| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Published online before print July 1, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Institut für Physiologische Chemie, Philipps-Universität Marburg, Germany; and
Zentrum Biochemie und Molekulare Zellbiologie, Abt. Biochemie II, Georg-August-Universität Göttingen, Germany
1Correspondence: Philipps-Universität Marburg, Institut für Physiologische Chemie, Karl-von-Frisch-Strasse 1, 35033 Marburg, Germany. E-mail: lemansky{at}home.staff.uni-marburg.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: chondroitin sulfate • serglycin • lactoferrin • cathepsin G • azurocidin
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Azurophilic granules contain several groups of biologically active molecules. The first group includes the microbicidal polypeptides, such as the defensins, lysozyme, bactericidal permeability-increasing protein, and MPO. Most, if not all, microbicidal polypeptides are cationic at physiologic pH and thus, are able to bind to the negatively charged surface of bacteria. For example, from the known amino acid sequence of human MPO [1 ], theoretical isoelectric points of the precursor and mature forms of MPO can be calculated as 9.26 and 9.22, respectively (Swiss-Prot data bank). The second group contains just one member, the proteoglycan serglycin, which is thought to act as a binding scaffold for basic polypeptides, thus balancing their positive charge. The third group consists of hydrolytic enzymes, which are characteristically found in lysosomes. Further, these organelles contain integral membrane proteins that are not dealt with in this study. As a result of their lack in the lysosome-associated membrane protein-1 (LAMP-1) and LAMP-2, azurophilic granules are not regarded as true lysosomes but rather as regulated, secretory granules [2 ].
Similar to lysosomal enzymes, MPO is synthesized as a larger precursor that is proteolytically converted to the mature protein upon reaching the granular compartment [3 , 4 ]. In this process, a 14-kDa propiece is removed, and the mature enzyme, consisting of a 60-kDa large chain and 13-kDa small chain, is generated. MPO contains five N-linked oligosaccharides, some of which may acquire the lysosomal recognition marker mannose 6-phosphate. It is interesting that the granular transport of MPO in HL-60 cells is independent of mannose 6-phosphate [3 , 5 ]. The transport of MPO to granules in myeloid cells has been subject to several recent studies [6 , 7 ] with the aim to elucidate the role of the propeptide in the diversion of MPO from the secretory route to granules. In these studies, MPO constructs were used, lacking the proregion or consisting of the proregion fused to secretory proteins. Unfortunately, the products were partially or completely retained in the endoplasmic reticulum (ER) and subsequently degraded by the quality-control mechanism involving proteasomes. This result showed the importance of the proregion for the folding of MPO; however, it contributed little to the elucidation of its role in granular targeting.
We therefore examined an alternative hypothesis, suggesting that targeting MPO to granules is mediated by serglycin. Toward this end, we studied the binding of human proMPO or MPO to chondroitin sulfate (CS)– or carboxymethyl (CM)–Sepharose and used the human promyelocytic HL-60 cell line, which has high contents of MPO and serglycin. In fact, serglycin was previously cloned and characterized using materials from this human cell line [8 , 9 ]. The carbohydrate side-chains of serglycin in HL-60 cells were shown to consist solely of CS-4 [10 ]. We labeled the proteoglycans of HL-60 cells with [35S]sulfate, which allowed us to monitor their transport and compared it with that of MPO. Moreover, we used the membrane-permeable, sulfhydryl-cleavable cross-linker dithiobis(succininmidylpropionate; DSP) to covalently cross-link macromolecular complexes within HL-60 cells after labeling with Tran35S-label (a mix of [35S]methionine and [35S]cysteine) or [35S]sulfate. Immunoprecipitation of MPO led to the detection of serglycin/MPO complexes en route to the azurophilic granules of these cells.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The rabbit antisera specific for human MPO and cathepsin D (CD) and the goat antiserum against human ß-N-acetylhexosaminidase (ß-Hex) were raised in our laboratory [3 , 11 ]. Goat anti-rabbit immunoglobulin G (IgG) antibodies coupled to horseradish peroxidase (HRP) were bought from Bio-Rad (Munich, Germany). A homogeneous preparation of mature human neutrophil MPO was a gift from Dr. Ragnar Olsen (University of Tromsö, Norway).
