Optimization of stem cell therapy after cardiovascular and renal injury depends on many factors, among which is stem cell donor health. The renin–angiotensin system (RAS) plays an important role in cardiovascular and renal homoeostasis and pathophysiology. It is becoming increasingly clear that the RAS affects the therapeutic performance of stem cells. In this issue of Clinical Science, Kankuri et al. dig deeper into the consequences of excessive angiotensin II signalling and reactive oxygen species (ROS) formation in the stem cell donor, applying a model of regenerative medicine after renal injury.
THE ROLE OF THE RENIN–ANGIOTENSIN SYSTEM ACTIVATION IN STEM CELL THERAPY
Stem and progenitor cells, either isolated from bone marrow or other tissues, have been the centre of attention since the discovery, almost two decades ago, of their participation in the formation and modulation of adult organ tissue. This opened a feverous quest to develop clinical protocols for regenerative medicine, among others, in cardiovascular and renal disease [1,2]. The initial assumption was that such progenitor cells, whether endogenously recruited or isolated and (re-)injected into the patient after an initial round of culture, would home to the damaged organ to undergo final differentiation and replace cells that were lost due to injury. Nowadays, it is assumed that these cells are improving regeneration or reducing damage thanks to their paracrine function . Stem cell therapy in the clinic has shown beneficial, but also detrimental, effects . This depends on many factors, such as age and health of the donor, isolation and selection methods, culture and injection methods . Among donor-related factors, excessive angiotensin II (AngII) signalling, a major pathophysiological mechanism in cardiovascular and renal disease, might be an important determinant .
In the study published in this issue of Clinical Science by Kankuri et al. , the therapeutic effect of so-called bone-marrow-derived stromal cells (BMSCs) isolated from Sprague–Dawley (SD) rats was compared with that of BMSCs derived from transgenic littermates producing excessive amounts of AngII due to transgenic overexpression of human angiotensinogen and renin (TG). BMSCs were isolated before the TG rats became hypertensive. In vitro, expanded BMSC were re-injected into SD rats shortly after induction of acute ischaemic kidney injury (AKI) by a temporary stop of the renal blood supply. Using two ‘dosages’, 2×106 and 6×106 cells, SD-BMSCs attenuated renal injury on several important markers. TG-BMSCs did not show an effect and even led to premature death when using the higher cell number. From microarray data it is suggested that TG-BMSCs display a pro-inflammatory phenotype, which might lead to the detrimental effects. However, this is not in agreement with the result that both SD- and TG-BMSCs reduce inflammatory markers. The authors suggest that TG-BMSCs increased the rate of the biphasic inflammation response normally present, but this was not assessed in that study. An alternative explanation might be that an anti-inflammatory effect alone is insufficient, and that successful BMSC therapy requires stimulation of a repair mechanism that is not elicited by the low dose and even counteracted by a high dose of TG-BSMCs.
Anyhow, induction of a pro-inflammatory state by bone-marrow-derived mononuclear cells that ‘ripened’ in an environment of excess AngII is a valid argument, and was also proposed in earlier publications. For example, a high concentration of AngII (1 μM) acutely increases oxidative stress in progenitor cells, implicating NADPH oxidase stimulation, which might in turn lead to a pro-inflammatory state at the places to which these cells home [5,6]. Kankuri et al.  now suggest that mitochondrial reactive oxygen species (ROS) governs this detrimental inflammatory effect of BMSCs. Hence, this study is the icing on the cake; in the negative use of the proverb that both of the main sources of ROS production are now recognized as mediators of detrimental AngII effects in progenitor cells, and in the positive use that this study importantly complements our understanding.
Rather complicated is their observation that AngII type 1 (AT1) and AT2 receptors, which normally oppose each other's function , both seem to suppress mitochondrial respiration rate in BMSCs. As an alternative to what the authors suggest, one might propose that this is an example of the functional switch of the AT2 receptor, which under pathological conditions supports instead of opposes the AT1 receptor . Until now, this switch has only been shown for AngII vasoconstrictions , and the findings by Kankuri et al.  might extend this to mitochondrial function. Another complicated observation is that the AT1 and AT2 receptor antagonists losartan and PD123319 have an effect in vitro, whereas on the other hand renin is not expressed in BMSCs. One might speculate that, due to its increased expression, spontaneous AT1 receptor activity occurs. Losartan, or its metabolite EXP3174, if formed, would act as an inverse agonist . Hypothetically, the functional switch of AT2 receptor might even facilitate spontaneous AT1 receptor activity, which would explain the effect of PD123319. Overnight incubation with 100 nM AngII paradoxically leads to similar effects as losartan, but this might be due to AT1 receptor down-regulation. Perhaps mitochondrial function analysis as described by Kankuri et al.  might be a tool to investigate this hypothetical receptor interaction.
Despite the importance of the study by Kankuri et al. , the reader must be alerted to the extreme condition of the renin–angiotensin system (RAS) in the TG mice used. The model is, as stated, humanized , but this is only true for the fact that the mice express human genes, and not in the sense that a certain human disease (state) is being mimicked. The expression levels of angiotensinogen and renin are extremely high in these animals, and it is doubtful whether such conditions can be found anywhere in humans. For patients, it seems more important to elucidate whether a mild chronic disturbance of the delicate AT1 receptors/AT2 receptors/ROS signalling balance will ultimately lead to similar changes as observed in these mice.
As mentioned in the first paragraph, the cell type that is used for therapy is important. The BMSCs used in the study of Kankuri et al.  are cultured adherent mononuclear cells, propagated in a type of medium that should keep mesenchymal stem cells in an undifferentiated state. However, due to the lack of characterization of the cells with cell-type-specific markers, it is unclear which cells are in the expanded isolate. Strikingly, the present study reports that the appearance of the BMSCs is different in SD as compared with TG. This might indicate that, in fact, there is a difference in composition of the cell population, which provides a potential further explanation of the differential effects.
In general, it is clear that RAS components may play a role in regulation of haemopoiesis and differentiation of bone-marrow-derived stem cells in various cell types [3,9]. However, it is not clear how the RAS is orchestrated in time and location (i.e. stem cells or niche/stroma cell) during differentiation into blood cells or cell types that are needed for tissue regeneration. In other words, the physiological function of RAS needs to be addressed more extensively. That this is relevant has very recently been demonstrated by the finding that AngII levels increase under conditions of vasculopathy to regulate haemopoietic stem and progenitor cell recruitment by bone marrow endothelial cells . Kankuri et al.  address the pathophysiological role of RAS. This also depends on the delicate regulation of RAS activity, i.e. avoidance of excessive AngII signalling, in progenitor cells or their surrounding tissues. In view of the results presented by Kankuri et al.  and those of previous studies described in this commentary, perhaps the time has come to consider exploration of RAS activity markers in donors, donor cells and recipients as selection markers for stem cell therapy.