Slow-cycling cells in renal papilla: stem cells awaken?

A fundamental goal of kidney stem cell models in regenerative medicine is to harness the kidney’s prodigious capacity for endogenous repair to develop new therapeutic strategies. Despite progress in understanding kidney injury and repair mechanisms, few regenerative therapies to treat human kidney injury are on the horizon. In part, this reflects the kidney’s complex architecture and cellular heterogeneity, a shortage of validated stem cell markers, slow homeostatic cell turnover, and the lack of robust transplantation assays. These factors combine to slow our ability to decipher cellular hierarchies in adult tissue repair and complicate the search for adult stem and progenitor cells. The renal papilla has attracted interest as a potential niche for kidney stem cell since Oliver et al.1 originally identified this kidney region as a site of slowly cycling cells, defined by retention of the DNA analogue 5-bromo-2-deoxyuridine (BrdU) after a short postnatal pulse followed by a long chase. Because subsequent cell divisions in the absence of label dilute the incorporated BrdU, only cells with the lowest replication profile are detected (label-retaining cells [LRCs]), and this approach has been used in other organs to enrich for stem cells that typically divide slowly. Importantly, the specificity and sensitivity of BrdU label retention is undefined for solid tissues and, in fact, is remarkably low for hematopoietic stem cells, the best characterized adult stem cell pool. For example, 6% of hematopoietic stem cells retain BrdU and 0.5% of BrdU-retaining hematopoietic cells are hematopoietic stem cells.2 Other factors recommend the papilla as a potential kidney stem cell niche beyond the residence of LRCs. Papilla has low oxygen tension and hypoxia may protect stem cells against DNA damage. There is preferential expression of genes known to regulate other stem cells in the renal papilla, including Wnt and Hedgehog ligands.3,4 In their previous study, the depletion of papillary LRCs after ischemia reperfusion injury and the ability of papillary cell populations enriched in LRCs to form “nephro-spheres” in culture provide additional, indirect evidence for the existence of stem or progenitor cells in renal papilla.1 In this issue of JASN, Oliver et al.5 present a welcome follow-up to their original work describing LRCs in the papilla. They use a powerful genetic model with doxycycline-inducible expression of a histone 2B-GFP fusion protein (H2BGFP) to identify and further characterize slow-cycling cells in renal papilla. Because the H2B-GFP protein is stable in vivo, long chase times are possible and this approach has distinct advantages over BrdU label retention. BrdU labeling requires that cells be in S phase at the time of the pulse, but the slowcycling nature of stem cells prevents a majority from labeling during that pulse. The H2B-GFP approach labels all cells, not just the fraction in S phase, and the superiority of this approach is proved in hematopoietic stem cells, where approximately 20% of hematopoietic stem cells retain the H2B-GFP label after a 24-wk chase compared with approximately 2% of hematopoietic stem cells retain BrdU after a similar chase.6 In addition, the H2B-GFP label is bright and does not require denaturing conditions for detection (like BrdU), allowing colabeling and FACS sorting and potentially recovery of viable cells for functional testing. Oliver et al.5 observe that, when pulsed with doxycycline during embryogenesis, newborn mice bigenic for both the tetracycline transactivator and a tet-O-H2B-GFP transgene have robust and ubiquitous kidney expression of H2B-GFP, but in adult mice, the H2B-GFP label is found only in the renal papilla, confirming the group’s previous findings. Costaining experiments allowed the authors to determine that the ratio of interstitial LRCs to tubular LRCs is almost equal in the upper papilla, but many more interstitial LRCs were found in the tip of the papilla, compared with tubular LRCs. Interestingly, interstitial LRCs express the neural stem cell marker nestin, an intermediate filament that is also linked to interstitial cells recruited after ischemic injury,7 a finding reminiscent of the ability of papillary cell clones to produce nestin-positive progeny.1 The LRCs also express the stromalderived factor 1 (SDF-1) receptor, CXCR7, which also expresses on renal progenitors isolated from Bowman’s capsule,8 and SDF-1 is a well-characterized chemokine for mesenchymal stromal cells. To address the functional properties of papillary LRCs, Oliver et al.5 examined whether these cells proliferate during homeostasis and after injury. During adult homeostasis, the authors found a small number of cells in the upper papilla, adjacent to the urinary space, that are positive for the proliferation marker Ki-67, and some of these Ki-67–positive cells also retain the H2B-GFP label. Moreover, just 12 h after ischemic injury and before outer medullary proliferation occurs, Published online ahead of print. Publication date available at www.jasn.org.