Culturing and metabolic labeling of cells
The human promyelocytic HL-60 cell line was cultured in RPMI medium containing 10% heat-inactivated fetal bovine serum (FBS; both from Gibco-BRL, Eggenstein, Germany) and supplemented with 100 units/mL penicillin and 100 µg/mL streptomycin.
Metabolic labeling was performed with Tran35S-label (11–18 MBq/ml) or [35S]sulfate (28.5–74 MBq/mL), purchased from ICN (Meckenheim, Germany) in methionine/cysteine- or sulfate-deficient RPMI medium. Both labeling media contained 10% heat-inactivated FBS, which was dialyzed against 0.9% NaCl. Sulfate-deficient medium contained 0.1 mg/mL ampicillin as an antibiotic. Before addition of the radioactive tracer, HL-60 cells were washed three times in deficient RPMI medium and kept in this medium for 0.5 and 3 h when labeled with Tran35S-label or [35S]sulfate, respectively. PMA was dissolved in dimethyl sulfoxide (DMSO) at a concentration of 0.1 mM before application to the cells (final concentration, 50 nM). Control cells were incubated with 0.05% (v/v) DMSO.
Cross-linking with DSP
HL-60 cells were washed twice with ice-cold 70 mM 3-(N-morpholino)-propanesulfonic acid (MOPS)/KOH, pH 6.8, 1 mM MgCl2, and 10 mM NaCl (MOPS buffer) and were then incubated with 1 mL MOPS buffer containing 2% DMSO, alone or containing 1 mM DSP for 10 min at 37°C. After one washing step, cells were processed for immunoprecipitation or subjected to subcellular fractionation followed by immunoprecipitation.
Subcellular fractionation of HL-60 cells by sucrose density centrifugation
HL-60 cells, suspended in MOPS buffer, were disintegrated by N2 cavitation, keeping them for 15 min under a pressure of 15 bar at 4°C. Sudden release of the pressure resulted in rupture of the plasma membranes, and cell organelles remained intact. Cell debris and nuclei were removed by a 10-min centrifugation step at 600 g at 4°C. The postnuclear supernatant was then mixed with MOPS buffer containing 67.2% (w/v) sucrose to yield a final sucrose content of 50% (w/v). This was layered under a linear 18–47% (w/v) sucrose gradient and centrifuged for 18 h at 200,000 g and 4°C. Intact organelles floated into the gradient until they reached the region of their inherent density. Cytosolic material and the contents of ruptured organelles remained at the bottom of the tube. After centrifugation, 0.9-mL fractions were collected, starting at the top of the centrifuge tube.
Immunoprecipitation
Labeling of HL-60 cells with Tran35S-label was followed by direct immunoprecipitation as described [3
], except that immunoprecipitates were washed according to Lorkowski et al. [11
]. After labeling with [35S]sulfate and subcellular fractionation, 200 µl subcellular fractions were mixed with 360 µl 50 mM Tris-HCl, pH 7.4, 145 mM NaCl (Tris-buffered saline), brought to 1 mg/mL bovine serum albumin (BSA), 2 mM phenylmethylsulfonylfluoride, 5 mM iodoacetamide (IAA), and 0.5% Triton X-100, and subjected to immunoprecipitation as described earlier [11
] with the following modifications: Each sample was preadsorbed only once with a resuspended pellet of 150 µL pansorbin. Washing the immune complexes was performed three times with 0.8 mL 10 mM Na phosphate, pH 7.4, 145 mM NaCl, 1% (v/v) Triton X-100, 0.5% (w/v) Na deoxycholate, 5 mg/ml BSA; two times with 0.8 mL 10 mM Tris-HCl, pH 8.6, 0.6 M NaCl, 0.1% (w/v) sodium dodecyl sulfate (SDS), 0.05% (v/v) Nonidet P-40; and once with 0.8 mL 5mM Tris-HCl, pH 7.4, 14 mM NaCl.
