipheral benzodiazepine receptor gene: cloning and characterizationof alternative splicing in normal tissues and in a patient with congenitallipoid adrenal hyperplasia. Genomics 18, 643–650
30. Ye, Q., and Worman, H. J. (1994) Primary structure analysis and lamin Band DNA binding of human LBR, an integral protein of the nuclearenvelope inner membrane. J. Biol. Chem. 269, 11306–11311
31. Malone, C. J., Fixsen, W. D., Horvitz, H. R., and Han, M. (1999) UNC-84localizes to the nuclear envelope and is required for nuclear migrationand anchoring during C. elegans development. Development 126,3171–3181
32. Dahlgren, C., Carlsson, S. R., Karlsson, A., Lundqvist, H., and Sjoeln, C.(1995) The lysosomal membrane glycoproteins Lamp-1 and Lamp-2 arepresent in mobilizable organelles, but are absent from the azurophilgranules of human neutrophils. Biochem. J. 311, 667–674
33. Deacona, S. W., and Gelfanda, V. I. (2001) Of yeast, mice, and men: Rabproteins and organelle transport. J. Cell Biol. 152, F21–F24
34. Zerial, M., and McBride, H. (2001) Rab proteins as membrane organizers.Nat. Rev. 2, 107–119
35. Hong-Geller, E., and Cerione, R. A. (2000) Cdc42 and Rac stimulate exo-cytosis of secretory granules by activating the IP3/calcium pathway inRBL-2H3 mast cells. J. Cell Biol. 148, 481–494
36. Izumi, T., Gomi, H., Kasai, K., Mizutani, S., and Torii, S. (2003) The roles ofRab27 and its effectors in the regulated secretory pathways. Cell Struct.Funct. 28, 465–474
37. Rothwell, S. W., Nath, J., and Wright, D. G. (1989) Interactions of cytoplas-mic granules with microtubules in human neutrophils. J. Cell Biol. 108,2313–2326
38. Pryzwansky, K. B., and Merricks, E. (1998) Chemotactic peptide-inducedchanges of intermediate filament organization in neutrophils during gran-ule secretion: role of cyclic guanosine monophosphate. Mol. Biol. Cell. 9,2933–2947
39. Gutkind, J. S., and Robbins, K. C. (1989) Translocation of the FGR protein-tyrosine kinase as a consequence of neutrophil activation. Proc. Natl.Acad. Sci. U. S. A. 86, 8783–8787
40. Mohn, H., Le Cabec, V., Fischer, S., and Maridonneau-Parini, I. (1995) Thesrc-family protein-tyrosine kinase p59hck is located on the secretorygranules in human neutrophils and translocates towards the phagosomeduring cell activation. Biochem. J. 309, 657–665
41. Mocsai, A., Banfi, B., Kapus, A., Farkas, G., Geiszt, M., Buday, L., Farago,A., and Ligeti, E. (1997) Differential effects of tyrosine kinases inhibitorsand an inhibitor of the mitogen-activated protein kinase cascade on
degranulation and superoxide production of human neutrophil granulo-cytes. Biochem. Pharmacol. 54, 781–789
42. Mocsai, A., Ligeti, E., Lowell, C. A., and Berton, G. (1999) Adhesion-de-pendent degranulation of neutrophils requires the Src family kinases Fgrand Hck. J. Immunol. 162, 1120–1126
43. Huynh, H., Bottini, N., Williams, S., Cherepanov, V., Musumeci1, L., Saito,K., Bruckner, S., Vachon, E., Wang, X., Kruger, J., Chow, C.-W., Pellec-chia, M., Monosov, E., Greer, P. A., Trimble, W., Downey, G. P., andMustelin, T. (2004) Control of vesicle fusion by a tyrosine phosphatase.Nat. Cell Biol. 6, 831–839
44. Chen, M., and Geng, J. G. (2001) Inhibition of protein tyrosine phosphata-ses suppresses P-selectin exocytosis in activated human platelets. Bio-chem. Biophys. Res. Commun. 286, 609–615
