| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,
* Laboratory for Immunohistochemistry and Immunopathology and
Department of Pathology, University of Oslo and Rikshospitalet University Hospital, Oslo, Norway;
Department of Molecular Biosciences, University of Oslo, Oslo, Norway; and
Department of Biomedicine, University of Bergen, Bergen, Norway
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
/CXCL1 and MCP-1/CCL2 reside in small granules that, similarly to the WPB, respond to secretagogue stimuli. In the present study, we report that GRO and MCP-1 colocalized in 50- to 100-nm granules, which occur throughout the cytoplasm and at the cell cortex. Immunofluorescence confocal microscopy revealed no colocalization with multimerin or tissue plasminogen activator, i.e., proteins that are released from small granules of ECs by regulated exocytosis. Moreover, the GRO/MCP-1-containing granules were Rab27-negative, contrasting the Rab27-positive, WPB. The secretagogues PMA, histamine, and forskolin triggered distinct dose and time-dependent responses of GRO release. Furthermore, GRO release was more sensitive than IL-8 release to inhibitors and activators of PKA and PKC but not to an activator of Epac, a cAMP-regulated GTPase exchange factor, indicating that GRO release is regulated by molecular adaptors different from those regulating exocytosis of the WPB. On the basis of these findings, we designated the GRO/MCP-1-containing compartment the type 2 granule of regulated secretion in ECs, considering the WPB the type 1 compartment. In conclusion, we propose that the GRO/MCP-1-containing type 2 granule shows preferential responsiveness to important mediators of EC activation, pointing to the existence of selective agonists that would allow differential release of selected chemokines. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Regulated exocytosis in ECs is induced by a number of agonists that can be separated into two distinct groups: those acting by elevating the level of cytosolic Ca2+ and those acting in a Ca2+-independent manner (reviewed in Refs.4 and 5). For example, thrombin (22, 23, 24), histamine (25), and purine nucleotides (26, 27) have been shown to induce granule release in a Ca2+-dependent manner. By contrast, epinephrine and forskolin elevate the level of cAMP, thereby activating PKA, and resulting in Ca2+-independent regulated exocytosis (28, 29). Regulated exocytosis might also be induced in a Ca2+-independent manner by phorbol esters, either by activation of PKC, RasGRP, or Munc13 (30, 31, 32), the latter being a SNARE-associated protein important for the secretory granule to become fusion competent. The molecular mechanisms of regulated exocytosis in ECs have been focused mainly on WPB release. For example, the monomeric GTPase Rab3D is suggested to play a role in WPB/granule formation (12), whereas Rab27A is associated with the mature pool of the WPBs (33). Furthermore, vesicle-associated membrane protein (VAMP)-3 and syntaxin4, two members of the SNARE core complex critical for intracellular fusion events, are involved in WPB exocytosis (34). Finally, the involvement of the small GTPase RalA in both thrombin- and epinephrine-induced WPB release suggests that there is convergence downstream of agonists activating distinct transduction pathways (29, 35, 36).
We have recently shown that the chemokines eotaxin-3 (CCL26) and IL-8 (CXCL8) are released from the WPB, whereas GRO (CXCL1) and MCP-1 (CCL2) are both stored in a compartment different from the WPB that is also sensitive to histamine and PMA (37). Use of distinct granules for regulated exocytosis might enable the cell to control the acute release of mediators, and evidence for differential release mechanisms in ECs have indeed been described previously (11, 12, 13, 38, 39). In this respect, we explored whether differential release might be a feature of regulated chemokine exocytosis. We first showed that GRO and MCP-1 colocalize in small granules distinct from the WPB, and found that the GRO/MCP-1-containing granules were Rab27-negative in contrast to the Rab27-positive WPBs. Moreover, the responsiveness of the GRO/MCP-1-containing granule to relevant secretagogues revealed distinct dose-dependent responses and release kinetics that were nevertheless similar to those obtained for the WPB. However, specific activation or inhibition of PKA or PKC revealed interesting differences between the GRO/MCP-1-containing compartment and the WPB, suggesting that different molecular mechanisms are involved in their exocytic machineries and pointing to the existence of agonists that allow selective release of either compartment.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Recombinant human IL-1
