Adhesion of Annexin 7 Deficient Erythrocytes to Endothelial Cells

Annexin 7 deficiency has previously been shown to foster suicidal death of erythrocytes or eryptosis, which is triggered by increase of intracellular Ca2+ concentration ([Ca2+]i) and characterized by cell shrinkage and cell membrane scrambling with subsequent phosphatidylserine exposure at the cell surface. Eryptosis following increase of [Ca2+]i by Ca2+ ionophore ionomycin, osmotic shock or energy depletion was more pronounced in erythrocytes from annexinA7-deficient mice (anxA7−/−) than in erythrocytes from wild type mice (anxA7+/+). As phosphatidylserine exposure is considered to mediate adhesion of erythrocytes to the vascular wall, the present study explored adhesion of erythrocytes from anx7−/− and anx7+/+-mice following increase of [Ca2+]i by Ca2+ ionophore ionomycin (1 µM for 30 min), hyperosmotic shock (addition of 550 mM sucrose for 2 hours) or energy depletion (removal of glucose for 12 hours). Phosphatidylserine exposing erythrocytes were identified by annexin V binding, cell volume estimated from forward scatter in FACS analysis and adhesion to human umbilical vein endothelial cells (HUVEC) utilizing a flow chamber. As a result, ionomycin, sucrose addition and glucose removal all triggered phosphatidylserine-exposure, decreased forward scatter and enhanced adhesion of erythrocytes to human umbilical vein endothelial cells (HUVEC), effects significantly more pronounced in anx7−/− than in anx7+/+-erythrocytes. Following ischemia, morphological renal injury was significantly higher in anx7−/− than in anx7+/+-mice. The present observations demonstrate that enhanced eryptosis of annexin7 deficient cells is paralleled by increased adhesion of erythrocytes to the vascular wall, an effect, which may impact on microcirculation during ischemia.


Introduction
Annexin A7 (or annexin VII, synexin), a member of Ca 2+ -and phospholipid-binding intracellular proteins [1][2][3] associates with secretory vesicles and serves as a Ca 2+ /GTP sensor in the regulation of secretion [4][5][6]. An initial attempt to create an annexin A7 knockout mouse yielded mice which were lethal at embryonic day 10, as anx7-deficient mutants died in utero from cerebral hemorrhage [7]. The heterozygous mice expressing only low levels of Anx7 protein were viable and fertile. The main molecular consequence of lower anx7 expression was a profound reduction in IP3 receptor expression and function in pancreatic islets [7]. Islet ß cell number and size were increased, presumably due to compensation of impaired insulin secretion by individual defective pancreatic ß cells [7]. In a further attempt, the homologous recombination of embryonic stem cells yielded a viable annexin A7 knockout mouse with mild phenotpye [8]. The phenotype included altered effects of heart rate on cardiac action potential [8] and enhanced glial cell proliferation [9]. Moreover, erythrocytes from annexin A7 knockout mice were shown to be more suseptible to eryptosis [10,11], the suicidal erythrocyte death, which is characterized by cell shrinkage and cell membrane scrambling with exposure of phosphatidylserine at the cell surface [12].
Phosphatidylerine exposing cells may adhere to endothelial cells [37] with resulting impairment of microcirculation [38][39][40][41][42] and, at least in theory, impact on reperfusion and tissue injury following transient ischemia. The present study thus explored, whether the enhanced susceptibility of annexin7 deficient erythrocytes to eryptosis is paralleled by enhanced adherence to human umbilical vein endothelial cells (HUVEC) and altered renal injury following ischemia.

Mice
Blood was drawn from the retroorbital plexus of gene-targeted mice lacking annexin A7 (anx7 2/2 ) and corresponding 129 SV wild type mice (anx7 +/+ ) into heparin-coated tubes. The erythrocytes were washed two times with Ringer solution containing 125 mM NaCl, 5 mM KCl, 1 mM MgSO 4 , 32 mM N-2hydroxyethylpiperazine-N-2-ethanesulfonic acid (HEPES), 5 mM Glucose, 1 mM CaCl 2 , pH = 7.4. The generation and properties of anx7 2/2 mice were described earlier [8]. The work was carried out in accordance with the code of ethics for experiments involving animals as well as the German law for the welfare of animals and has been approved by the respective authority (Regierungsprä sidium Tübingen).