Isolation of membrane-associated proteins from dense organelles isolated from human placenta
Preparation of dense organelles from human placenta, determination of the activity of soluble- and membrane-associated lysosomal markers (ß-Hex and acid ß-glucosidase), and the isolation of membranes by ultracentrifugation were performed as described earlier [12
]. As an organ that is well supplied with blood, human placenta contains blood cells, which give rise to leukocytic granules in dense organelle preparations as detected by MPO activity. Membranes of these organelles were washed once with 500 mM NaCl, centrifuged for 1 h at 100,000 g, and extracted with 50 mM Tris-HCl, pH 7.4, 100 mM NaCl, 1% (w/v) 3-[(3-cholamidopropyl) dimethylammonio]-1-propansulfonate (CHAPS), 5 mM EDTA. The extract was diluted 1:10 with the same buffer containing 1% (v/v) Triton X-100 instead of 1% (w/v) CHAPS and subjected to CS-affinity chromatography.
Gel electrophoresis-related techniques and trichloroacetic acid precipitation of polypeptides
SDS-polyacrylamide gel electrophoresis (PAGE) was performed according to Laemmli [13
] followed by staining with colloidal Coomassie blue G-250, according to the manufacturer’s instructions (Carl Roth, Karlsruhe, Germany), or fluorography [14
]. Radioactively labeled secretions or cell lysates [lysis buffer: 50 mM Tris-HCl, pH 7.4, 145 mM NaCl, 0.5% (v/v) Triton X-100, and 5 mM IAA] were adjusted to 10% (w/v) trichloroacetic acid (TCA) and were kept on ice for at least 30 min. After a 2-min centrifugation at 14,000 g, pellets were washed three times with 0.5 mL ice-cold 5% (w/v) TCA and dissolved in 90 µL reducing Laemmli sample buffer and 5 µL 1 M Tris base. SDS-PAGE and liquid scintillation counting analyzed aliquots of this solution.
o-Dianisidine assay of MPO activity
Samples (40 µL) were incubated with 160 µL 0.8 mM o-dianisidine in 0.1 M sodium citrate, pH 5.5, containing 0.125% (w/v) Triton X-100 and 0.1 mM H2O2 at room temperature in 96-well microtiter plates. The assay was performed in duplicate and where applicable, with and without inhibition by 0.01% (w/v) NaN3, because of a peroxidase activity contained in the fetal calf serum, which was insensitive to NaN3. MPO activity was defined as NaN3-inhibitable activity and expressed in relative or absolute units. The latter are defined by the peroxidation of 1 µmol o-dianisidine per minute at 25°C, as measured at a wavelength of 450 nm.
Chromatography with CS– or CM–Sepharose CL-4B
CS–Sepharose CL-4B columns (5x8 mm) were equilibrated with 5 mM Tris-HCl, pH 7.4, and were allowed to bind pure hen egg lysozyme or human neutrophil MPO for 15 min at room temperature. The purity of both enzymes was verified by SDS-PAGE followed by silver-staining. The columns were then washed extensively with 5 mM Tris-HCl, pH 7.4, until no protein was detected in the washing buffer. Elution was performed with stepwise application of 5 mM Tris-HCl, pH 7.4, containing increasing concentrations of NaCl. The contents of lysozyme and MPO in the fractions were analyzed by photometric detection of protein at 280 nm and the o-dianisidine activity assay, respectively.