, recombinant human epidermal growth factor, recombinant human basic fibroblast growth factor, recombinant human platelet factor 4, and matched Ab pairs for ELISAs of human IL-8 and GRO were obtained from R&D Systems or PeproTech. FBS, gentamicin, fungizone, L-glutamine, MCDB 131, lipofectin reagent, and Opti-MEM I were purchased from Invitrogen Life Technologies, trypsin-EDTA was purchased from BioWhittaker, and alkaline phosphatase- or HRP-conjugated streptavidin was purchased from Southern Biotechnology Associates and R&D Systems, respectively. The tetramethylbenzidine microwell peroxidase substrate system was purchased from Kirkegaard & Perry Laboratories and protein A-coated colloidal gold particles were obtained from G. Posthuma (University Medical Center, Utrecht, The Netherlands). LPS, sodium butyrate, 3-isobutyl-1-methylxanthine (IBMX), and (±)-epinephrine were from the Sigma-Aldrich; 8-(4-Chlorophenylthio)-2'-O-methyladenosine-3',5'-cyclic mono-phosphate (8-pCPT-2'-O-Me-cAMP) and N6-benzoyladenosine-3', 5'-cyclic monophosphate (6-Bnz-cAMP) were purchased from BioLog; and H-89, chelerythrine chloride, and Gf109203x/bisindolylmaleimide I were purchased from Calbiochem. Tacrolimus/FK506 was purchased from Fujisawa. The primary Abs used for immunostaining are listed in Table I. Unconjugated rabbit anti-mouse IgG+M and tetramethylrhodamine isothiocyanate-labeled swine anti-rabbit IgG were purchased from DakoCytomation, biotinylated horse anti-mouse IgG from Vector Laboratories, streptavidin-Cy2 conjugate was purchased from Amersham Biosciences, Alexa-488 goat anti-rabbit IgG was purchased from Molecular Probes, and Cy3-conjugated donkey anti-mouse IgG, Cy3-conjugated donkey anti-rabbit IgG, and Cy3-conjugated streptavidin were purchased from The Jackson Laboratory.
|
Umbilical cords were obtained from the Department of Gynecology and Obstetrics, Rikshospitalet, and HUVECs were isolated as described by Jaffe et al. (40) and cultured in MCDB 131 containing 7.5% FBS, 10 ng/ml recombinant human epidermal growth factor, 1 ng/ml recombinant human basic fibroblast growth factor, 1 µg/ml hydrocortisone, 50 µg/ml gentamicin, and 250 ng/ml fungizone. The cells were maintained at 37°C in humid 95% air/5% CO2 atmosphere and split at ratio 1:3. The cultures were used at passage level one to six.
Transfection
Transient transfections into HUVECs were performed by lipofection according to the manufacturer’s instruction (Invitrogen Life Technologies), using 0.6 µg of DNA mixed with 1.2 µl of lipofectin/well in 8-well Lab-Tek chamber slides (Nalge Nunc International). Alternatively, the HUVECs were transfected by electroporation according to the protocol 0394 from BTX.
Immunoelectron microscopy
IL-1-treated cells were fixed for 3 h at room temperature in 0.1 M phosphate buffer containing 4% paraformaldehyde or 0.1% glutaraldehyde. After washing in 1x PBS, the cells were scraped off and spun down. Cell pellets were embedded in 1x PBS with 12% gelatin, infiltrated with 2.3 M sucrose overnight at 4°C (41), and then cut in small blocks which were mounted on pins and frozen in liquid nitrogen. Ultrathin cryosections of 
60- to 70-nm thickness were obtained by cutting at –120°C with a Reichert Ultracut S ultracryomicrotome from Leica and were picked up in a 1:1 mixture of 2% methylcellulose and 2.3 M sucrose. Cryosections were then single or double immunostained and the labeling detected using protein A-coated colloidal gold particles (PAGs) of different sizes. For the single labeling, cryosections were sequentially incubated for 30 min at room temperature with the rabbit anti-MCP-1, then PAG in 1x PBS with 1% BSA. For the double labeling, cryosections were sequentially incubated with the rabbit anti-GRO and PAG, then postfixed in 1x PBS containing 1% glutaraldehyde before performing the second labeling with the mouse anti-MCP-1, followed by the rabbit anti-mouse IgG/IgM, and PAG. Finally, cryosections were contrasted with a 1:9 mixture of 3% uranyl-acetate and 2% methylcellulose and examined in a Philips CM100 transmission electron microscope from FEI. Pictures were obtained with a MegaView III 1000*1000 pixel digital camera from Soft Imaging System (SIS).