45. Ostenson, C. G., Sandberg-Nordqvist, A. C., Chen, J., Hallbrink, M., Rotin,D., Langel, U., and Efendic, S. (2002) Overexpression of protein-tyrosinephosphatase PTP sigma is linked to impaired glucose-induced insulinsecretion in hereditary diabetic Goto-Kakizaki rats. Biochem. Biophys.Res. Commun. 291, 945–950
46. Hermel, J. M., Dirkx, R., and Solimena, M. (1999) Post-translational modi-fications of ICA512, a receptor tyrosine phosphatase-like protein ofsecretory granules. Eur. J. Neurosci. 11, 2609–2620
47. Michalak, M., Corbett, E. F., Mesaeli, N., Nakamura, K., and Opas, M.(1999) Calreticulin: one protein, one gene, many functions. Biochem. J.344, 281–292
48. Nauseef, W. M., McCormick, S. J., and Clark, R. A. (1995) Calreticulinfunctions as a molecular chaperone in the biosynthesis of myeloperoxi-dase. J. Biol. Chem. 270, 4741–4747
49. Ghiran, I., Klickstein, L. B., and Nicholson-Weller, A. (2003) Calreticulin is atthe surface of circulating neutrophils and uses CD59 as an adaptormolecule. J. Biol. Chem. 278, 21024–21031
50. Sipione, S., Ewen, K., Shostak, I., Michalak, M., and Bleackley, R. C. (2005)Impaired cytolytic activity in calreticulin-deficient CTLs. J. Immunol. 174,3212–3219
51. Jahn, R., and Sudhof, T. C. (1999) Membrane fusion and exocytosis. Annu.Rev. Biochem. 68, 863–911
52. Burgoyne, R. D., and Morgan, A. (2002) Secretory granule exocytosis.Physiol. Rev. 83, 581–632
53. Rizo, J., and Sudhof, T. C. (1998) C2-domains, structure and function of auniversal Ca2�-binding domain. J. Biol. Chem. 273, 15879–15882
54. Corthals, G. L., Wasigner, V. C., Hochstrasser, D. F., and Sanchez, J. C.(2000) The dynamic range of protein expression: a challenge for pro-teomic research. Electrophoresis 21, 1104–1115
55. Gygi, S. P., Corthals, G. L., Zhang, Y., Rochon, Y., and Aebersold, R. (2000)Evaluation of two-dimensional gel electrophoresis-based proteome anal-ysis technology. Proc. Natl. Acad. Sci. U. S. A. 97, 9390–9395
56. Olsson, I. (1969) The intracellular transport of glucosaminoglycans (muco-polysaccharides) in human leukocytes. Exp. Cell Res. 54, 318–325
57. Washburn, M. P., Wolters, D., and Yates, J. R. (2001) Large-scale analysisof the yeast proteome by multidimensional protein identification technol-ogy. Nat. Biotechnol. 19, 242–247
58. Wu, C. C., MacCoss, M. J., Howell, K. E., and Yates, J. R. (2003) A methodfor the comprehensive proteomic analysis of membrane proteins. Nat.Biotechnol. 21, 532–538
59. Krungkrai, J., Krungkrai, S. R., and Bhumiratana, A. (1993) Plasmodiumberghei: partial purification and characterization of mitochondrial cyto-chrome c oxidase. Exp. Parasitol. 77, 136–146
60. Brinkmann, V., Reichard, U., Goosmann, C., Fauler, B., Uhlemann, Y.,Weiss, D. S., Weinrauch, Y., and Zychlinsky, A. (2004) Neutrophil extra-cellular traps kill bacteria. Science 303, 1532–1535
61. Hiempstra, P. S., Eisenhauer, P. B., Harwig, S. S., van den Barselaar, M. T.,van Furth, R., and Lehrer, R. I. (1993) Antimicrobial proteins of murinemacrophages. Infect. Immun. 61, 3038–3046
62. Kim, H. S., Cho, J. J., Park, H. W., Yoon, H., Kim, M. S., and Kim, S. C.(2002) Endotoxin-neutralizing antimicrobial proteins of the human pla-centa. J. Immunol. 168, 2356–2364
63. Orci, L., Gabbay, K. H., and Malaisse, W. J. (1972) Pancreatic beta-cellweb: its possible role in insulin secretion. Science 175, 1128–1130