FACS Analysis of Annexin V-binding and Forward Scatter
After incubation under the respective experimental conditions, 50 ml cell suspension were washed in Ringer solution containing 5 mM CaCl 2 and then stained for 20 minutes with Annexin-V-Fluos (1:500 dilution; Roche, Mannheim, Germany) under protection from light [43]. In the following, the forward scatter (FSC) of the cells was determined and annexin V fluorescence intensity was measured in FL-1 with an excitation wavelength of 488 nm and an emission wavelength of 530 nm on a FACS Calibur (BD, Heidelberg, Germany).

Cell Culture of Human Umbilical Vein Endothelial Cells (HUVEC)
Human umbilical vein endothelial cells (HUVEC) from Promocell (Heidelberg, Germany), listed as CRL 1730 by the American Type Culture Collection database, were grown to confluency in complete endothelial cell basal medium from Lifeline cell-system (Kirkland, USA) containing growth factors and 10% fetal bovine serum from Lifeline Cell-Systems (Kirkland, USA). Erythrocyte adhesion to HUVEC cells was determined as described previously [37].

Dynamic Erythrocyte Adhesion to Endothelium in vitro
Cultured HUVEC (5610 5 ) were attached on sterile coverslips coated with 0.2% gelatine (Sigma-Aldrich) by overnight incubation in complete endothelial cell basal medium under cell culture conditions for 24 hours. Erythrocytes prepared as indicated were perfused on a HUVEC monolayer in a flow chamber (Harvard, USA) at arterial shear rates (1200 2s ). The interaction events were recorded with a CCD camera (Carl Zeiss) with 206magnification, followed by analysis of the number adherent erythrocytes per high powerfield. Enhanced eryptosis and adhesion of erythrocytes from annexin7-deficient mice following glucose depletion. A. Histogram of annexin V-binding reflecting phosphatidylserine exposure in a representative experiment of erythrocytes from annexin7-deficient mice (anx7 2/2 ) and their wild type control mice (anx7 +/+ ) exposed for 12 hours to glucose-depleted Ringer. B. Arithmetic means 6 SEM (n = 829) of the percentage of annexin V-binding erythrocytes from annexin7-deficient mice (anx7 2/2 , black bars) and their wild type control mice (anx7 +/+ , white bars) exposed for 12 hours to glucose-containing (left bars) or glucose-depleted (right bars) Ringer. ### significant (p,0.001) difference from glucose-containing Ringer, *** significant difference (p,0.001) from anx7 +/+ erythrocytes (ANOVA). C. Arithmetic means 6 SEM (n = 829) of the forward scatter of erythrocytes from annexin7-deficient mice (anx7 2/2 , black bars) and their wild type control mice (anx7 +/+ , white bars) exposed for 12 hours to glucose-containing (left bars) or glucose-depleted (right bars) Ringer. ### significant (p,0.001) difference from glucose-containing Ringer, *** significant difference (p,0.001) from anx7 +/+ erythrocytes (ANOVA). D. Arithmetic means 6 SEM (n = 6) of the number of erythrocytes from annexin7deficient mice (anx7 2/2 , black bars) and their wild type control mice (anx7 +/+ , white bars) adhering to human umbilical vein endothelial cells (HUVEC) following exposure of the erythrocytes for 12 hours to glucose-containing (left bars) or glucose-depleted (right bars) Ringer. ### significant (p,0.001) difference from glucose-containing Ringer, *** significant difference (p,0.001) from anx7 +/+ erythrocytes (ANOVA). doi:10.1371/journal.pone.0056650.g002 Erythrocyte Adhesion and Tissue Injury Following Acute Renal Failure Annexin A7 knockout (anx7 2/2 ) and wild type mice (anx7 +/+ ) were kept on a regular 12 h dark-light cycle with free access to rodent chow and tap water. Animals were anesthetized and placed on a temperature-controlled heating table (RT, Effenberg, Munich, Germany) to maintain body temperature at 37.0uC for renal ischemia/reperfusion procedure as described [44]. Briefly, right flank incision was performed with help of a coagulation electrode (Erbe, ICC50, Tübingen, Germany) to prevent bleeding of muscle and skin vessels. The renal pedicle, including the renal artery and renal vein, were ligated using a 4/0 silk suture, and the right kidney was removed without interfering with the adrenal vessels. The surgical wound was closed using 5/0 nylon sutures. Animals were then placed in a right lateral decubitus position and a left flank incision was performed. Operations were performed under an upright dissecting microscope (Leica, MZ95, Bensheim, Germany). The left kidney was carefully removed from connective tissues, avoiding the adrenal gland and vessels and the kidney positioned in a lucite cup. The kidney was kept wet and warm with a wet swab soaked with water at 37.0uC. The vessel was dissected from adjacent tissues, close to its takeoff from the abdominal aorta, then an 8/0 nylon suture (Ethicon, Norderstedt, Germany) was placed around the artery. This technique allows for interruption of only the arterial blood flow to the kidney without compression of the renal vein. The suture was placed over a small pole and a weight of 1 g was attached to each end immediately followed by occlusion of the artery. Successful occlusion was confirmed by a change in color from red to white. After the experimental procedure (45 minutes ischemia), the left kidney was reperfused by removal of the hanging weights and returned into its retroperitoneal position. The surgical wound was closed using 5/0 nylon sutures. Following 24 hours after reperfusion mice were sacrificed and kidneys were perfused with PBS followed by perfusion with 4% paraformaldehyde. . Enhanced eryptosis and adhesion of erythrocytes from annexin7-deficient mice following osmotic shock. A. Histogram of annexin V-binding reflecting phosphatidylserine exposure in a representative experiment of erythrocytes from annexin7-deficient mice (anx7 2/2 ) and their wild type control mice (anx7 +/+ ) exposed for 2 hours to hyperosmotic shock (550 mM sucrose added). B. Arithmetic means 6 SEM (n = 829) of the percentage of annexin V-binding erythrocytes from annexin7-deficient mice (anx7 2/2 , black bars) and their wild type control mice (anx7 +/+ , white bars) exposed for 2 hours to isotonic (left bars) or hyperosmotic (550 mM sucrose added, right bars) Ringer. ### significant (p,0.001) difference from isotonic Ringer, *** significant difference (p,0.001) from anx7 +/+ erythrocytes (ANOVA). C. Arithmetic means 6 SEM (n = 829) of the forward scatter of erythrocytes from annexin7-deficient mice (anx7 2/2 , black bars) and their wild type control mice (anx7 +/+ , white bars) exposed for 2 hours to isotonic (left bars) or hyperosmotic (550 mM sucrose added, right bars) Ringer. ### significant (p,0.001) difference from isotonic Ringer, *** significant difference (p,0.001) from anx7 +/+ erythrocytes (ANOVA). D. Arithmetic means 6 SEM (n = 6) of the number of erythrocytes from annexin7deficient mice (anx7 2/2 , black bars) and their wild type control mice (anx7 +/+ , white bars) adhering to human umbilical vein endothelial cells (HUVEC) following exposure of the erythrocytes for 2 hours to isotonic (left bars) or hyperosmotic (550 mM sucrose added, right bars) Ringer. ### significant (p,0.001) difference from isotonic Ringer, *** significant difference (p,0.001) from anx7 +/+ erythrocytes (ANOVA). doi:10.1371/journal.pone.0056650.g003