Immunostaining protocols
Monolayers of HUVECs grown on Lab-Tek chamber slides (Nalge Nunc International) coated with 1% (w/v) gelatin type A from porcine skin, were briefly submerged in prewarmed PBS (37°C) and fixed in prewarmed 4% paraformaldehyde (37°C, pH 7.4) for 10 min before washing 2 x 5 min in PBS. For immunostaining, the fixed monolayers were incubated with the primary mouse Ab (see Table I) overnight at 4°C; then with biotinylated horse anti-mouse IgG (1/200) combined with primary rabbit IgG (Table I) for 1.5 h at room temperature; and finally with streptavidin-Cy2 conjugate (1/1000) combined with tetramethylrhodamine isothiocyanate-labeled swine anti-rabbit IgG (1/80) or Cy3-conjugated donkey anti-rabbit IgG for 1 h at room temperature. In an alternative protocol, the fixed monolayers were incubated with the primary mouse Ab overnight at 4°C before incubation with biotinylated horse anti-mouse IgG (1/200) combined with primary rabbit IgG for 1.5 h at room temperature visualized by streptavidin-Cy3 conjugate (1/1000) combined with Alexa 488-labeled swine anti-rabbit (1/800). Saponin (0.1%) was used for permeabilization in all steps during immunostaining, except in the last washing. Irrelevant, concentration-matched primary Abs were used as negative controls. The immunostained cells were examined by a confocal laser scanning microscope (Leica TCS) equipped with an Ar (488 nm) and a He/Ne (543 and 633 nm) laser. A Plan apochromat x100/1.4 oil objective was used, and the fluorochromes were excited and detected sequentially.
Secretion experiments
HUVECs were seeded out at confluence (1.6 x 104 cells/well in 96-well trays; BD Biosciences) and cultivated for at least two days before being stimulated by 1 ng/ml IL-1 for analysis of IL-8 and GRO. To maintain good culture conditions, we added fresh medium daily. Approximately 35 h after adding the cytokine, the monolayers were washed twice in prewarmed PBS and incubated in fresh medium without cytokine, containing 100 µg/ml cycloheximide for the next 3–4 h. Subsequently, the cells were washed twice in prewarmed PBS and incubated in growth medium with 100 µg/ml cycloheximide and the different agonists. IBMX was added together with forskolin to inhibit phosphodiesterase activity. All secretion experiments were performed in duplicate (dose responsiveness of PMA) or triplicate wells of microtiter plates from which supernatants were harvested for measurement of GRO and IL-8.
ELISA
The chemokines were analyzed by DuoSet ELISA kits or matched Ab pairs (R&D Systems or PeproTech) according to the recommendations of the manufacturer’s with the following modifications: microtiter plates were incubated overnight with the coat Abs diluted in PBS (all steps were performed at room temperature), washed in H2O or PBS with 0.05% Tween 20, and blocked by 1% (w/v) BSA in PBS for 2 h. Before each of the following incubation steps, the plates were washed four times in PBS with 0.05% Tween 20. Samples (50 µl/well) were incubated overnight followed by detection Abs (1.5 h) and alkaline phosphatase-conjugated streptavidin (1/3000, 1.5 h) or HRP-conjugated streptavidin (1/200, 1.5 h). p-Nitrophenyl phosphate in diethanolamine buffer or peroxidase substrate was developed for 5–40 min, and the absorbance was measured at 405 or 450 nm, respectively, with a Tecan Sunrise Microplate Reader (Tecan Austria Gesellschaft). Standard curves were generated from 3-fold dilutions of recombinant chemokines (R&D Systems). Sigmoidal dose-response curves were fitted to the Hill equation by means of Graph Pad Prism to estimate EC50.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
We have recently shown that the chemokines GRO and MCP-1 are released from HUVECs in response to the secretagogues histamine and PMA, indicating that they are stored in a compartment for regulated secretion (37). Interestingly, immunocytochemistry and confocal analysis revealed that both chemokines localized in a compartment different from the WBP. Moreover, this compartment did not belong to the endocytic pathway and we concluded that GRO and MCP-1 were sorted to a histamine- and PMA-sensitive intracellular storage granule distinct from the WPB (37).
