Renal Intercalated Cell Differentiation and Acid–Base Regulation

The collecting duct system is made of a complex epithelium consisting of several cell types with distinct functions. Collecting ducts fine-tune urinary excretion and are critical for the kidney’s ability to adapt to ever-changing environmental conditions, dietary intake, metabolism, physical activity, and health or disease states. The importance of the collecting duct system for clinical conditions is self-evident from the various disorders caused by either inherited or acquired disorders primarily originating from the collecting duct. These disorders include distal renal tubular acidosis (RTA), diabetes insipidus, and forms of salt-sensitive hypertension or salt-losing nephropathies. At least three molecularly and functionally distinct cell types contribute to functions of the collecting duct. Principal cells are the majority cell type and are classically associated with salt, potassium, and water balance. Type A intercalated cells are critical for urinary acidification, while type B intercalated cells excrete bicarbonate. On a molecular level, principal cells are characterized by the presence of the aquaporin 2 (AQP2) water channel and the subunits of the epithelial sodium channel. By contrast, all subtypes of intercalated cells express carbonic anhydrase II (CAII) and various subunits of the H⁺-ATPase. Furthermore, type A intercalated cells express the kidney isoform of the basolateral anion exchanger AE1, while type B intercalated cells have the luminally localized anion exchanger pendrin.^1,2

This simplifying classification of cell types, however, has been challenged by many findings over the past 1 or 2 decades showing that intercalated cells also contribute to salt and potassium homeostasis and that multiple pathways exist through which the different cell types communicate with each other. Furthermore, the relative abundance of these three major cell types along the different segments of the collecting duct system changes under various conditions, such as use of diuretics, changes in acid–base status, or with drugs such as lithium, revealing a large plasticity of the collecting duct system.² The mechanisms underlying this plasticity are not entirely clarified: Two major hypotheses exist that may not exclude each other. Plasticity may be achieved either by the ability of collecting duct cells to change their identity, a concept called interconversion, or may be due to the selective proliferation and removal of cells. In both scenarios, pathways and molecules must exist that sense changes in acid–base or electrolyte status and that orchestrate adaption of collecting duct cells. Several key players have been identified that include GPR4, GDF15, hensin, galectin-3, β1-integrin, the hypoxia-inducible factor 1 alpha, the stromal-derived factor 1 (also known as CXCL12), and the C-X-C chemokine receptor type 4.

The question of how plasticity of the collecting duct is achieved in an adult organism overlaps with the interest to understand how this complex epithelium is formed during kidney organogenesis as similar mechanisms may operate to drive the differentiation of distinct cell types and to adapt the mature collecting duct. A network of transcription factors has been identified that orchestrates the differentiation of cells along the collecting system. The current model predicts that AQP2⁺ precursor cells differentiate into principal cells and intercalated cells (Figure 1). This step depends on NOTCH1/2 signaling and its ligand JAG1. In the presence of NOTCH1/2 signaling activated by JAG1 and involving a γ-secretase and ADAM10 proteases, JAG1 formation is suppressed, and a cascade of signals is activated including E26 transformation-specific–related transcription factor 5 (ELF5), the histone-lysine N-methyltransferase DOTl1, and transcription factor hairy and enhancer of split-1 that ultimately leads to the suppression of the transcription factors FOXI1 and stimulates the transcription of principal cell–specific transcripts, such as epithelial sodium channel or AQP2. Thus, NOTCH1/2 signaling drives differentiation toward principal cells. By contrast, in the absence of NOTCH1/2 signals, cell differentiate into intercalated cells, which requires secretion of JAG stimulated by the E3 ubiquitin ligase MIB1 and the transcription factors CP2-like protein 1 and FOXI1. FOXI1 also induces transcription of intercalated cell–specific genes including kidney-isoform of the basolateral anion exchanger AE1, pendrin, CAII, and various subunits of the H⁺-ATPase.^2–4

Figure 1.

Model of the regulatory network that controls differentiation of cells in the collecting system. See text for details. Modified from ref. 2 with permission. AQP2, aquaporin 2; CAII, carbonic anhydrase II; HES1, hairy and enhancer of split-1.

