From the ureteric bud to the penome

Protein kinase A regulates GDNF/RET-dependent but not GDNF/Ret-independent ureteric bud outgrowth from the Wolffian duct

Embryonic kidney development begins with the outgrowth of the ureteric bud (UB) from the Wolffian duct (WD) into the adjacent metanephric mesenchyme (MM). Both a GDNF-dependent and GDNF-independent (Maeshima et al., 2007) pathway have been identified. In vivo and in vitro, the GDNF-dependent pathway is inhibited by BMPs, one of the factors invoked to explain the limitation of UB formation in the unbudded regions of the WD surrounding the UB. However, the exact mechanism remains unknown. Here a previously described in vitro system that models UB budding from the WD was utilized to study this process. Because Protein kinase A (PKA) activation has been shown to prevent migration, morphogenesis and tubulogenesis of epithelial cells (Santos et al., 1993), its activity in budded and non-budded portions of the GDNF-induced WD was analyzed. The level of PKA activity was 15-fold higher in the unbudded portions of the WD compared to budded portions, suggesting that PKA activity plays a key role in controlling the site of UB emergence. Using well-characterized PKA agonists and antagonists, we demonstrated that at various levels of the PKA-signaling hierarchy, PKA regulates UB outgrowth from the WD by suppressing budding events. This process appeared to be PKA-2 isoform specific, and mediated by changes in the duct rather than the surrounding mesenchyme. In addition, it was not due to changes in either the sorting of junctional proteins, cell death, or cell proliferation. Furthermore, the suppressive effect of cAMP on budding did not appear to be mediated by spread to adjacent cells via gap junctions. Conversely, antagonism of PKA activity stimulated UB outgrowth from the WD and resulted in both an increase in the number of buds per unit length of WD as well as a larger surface area per bud. Using microarrays, analysis of gene expression in GDNF-treated WDs in which the PKA pathway had been activated revealed a nearly 14-fold decrease in Ret, a receptor for GDNF. A smaller decrease in GFRα1. a co-receptor for GDNF, was also observed. Using Ret-null WDs, we were able to demonstrate that PKA regulated GDNF-dependent budding but not GDNF-independent pathway for WD budding. We also found that BMP2 was higher in unbudded regions of the GDNF-stimulated WD. Treatment of isolated WDs with BMP2 suppressed budding and resulted in a 3-fold increase in PKA activity. The data suggests that the suppression of budding by BMPs and possibly other factors in non-budded zones of the WD may be regulated in part by increased PKA activity, probably partially through downregulation of Ret/GFRα1 coreceptor expression.

Neuropeptide Y functions as a facilitator of GDNF-induced budding of the Wolffian duct

Ureteric bud (UB) emergence from the Wolffian duct (WD), the initiating step in metanephric kidney morphogenesis, is dependent on GDNF; however, GDNF by itself is generally insufficient to induce robust budding of the isolated WD in culture. Thus, additional factors, presumably peptides or polypeptide growth factors, might be involved. Microarray data from in vivo budding and non-budding conditions were analyzed using non-negative matrix factorization followed by gene ontology filtering and network analysis to identify sets of genes that are highly regulated during budding. These included the GDNF co-receptors GFRalpha1 and RET, as well as neuropeptide Y (NPY). By using ANOVA with pattern matching, NPY was also found to correlate most significantly to the budded condition with a high degree of connectedness to genes with developmental roles. Exogenous NPY [as well as its homolog, peptide YY (PYY)] augmented GDNF-dependent budding in the isolated WD culture; conversely, inhibition of NPY signaling or perturbation of NPY expression inhibited budding, confirming that NPY facilitates this process. NPY was also found to reverse the decreased budding, the downregulation of RET expression, the mislocalization of GFRalpha1, and the inhibition of AKT phosphorylation that resulted from the addition of BMP4 to the isolated WD cultures, suggesting that NPY acts through the budding pathway and is reciprocally regulated by GDNF and BMP4. Thus, the outgrowth of the UB from the WD might result from a combination of the upregulation of the GDNF receptors together with genes that support GDNF signaling in a feed-forward loop and/or counteraction of the inhibitory pathway regulated by BMP4.

How does the ureteric bud branch?

Many genes that modulate kidney development have been identified; however, the molecular interactions that direct arborization of the ureteric bud (UB) remain incompletely understood. This article discusses how "systems" approaches may shed light on the structure of the gene network during UB branching morphogenesis and the mechanisms involved in the formation of a branched collecting system from a straight epithelial tube in the context of a stage model. In vitro and genetic studies suggest that the stages seem to be governed by a conserved network of genes that establish a "tip-stalk generator"; these genes sustain iterative UB branching tubulogenesis through minimal alterations in the network architecture as a budding system shifts to one that autocatalytically branches through budding. The differential expression of stage-specific positive and inhibitory factors in the mesenchyme, likely presented in the context of heparan sulfate proteoglycans, and effector molecules in the epithelium seems to regulate advancement between stages; similar principles may apply to other branching epithelia such as the lung, salivary gland, pancreas, mammary gland, and prostate. Active mesenchymal interactions with the UB seem to govern vectorial arborization and tapering of the collecting system and its terminal differentiation. Cessation of branching correlates with induction of mesenchyme as well as local extracellular matrix changes. Perturbations of these mechanisms and/or single-nucleotide polymorphisms in genes regulating UB branching may predispose to a variety of renal diseases (e.g., hypertension and chronic kidney disease) by altering nephron number. Decentralization of the gene-protein interaction network may explain the relative paucity of branching phenotypes in mutant mice and in human disease.