To characterize the GRO- and MCP-1-containing granules in more detail, we performed paired immunostaining and confocal analysis, finding that the two chemokines colocalized in the Golgi and in the majority of granules (Fig. 1A). Next, we analyzed the ultrastructure of these granules by means of immunoelectron microscopy on ultrathin cryosections of IL-1-treated HUVECs. MCP-1 was found in granules close to the cell membrane as well as in the trans-Golgi network and tubulovesicular structures (Fig. 1B). In contrast to the large and elongated WPBs (see inset Fig. 1C, upper right corner), the MCP-1-containing granules were round, 50–100 nm in diameter, and moderately electron dense. In general, no coat was observed on these granules. The same intracellular distribution was found for GRO (Fig. 1C) and double labeling for GRO and MCP-1 confirmed colocalization in such granules as well as in the Golgi stack (Fig. 1C). By contrast, the mononuclear cell-recruiting chemokine SDF-1/CXCL12, that was also found in a punctuate, perinuclear pattern, failed to colocalize with GRO (Fig. 1D). Based on the colocalization of GRO and MCP-1, we focused on GRO detection in the further studies of this novel chemokine-storage granule.
|
synthesis, and that LPS-activated cells contained GRO and MCP-1-positive granules of similar size and distribution as those induced by IL-1 (data not shown). Moreover, LPS-induced granules were released in response to PMA in a similar manner (data not shown).
Furthermore, we costained for GRO and multimerin or tPA, two proteins reported to be stored by ECs in secretory granules distinct from the WPB (10, 11, 12, 13, 15). Multimerin was found in rod-shaped structures and small, round granules consistent with a previous report (15) but did not colocalize with GRO (Fig. 1E). To detect tPA, HUVECs were either treated with 3 mM sodium n-butyrate for 20 h to increase the endogenous expression (data not shown) or transfected with pcDNA3-tPA (Fig. 1F). These treatments enabled immunodetection of tPA in small granules that were mainly perinuclear after n-butyrate stimulation but were distributed throughout the whole cell after transfection, probable reflecting differences in the expression levels. In addition, tPA was found diffusely throughout the cytoplasm. With the exception of a very few double positive vesicles, we did not observe overt colocalization of GRO and tPA (Fig. 1F) nor did we observe colocalization with vWf (data not shown). On the basis of these data, we concluded that GRO and MCP-1 were stored in a hitherto undescribed granule in HUVECs designated for regulated secretion and propose to refer to it as the type 2 compartment for regulated secretion in ECs, considering the WPB the type 1 compartment.
The GRO/MCP-1-containing type 2 granule is Rab27 negative in contrast to the WPB
To further explore whether there were differential mechanisms involved in GRO and IL-8 release, we examined the intracellular distribution of SNARE core complex-associated proteins relative to that of GRO and vWf. We chose to study adaptors known to be involved in regulated secretory granule release but excluded candidates that were not expressed at substantial mRNA levels in cultured HUVECs (M. Veuger and G. Haraldsen, unpublished observations).
Due to lack of a well-working Ab to Rab4, which is involved in release of -granules but not dense granules from platelets (43, 44), we transfected Rab4 fused to YFP into HUVECs. We observed yellow fluorescence in the Golgi region, in structures outside the Golgi and in a punctate pattern in the periphery of the cell. However, immunostaining showed no colocalization with GRO (Fig. 2A). Instead Rab4 colocalized with EEA1 and partially with LAMP-2 (data not shown). Furthermore, we evaluated Rab8, Rab27, and Rab32 by paired immunostaining and confocal analysis or conventional fluorescence microscopy. Rab8 is associated with the -granule of platelets (45) and we observed a fluorescent signal for this protein in the Golgi region, in scattered granules throughout the cytoplasm, as well as in elongated structures at the cell periphery. However, neither GRO nor vWf colocalized with Rab8 outside the Golgi region (Fig. 2B and data not shown). We next costained for GRO and Rab27, which is associated with regulated exocytosis in a number of cells (reviewed in ref (46). Because HUVECs express mRNA encoding Rab27a but not Rab27b (33), we considered our anti-Rab27 reagent to detect Rab27a. For the latter protein, we observed staining in the nucleus, in cigar-shaped structures, as well as in a punctate pattern throughout the cytoplasm, but there was no colocalization with GRO (Fig. 2C). Consistent with a previous report (33), the Rab27a-positive cigar-shaped structures were vWf-positive (Fig. 2C and data not shown). Finally, we stained for Rab32, which is associated with platelet granules and melanosomes (47, 48), observing Rab32 in cytoplasmic peripheral granules negative for GRO (Fig. 2D).