Conditioned medium from HL-60 cells containing proMPO was diluted 1:4 with 5 mM Tris-HCl, pH 7.4, and subjected to CS-affinity chromatography, as described above for human neutrophil MPO. When applied to CM–Sepharose CL-4B (5x8 mm column), PMA-induced secretions of HL-60 cells were diluted 1:4 with 10 mM Tris-HCl, pH 7.4. The chromatographic procedure was performed with this buffer as described above.
CS-affinity chromatography of placental granular membrane-associated proteins was performed in the presence of 50 mM Tris-HCl, pH 7.4, 100 mM NaCl, 1% (w/v) Triton X-100, 5 mM EDTA. Bound proteins were eluted in the same buffer containing increasing concentrations of up to 1 M NaCl.
Immunoblotting and ECL detection
HL-60 cells were extracted with lysis buffer (see TCA precipitation) for 10 min at room temperature. Nuclei were eliminated by centrifugation at 600 g, and aliquots of these lysates or secretions were mixed with 1/4 vol of fivefold-concentrated Laemmli sample buffer, boiled for 5 min at 95°C and subjected to SDS-PAGE. The separated polypeptides were transferred to nitrocellulose in a semidry blotting device (Bio-Rad) at 1 mA/cm2 for 2 h and were blocked with 5% (w/v) skim milk powder in 50 mM Tris-HCl, pH 7.8, 0.05% (v/v) Triton X-100, 0.01% (v/v) antifoam A, 2 mM CaCl2, followed by ECL detection of MPO, according to the manufacturer’s instructions (Amersham Pharmacia). Dilution of the anti-MPO antiserum and the goat anti-rabbit IgG antibody coupled to HRP was 1:3000.
Mass spectrometry
Proteins were reduced and carboxamidomethylated before SDS-PAGE. Following electrophoresis, Coomassie blue-stained proteins were excised from the gel and subjected to in gel digestion with trypsin. Peptides were eluted with 1% (v/v) TFA in 50% acetonitrile/water, dried in a Heraeus (Karlsruhe, Germany) speed-vac, and analyzed by matrix-assisted laser-adsorption time-of-flight mass spectrometry in a Reflex III instrument from Bruker Daltonik (Leipzig, Germany). The Bruker X tof 5.1.1 program version was used for the analysis of spectra, followed by identification through Mascot analysis tools (http:\\www.matrixscience.com; Matrix Science, London, UK).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Preparations of pure lysozyme and mature MPO were allowed to bind to CS–Sepharose under low ionic conditions (5 mM Tris-HCl, pH 7.4) for 15 min at room temperature. After washing, the CS–Sepharose columns were eluted with the same buffer, containing increasing concentrations of NaCl. Finally, the contents of lysozyme and MPO were determined in the fractions (Fig. 1A ). Lysozyme eluted as a single, sharp peak at 200 mM NaCl. MPO eluted as a much broader peak between 200 and 400 mM NaCl. A similar binding pattern was observed for proMPO secreted from HL-60 cells (Fig. 1A) . The results of this experiment show that binding of MPO to CS is stronger than that of lysozyme, and only a minor portion of the bound enzyme eluted from the column at the physiological salt concentration. Therefore, the contact between MPO and serglycin in a cellular compartment is likely to result in a complex formation of the two with each other.
|
Binding of MPO to CS–Sepharose was completely abolished when affinity chromatography was performed at pH 12 (not shown), which is consistent with lysines and arginines of MPO mediating the ionic interaction with CS. pH-induced denaturation of MPO as an explanation for this behavior is unlikely, as practically all MPO activity was found in the unbound fraction after neutralization.
HL-60 cells synthesize a proteoglycan that is almost completely secreted when PMA is present
To examine the fate of [35S]sulfate-labeled proteoglycans in HL-60 cells, a pulse (5 min) and chase (up to 1 h) labeling was performed with [35S]sulfate in the absence and presence of 50 nM PMA. Polypeptides from corresponding aliquots of cell lysate and the medium were precipitated with TCA. The precipitates were dissolved in sample buffer, and liquid scintillation counting analyzed one-half of each sample and SDS-PAGE and fluorography, the other.