In this issue of JASN, Wu and colleagues add another interesting player to this network, the transcription factor FOXP1.⁵ Single-cell transcriptomics had identified Foxp1 to be highly enriched in murine intercalated cells and podocytes. Therefore, the authors deleted Foxp1 only in the distal nephron and found that these mice had a highly reduced number of all types of intercalated cells and distal renal tubular acidosis. Surprisingly, older Foxp1-depleted mice regained expression of the intercalated cell marker CAII in AQP2⁺ cells, which is highly reminiscent of the undifferentiated cellular phenotype of mice lacking Foxi1.⁴ Chromatin immunoprecipitation DNA sequencing identified potential target genes of Foxp1, among them two other transcription factors Hmx2 and Dmrt2. Of note, single-cell transcriptome data from mouse kidney found Hmx2 to be selectively expressed in type B intercalated cells and Dmrt2 only in type A intercalated cells, suggesting that these transcription factors may be involved in further determining the differentiation of intercalated cell subtypes. To test the relevance of these transcription factors, the authors took the example of Dmrt2 and deleted it from the distal nephron, leading to a selective loss of type A intercalated cells and distal RTA. Next, the authors wanted to clarify the order of signaling between Notch and Foxp1 using mice devoid of Notch2 signaling in the distal nephron, which led to the expected increase in intercalated cells and reduction in principal cells. Deleting Foxp1 in Notch-deficient mice showed the same loss of intercalated cells as in mice lacking Foxp1 alone, indicating that Notch acts upstream of Foxp1. Conversely, when authors generated mice overexpressing Foxp1, Dmrt2 and Hmx2 were upregulated but the abundance of intercalated and principal cells were not altered, suggesting that Foxp1 is needed for intercalated cell differentiation but that other factors co-control this step. Because Foxi1 shows a similar effect in intercalated cell differentiation as Foxp1, it remains to be clarified what the exact relationship is between these two transcription factors. However, several observations suggest that both transcription factors may act in parallel: Foxp1 seems not to bind to Foxi1 (and vice versa), Foxi1 is still expressed in Foxp1 knock-out kidneys, but Foxp1 seems to be necessary for full differentiation of Foxi1-positive cells, and Notch seems to suppress both Foxi1 and Foxp1. Further experiments will be needed to clearly establish the relationship between these two transcription factors. However, there are also differences: Foxi1 has several target genes that are present either in type A or type B intercalated cells, such as pendrin, while Foxp1 apparently drives the two transcription factors Dmrt2 and Hmx2 (Figure 1). The latter may be necessary to further differentiate intercalated cells to become type B intercalated cells. This finding, together with the suggestion that Dmrt2 may do the same for type A intercalated cells, will need further experimental proof but would elucidate how the different types of intercalated cells are eventually differentiated. Similarly, it would be interesting to test if the same factors are also involved in the plasticity of the adult collecting duct.

Understanding the complex networks forming functional collecting ducts is clinically relevant for disorders affecting kidney development and for forms of inherited and acquired distal RTA. Biallelic loss-of-function variants in FOXI1 cause distal RTA with sensorineural deafness in humans.³ In addition, patients with monoallelic loss-of-function variants in FOXP1 have been identified in patients with congenital anomalies of the kidney and urinary tract.⁶ Similarly, variants in the transcriptional corepressor ZMYM2 cause congenital anomalies of the kidney and urinary tract and ZMYM2 binds to FOXP1, suggesting that they act in a common pathway.⁷ Acquired forms of nephrogenic diabetes insipidus or distal RTA can be caused by lithium or are associated with autoimmune disorders such as Sjögrens syndrome, respectively. In rats treated with lithium, the relative abundance of principal cells is decreased while more type A intercalated cells are found.⁸ In patients with distal RTA and types of autoimmune disorders including Sjögrens syndrome, the absence of markers of intercalated cells was noted, which might be due either to selective loss of intercalated cells or a differentiation defect induced by the autoimmune disorder.^9,10

The identification of Foxp1, Dmrt2, and Hmx2 as novel players in collecting duct differentiation further reveals the complexity of the network controlling the unique architecture and function of the collecting system. It may also contribute to a better molecular understanding of diseases originating from this important nephron segment.

Supplementary Material

jasn-35-519-s001.pdf ^{(1.3MB, pdf)}

Disclosures

Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/JSN/E605.

Funding

C.A. Wagner: Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung.

Author Contributions

Conceptualization: Carsten A. Wagner.

Funding acquisition: Carsten A. Wagner.

Writing – original draft: Carsten A. Wagner.

Writing – review & editing: Carsten A. Wagner.