|
(Fig. 2E) or vWf (data not shown). Furthermore, two primary Abs were used for detection of VAMP-8. The first Ab, from Synaptic Systems, stained perinuclear round vesicles, while the second (kind gift of Dr. R. C. Piper, University of Iowa, Iowa City, IA) stained the nucleus, donut-shaped structures mainly in the perinuclear area, and elongated structures in the periphery of the cells. However, none of the VAMP-8 patterns colocalized with GRO (Fig. 2F and data not shown). Finally, we examined the subcellular distribution of syntaxin4 with an Ab that has been shown to affect WPB release in a streptolysin-O assay (34). Staining against syntaxin4 resulted in a diffuse signal throughout the cytoplasm as well as a weak granular signal, but revealed no colocalization with GRO (Fig. 2G). Secretagogue-specific release kinetics of the type 2 granule
Use of distinct granules for storage would enable the EC to control the release of molecules in a differential manner. Such differential release is known for secretory granules in other cell types and has been described for vWf and t-PA in ECs (11, 12, 13, 39), as well as for the WPB and small granules of unknown identity (38). To examine whether the GRO- and MCP-1-containing granule and the IL-8-containing WPB are differentially released, we compared the release kinetics of GRO and IL-8 in response to PMA, histamine, and forskolin, each known to induce release of the WPB by different signal transduction pathways (4, 5, 29). First, we examined the response to 100 ng/ml PMA (Fig. 3), observing release of both GRO and IL-8 as early as 2 min poststimulation and finding that about half of the releasable chemokines were secreted within the first 10 min (45% of GRO, 38–53% of IL-8, n = 3). When examining the release kinetics in the presence of 100 µM histamine, we observed an immediate, faster release of both chemokines compared with that in response to PMA (Fig. 3). Within 2 and 10 min more than 40 and 60% of both chemokines were released relative to the amount released after 30 min, respectively. Finally, we examined the response to the cAMP-elevating mediators epinephrine and forskolin. In initial experiments we observed that epinephrine combined with IBMX (a nonspecific inhibitor of phosphodiesterases) gave variable but weak responses (data not shown), consistent with a previous study examining the release of vWf (28). Thus, we proceeded with forskolin combined with IBMX, and observed a substantially slower time course of release for both chemokines that only became apparent after 10 min. Therefore, we concluded that the GRO-containing compartment and the WPB showed similar release kinetics in response to all three secretagogues, but each secretagogue nevertheless induced distinct release kinetics because histamine clearly induced the fastest response while forskolin induced a slower and smaller response than the other agonists.
|
We next assessed the dose response to PMA, histamine, and forskolin. When examining chemokine release in response to increasing concentrations of PMA (Fig. 4), we found that the half-maximum release (EC50) was reached at 1.5 ng/ml (95% confidence interval (CI): 0.4–6.0) for GRO. By analyzing another fraction of the same supernatants, we found EC50 for IL-8 to be reached at a somewhat higher concentration of PMA (2.6 ng/ml, 95% CI: 0.4–17.4). However, the threshold for release of GRO and IL-8, defined as the intercept with the abscissa of the extrapolated slope of the dose-response curve (indicated by the stippled line), was 0.5 ng/ml for GRO and 0.1 ng/ml for IL-8. Moreover, while plateau and maximum release was obtained at 10–100 ng/ml for GRO, release of IL-8 did not reach a plateau at the tested concentrations of PMA. In addition, the released amount of GRO differed from that of IL-8. Average release above the level of constitutive secretion was 298 pg/ml (range, 188–454 pg/ml, n = 4) for GRO and 502 pg/ml (range, 407–691 pg/ml, n = 4) for IL-8 at 100 ng/ml PMA, showing that the releasable pool of GRO was smaller than that of IL-8 in response to PMA.
|
and to 5.0 µM (95% CI: 2.6–9.6) for IL-8, thus indicating a trend to a higher agonist sensitivity of the GRO-granules. Moreover, the threshold of release in response to histamine was 0.08 µM for GRO and 0.7 µM for IL-8. Thus, GRO release appeared to be triggered at lower concentrations of histamine than IL-8 release. Furthermore, both chemokines were released to a lesser extent by histamine than by PMA and similar to PMA, histamine released less GRO than IL-8 (Fig. 4). Therefore, we concluded that a smaller pool of both chemokines was responsive to histamine compared with PMA.
