Unraveling the Complexity of the Renal Mononuclear Phagocyte System by Genetic Cell Lineage Tracing

• immunology
• embryology
• macrophages

The renal mononuclear phagocyte (MoPh) system is composed of a complex network of macrophages (MØs) and dendritic cells (DCs). A study in this issue of JASN used genetic cell lineage tracing to shed new light on the ontogeny of these cells and reveals their great functional and phenotypic plasticity in kidney homeostasis and inflammation.¹

The renal MoPh system is critical for kidney homeostasis, defense against pathogens, and tissue repair after injury, but it may also contribute to the progression of inflammatory kidney disease. Renal MoPhs are traditionally divided into MØs and DCs, which possess distinct hallmark functions. MØs are thought to be innate immune effector cells, and DCs are thought to prime naïve T cells.^2,3 However, there is considerable overlap in their functionality and phenotype. Modern technologies, especially single-cell transcriptomics, have been used to define multiple subpopulations of DCs and MØs. Traditional classification approaches are on the basis of morphology and the expression of surface markers, such as F4/80, CD11b, and CD64 for MØs and MHCII and CD11c or XCR1 for DCs.³ The chemokine receptor XCR1 is specific for DCs, albeit only for one subset that is rare in the kidney. Recently, transcription factors are increasingly used by immunologists for cell type classification, such as MAFB for MØs or ZBTB46 for DCs.⁴ However, none of these molecules are expressed specifically by one cell type, their expression levels may vary, and the organ microenvironment affects marker expression as well.³

A further classification approach is by genetic cell lineage tracing (also known as fate mapping), which can reveal whether a cell of interest is derived from a defined precursor cell. This method uses transgenic mice expressing a fluorochrome gene after a transcriptional stop sequence embedded by flox sequences. When such mice are crossed to mice expressing CRE recombinase under the control of the promoter of a given gene, any cell will become irreversibly fluorescent after it expresses that gene. In combination with conditional gene enhancers, the fluorescence signal can be activated on demand. This technique has been used, for example, to demonstrate that dying podocytes can be replaced by cells derived from parietal glomerular epithelial cells.⁵

In immunology, genetic lineage tracing is often used to identify the offspring of hematopoietic precursors. During embryonic development, MØ progenitors first spawn from the yolk sac and seed tissues. Subsequently, hematopoietic stem cells initiate fetal hematopoiesis in the liver, which shifts to the bone marrow where most MoPhs are generated after birth. Lineage tracing for chemokine receptor CX₃CR1 in embryos has shown that yolk sac–derived MoPhs are maintained in some tissues by local proliferation over a long time.^6,7 Lineage tracing for the transcription factor Ms4a3, which reveals monocyte-derived MoPhs, indicated that embryonic MØs are gradually replaced by bone marrow–derived MØs, with great variability between organs. In the case of the kidney, about 40% of the MØs were replaced in 12-week-old mice.⁸

The offspring of the common dendritic precursor (CDP) can be identified by lineage tracing for Clec9A (also known as DNGR-1 or CD370), which is not expressed by putative monocyte/MØ precursors. Previously, Schraml et al.⁹ reported that CDP-derived cells display the typical DC phenotype in most tissues, except the kidney, where some CDP-derived cells expressed CD64. These findings challenged the specificity of this marker for MØs, at least in the kidney.

In this issue of JASN, Salei et al.¹ elaborated their Clec9A lineage-tracing approach and identified four CDP-derived subsets in the kidney that differed in phenotype and transcriptome. These populations were XCR1⁺ type 1 DCs, CD64⁻ CD11b⁺ type 2 DCs, and two subsets expressing CD64 and high levels of either CD11b or F4/80 (Figure 1). The F4/80-expressing cells were phenotypically almost identical to embryonic-derived MØs, which are a non-CDP–derived population and therefore, do not exhibit Clec9A expression history. Such Clec9A lineage-negative F4/80^hi MoPhs originated in the yolk sac and were first detectable on day 16.5 after conception, and they lacked MHCII expression. These dominated the kidney at birth but declined in numbers as the mice aged. Eventually, they were replaced by Clec9A lineage-positive (CDP-derived) F4/80^hi MHCII⁺ c-Myb⁺ CD11b^hi MoPhs, which were first detectable on postnatal day 2. These cells steadily increased in numbers with postnatal age. At 10-12 weeks after birth, they constituted the majority of MHCII⁺ MoPhs in the mouse kidney.¹

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Common DC precursors give rise to four different types of renal mononuclear phagocytes according to Salei et al.¹

In a model of renal inflammation induced by cisplatin toxicity, F4/80^hi MoPhs were found to express very little MHCII. This raised the possibility that embryonical MoPhs had reappeared. With the conventional approach of surface marker discrimination, this question could not be answered. Lineage tracing indicated that those cells were not of embryonic origin and in consequence, must have been F4/80^hi MHCII⁺ MoPhs that then must have downregulated MHCII expression in response to inflammation.

These results challenge a recent study published in JASN, which proposed that MoPhs expressing F4/80 and CD64 are MØs, consistent with the immunological nomenclature used in lymphatic tissues.⁴ Their classification as DCs had been mainly on the basis of the expression of the transcription factor ZBTB46.⁴ However, ZBTB46 is not specific for DCs, being also expressed for example in endothelial cells.¹⁰

In summary, this study by Salei et al.¹ demonstrated that some CD64⁺ and F4/80⁺ renal MoPhs are CDP derived. This conclusion of course depends on whether Clec9A is a specific and sensitive CDP marker. Within that caveat, these findings are important for understanding the development of the renal MoPh system and for the current debate regarding the classification of DCs and MØs. Clinical nephrologists may also be interested in whether cell types defined by lineage tracing possess distinct functions: for example, in diseases. Such information may shed new light on kidney disease pathogenesis and may ultimately help in developing therapies that target pathogenic components of the renal MoPh system.

• Copyright © 2020 by the American Society of Nephrology