Cutting to the chase: taking the pulse of label-retaining cells in kidney

one of the oldest assays for identifying stem cells, DNA analog pulse-chase studies date back more than 60 years (2). The concept derives from the observation that stem cells divide only rarely to preserve their proliferative potential and reduce DNA errors that occur during chromosome duplication (1). The assay consists of two parts: a pulse of DNA analog [bromodeoxyuridine (BrdU) or tritiated nucleotide, for example] followed by a chase period. During the pulse, all proliferating cells are labeled. During the chase, the analog is absent, so fast cycling cells will dilute the label by half with each cell division until the label is undetectable (in practice, 3–4 divisions for BrdU). Slow-cycling cells proliferate only rarely during the chase, retaining their label and identifiable as label-retaining cells (LRC) at the end of the chase.

These label-retaining assays work very well to identify the cell cycle properties of a mixed population of cells but care is required for their proper interpretation. Most importantly, label retention itself does not define stemness; it simply identifies a cell that divided during the pulse and did not divide during the chase. Second, very slowly cycling cells may not be labeled at all during the pulse, particularly if the pulse is short. Third, the interpretation of a label retention study is different in a tissue with normally low rates of proliferation (such as adult kidney) than in one with normally high rates (such as skin or intestine). In a slow-cycling tissue, unless both the pulse and the chase are sufficiently long, there will be few cells labeled during the pulse and little subsequent dilution of label during the chase because so few cells are proliferating.

To define stem cells in kidney, a number of groups employed pulse-chase approaches to identify LRC. A major problem in comparing the results of these studies has been the different pulse-chase protocols employed that have investigated different age mice and rats with variable pulse times and chase durations. With this backdrop, the paper by the Curtis lab in an issue of the American Journal of Physiology-Renal Physiology is a very welcome addition to the field (6a). The authors utilized three different pulse times–early neonatal, late neonatal, and adult–and compared the distribution of LRC between kidney regions and compartments. A particular strength is their use of sequential pulses with different deoxyuridine analogs [5-chloro-2-deoxyuridine (CldU) and 5-iodo-2-deoxyuridine (IdU)] that can be distinguished by different monoclonal antibodies, allowing two different pulse protocols to be compared in the same animal.

The authors first observation was of substantial heterogeneity in LRC number according to both gross kidney region and according to the time of the pulse. For example, papilla contained the most LRC with either an early or late neonatal pulse, but an adult pulse identified the most LRC in the cortex. Importantly, there was very little overlap between CldU and IdU, indicating that the various LRC populations are distinct from one another. These differences in postnatal kidney LRC abundance according to age at pulse are reminiscent of regional differences in both cell cycle length and proliferative rates during nephrogenesis (8). They also correspond well to previous reports of either papillary or cortical LRC, depending on whether the pulse was neonatal or adult (reviewed in Ref. 6a).

Careful quantitation of differences in LRC distribution between epithelial and interstitial compartments revealed intriguing differences in interstitial cell LRC abundance. Rangarajan et al. (6a) observed the highest percentage of interstitial LRC in papilla. This finding brings to mind another very recent study that identified a unique interstitial papillary cell population defined by low expression of HoxB7 and strong expression of Wnt4 during late gestation. Li and colleagues (5) isolated these cells and showed that they were clonogenic, long-term self-renewing and possessed mesenchymal stem cell-like properties ex vivo. When these cells were injected into a neonatal kidney, they preferentially incorporated into the collecting duct where they acquired a collecting duct phenotype. It is very reasonable to speculate that the interstitial papillary LRC identified by Rangarajan et al. (6a) may be the same progenitors identified by Li et al. (5). Supporting this interpretation, label retention has recently been used to identify a separate mesenchymal stem cell-like population in adult thymus (6).

This study points to two important remaining questions. First, do all of these LRC represent slow-cycling cells? Particularly with the early and late neonatal pulses, it seems likely that at least some of these LRC are truly slow cycling compared with their neighbors. On the other hand, the basal proliferative rate in adult kidney is so low that it is likely that LRC from this protocol may not reflect slow-cycling cells, but rather that very small fraction of cells that happened to be proliferating at the time of the pulse–as discussed by Rangarajan et al. This leads to the second question: which of these LRC represent stem cells (if any)? Ultimately, genetic lineage analysis will be required to demonstrate the long-term kidney-regenerating capacity of these putative stem cells (3). Since the evidence strongly points against a prespecified stem cell population in the adult proximal tubule (4), it is highly unlikely that proximal tubule LRC reflect a stem cell population. However, the mechanisms of homeostasis and repair in other nephron segments are poorly defined. One recent study suggests that each nephron segment is governed by segment-specific repair mechanisms (7), meaning that a distal tubule cell does not repair a connecting segment for example, but whether the LRC identified in each segment might represent a progenitor population requires detailed analysis, and genetic lineage tracing.

By reconciling results from previous disparate pulse-chase protocols, this work now allows the field to focus on a particular LRC population (vascular, interstitial, epithelial, nephron segment-specific) and cut to the chase of whether it contains a stem cell.

GRANTS

This work was supported by grants from the National Institutes of Health/National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK; DK088923 and DK103050), the NIDDK Diabetic Complications Consortium (DK076169), and an Established Investigator Award from the American Heart Association.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: B.D.H. drafted manuscript; B.D.H. edited and revised manuscript; B.D.H. approved final version of manuscript.