Next, we assessed the dose response of forskolin. Although forskolin yielded more reproducible results than epinephrine, we nevertheless observed only weak responsiveness in the range of 0.001–100 µM (Fig. 4). Moreover, the dose response did not reach sigmoid saturation but appeared to transiently peak at 5 µM for both chemokines. The reduced release observed at higher concentrations of forskolin was not caused by cell damage as the monolayer remained trypan blue-impermeable (data not shown). Instead the bell shaped dose response was similar to that reported for tPA and vWf release in response to the cAMP-elevating, -adrenergic isoproterenol (50). Although forskolin appeared to release even less chemokine than histamine, we again found that more IL-8 than GRO was released (Fig. 4).
Finally, we analyzed the possible effect of TNF- on regulated EC secretion as this cytokine is a rapidly releasable agonist stored in mast cells and therefore of putative importance in inducing rapid cellular responses in analogy to for example mast cell-derived histamine. However, TNF- did not promote substantial secretion of either chemokine in the range of 1–100 ng/ml over 1 h, despite readily detectable effects in the detection of E-selectin and VCAM-1 (data not shown).
Additive agonist effects on GRO/MCP-1 release
Synergism between cAMP activating and calcium-raising agents has been described in for example pancreatic cells (51, 52, 53). To examine whether such mechanisms are involved in release of the GRO/MCP-1-compartment, we combined forskolin (10 µM) and IBMX (100 µM) with increasing doses of histamine (1–100 µM) for 30 min (Fig. 5, top panels). The release of both GRO and IL-8 in the presence of both agonists were similar to the sum of released chemokine in response to either agonist alone, suggesting that histamine and forskolin mediate exocytosis of GRO and IL-8 through additive independent, noninteracting signaling pathways. In a separate experiment, we added forskolin 25 h before stimulation with histamine (data not shown), but a priming effect of forskolin was not observed.
|
and IL-8 induced by FK-506 alone that was further enhanced in an additive manner by coincubation with forskolin (5 or 10 µM, Fig. 5, bottom panels). PKA and PKC are potential regulators of differential chemokine release
Finally, we compared the extent to which Epac, PKA, or PKC were involved in the regulated exocytosis of the type 1 and 2 compartment. Epac, a cAMP-regulated exchange factor for the small GTPases Rap1 and Rap2 is a PKA-independent activator of regulated exocytosis in both pancreatic cells (reviewed in Ref.58) and ECs (29). In the presence of the Epac-specific activator 8-pCPT-2'-O-Me-cAMP (59), GRO and IL-8 were secreted at levels similar to those observed in response to forskolin (Fig. 6A). Also, the PKA-specific activator 6-Bnz-cAMP induced release of GRO and IL-8 (Fig. 6B), suggesting that both Epac and PKA mediate regulated exocytosis of GRO and IL-8. However, 6-Bnz-cAMP released substantially less IL-8 than GRO (relative to the amount released by forskolin), suggesting a stronger involvement of PKA in GRO release than in IL-8 release (Fig. 6B). This possibility was further supported by the observation that the PKA inhibitor H-89 affected the release of GRO more efficiently than that of IL-8 (Fig. 6C). Incubation with the inhibitor alone did not decrease the levels of constitutively released GRO or IL-8 (data not shown).