[35S]Sulfate was readily incorporated into a high molecular mass species with most of the material, having an apparent molecular mass higher than 220 kDa (Fig. 2
). The label was associated with a broad band, as expected of a proteoglycan with varied lengths of the glycosaminoglycan chains. In the period of 30-min chase, the secretion reached its maximum, which corresponded to 
50% of the pulse-labeled material. The cellular contents of proteoglycans were rapidly depleted. The loss of total radioactivity associated with the cellular and extracellular proteoglycan indicated a rapid degradation of approximately half of the initially labeled material. Intracellular 35S-labeled proteoglycans had a half-life of 15 min (estimated from three experiments). This suggests that transport from the trans-Golgi network to a degradative compartment, likely the azurophilic granule or its precursor, is a rather rapid process.
|
Secretion of MPO is greatly enhanced by PMA
In the following experiment, we used PMA to interfere with the granular transport of proteoglycans and NaClO3, an inhibitor of sulfation [15
], to diminish the negative charge of proteoglycans. In both cases, we were interested to learn whether granular targeting of MPO would be altered.
HL-60 cells (1x106 cells/mL) were incubated with or without 50 nM PMA or 20 mM NaClO3 for 16 h. Cells were then lysed, and corresponding aliquots of cells and media were subjected to SDS-PAGE and MPO-immunoblotting.
As can be seen in Figure 3 , MPO was synthesized as a 85-kDa precursor, which was secreted as such or processed via an intermediate form to the mature large and small chains upon reaching the granule [3 , 4 ]. In addition, fragments with molecular masses of 43–47 kDa were formed by autocatalysis during warming in sample buffer [16 ], which added to the complexity of the MPO polypeptide pattern.
|
In the presence of NaClO3, little or no inhibition of the granular targeting of MPO in HL-60 cells was observed. This may indicate that the negative charge of CS provided by the carboxylic groups of glucuronic acid is strong enough to bind MPO and to mediate its delivery to azurophilic granules. This interpretation is corroborated by chromatography of PMA-induced HL-60 secretions with CM–Sepharose, which resulted in binding proMPO and a similar elution pattern as with CS–Sepharose (Fig. 1B) .
A quantification of the PMA effect on MPO secretion was performed in the following experiment: Cultures of HL-60 cells (2x106 cells/mL) were incubated in the absence and presence of 50 nM PMA for 4.25 h. Cells were then washed and lysed, and MPO activity was measured in cell lysates and secretions. The relative secretion rate was 6.6 ± 0.7% (n=10) for control cells and 10.8 ± 0.7% (n=10) for PMA-treated cells (Fig. 4 ) or 1.52 ± 0.16 mU (n=10) secreted versus 21.4 mU cellular MPO in control cells and 2.68 ± 0.17 mU (n=10) secreted versus 21.8 mU cellular MPO in PMA-treated cells. Thus, treatment of HL-60 cells with PMA induced a 1.63-fold rise in the secretory release of MPO. This observation matches well with the PMA-induced 1.8-fold increase in proteoglycan secretion (Fig. 2) , indicating that there is a combined diversion of 0.63/1.63 = 39% of newly synthesized MPO and 0.8/1.8 = 44% of newly synthesized proteoglycan from the granular to the secretory route of HL-60 cells.
|
Cross-linking MPO and serglycin
If it is true that MPO is delivered to granules as a complex with serglycin, this complex should form in the trans-Golgi region, the site of glycosaminoglycan biosynthesis, and it should be cross-linkable by a proper cross-linker.