Morphologic Evaluation of Renal Histology
For histological analyses, 4% paraformaldehyde fixed kidneys were with, dehydrated in ethanol and xylol followed by embedding in paraffin. Paraffin-embedded tissues were cut into sections of 2 mm thickness and stained with periodic acid Schiff (PAS) and Sirius red stain. Renal morphology was investigated by light microscopy as described below with the investigator being blinded to the genotype of the mice. Tubulointerstitial damage, i.e. tubular necrosis, tubular atrophy and tubular dilation was assessed on PAS-stained paraffin sections at a magnification of x200 using a semiquantitative scoring system (acute tubular necrosis score, ATN). For determination of ATN, 15 fields per kidney were randomly sampled and graded as follows: grade 0, no change; grade 1, necrosis involving less than 25% of the area; grade 2, necrosis affecting 25-50%; grade 3, necrosis involving more than 50%, and grade 4 involving (almost) the entire area. Erythrocyte attachment was monitored in the inner stripe of diseased kidneys in 12 fields at 2006 magnification and graded using a scoring system from 0 to 4. (Grade 0, no Erythrocytes attached; grade 1, sporadic erythrocytes attached; grade 2, attached erythrocytes in more than 25% of the capillaries; grade 3, attached erythrocytes in more than 50% of the capillaries; grade 4, attached erythrocytes in more than 75% of the capillaries.).

