Background

The metanephric (permanent) kidneys of mammals develop from simple rudiments located at the caudal end of the intermediate mesoderm. In mice, at about embryonic day (‘E’) 10 these rudiments form and consist of two morphologically distinguishable components; an epithelial ureteric bud that arises as a diverticulum of the Wolffian (nephric) duct, and a metanephrogenic mesenchyme that forms next to the duct. As development progresses, the ureteric bud enters the metanephrogenic mesenchyme and undergoes many successive rounds of growth and branching to make a ‘tree’; this later remodels to produce a mature collecting duct system in which tubules radiate from a central cavity, the renal pelvis (Lindstrom et al., 2015). The renal pelvis drains to the ureter, which forms from the original stalk of the ureteric bud. As the ureteric bud develops, it induces cells from the metanephrogenic mesenchyme to condense around each of its tips to form a ‘cap mesenchyme’ (Schreiner, 1902; Reinhoff, 1922). The cap mesenchymes are stem cell populations that divide as the tips divide, so that each tip formed by bifurcation of an existing branch inherits its own cap (reviewed by Hendry et al., 2011). Cells at the more distal ends of the caps differentiate to form excretory nephrons, which connect to the ureteric bud branch from whose cap they formed (Georgas et al., 2009). Blood vessels invade from the base of the metanephros and follow the ureteric bud, making a network around (but never entering) the cap mesenchymes (Munro et al., 2017): later, these vessels will serve glomeruli and other parts of the kidney.

In vitro culture of metanephric kidney rudiments has a long history. Indeed, these were among the first embryonic organs to show continued development outside the body (Carrel and Burrows, 1910). The continued development of kidney rudiments outside of the body was a scientific, as well as a technical, advance: the autonomous development of isolated organs demonstrated that the information required to build them was ‘local’ and did not depend on the rest of the embryo. This observation added considerable support to the idea that architecture of the body is hierarchical, with modules (organs) that largely look after themselves and interact with the rest of the body only at specific functional interfaces. The earliest culture methods used rather complex media and culture supports, such as Grobstein’s use of clotted avian plasma and chick embryo extract (Grobstein, 1953). These systems were necessary because simply placing a kidney rudiment in a glass dish or flask resulted in its breakdown because cells adhered to the substrate and spread out to form a monolayer. Immersion in medium in non-adhesive dishes prevents cell dispersal but does not result in proper development (see ‘Data analysis’ section). Both imaging and reproducibility were greatly improved by Saxén’s adoption of a culture method developed by Trowell for culture of rat lymph nodes (Trowell, 1954). Saxén placed embryonic kidney rudiments, isolated from mice at E11, on filters that were supported by a stainless steel grid at the interface between gas and medium (Saxén et al., 1962).

The Trowell method has been a mainstay of research in kidney development for many decades. It allows for significant ureteric bud growth and branching, formation of nephrons, differentiation of their separate proximal and distal domains and connection of nephrons to ureteric bud branches. Rudiments grow flat enough to facilitate confocal microscopy of fixed specimens without the need for ‘clearing’ techniques, and even conventional epifluorescence microscopy is adequate for many purposes (e.g., Davies et al., 2014). Another advantage is that diffusion paths have proved to be short enough and open enough to enable mechanisms of development to be investigated in this method with drugs (e.g., Fisher et al., 2001), growth factors (e.g., Piscione et al., 1997), function-blocking antibodies (e.g., Falk et al., 1996) and, to a limited extent, siRNAs (e.g., Davies et al., 2004). The Trowell system, together with variants that place rudiments at the air-medium interface using use Transwell filters instead of stainless steel supports, remains very common in studies of kidney development.

Useful as it is, the Trowell culture system has a few problems. One is that the filter itself interferes with bright-field/phase contrast imaging because the filter pores appear, out of focus, in the image (though some Transwell systems do achieve good optical clarity). Another is that the tissue is too thick for reliable high-resolution, unattended time-lapse photography, because tubules leave the focal plane. A third problem is that some aspects of renal development, such as the formation of distinct cortex and medulla, and extension of nephrons’ loops of Henle from the cortex to the medulla, occur poorly if at all. In an attempt to address these issues, we developed an alternative culture system that uses the surface tension of very shallow medium to hold a kidney rudiment on to a clear glass substrate. To our surprise, the system not only solved the imaging problem, but it also allowed the organ rudiments to develop clear cortico-medullary zonation and nephron maturation proceeded as far as the production of clear and elongated loops of Henle (Sebinger et al., 2010) (Figure 1). The enhanced development is seen in cultures made from natural kidneys isolated from mouse embryos, and also in organoids engineered from suspensions of stem cells (Chang and Davies, 2012).


Figure 1. Kidneys cultured in the Sebinger system. A. Bright field images for E11.5 kidney grown in culture for 0, 3 and 7 days. Scale bars = 0.5 mm. B-D. E11.5 kidneys cultured for 7 days in the Sebinger system and stained for different renal markers to show maturation. B. Stained for the ureteric bud marker CALB (shown in green) and the basement membrane marker Laminin (shown in red). The red channel shows the presence of loops of Henle dipping into the medulla; C. Stained for CALB (green) and the proximal tubular marker LTL (red); D. Stained for ECAD (ureteric bud and distal tubular marker; shown in green) and WT1 (podocyte and cap mesenchyme marker; shown in red). Scale bars = 100 μm.