GPR17: a receptor preferentially expressed in resident stromal cells before ischaemia, but shared in inflammatory cells and cardiac mesenchymal cells with MF progenitor characteristics after infarction
GPR17 is a P2Y-like receptor responding to either uracil-nucleotides and Cys-LTs whose presence characterizes various organs susceptible to ischaemic damage (e.g. brain, kidney and heart) . Its peculiar pharmacological profile, initially unveiled in heterologous expressing systems, has been independently confirmed by some [36, 37], but not other Authors . These contrasting data may be as a result of dependence of the activity tests on the conditions adopted in recombinant ‘artificial’ systems, which potentially give rise to artefacts, especially in the case of constitutive GPR17 activation . On the other hand, GPR17 responses have been more recently fully confirmed by us [18, 40, 41] and other authors  in cells that natively express the receptors, like oligodendroglia [15, 18, 42] and PC12 cells . GPR17 has been also reported to act as a negative regulator of the Cys-LT1 receptor [36, 38]. This suggests that, in addition to responding to cys-LTs, depending on specific pathophysiological conditions, GPR17 can interact with other closely related receptors.
GPR17 functions have been mostly addressed in the brain and the central nervous system, particularly in brain ischaemia or spinal cord injury models [13, 43-45]. Interestingly, it was found that in vivo modulation of GPR17 by pharmacological or genetic targeting improves the outcome of these injuries, suggesting a central role of the receptor in nervous tissue damage sensing, healing and repair [13, 45]. By contrast, except for studies addressing the expression at transcriptional level , no further indication has been provided concerning the identity of cells expressing the receptor in the normal heart and, most importantly, whether GPR17 is modulated following induction of MI.
The first part of the present study was therefore designed to address this crucial point. We found that before MI induction, the most abundant cells expressing GPR17 were located in the interstitium between adjacent cardiac myocytes. These cells were distributed in the ventricular myocardium without a preferential location and were not grouped into cellular clusters resembling the typical cardiac stem cell ‘niches’ [46, 47]. As the majority of cells expressing GPR17 in non-injured hearts displayed Sca-1 but not Iso-B4 Lectin staining (Figs 1A, B and 2A), we conclude that in normal conditions, the receptor characterizes a unique population of cardiac-resident stromal cells that does not display endothelial or myeloid characteristics. This establishes a striking similarity between the heart and the central nervous system, where GPR17+ cells may have a role of damage ‘sensor’ able to activate healing programmes .
After the induction of ischaemia, GPR17 was found in at least two distinct populations. The first is represented by Iso-B4+ or CD45+ myeloid cells, probably recruited from peripheral circulation to the ischaemic tissue (Figs 1C, D and 3A); the other, characterized by Sca-1 and negative for CD31, probably deriving from an intra-cardiac stromal cells store (Fig. 4). This hypothesis is supported by the previous observations that myocardial infarction determines an increase in the amount of Sca-1+/CD31− cells , and recruitment of a multipotent cardiac MSC population which contributes to cardiac fibrosis . Particularly interesting appeared, in this respect, also the expression of CD44 in GPR17+ cells in the infarct tissue (Fig. 3B) and the lack of mature MF marker αSMA at early times after MI, suggesting an immature phenotype. In fact, expression of CD44, in the presence or the absence of CD45, was the principal criterion to discriminate between inflammatory cells homing in the infarcted myocardium from peripheral circulation and stromal progenitors recruited from heart-resident pools . For technical reasons (need of performing indirect GPR17 staining of living cells in conjunction with multicolour membrane marker analysis), it was not possible to proceed with a direct flow-sorting experiment of GPR17+/Sca-1+/CD44+/CD31−/CD45− cells from infarcted hearts to resolve the cardiac MSC phenotype and the MF differentiation potency of the GPR17+ expressing cells in the infarct. On the other hand, the finding that Sca-1+/CD45− cells sorted from infarcted mice (Fig. 5) and that GPR17+/Sca-1+/CD31−/CD45− cells cloned from the normal hearts (Fig. 6) expressed high levels of CD44 and were induced by TGF-β treatment to differentiate into cells with enhanced expression and intracellular polymerization of αSMA protein, strongly suggest the identity of the observed GPR17+/CD44+ cells in the ischaemic myocardium (Fig. 3B) as heart-resident MSCs, endowed with MF differentiation potency. This latter conclusion is also supported by the absence of c-kit, a marker typical of cardiac-resident stem cells in mice and humans [26, 32] and of other stromal cells such as telocytes [28, 30].
Possible function of GPR17 in cardiac-resident MF progenitors
Cys-LTs and purines are well-established initiators of the inflammatory responses. Their role is that of a ‘find-me signal’ attracting the phagocytes to sites of apoptotic cells clearance  and inducing a chemokine ‘milieu’ necessary for sequential chemo-attractant cascades involved in pro- inflammatory phases associated with tissue damage, innate host-defence, and autoimmunity . Prompted by our preliminary investigations showing that blockade of Cys-LTs signalling alleviates the extent of hypoxic stress-related cardiac fibrosis , and other reports suggesting the relevance of Adenosine signalling in the protective function of cardiac stromal cells , we determined the function of GPR17 in the cloned cardiac Sca-1+/CD31− line. While the treatment of the cells with known Cys-LT/purinergic agonists (LTD4 and UDP-gluc) in vitro did not have an effect on proliferation and hypoxia-induced apoptosis (data not shown), both ligands exerted a potent chemotactic effect via GPR17 activation. This was shown by reversion of the migratory effect by co-treatment with known GPR17 pharmacological antagonists such as Montelukast and Cangrelor, as well as by a more specific siRNA knockout approach (Fig. 7). In our experiments, we also investigated whether treatment with the two ligands determined modifications in the TGF-β-induced MF differentiation programme. However, we did not observe remarkable differences (data not shown). Altogether, these findings point to a specific GPR17 role in chemotactic guidance of stromal cells towards the ischaemic sites.
A preliminary assessment of a possible in vivo GPR17 function was provided by treating infarcted mice with systemic and intra-myocardial injection of the GPR17 antagonist Cangrelor, followed by analysis of the principal cell types recruited in the forming scar (Fig. 7). The blockade of purinergic signalling transmission by Cangrelor led to an interesting and unexpected imbalance between the amount of the recruited cells in control versus treated mice. In fact, Cangrelor treatment determined a significant reduction in the number of GPR17+ cells and myeloid cells, characterized by CD45 expression, but not that of CD44+ cells (Fig. 7). As Cangrelor targets the purinergic, but not the Cys-LT GPR17 signal transduction pathway, this suggests the existence of redundant mechanisms, possibly linked to the dual receptor GPR17 nature. Alternatively, an antagonistic function of GPR17 signalling in distinct stromal and myeloid cells may justify the observed imbalance between cardiac-derived versus blood-derived infiltrating cells.
In summary, our results suggest a potential role of the Purinergic and Cys-LT GPR17 receptor in the early response of cardiac stromal cells to ischaemia. This is in agreement with a very recent study showing the involvement of the Adenosine A2B receptor in cardioprotection by Sca-1+ stromal cells . Whether selective modulation of GPR17 signalling in cardiac-resident stromal cells translates into beneficial treatments to reduce the extent of myocardial fibrosis and to limit the functional consequences of heart ischaemia is still a matter of speculation, and it is the subject of our current analyses.