Modulators of Nitric Oxide-Dependent Osteoinductive Activity of Human Red Blood Cells

not applicable

cation to a level similar to the basal activity using OM alone.
In further experiments (►Fig. 1B), we demonstrated the variability of the osteoinductive potential of RBC membranes (ghosts) from healthy individuals (n ¼ 12), using as a reference the effect of NO released by 400 μM DETA-NONOate (concentration chosen after a series of titration assays); this NO donor is characterized by slow release, which might better simulate the NO release by RBC ghosts in this specific assay.
In previous experiments, we compared distinct fractions of RBC and found that an amount of RBC lysates equivalent, in terms of protein content, to RBC ghosts did not induce calcification of AoSMCs; we also showed that the osteoinductive activity of ghosts was diminished when they were used together with an amount of heme corresponding to that isolated from the same RBC preparation. 1 These findings suggested a NO-scavenging role for oxygenated hemoglobin (OxyHb(Fe 2þ )) in the cytosol of RBC from collected blood. 4,5 OxyHb is preserved in this form by a functional redox regulation system in intact RBCs. 6 However, upon RBC extravasation and cell lysis, it is prone to oxidation to methemoglobin (metHb(Fe 3þ )) or other forms, 7,8 which are incapable of scavenging NO. 6 These reactions are promoted by the oxidative environment reported to exist under conditions of intraplaque hemorrhage. [7][8][9][10][11] Thus, to examine the changes of the osteoinductive potential of RBC lysates in dependence of the hemoglobin oxidation status, we tested the effect of RBC lysates isolated from blood, in which Hb exists in the oxyHb form, 12,13 against that of RBC lysates treated with NaNO 2 , which converts oxyHb to metHb. 14 Conversion was visible as a color change from bright red to brown by visual observation and confirmed spectrophotometrically. 15 The osteoinductive effect on AoSMC by different amounts of RBC lysates (co-incubated with a standard amount of 5 μg of ghost protein from the same donor) was compared with that of 5 μg of ghost protein alone. In these experiments, RBC lysates corresponding to 1/5-1/10 of 5 μg of ghost protein (based on the number of intact RBCs needed  1,000 Â ). Relative arginase content was measured in RBC from healthy donors (n ¼ 7) by means of band densitometry of immunoblots of respective samples, normalized to the housekeeping GAPDH protein and also to the RBC number per sample run. Analyses were performed using the Wilcoxon matched-pairs signed rank test. Medians with IQR are shown. The insert shows a representative image of a Western blot with two samples. (F) Arginase inhibition enhances the osteoinductive potential of RBC lysates, and these effects are additive to those of Hb oxidation. CLs from healthy donors (n ¼ 4) were incubated with AoSMC under osteogenic conditions in the presence or absence of the arginase inhibitor L-norvaline (L-NORVA). Enhancement of osteoinductive activity in the presence of L-NORVA was observed only following conversion of oxyHb to metHb with NaNO 2 . Means AE SD are shown. Ã p < 0.05 for the difference between CL þ NaNO 2 þ L-NORVA and each one of the other conditions (CL, CL þ L-NORVA, CL þ NaNO 2 ). AoSMCs, human aortic smooth muscle cell(s); CL, erythrocyte cell lysate(s); GH, (red blood cell membrane) ghosts; Hb, hemoglobin; IQR, interquartile range; NO, nitric oxide; OD, optical density; OM, osteogenic medium; RBC, red blood cell(s); SD, standard deviation.
to be lysed to obtain the given ghost protein amount) exerted an inhibitory effect on calcification. This was in contrast to NaNO 2 -treated lysates containing metHb, which enhanced the osteoinductive activity of ghosts. The results from four donors are summarized and representative results from one donor are shown in ►Fig. 1(C). Of note, and as also shown in ►Fig. 1(C), RBC lysate concentrations of 1/50 or less had no effect compared with untreated cells. Consistently, RBC lysates alone (in the absence of ghosts) exerted osteoinductive activity on AoSMC only after conversion of Hb to metHb with NaNO 2 treatment (n ¼ 4; representative findings from one donor displayed in ►Fig. 1D). Statistical analysis comparing the osteoinductive effect of untreated with that of NaNO 2 -treated lysates (focusing on the 1/10 concentration; paired t-test for four donors), yielded a p-value of 0.006 for the augmentation in the presence of NaNO 2 when RBC lysates were combined with ghosts, and 0.001 when lysates were used alone.
Another factor that may interfere with NO availability is arginase, which competes with eNOS for their common substrate, L-arginine. 16 First, we quantitated the amount of arginase in ghosts and RBC lysates, and subsequently assessed the effect of arginase on osteoinductive activity using a specific inhibitor, L-norvaline, in lysates from the same donors. Using immunoblot band densitometry, the amount of arginase corresponding to a given number of RBC was 1,000-fold higher in RBC lysates compared with ghosts (►Fig. 1E)). Inhibition of arginase by L-norvaline enhanced both the osteoinductive effect of RBC ghosts (confirming our previous results 1 ) and that of lysates (►Fig. 1F). This effect varied, depending on the levels of arginase in the samples from different donors, which may explain, at least in part, the variability observed in ►Fig. 1(B).
Taken together, our results suggest that both membrane and lysate fractions of RBCs exert NO-dependent osteoinductive activity, which requires a close contact of these components with target cells and is enhanced upon hemoglobin oxidation. Variability of NO bioavailability and RBC lysate-mediated calcification may be due to interindividual differences in the levels of arginase which is present in human erythrocytes and differentially regulated in cardiovascular disease states. 17