DSP is a membrane-permeable cross-linker, which can be used in living cells and reacts with free amino groups. After attraction of the MPO by the negatively charged CS chains of serglycin, the MPO is likely to become juxtaposed to the protein moiety of the proteoglycan within the reach of DSP-succinimidyl groups, which then facilitate cross-linking. These cross-links can later be cleaved by reducing agents, thus releasing the two cross-linked partners from each other.
In a first attempt, HL-60 cells were labeled with Tran35S-label for 19 h and treated with or without 1 mM DSP for 10 min at 37°C. After lysis of the cells, MPO was immunoprecipitated and analyzed by SDS-PAGE under reducing and nonreducing conditions (Fig. 5 ). Immunoprecipitation of MPO in the absence of cross-linking showed that most of the MPO was present as proMPO (pMPO), which primarily resides in the ER, Golgi, and transport vesicles deriving thereof. Only a small portion of the MPO had reached the granular compartment, where it was proteolytically processed to yield the large, mature chain (lmMPO). The above-mentioned autocatalytic fragments of the MPO (fMPO) are also visible. The mobility of all polypeptides was lower under reducing conditions, resulting from cleavage of intramolecular disulfide bonds and maximal unfolding of the polypeptide chains.
|
It is interesting that after DSP cross-linking, the autocatalytic MPO fragments were formed even under nonreducing conditions, indicating that this catalytic activity was able to release a part of the MPO from the cross-linked complex.
To verify serglycin as a cross-linked partner of MPO, HL-60 cells were labeled for 10 min with [35S]sulfate and were divided into two aliquots. DSP (1 mM) cross-linked one aliquot of HL-60 cells for 10 min at 37°C but not the other. Cells were then disrupted by N2cavitation, and postnuclear supernatants were prepared and subjected to sucrose density centrifugation at 200,000 g overnight. Fourteen fractions were collected from each gradient, and the TCA-precipitable radioactivity and MPO activity were determined from each fraction (Fig. 6A ).
|
Fractions 5 and 6 from the control gradient and fractions 6 and 7 from the DSP gradient were combined, and CD, ß-Hex, and MPO were immunoprecipitated from corresponding aliquots of these fractions. A negative control was included, receiving the same amount of nonimmune serum. The immunoprecipitates were analyzed by nonreducing SDS-PAGE and fluorography (Fig. 6B) . A small aliquot of total TCA-precipitable material (100 Bq each) was also analyzed.
The [35S]sulfate label was almost exclusively incorporated into proteoglycans, leaving the immunoprecipitated enzymes invisible. After cross-linking and immunoprecipitation of MPO, a strong signal of coimmunoprecipitated serglycin became visible, indicating that serglycin and MPO form cross-linkable aggregates early after biosynthesis of the proteoglycan. A small portion of serglycin was immunoprecipitated along with MPO even after omission of cross-linking. Immunoprecipitation of the noncationic granular enzymes CD and ß-Hex did not result in any detectable coimmunoprecipitation of serglycin, regardless of whether cross-linking was performed or not. These immunoprecipitates were similar to control immunoprecipitations. Reduction of the cross-linked MPO-immunoprecipitate with DTT did not yield any visible molecular mass shift (not shown), as expected from the result in Figure 5 .
Analysis of the total TCA-precipitable material revealed that only a small percentage of the DSP-cross-linked proteoglycans shifted toward higher molecular mass aggregates that were resolvable by the gel. As serglycin shows apparent molecular masses significantly higher than 220 kDa, this is thought to reflect linkage of two serglycins with each other directly or via a protein forming a linker between two serglycins. Cross-linking proteins with a molecular mass of MPO would not result in a detectable shift in this region of the gel (see Fig. 5 ).
Identification of CS-binding granular proteins from human placental leukocytes
We were interested to know whether cationic polypeptides other than MPO, known to be targeted to azurophilic granules [18
] can bind to CS–Sepharose. After performing affinity chromatography with PMA-induced [35S]methionine/cysteine-labeled secretory proteins from HL-60 cells, several polypeptides including MPO bound specifically to CS–Sepharose (not shown).
