Physiological role of Prion Protein in Copper homeostasis and angiogenic mechanisms of endothelial cells

The endothelial cell model used in the present study is represented by the HUVEC (Human Umbilical Vein Endothelial Cells) primary cell line. Cells were isolated by digestion of endothelial fractions derived from the umbilical cord vein collecting in the Gynecology and Obstetrics unit of the “Vito Fazzi” hospital in Lecce, within 24 hours from their withdrawal. Informed consent was obtained from each subject, according to the declaration of Helsinki. The umbilical cords have been preliminarily treated with a disinfectant solution, then washed with PBS (Sigma-Aldrich, Milan) containing antibiotics (Penicillin/Streptomycin; Sigma-Aldrich) and an antimycotic (Amphotericin B) and then perfused by cannulation to remove blood residues. Collagenase IV solution (0.1% in PBS, Sigma-Aldrich) was then injected into the vein. Umbilical cords were transferred to the incubator (37 ° C) for 20 minutes, in order to allow the enzymatic digestion. The endothelial fraction was collected in a Falcon tube containing bovine fetal serum (FBS; Sigma- Aldrich) and cell suspension was centrifuged for 8 minutes at 1000 g. The pellet was washed twice in PBS and cells were resuspended in fresh culture medium (M199 Euroclone, Life Science Division) and seeded in pre-treated 0.2% gelatin/PBS T25 flasks. For experimental tests, endothelial cells were used within two and four passages.

Cells were plated into 96-well trays and 0.5 mg/ml MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide; Sigma-Aldrich] was added to cell cultures for 3 h at 37°C. The MTT formazan product was released from cells by adding dimethylsulfoxide and measured spectrophotometrically at 570 nm (44). The percent of survival was assessed by comparison with untreated cultures. The estimation of viable cells is expressed as a percentage of the optical density values acquired compared to the controls.

RNA extraction from cells grown at 70–80% confluence was performed by using the TRIzol reagent (Invitrogen, Milan, Italy), accordingly to the manufacturer’s instructions. Total RNA (0.5 μg) was reverse-transcribed with random hexamers in a 20 μl reaction volume by the GeneAmp Gold RNA PCR kit (Perkin-Elmer, Monza, Italy). Amounts of cDNA template from 0.2 to 0.5 μg were needed for real-time PCR analysis. Primers and annealing sequences are listed in Table 1. Reactions were carried out into a 25-μl mixture volume using the SmartCycler System (Cepheid, Euroclone, Milan, Italy) and the threshold cycles (Ct) were averaged from three samples, each analysed in duplicate. The average expression ratio of genes with respect to the control condition was computed by Pfaffl’s formula (45).

The cell lysis protocol was applied to semi-confluent monolayers, performing all operations on ice. Preliminarily two washes with PBS were carried out, after which the cell lysis was performed in RIPA buffer (Tris-HCl 50 mM, pH 7.4, NP-40 1%, Sodium-deoxycholate 0.25%, NaCl 150 mM, EDTA 1 mM, PMSF 1 mM, Na₃VO₄ 1 mM; 0.5 ml for 10⁶ cells). Protein extracts (~30 μg) were boiled in Laemmli sample buffer (Sigma-Aldrich) for 5 min, resolved on 10% SDS-polyacrylamide gels and transferred onto nitrocellulose membranes at 190 mA for 1.5 h. Membranes were blocked for 1 h in Tris-buffered saline (TBS), 0.05% Tween-20, 5% non-fat dry milk, followed by overnight incubation with specific primary antibodies diluted in the same buffer. The primary antibodies used are: anti-PrP (1:8000, Sigma-Aldrich) and anti-β-actin (1:8000, Sigma-Aldrich). After washing with 0.1% Tween in TBS, membranes were incubated with a peroxidase-conjugated secondary antibody for 1 h, washed and developed using the ECL chemiluminescent detection system (Clarity™ Western ECL Substrate Biorad). The densitometric analyses of blots were performed by a computerized image processing system (Image J, 1.0 version).