|
, I, II, 
, 
, and 
, or by means of chelerythrine chloride, which inhibits all PKC isoforms. Gf109203x inhibited PMA-induced GRO release in a dose-dependent manner and at 10 µM the release was reduced by 32% (Fig. 6D). By contrast, we observed a 20% reduction of IL-8 release at all concentrations of Gf109203x tested in these experiments (0.1, 1, and 10 µM) (Fig. 6D). Cell viability was at all doses excellent as judged by trypan blue. In the presence of 10 µM chelerythrine chloride, PMA-induced release of GRO was more strongly reduced (50%) compared with that of IL-8 (25%) (Fig. 6E). Moreover, chelerythrine chloride reduced the constitutive release of GRO by 15% but did not affect the constitutive IL-8 release (data not shown) while Gf109203x alone affected release of neither (data not shown). |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
and MCP-1 are stored in ECs and released from intracellular compartments distinct from the IL-8/eotaxin-3-containing WPB (37). Because the GRO/MCP-1-containing granules failed to exhibit markers of the endocytic pathway and disappeared after PMA stimulation, we concluded that they belong to the secretory pathway (37). Here, we found that both chemokines are sorted to the same compartment of small, Rab27-negative, coatless granules, 50–100 nm in diameter that occur throughout the cytosol and close to the plasma membrane. These structures resemble secretory granules previously reported to harbor other EC mediators, i.e., protein S (14), tPA (10, 11, 12, 13), and multimerin (15). We found that HUVECs expressed only low levels of mRNA transcripts for protein-S (M. Veuger and G. Haraldsen, unpublished observation), and efforts to immunostain these monolayers with three different Abs were unsuccessful (I. Øynebråten, M. Veuger, and G. Haraldsen, unpublished data). However, the transient release reported for protein S (14) is different from the release kinetics observed for GRO (see below), and it is therefore unlikely that protein S and GRO originate from the same storage compartment. Moreover, tPA has been reported by some groups to localize in the WPB of cultured HUVECs (16, 17, 18) and by others in small granules (10, 11, 12, 13). We observed tPA in small granules of both primary and advanced-passage level cultures of HUVECs, thus supporting the studies by Emeis and Knop (10, 11, 12, 13) with no evidence for vWf or GRO colocalization. Given that multimerin also failed to colocalize with GRO, our data suggested that GRO and MCP-1 are released from a novel regulated secretory granule in HUVECs. Finally, because tissue factor pathway inhibitor is found in caveoli rather than coatless granules (19, 20, 21), it is interesting to note that chemokines stored in ECs do not appear to reside in compartments containing anticoagulant proteins.
Differential release of various granule populations is a well-known feature of certain regulated secretory cells (reviewed in Refs.6 and 60), and it has also been described for ECs with respect to the WPB and a small tPA-containing granule (11, 12, 13, 39) as well as for large vs small vesicles measured by whole-cell patch clamp technique (38). In the present study, interesting trends were observed for GRO and IL-8 in terms of EC50 and threshold concentrations of release in response to both PMA and histamine. However, the most apparent differences were observed at the level of intracellular mediators. Thus, release of GRO varied substantially from that of IL-8 in response to the PKA-specific activator 6-Bnz-cAMP and the PKA inhibitor H-89. Consistent with previous work (61), PKA was involved in forskolin-mediated WPB exocytosis as tested by the release of IL-8, but our data suggested that PKA was even more crucially involved in release of the GRO-containing compartment. In a more general context, this finding points to a novel pathway in differential release of secretory granules, which until now has been associated with regulation at the level of calcium signaling (reviewed in Ref.6).
We also observed a stronger effect on PMA-induced GRO release compared with that of IL-8 release in the presence of the broadly acting PKC inhibitor Gf109203x or chelerythrine chloride, suggesting another level of regulating differential release of secretory granules from ECs. However, although PKC has been shown to be important in WPB release (24), our data clearly imply that the diacylglycerol-mimetic phorbol ester PMA also exhibits PKC-independent effects in regulated secretion in ECs. Such possible mechanisms may include, but not be limited to, the diacylglycerol/PMA-responsive Munc13, a crucial protein for phorbol ester-induced PKC-independent presynaptic neurotransmitter release (30). Moreover, the known link between Munc13-4 and Rab27 in platelets (62) and basophils (63), makes it tempting to assume that Rab27 on WPB provide a less PKC-dependent molecular machinery of secretion compared with the type 2 Rab27-negative granule. Thus, a more careful dissection of the PKC involvement in GRO release as well as other diacylglycerol targets deserves further attention. Moreover, the differential effects of both PKA and PKC on GRO and IL-8 release, makes it tempting to suggest possible molecular mechanisms of convergence, such as the binding of both kinases to A kinase-anchoring proteins (AKAPs) (64, 65), one of which (AKAP79) has been demonstrated to regulate release of insulin from pancreatic cells (54, 66). Furthermore, because molecules such as AKAPs contribute to target effects of second messengers to specific compartments, it may also be assumed that such anchoring molecules are differentially distributed among subsets of granules.