To identify some of these polypeptides, we used human term placenta, as in preliminary experiments (not shown), we observed that extracts from this organ are rich in myeloperoxidase activity. We assume that the trophoblastic tissue contained leukocytes. We prepared isotonic extracts and isolated dense organelles (granules and lysosomes) by Percoll-density centrifugation on a milligram scale. Membrane-associated proteins of these organelles were subjected to CS-affinity chromatography and were analyzed by SDS-PAGE and Coomassie blue staining (Fig. 7 ).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
50% of the proteoglycans were secreted, whereas the remainder was retained in the cells and degraded. We hypothesize that MPO is synthesized as a precursor that is delivered to the trans-Golgi compartment in which it forms complexes with proteoglycans. These complexes are then partially diverted from the secretory route of HL-60 cells to azurophilic granules. The hypothesis is supported by the following observations: i) MPO binds to CS by ionic interactions, which are disrupted by 200–500 mM NaCl. (ii) PMA-induced redirection of proteoglycans to the extracellular compartment is accompanied by a similar redirection of newly synthesized precursor MPO. (iii) The fractional PMA-sensitive granular targeting of MPO in HL-60 cells (39%) mirrors that of proteoglycans (44%). (iv) Proteoglycans form cross-linkable complexes with MPO shortly after biosynthesis, most likely in the trans-Golgi, whereas other granular proteins (CD, ß-Hex), relying on other transport systems (mannose-6-phosphate receptors), do not.
These results indicate that serglycin and MPO are transported toward the granule as complexes within the same vesicles.
A similar, PMA-triggered secretion of proteoglycans was observed in the human promonocytic U937 cell line [17 ]. In this cell line, the lysosomal transport of 80% of the proteoglycans was accompanied by that of 33% of lysozyme, indicating that transport efficiency of proteoglycans for lysozyme is less effective than for MPO. This correlates well with the fact that ionic interactions of lysozyme with CS–Sepharose are weaker than those of MPO (see Fig. 1A ).
From the literature, it is known that azurophilic granules contain various cationic bactericidal polypeptides [18 19 20 ], such as the defensins, bactericidial permeability-increasing protein, lysozyme, MPO, cathepsin G, and azurocidin, to name just a few. The cationic nature of these polypeptides, which is thought to aid binding to negatively charged bacterial cell walls, is likely to serve another purpose: the diversion of the enzymes from the secretory route to azurophilic granules upon binding to serglycin. This was corroborated by the finding that in the PMA-induced secretions of HL-60 cells, besides MPO, many different polypeptides bound to CS–Sepharose (not shown) and that the CS-binding proteins from human placental dense organelles, which apparently contain leukocytic granules, were identified as the cationic enzymes MPO, lactoferrin, cathepsin G, and azurocidin/CAP37 (Fig. 7) .
Studies with PMA-induced secretions from U937 cells revealed that besides lysozyme, several other proteins bound to CS–Sepharose [21 ], indicating that also in U937 cells, proteoglycans may deliver a defined subset of cationic proteins to lysosomes. A similar function for serglycin and other proteoglycans was observed in targeting proteinases to granules in mast cells [22 , 23 ] of tissue plasminogen activator to secretory granules in human endothelial cells [24 ] and of secretory proteins to zymogen granules in pancreatic acinar cells [25 , 26 ].
The hydrophilic nature of proteoglycans suggests that these molecules represent a binding module, which is likely to interact with one or several membrane components responsible for packaging the complexes into granular transport vesicles, as outlined in the review by Gullberg et al. [27 ]. The nature and identity of these membrane components are not known yet.