PrP-targeting siRNAs have been provided from Ambion (Life Technologies Italia, Monza, MB) in the Silencer® Select Pre-Designed & Validated formulation. In particular, we used two different siRNAs (ID s11212, s11213) for the target gene (PRNP), one scrambled siRNA as negative control and the transfectant agent siPORT NeoFXTM (Ambion), preliminarily diluted in OPTI-MEM® I (Invitrogen), at reduced culture medium serum content, in order to vehicle them into the cell.

The gene silencing experiments were set up in 12-well plates and each of the experimental conditions listed as follow were tested in duplicate: control (incubation with the transfectant agent in the absence of siRNA), negative control (transfection with 30 nM scrambled siRNA), transfection with siRNA ID 11212 30 nM, transfection with siRNA ID 11213 30 nM, simultaneous transfection with the two siRNAs, both tested at concentration of 15 nM. After a short incubation period, the siRNA/siPORT NeoFX complexes transfection agents were distributed in the multi-well plate (100μL/well) immediately before cell seeding. The expression levels of the prion protein were analyzed by Western Blotting after 48 h from transfection.

Confluent HUVEC cell cultures seeded in a 12-well plate have been used for membrane Cu²⁺-reductase activity assay. Monolayer cells have been washed with PBS and incubated for 3-15 min at 37°C, with HBSS buffer containing Cu-His₂ (10-50 μM) and the copper ions chelation agent bathocuproine-disulfonic acid disodium salt (BCS; Sigma-Aldrich, Milano) 200 μM. The BCS-Cu⁺ complex formation (molar extinction coefficient ε = 12.25 mM^-1 cm^-1) within the cell culture medium has been assessed by spectrophotometer reading at λ = 482 nm. Data have been normalized on the cell lysate protein content of each well.

The kinetics of Cu intake were studied in HUVEC control and PrPKD cells using the green-fluorescent heavy metal indicator Phen Green SK diacetate (PG SK) λ_exc = 506 nm, λ_em = 530 nm; Molecular Probes, Invitrogen). Cells were harvested and then dye-loaded by incubation for 10 min at 37°C in HBSS (in mM: 140 NaCl, 5 KCl, 1 Na₂HPO₄, 1 CaCl₂, 0.5 MgCl2, 5 D-glucose, 10 Hepes; pH 7.4) containing 1 μM PG SK. After loading, cells were washed twice by centrifugation and then resuspended in the physiological buffer cited above. For each evaluation, an aliquot of cells (3 x 104) was introduced in a cuvette housed in a LS-50B Perkin-Elmer spectrofluorometer, after the preheating of the sample compartment at 37°C. Once the emission signal of the probe ‘‘entrapped’’ within the cells was stable, Cu chloride was added at different concentrations (0–5 μM) and the fluorescence quenching (ΔF a.u. s^-1 nr cells^-1 10^-5) was monitored as an indicator of Cu influx rates into the cytoplasm. A set of experiments was conducted on both untreated and Cu-depleted cells, before and after the enzymatic removal of surface PrP. In the last case, cultures were washed twice with phosphate-buffered salt solution and then pre-incubated in serum-free medium containing 5 mU/cm² phosphatidylinositol-specific phospholipase C (PI-PLC) from B. Cereus (Sigma-Aldrich) for 2.5 h at 37°C. Cells were dye-loaded as reported above and Cu transport assays were performed by testing single extracellular Cu concentrations in separate experiments. The absolute Cu intake, expressed as a percentage of control, was calculated as ΔF_max/F₀ ratio, where ΔF_max is the maximum quencing in fluorescence observed after the addition of Cu to cells and F₀ is the baseline signal.

