Table 1.
Antibodies.
Table 2.
TaqMan gene expression assays used in qRT-PCR analysis.
Fig 1.
HUVEC were cultured in low (sparse) and high (dense) densities. (A) Illustration of the morphology of sparse (left panel) and dense (right panel) HUVEC. (B) Gene expression of quiescence marker IL33 and proliferation marker MKI67 in sparse relative to dense cell cultures determined by qRT-PCR in material from six donors. (C) Staining for IL-33 in dense culture (right panel) and sparse culture (left panel) visible as magenta, costained with DAPI nuclear staining visible as blue. The pictures were acquired using a confocal microscope at 60X magnification. (D) The proliferation rate was assessed by applying the MTS-assay on cell cultures of increasing densities from four donors. (E) Cell mortality rate was determined with the LDH-assay for 5 donors in both sparse (upper panel) and dense (lower panel) cell cultures. All data are presented as means with SEM, and statistical significant differences (p < 0.05) was tested with Students paired t-test and denoted by *. The scale bars indicates 100 and 50 μm respectively.
Fig 2.
35S-proteoglycan secretion and cellular density.
Sparse and dense cultures of HUVEC were metabolically labeled for 24 hours with 35S-sulfate, and 35S- macromolecules were recovered. (A) The amount of secreted de novo synthesized 35S-PGs from 6 donors was determined by scintillation counting and normalized to protein. The presented results are relative to the mean of the dense values. (B) The % distribution of 35S-PGs related to protein detected in cell lysate (cell) and in conditioned medium (medium). Mean values from both sparse and dense cultures originating from 5 donors are shown. Results are presented as mean with SEM denoted by vertical bars, and p-values * < 0.05 were taken as a significant difference between the sparse relative to dense cells using the Students paired t-test. (C) Secreted 35S-PGs were harvested from cell cultures of increasing densities, indicated by arrows. Equal amounts (cpm) were loaded in each lane and separated using SDS-PAGE. We here show one representative of 6 individual experiments. The migration positions of molecular mass markers are given in kDa. (D The intensity of the bands was quantified using Image J and mean ± SEM is presented.
Fig 3.
Major PGs expressed and secreted by HUVEC.
(A) Gene expression levels of perlecan, biglycan, serglycin, syndecan-4 and versican in sparse cultures of HUVEC determined by qRT-PCR was expressed as the inverse of the Ct value relative to the Ct value of the endogenous control gene RPL30. The number of different primary cultures was 7 for all genes except for biglycan and versican resulting from 3 cultures. Differences relative to versican are tested for statistical significanse. (B) Gene expression of perlecan, biglycan, serglycin, syndecan-4 and versican in HUVEC cultured in low (sparse) compared to high (dense) cell densities was determined by qRT-PCR. Values are expressed as the fold change in sparse relative to dense cells. The number of primary cell culture donors was 7, 8, 8, 10 and 6 respectively. (C) The secretion of perlecan, biglycan, serglycin, and versican from dense (d) and sparse (s) HUVEC cultures was compared performing western blotting (perlecan and biglycan) or immunoprecipitation (serglycin and versican). The sample size was adjusted to the protein content of the corresponding cell lysate, and shown are representative results from four donors. Similarly, syndecan-4 shedded to the medium was measured by ELISA in sparse and dense cultures and presented as syndecan-4 per cell relative to the mean of the sparse cells. The results are from five donors. Results are presented as mean with SEM denoted by vertical bars, and p-values < 0.05 were taken as a significant difference between the gene expression in sparse relative to dense cells using the Students paied t-test. * p<0.05, ** p<0.01, *** p<0.001. These results are presented as densitometric measurements in (D).
Fig 4.
Characterization of the high molecular weight CS/DS-proteoglycan component.
Left panel: Equal amounts (cpm) of untreated (u) and trypsin treated (t) 35S-PGs secreted from dense and sparse cells were separated by SDS-PAGE. Right panel: Immunoprecipitated (IP) 35S-labeled serglycin secreted from dense (d) and sparse (s) HUVEC. Sample size was adjusted to the protein content of the corresponding cell lysate. These results are from one representative culture of three donors analyzed. The migration positions of molecular mass markers are given in kDa.
Fig 5.
Intracellular distribution of serglycin, Golgi-marker and vWF in sparse and dense HUVEC.
The left panel shows dense and the right panel shows sparse cultures of HUVEC. The cultures were fixed and stained for serglycin (green) and Golgi marker GM130 (A) in red or the endothelial marker vWF in red (B), with co-localization of the two visible as yellow. Blue color indicates DAPI nuclear staining. The pictures were acquired using a confocal microscope with 60 times magnification; scale bars shows 50 μm.
Fig 6.
Serglycin in functional assays.
