Angiotensin II That Reduces the Colony‐Forming Ability of Hematopoietic Progenitors in Serum Free Medium Has an Inverse Effect in Serum‐Supplemented Medium

Hematopoiesis is a complex process regulated by multiple factors: cellular interactions, cytokines, chemokines, and numerous other molecules including vasoactive peptides (VAP). Angiotensin II (AII) and substance P (SP) were recently described for their stimulatory effect on hematopoiesis [1–3]. Recently, Rodgers et al. [4] reported in Stem Cells that the vasoactive peptide, AII, increases the colony-forming ability of human cord blood CD34+CD38− cells cultured in methylcellulose 1% in Iscove's modified Dulbecco's medium supplemented with 30% fetal calf serum (FCS). Vasoactive peptides are products (AII) or substrates (SP) of angiotensin-converting enzyme (ACE) [5, 6], which is present in cell culture sera [7]. Our interest in the role of AII in the regulation of hematopoiesis and the data mentioned above led us to explore the effect of AII on the proliferation and colony formation of human cord blood CD34+ cells in serum-free and serum-containing media. Cord blood CD34+ cells were isolated by an immunomagnetic technique (Miltenyi Biotech Kit; Paris, France; http://www.miltenyibiotec.com) and cultured (500 cells/ml in a 35-mm Petri dish) for 14 days at 37°C and 5% CO2 in semisolid medium (1% methylcellulose) (Stem αII; Stem α; Saint Clément les Places, France) containing erythropoietin 3 U/ml, stem cell factor 50 ng/ml, and GM-CSF 10 ng/ml with or without AII 10−4 M and with or without 2%, 10%, or 20% of FCS. AII was used at 10−4 M since initial experiments testing a range of concentrations from 10−10 to 10−4 M showed that the maximal effect, observed at 10−4 M, was reduced at 10−7 M and disappeared at 10−10 M. Similarly to Rodgers et al. [4], who performed their cultures with 30% FCS, we found that AII increased the colony-forming unit-granulocyte-macrophage (CFU-GM) growth by about 50% when CD34+ cells were cultured in methylcellulose culture medium containing 20% FCS (Fig. 1). In contrast, AII in serum-free medium had a major inhibitory effect (70% reduction) on CFU-GM colony formation and a lesser inhibitory effect (27% reduction) on BFU-E colony formation. The addition of increasing concentrations of FCS (2% or 10%) to the culture medium (Fig. 1) showed that AII still reduced the formation of CFU-GM by 20% in the presence of 2% FCS, but lost its inhibitory effect when the culture was done with 10% FCS. BFU-E and CFU-GM colony formations were studied in culture media with 0%, 2%, 10%, or 20% FCS and with AII 10−4 M. Results of one representative experiment out of three are shown as a percentage of the control culture, which was performed in culture media supplemented with 0%, 2%, 10%, or 20% FCS and without AII (dashed vertical line). Abbreviations: CFU-GM = colony-forming unit-granulocyte-macrophage; FCS = fetal calf serum; AII = angiotensin II. Our data firstly confirm that the addition of AII stimulates colony formation of human cord blood CD34+ cells (by 47% and 30% for CFU-GM and BFU-E, respectively) in the presence of 20% FCS. Interestingly, we also showed, for the first time, that in serum-free medium AII has an inhibitory effect on colony formation and that the shift from a negative (at ≤2% FCS) to a positive (at ≥20% FCS) effect depended on the concentration of serum. Two hypotheses might explain these contradictory results. First, the effect observed in serum-containing medium, in fact, may be related not to AII itself, but to its catabolites produced in culture. Indeed, infusion of exogenous AII increased hematocrit of anemic ACE knockout mice [8], providing evidence that the renin-angiotensin system plays a role in erythropoiesis. However, these results suggest that AII could be the active molecule, but they cannot exclude the hypothesis that other enzymes present in the serum degrade AII into angiotensin III (aminopeptidase A) and then into angiotensin IV (aminopeptidase N) or into angiotensin II (1-7) (neutral endopeptidase) (Fig. 2). These peptides, which either share AT1 and AT2 receptors with AII (angiotensin III) or which bind to other receptors (angiotensin IV and angiotensin II [1–7]), exert similar or contrary biological effects on several cell types (Fig. 2) [9]. Bioactive metabolites of the renin-angiotensin system, their receptors, and functions. Modified from Chansel et al. [9]. The second hypothesis is that the presence of FCS in the culture medium might downregulate AT2 receptors at the surface of CD34+ cells, as already shown on R3T3 cells [10]. This would lead AII to act predominantly on AT1 receptors, which induce proliferation of several cell types [11], without being antagonized by simultaneous activation of AT2 receptors, as previously shown [12]. Taken together, these data suggest that the regulation of hematopoiesis by AII is complex and probably dependent on the respective concentrations of AII and its catabolites as well as on the types of receptors at the surface of CD34+ cells. Indeed, most results of cultures in serum-containing media could reflect the effects of active metabolites of AII and other VAP, rather than the direct effect of these peptides on the colony-forming cells, and/or the effects of serum on AT2-receptor expression. Testing the effects of peptides in serum-free media would avoid this bias. Even if serum-free cultures do not reflect the more complex in vivo physiological situations, they provide important information about the direct effects of the tested peptides, and they probably reflect the effects observed in paracrine and autocrine regulation by AII. Our data also suggest that in vivo experiments with molecules such as VAP, whose metabolites are also bioactive, should be analyzed carefully since they can reflect the results of different, and possibly inverse, effects of the substrate and its metabolites.