Both GRO and IL-8 release responded more rapidly to histamine than to PMA and forskolin. For IL-8, this difference accorded with recently published real-time imaging data of the WPB (67). The late response to PMA compared with that of histamine could reflect a delay in the intracellular mechanisms involved in exocytosis, or alternatively, suggest that PMA must penetrate the cell membrane to induce effector mechanisms. The substantially slower response to forskolin has been suggested to reflect a difference in the physiological function between Ca2+-raising and cAMP-elevating agonists. Although, for instance, the Ca2+-raising agonist thrombin might act as a local emergency signal, the cAMP-raising agonist epinephrine could act more systemically in respect to vWf release, perhaps regulating the vWf plasma levels (28, 68). Moreover, histamine released more GRO and IL-8 than forskolin. This was consistent with a previous report finding that while histamine releases both peripheral and centrally located granules, forskolin primarily involves release of vesicles located in the periphery (68). In release of, for instance, amylase and insulin from exocrine and endocrine pancreatic cells, respectively, calcium-raising agents and hormones raising the level of cAMP have been described to have synergistic effects (51, 52, 53). However, when forskolin was added together with histamine we only observed additive effects related to GRO and IL-8 release and our data do not support the existence of synergistic, interacting cAMP- and histamine/calcium-mediated pathways in ECs. This is consistent with results reported for vWf release from HUVECs in the presence of thrombin combined with forskolin or epinephrine (28). Furthermore, cAMP-mediated release might be regulated by the protein phosphatase 2B/calcineurin. To inhibit the potential activity of this phosphatase, we used FK-506/tacrolimus that alone induced release of both GRO and IL-8. This is well in line with the insulin release stimulated by short-term exposure to cyclosporine A (another inhibitor of calcineurin) in pancreatic cells (69). The enhancing effect of FK-506 on forskolin-induced chemokine secretion is also similar to the response reported in cells (54). Furthermore, while FK-506 and forskolin appear to act in a synergistic fashion in insulin release, we found only additive-effects on chemokine release, not allowing us to speculate on common targets as has been demonstrated in cells.
In conclusion, we have characterized the ultrastructure and release properties of a novel chemokine-containing granule in ECs that appears to be distinct from compartments containing anticoagulant proteins. Furthermore, this GRO/MCP-1-containing granule showed distinct release properties related to both PKA and PKC compared with that of the IL-8/eotaxin-3-containing WPB, suggesting that these two granules may be released differentially in response to currently unknown extracellular signals.
|
Acknowledgments |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This study was supported by the Norwegian Cancer Society Grant B02085, Research Council of Norway Grant 133924/300 and Anders Jahrés Fund. I.Ø. is a Research Fellow of the Norwegian Cancer Society. F.-E.J. and G.H. are Career Investigators of the Research Council of Norway. 
2 Address correspondence and reprint requests to Inger Øynebråten, Laboratory for Immunohistochemistry and Immunopathology, Institute of Pathology, Rikshospitalet, N-0027 Oslo, Norway. E-mail address: inger.oynebraten{at}labmed.uio.no
3 Abbreviations used in this paper: EC; endothelial cell; AKAP, A kinase-anchoring protein; 6-Bnz-cAMP, N6-benzoyladenosine-3',5'-cyclic monophosphate; CI, confidence interval; IBMX, 3-isobutyl-1-methylxanthine; PAG, protein A-coated colloidal gold particle; 8-pCPT-2'-O-Me-cAMP, 8-(4-chlorophenylthio)-2'-O-methyladenosine-3',5'-cyclic monophosphate; PKA, protein kinase A; PKC, protein kinase C; tPA, tissue plasminogen activator; WPB, Weibel-Palade body; VAMP, vesicle-associated membrane protein; vWf, von Willebrand factor.
Received for publication December 14, 2004. Accepted for publication July 28, 2005.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
-dependent phosphorylation of syntaxin 4 and Munc18c: role in signaling P-selectin expression. J. Biol. Chem. 280:3178.-3184. Phorbol ester- and diacylglycerol-induced augmentation of transmitter release is mediated by Munc13s and not by PKCs. Cell 108:121.-133. [Medline]
-granule secretion in platelets. J. Biol. Chem. 275:33844.-33849. - and dense-core granule secretion of platelets. Ann. NY Acad. Sci. 947:403.-406. cells. Nature 363:356.-358. [Medline]
cells. J. Biol. Chem. 279:45455.-45461. 