The HUVEC control and silenced cells (since now on referred to as PrPKD cells) at the same number of passages were trypsinized and resuspended in the M199 culture medium at a density of 70.000 cells/ml. The “hanging drop method” (46) was used to generate the “spheroids pendant”. 5 ml of sterile PBS were introduced on the bottom of 60 mm plates in order to create “wet rooms”. Spheroids growth was monitored and images were acquired by an inverted optical microscope in contrast phase mode, at regular intervals of time (0-48 h).

50 μL of Matrigel (BD Biosciences) with reduced growth factor content was deposited on the bottom of pre-cooled 96-well plates. In order to allow matrix polymerization, the plates were transferred to 37° C for 2 hours. HUVEC control and PrPKD cells were, therefore, seeded at a density of 5000 cells/cm² and stimulated with VEGF 10ng/ml for 16h. The formation of capillary-like structures was evaluated by observation via an inverted optical microscope in contrast phase mode. The characteristics of the networks (percentage area covered by cells, tubule extension, number of nodes) were analyzed using the WimTube software (Wimasis GmbH, Munich, Germany).

Results are presented as means ± SE and each experiment has been done at least in triplicate. Statistical comparisons have been made by Student’s t test and ANOVA followed by Dunnett’s post-test. Significance was demonstrated at p<0.05 (*), p <0.01 (**).

In order to understand the contribution of PrP in the processes of endothelial activation we used a RNA interference (RNAi) protocol.

The experimental procedure associated with RNAi could cause cellular stress and cytotoxicity, masking the specific effect of the PrP gene down-regulation. In order to critically evaluate this aspect and to avoid the use of toxic amount of transfectant agent, cellular viability assays have been performed. As shown in Figure 1, the administration of scalar quantities of transfectant agent (0.15-75 μl/well) did not cause any significant alterations of cell viability with respect to the control condition (“no siRNA”, siPORT NeoFX 0 μl/well). Furthermore, the combined use of siRNAs did not produce any significant toxic effect even after 72h of incubation time.

Figure 1

After these preliminary assessments, we proceeded to the preparation of the knockdown cultures used for subsequent evaluations. HUVEC cells were transfected with siRNA s11212 and s11213 (used individually or mixed according to the concentrations already mentioned), capable of recognizing different sequences of the PrP transcript. The silencing efficiency rate was checked 2 days after transfection by Real Time PCR analysis (Fig.2a). At any tested experimental condition, the transcript levels were found to be lower than those measured in the control (Fig.2a); in particular, the highest silencing efficiency rate was obtained simultaneously transfecting both siRNAs. In order to quantify the persistence level of the post-transcriptional silencing effect, the values of the transcript expression were measured at 48h (Fig. 2c).

Figure 2

In parallel, immunoblotting assay was performed to check the reduction of the PrP protein expression levels (Fig.2b, c).

The HUVEC control and PrPKD cells were exposed for 24 hours to increasing copper concentrations and cell viability was measured by MTT assay. The cytotoxicity profiles reported in Figure 3 shows how both cultures are sensitive to the dose-dependent effect, with a higher extent for PrPKD cells. In fact, the EC₅₀ values for copper-induced cytotoxicity are 50.78 ± 1.29 μM for control cells and 26.06 ± 1.25 μM for PrPKD cells, respectively.

Figure 3

According to current hypothesis, copper ions are reduced (Cu⁺) before being translocated through the membrane. In the light of in vitro studies showing how PrP N-terminal octarepeat region is able to exert a reductase activity, we decided to test PrP enzymatic function also in the HUVEC cells using a colorimetric assay (Fig.4). HUVEC control and PrPKD cells have been exposed to a high concentration of Cu²⁺ (50 μM) for 15 min together with Cu⁺ chelating BCS.

The absorbance values of the culture medium after incubation showed a marked but still not significant reduction of about 30% in HUVEC PrPKD with respect to the control cells. As reported in Figure 4 “Cu/BCS”, no spontaneous conversion phenomenon of copper ions has been observed. Thus, PrP shows to have a role in copper ions transport facilitating their reduction, a necessary step needed for transmembrane translocation.