(A) Proliferative capacity of serglycin knockdown cells (siSRGN) compared to scrambled controls (siCtr), n = 5. (B) HUVEC were subjected to 10% cyclic stretch for 4 and 24 hours respectively, and the response in serglycin expression was determined by qRT-PCR (FlexCell) and compared to controls (Ctr), n = 3–11. (C) Serglycin mRNA expression (qRT-PCR) and protein (ICC) in serglycin knockdown (siSRGN) in HUVEC was compared to scrambled controls (siCtr), n = 6. (D) Control and serglycin knockdown cells were subjected to an in vitro angiogenesis assay. Left panel show the quantification of tube length and loop numbers by Wimasis Image Analysis. The right panels show a representative picture of in vitro angiogenesis assay showing tube formation capacity on BME gel in siSRGN HUVEC and siCtr, n = 4. (E) Gene expression of SRGN (left) and ANG2 (right) in hypoxia compared to normoxia in triplicates from each of two cell donors. (F) Closure of scratch wound in control cells compared to siSRGN cells. Left panel show percent wound closure after 6 hours, and right panel show representative phase-contrast images of scratch wound at 0 and 6 hours after wounding, n = 3. Results are presented as mean with SEM denoted by vertical bars, and p-values < 0.05 were taken as a significant difference using the Students paired t-test. * p<0.05, ** p<0.01, *** p<0.001, ns: not significant.
Fig 7.
Effect of IL-1β on sparse compared to dense cell cultures.
HUVEC were cultured in high (dense) or low (sparse) cell density for 24 hours with or without IL-1β. (A) HUVEC from 6 donors were metabolically labeled with 35S-sulfate for 24 hours and the 35S-proteoglycan secretion was determined by scintillation counting and related to protein content. (B) SRGN mRNA expression in response to IL-1β activation for dense and sparse cultures (n = 6). (C) Serglycin secretion was determined by immunoprecipitation of 35S-serglycin from conditioned medium. These results are from one representative culture of three donors analyzed (upper panel) and mean densitometric measurements of results from all donors (lower panel). The migration positions of molecular mass markers are given in kDa. (D) Confluent (dense) and sparse HUVEC cultures were cultured on chamber slides, fixed and stained for serglycin, and pictures were acquired using a Olympus Fluo View FV1000 confocal microscope with 60x magnification. The intracellular localization and expression of serglycin after IL-1β stimulation is compared to unstimulated control cells in both dense (left panel) and sparse (middle panel) cultures. The right panel show a 3.7 times magnification of the indicated selected areas and the scale bars indicate 20 and 50 μm respectively. Results are presented as mean with SEM and differences with p<0.05 was regarded as statistically significant using the Students paired t-test. * p<0.05, ** p<0.01, ns: not significant.
Fig 8.
CCL2 and Ang2 response to inflammation and to serglycin knockdown.
The secretion of Ang2 (A, n = 6) in control cells was compared to IL-1β stimulated cells. Also, secretion (B, n = 6), intracellular levels (C, n = 4) and gene expression (D, n = 4) of CCL2 in control cells was compared to IL-1β stimulated cells. Response in Ang2 secretion (E, n = 6) and CCL2 secretion (F, n = 6), intracellular levels (G, n = 4) and gene expression (H, n = 4) in serglycin knockdown cells (siSRGN) were compared to control cells (siCtr). Results are presented as mean with SEM and differences with p<0.05 was regarded as statistically significant using the Students paired t-test. * p<0.05, ns: not significant.
Fig 9.
CCL2 secretion from polarized cells.
HUVEC from three different donors were polarized on semipermeable filters. Confluent monolayers were untreated (Ctr) or incubated with IL-1β for 24 hours. CCL2 content in apical (AP) and basolateral (BL) conditioned media was compared.
Fig 10.
Serglycin and CCL2 distribution in unstimulated and stimulated cells.
HUVEC were cultured on chamber slides for 24 hours without (Ctr) or with IL-1β. The cells were fixed and stained, and pictures were acquired using a Zeiss LSM 700 confocal laser scanning microscope at 62x magnification. These are representative results from cells isolated from three different donors. (A) The left panel show cells stained for serglycin in red and CCL2 in green. The right panel show the negative control stained with secondary antibodies only. These pictures are Z-stacks showing the distribution of the target proteins throughout the cells. (B) The panels show untreated (top) and IL-1β stimulated (bottom) HUVEC cultures, fixed and immunostained for serglycin (green) and CCL2 (red). The left and middle panels show serglycin and CCL2 staining, respectively. The right panel shows the overlay (merge) of serglycin and CCL2. These pictures are the view from the middle of the cells, were both target proteins are expressed equally. The scale bar indicates 10 μm.
Fig 11.
Possible mechanisms for the role of serglycin in regulation of proliferation.
In endothelium, serglycin is subject to constitutive (A) and regulated (B) secretion. Constitutive secretion occurs mainly to the apical side, facing the bloodstream, but also to the basolateral side. Vesicles for regulated secretion might be stored intracellularly and serglycin is secreted mainly to the apical side upon stimulation. Through the glycosaminoglycan (GAG)-chains, serglycin has the ability to bind partner molecules including chemokines, growth factors and proteases, and thus secreted serglycin can offer protection, transport and presentation of these molecules (C). Hence, serglycin can contribute to the formation of chemotactic gradients [72] of both bound and dissociated effector molecules (D). This can have autocrine effects (E) as well as effects on target cells (F). CCL2 is recognized as an angiogenic chemokine [45], and in serglycin knockdown cells we observe an increase in CCL2 secretion. Lack of serglycin has several possible outcomes. It could result in a lack of gradient formation, as well as reduced transport and presentation affecting target cells and autocrine signaling. Further, absence of serglycin could affect storage and secretion, resulting in dysregulated secretion exemplified by the increased levels of CCL2 observed here (G). This can have consequences for the autocrine effect of CCL2 on endothelium, suggesting that serglycin is necessary for chemokine presentation to its receptor.