Statistics
Data are expressed as arithmetic means 6 SEM. Statistical analysis was made using t test or paired ANOVA with Tukey's test as post-test, as indicated in the figure legends. n denotes the Figure 4. Role of phosphatidylserine exposure in dynamic adhesion of annexin-treated erythrocytes to endothelial cells under arterial shear stress. A. Arithmetic means 6 SEM (n = 6) of erythrocytes from annexin7-deficient mice (anx7 2/2 , black bars) and their wild type control mice (anx7 +/+ , white bars) binding to human umbilical vein endothelial cells (HUVEC) under flow following exposure for 30 minutes to Ringer solution without (control) or with 1 mM ionomycin without or with a prior 30 min treatment with Annexin V (5 ml/ml). ***(p,0.001) indicates statistically significant difference from absence of ionomycin, ###(p,0.001) indicates statistically significant difference from absence of annexin V (ANOVA). B. Arithmetic means 6 SEM (n = 6) of erythrocytes from annexin7-deficient mice (anx7 2/2 , black bars) and their wild type control mice (anx7 +/+ , white bars) binding to human umbilical vein endothelial cells (HUVEC) under flow following a 12 hours treatment with Ringer solution with (control), or without (-Glucose) glucose without or with a prior 30 min treatment with Annexin V (5 ml/ml). ***(p,0.001) indicates statistically significant difference from absence of glucose, ###(p,0.001) indicates statistically significant difference from absence of annexin V (ANOVA). C. Arithmetic means 6 SEM (n = 6) of erythrocytes from annexin7-deficient mice (anx7 2/2 , black bars) and their wild type control mice (anx7 +/+ , white bars) binding to human umbilical vein endothelial cells (HUVEC) under flow following a 2 hours exposure to isotonic (control) or hypertonic (550 mM sucrose) Ringer solution without or with a prior 30 min treatment with Annexin V (5 ml/ml). ***(p,0.001) indicates statistically significant difference from absence of hyperosmotic shock, ###(p,0.001) indicates statistically significant difference from absence of annexin V (ANOVA). doi:10.1371/journal.pone.0056650.g004 number of different erythrocyte specimens studied. The batches of erythrocytes differed moderately in their susceptibility to eryptosis. Thus, the control values were not identical in all series of experiments. To avoid any bias potentially introduced by the use of different erythrocyte batches, comparisons were always made within a given erythrocyte batch.
Annexin7 deficiency further sensitized erythrocytes to the eryptotic effects of energy depletion. As shown in Fig. 2, energy depletion by removal of glucose resulted within 12 hours in a significant increase of the percentage of annexin V-binding erythrocytes reflecting erythrocytes exposing phosphatidylserine at their surface ( Fig. 2A,B). The effect was slightly, but significantly, more pronounced in anx7 2/2 than in anx7 +/+ erythrocytes. Glucose removal further decreased the forward scatter, reflecting erythrocyte shrinkage (Fig. 2C). Again, the effect was significantly more pronounced in anx7 2/2 than in anx7 +/+ erythrocytes. As shown in Fig. 2D, glucose depletion further enhanced the percentage of erythrocytes adhering to HUVEC, an effect again significantly more pronounced in anx7 2/2 than in anx7 +/+ erythrocytes.
Similar observations were made under hyperosmotic shock, which was induced by addition of 550 mM sucrose to isotonic Ringer solution. As shown in Fig. 3, hyperosmotic shock was within 2 hours followed by a significant increase of the percentage of annexin V-binding erythrocytes (Fig. 3A,B), an effect significantly more pronounced in anx7 2/2 than in anx7 +/+ erythrocytes. Hyperosmotic shock further decreased forward scatter (Fig. 2C), an effect again significantly more pronounced in anx7 2/2 than in anx7 +/+ erythrocytes. Osmotic shock further enhanced the percentage of erythrocytes adhering to HUVEC, an effect again significantly more pronounced in anx7 2/2 than in anx7 +/+ erythrocytes (Fig. 3D).
Additional experiments were performed to investigate whether enhanced vascular adhesion of Annexin A7 erythrocytes following ionomycin-treatment requires phosphatidylserine exposure at the cell surface. To this end, phosphatidylserine on the erythrocyte surface was blocked with annexin-V, which firmly binds to and thus masks phosphatidylserine. As illustrated in (Fig. 4A), the increased adhesion of ionomycin-treated erythrocytes to HUVEC under flow at shear rates of 1200 2s was significantly attenuated in the presence of annexin-V (5 ml/ml). Similar observations were made following exposure of the erythrocytes to energy depletion (Fig. 4B) and following exposure of the erythrocytes to hyperosmotic shock (Fig. 4C).
Since CXCL16 has previously been shown to bind phosphatidylserine and to be thus involved in cell adhesion [37,45,46], further experiments were performed to test whether the binding of ionomycin-treated erythrocytes to endothelium under flow conditions involves CXCL16. Therefore, HUVEC were exposed to an antibody directed against CXCL16 (4 mg/ml). As shown in Fig. 5A, the endothelial adhesion of ionomyin-treated erythrocytes was significantly less pronounced following exposure of HUVEC cells to CXCL16-blocking antibody than following treatment of HUVEC with isotype control antibody of the same concentration. Again, similar observations were made following exposure of the erythrocytes to energy depletion (Fig. 5B) and following exposure of the erythrocytes to hyperosmotic shock (Fig. 5C).
Plotting of erythrocyte adhesion to HUVEC against the percentage of phosphatidylserine exposing erythrocytes (Fig. 6) revealed that at any given phosphatidylserine exposure the adhesion of anx7 2/2 and anx7 +/+ erythrocytes was similar. Thus, increased adhesion of annexin7 deficient erythrocytes was fully explained by their enhanced susceptibility towards triggers of cell membrane scrambling.
To test for a potential consequence of erythrocyte adhesion for microcirculation in vivo, kidneys were subjected to ischemia (45 min) and subsequently analysed by light microscopy. 24 hours after induction of ischemia/reperfusion model abundance of erythrocytes tended to be increased in anx7 2/2 in ischemic kidneys at the inner stripe ( Fig. 7A; E arrows) compared to anx7 +/+ mice ( Fig. 7A; D arrows). This effect did not reach statistical significance due to high variability. However, kidneys from anx7 +/+ mice showed only mild tubular necrosis with partial loss of tubular brush border (Fig. 7F). In contrast, ischemia/reperfision in anx7 2/ 2 mice resulted in significantly enhanced tubular necrosis with cast formation and more pronounced loss of the tubular brush border ( Fig. 7B; C; G).