Figure 4

PrP usually undergoes recycling between the plasmatic membrane and the endosomal district; so, we decided to investigate any possible association between protein internalization process and copper transport in HUVEC cells. These cells were pre-treated with two different endocytosis inhibitors, Cytochalasin B and Chlorpromazine separately, and exposed for a short time to hypertonic shock in order to evaluate the copper transport activity in different experimental conditions. The cells previously incubated with the Phen Green SK fluorescent dye and were exposed to Cu²⁺ 20 μM (Fig. 5), making it possible to discriminate between the amount of the copper endothelial transport mediated by either dependent or independent endocytosis mechanisms. As shown in Figure 5a, only Cytochalasine B is able to significantly inhibit copper transport. Testing scalar concentrations of Cytochalasin B (0-100 μM) we found that the highest dose (100 μM) represented the most feasible condition to evaluate the PrP silencing effect on the endocytotic copper translocation (Fig. 5b). The endocytosis inhibitor produced a 50% reduction of the copper transport in Ctrl cells, while it did not exert any detectable effect on PrPKD cells (Fig. 5c). We can then conclude that copper endocytosis-dependent transport in HUVEC cells is regulated by the prion protein, given also by the high velocity of the experimentally observed copper uptake compatible with a fast-kinetic endocytosis process.

Figure 5

The use of a cell culture protocol in suspension for the generation of spheroids, allowed us to evaluate the ability of PrPKD cells to establish stable interactions with other cellular and extracellular matrix elements. As shown in Figure 6, the morphological analysis of spheroids growth after 16h showed an evident irregularity in the aggregates obtained from PrPKD cells. After 48h, these structures manifested their fragility, showing signs of disaggregation comparable to those of spheroids derived from the control cells. The spheroids are heterogeneous and, because of their architecture, contain a proliferating cell population on their surface and quiescent cells in the centre due to limited access to nutrient and oxygen (47). For this reason, we calculated the percentage of viable cells in control and PrPKD spheroids. These cellular structures underwent mechanical dissociation and the number of released cells was counted. Cell viability levels were significantly lower for PrPKD derived spheroids (Fig. 6). Although this experimental result does not allow to discriminate between necrosis and apoptosis phases, it suggests however, a net loss of viability for PrPKD cultures, probably due to a higher sensitivity to a nutrient and/or oxygen deficiency. In details, the values of the cellular vitality are comparable to those observed after 16h of cultured cells (Fig. 7). After 24h, a 20% reduction in PrPKD cultures viability was shown, despite no significant morphological difference emerged between the two conditions. After 48h, the cell viability reduction reaches approximately 30% compared to the control (Fig. 7). Indeed, in this phase the PrPKD spheroids resulted fragmented. The presence of blebs around the margin of the PrPKD aggregates as shown in Figure 6 was used as a feature of apoptotic events.

Figure 6

Figure 7

HUVEC control and PrPKD cells were seeded at the same density on a thin layer of Matrigel without growth factors and then stimulated with VEGF 10 ng/ml for 16h to induce the formation of a vascular-like network. Figure 8 shows clearly how PrPKD cells are unable to establish contacts and to shape a complex network: only a few and brief cellular extroflections (sprouts) are noted, indicating a reduced invasive capacity. On the other hand, cells transfected with scrambled siRNAs are able to form mature networks, with a fair number of nodes and branches, as illustrated by using the WimTube image analysis software (Wimasis Gmbh, Munich, Germany). This data clearly suggests PrP ability to interact with several extracellular matrix elements (data being confirmed). Since PrP regulates many different signal transduction pathways with effects on the cellular transcriptional activity, we evaluated Flt-1 and KDR receptors expression levels that may be involved in the cell response mechanism to VEGF growth factor. RT-PCR analysis highlights no significative differences between the receptor transcript levels in Ctrl and PrPKD cells as reported in Figure 9, strengthening the hypothesis of PrP VEGF-independent role in the endothelial cells morphogenesis mechanisms.

Figure 8

Figure 9