In transfected mouse myeloid 32D cells, it was observed that a mutant human MPO
pro, lacking the propiece, was secreted, whereas wild-type human proMPO was partially targeted to granules, suggesting a role for the MPO propiece in the targeting to granules [6
]. In the present study, affinity chromatography of proMPO and MPO on CS–Sepharose showed that both forms of the enzyme bound well to the affinity matrix (Fig. 1A)
and therefore should be transported by serglycin with similar efficiency. Because both forms were still in their native conformation, as judged from their enzymatic activity, these controversial results may suggest that besides positive charges, the native conformation of the enzyme is important for binding to serglycin. If the entire, rather large propiece of MPO is deleted by recombinant techniques, the enzyme may fold incorrectly in the ER, causing problems in every step of its biosynthesis and transport. In fact, we could measure that binding of partially SDS-denatured MPO to CS–Sepharose was significantly weaker than that of native MPO (not shown).
In the granules of HL-60 cells, degradation of the proteoglycan’s protein backbone rendered the glycan resistant to TCA precipitation with a half-life of 15 min (Fig. 2)
. Precipitation of the same [35S]sulfate-labeled material with cetyl-trimethlyl-ammonium bromide, an agent that precipitates proteoglycans regardless of the integrity of their protein backbone, indicated that in azurophilic granules, the half-life of proteoglycans is in the range of several hours (not shown). This finding is compatible with the presence of proteoglycans in immature azurophilic granules, as detected by the reaction with cationic colloidal gold in electron micrographs and the absence of this reaction in mature azurophilic granules [28
]. It appears that shortly after the granular targeting, the exposed protein portions and somewhat later, the glycosaminoglycan moieties and the linkage region of the proteoglycan become degraded.
The degradation of the proteoglycan after the targeting of the complexes to granules may be a prerequisite for an eventual binding of the azurophilic granule proteins to bacteria following degranulation or phagocytosis, assuming that the same groups in the cationic proteins participate in binding to serglycin or bacteria. This may explain why in mucopolysaccharidosis type VII (Sly disease), which is caused by a deficiency of ß-glucuronidase, granulocytes and monocytes show excessive granulation in all patients examined [29 , 30 ] (similar observations were made in the mouse model of this disease [31 ]) and patients present with repeated episodes of pneumonia [29 ]. Without doubt, further work is needed to prove a causal relationship between degradation of granular serglycin and immunological efficiency of granulocytes and monocytes.
|
ACKNOWLEDGEMENTS |
|---|
Received December 19, 2002; revised April 23, 2003; accepted April 25, 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  E. P. Dick, L. R. Prince, and I. Sabroe Ex Vivo-Expanded Bone Marrow CD34+ Derived Neutrophils Have Limited Bactericidal Ability Stem Cells, October 1, 2008; 26(10): 2552 - 2563. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. U. Niemann, M. Abrink, G. Pejler, R. L. Fischer, E. I. Christensen, S. D. Knight, and N. Borregaard Neutrophil elastase depends on serglycin proteoglycan for localization in granules Blood, May 15, 2007; 109(10): 4478 - 4486. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Lemansky, I. Fester, E. Smolenova, C. Uhlander, and A. Hasilik The cation-independent mannose 6-phosphate receptor is involved in lysosomal delivery of serglycin J. Leukoc. Biol., April 1, 2007; 81(4): 1149 - 1158. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Zernichow, M. Abrink, J. Hallgren, M. Grujic, G. Pejler, and S. O. Kolset Serglycin Is the Major Secreted Proteoglycan in Macrophages and Has a Role in the Regulation of Macrophage Tumor Necrosis Factor-{alpha} Secretion in Response to Lipopolysaccharide J. Biol. Chem., September 15, 2006; 281(37): 26792 - 26801. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. U. Niemann, J. B. Cowland, P. Klausen, J. Askaa, J. Calafat, and N. Borregaard Localization of serglycin in human neutrophil granulocytes and their precursors J. Leukoc. Biol., August 1, 2004; 76(2): 406 - 415. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, control cells; 
, PMA cells; 
, control medium; 
, PMA medium), the other.