Discussion
The present study confirms previous observations revealing enhanced susceptibility of annexin7 deficient erythrocytes to stimulators of eryptosis [10,11]. More importantly, the present observations reveal that the enhanced eryptosis of annexin7 deficient erythrocytes is paralleled by increased adhesion of affected erythrocytes to endothelial cells.
As shown earlier [10], PGE 2 -formation, cation currents, increase of cytosolic Ca 2+ concentration ([Ca 2+ ] i ), and cell membrane scrambling were all more pronounced in anx7 2/2 than in anx7 +/+ -erythrocytes. The difference was blunted following inhibition of cyclooxygenase by aspirin or diclofenac [10]. Increase of cytosolic Ca 2+ concentration is well known to trigger erythrocyte cell membrane scrambling with subsequent phosphatidylserine exposure at the cell surface [26,28]. Cytosolic Ca 2+ is further known to activate Ca 2+ sensitive K + channels [23,24] resulting in cell shrinkage due to exit of K + , hyperpolarisation of the cell membrane, exit of Cl 2 and thus cellular loss of KCl with osmotically obliged water [25].
Coating of phosphatidylserine at the erythrocyte surface by annexin V interfered with the engulfment of the eryptotic cells by macrophages [11] and interfered with binding of phosphatidylserine exposing erythrocytes to human endothelial cells (HUVEC) [37].
At least in theory, adhesion of eryptotic erythrocyte in renal medulla could contribute to acute ischemic renal failure. The high Cl 2 and urea concentrations prevailing in renal medulla counteract erythrocyte phosphatidylserine exposure [87] and at normal renal blood flow the contact time of erythocytes within the renal medulla is too short to trigger significant phosphatidylserine exposure. The contact time is, however, markedly increased in acute renal failure which is typically paralleled by trapping of erythrocytes in the transition from medulla to cortex [88]. Following ischemia, the abundance of erythrocytes tended to be higher in anx7 2/2 mice and anx7 +/+ mice., a difference, however, not reaching statistical significance. Ischemia was followed by tubular necrosis, which was significantly more pronounced in anx7 2/2 mice than in anx7 +/+ mice. It remains uncertain, however, whether this difference on renal injury is due to impact of erythrocytes on microcirculation, due to altered function of other blood cells or due to differences in renal epithelial cells.
In conclusion, the present observations show that the enhanced susceptibility of annexin7 deficient erythrocytes is paralleled by enhanced adhesion to human umbilical vein endothelial cells. The enhanced adhesion of eryptotic erythrocytes is expected to compromize microcirculation and thus to participate in the pathophysiology of disorders with excessive eryptosis.