The developmental nephrome: systems biology in the developing kidney

PURPOSE OF REVIEW

A set of important genes and signaling pathways involved in kidney development is emerging from analyses of mutant mice, in-vitro models, and global gene expression patterns. Conversion of data into dynamic models or networks through the synthesis of information at multiple levels is crucial for a better understanding of kidney development.

RECENT FINDINGS

Genetic and in-vitro evidence is beginning to provide a limited sense of the network topology in stages of kidney development. Intriguing data from other fields suggest how, with the aid of large-scale gene expression studies, these stages might be represented as dynamic attractor states. It is also suggested how branching morphogenesis of the epithelial ureteric bud may be sustained by an autocatalytic set of proteins whose interactions lead to repeated rounds of tip and stalk generation. Accumulating data in lower organisms suggest network topologies may be quite flexible, and the implications of these results for varieties of tubulogenesis and renal regeneration after acute injury are discussed.

SUMMARY

Currently it may be feasible to build tentative dynamic multistage models of nephrogenesis that facilitate experimental thinking. As data accumulate, it may become possible to test their predictive value.

Implications of gene networks for understanding resilience and vulnerability in the kidney branching program

Branching morphogenesis in the kidney is tightly regulated. Whereas disruption of certain pathways produces catastrophic effects, numerous instances exist in which mutation of ostensibly key molecules has minimal apparent phenotypic consequence. We suggest how the network structure of gene interactions in the branching program might explain these findings as well as apparant discrepancies between in vivo and in vitro studies. Emerging genetic, cell-biological, and microarray data should help test and/or clarify these ideas.

Spatiotemporal regulation of morphogenetic molecules during in vitro branching of the isolated ureteric bud: toward a model of branching through budding in the developing kidney

In search of guiding principles involved in the branching of epithelial tubes in the developing kidney, we analyzed branching of the ureteric bud (UB) in whole kidney culture as well as in isolated UB culture independent of mesenchyme but in the presence of mesenchymally derived soluble factors. Microinjection of the UB lumen (both in the isolated UB and in the whole kidney) with fluorescently labeled dextran sulfate demonstrated that branching occurred via smooth tubular epithelial outpouches with a lumen continuous with that of the original structure. Epithelial cells within these outpouches cells were wedge-shaped with actin, myosin-2 and ezrin localized to the luminal side, raising the possibility of a "purse-string" mechanism. Electron microscopy and decoration of heparan sulfates with biotinylated FGF2 revealed that the basolateral surface of the cells remained intact, without the type of cytoplasmic extensions (invadopodia) that are seen in three-dimensional MDCK, mIMCD, and UB cell culture models of branching tubulogenesis. Several growth factor receptors (i.e., FGFR1, FGFR2, c-Ret) and metalloproteases (i.e., MT1-MMP) were localized toward branching UB tips. A large survey of markers revealed the ER chaperone BiP to be highly expressed at UB tips, which, by electron microscopy, are enriched in rough endoplasmic reticulum and Golgi, supporting high activity in the synthesis of transmembrane and secretory proteins at UB tips. After early diffuse proliferation, proliferating and mitotic cells were mostly found within the branching ampullae, whereas apoptotic cells were mostly found in stalks. Gene array experiments, together with protein expression analysis by immunoblotting, revealed a differential spatiotemporal distribution of several proteins associated with epithelial maturation and polarization, including intercellular junctional proteins (e.g., ZO-1, claudin-3, E-cadherin) and the subapical cytoskeletal/microvillar protein ezrin. In addition, Ksp-cadherin was found at UB ampullary cells next to developing outpouches, suggesting a role in epithelial-mesenchymal interactions. These data from the isolated UB culture system support a model where UB branching occurs through outpouching possibly mediated by wedge-shaped cells created through an apical cytoskeletal purse-string mechanism. Additional potential mechanisms include (1) differential localization of growth factor receptors and metalloproteases at tips relative to stalks; (2) creation of a secretory epithelium, in part manifested by increased expression of the ER chaperone BiP, at tips relative to stalks; (3) after initial diffuse proliferation, coexistence of a balance of proliferation vs. apoptosis favoring tip growth with a very different balance in elongating stalks; and (4) differential maturation of the tight and adherens junctions as the structures develop. Because, without mesenchyme, both lateral and bifid branching occurs (including the ureter), the mesenchyme probably restricts lateral branching and provides guidance cues in vivo for directional branching and elongation as well as functioning to modulate tubular caliber and induce differentiation. Selective cadherin, claudin, and microvillar protein expression as the UB matures likely enables the formation of a tight, polarized differentiated epithelium. Although, in vivo, metanephric mesenchyme development occurs simultaneously with UB branching, these studies shed light on how (mesenchymally derived) soluble factors alone regulate spatial and temporal expression of morphogenetic molecules and processes (proliferation, apoptosis, etc.) postulated to be essential to the UB branching program as it forms an arborized structure with a continuous lumen.

Regulation of ureteric bud branching morphogenesis by sulfated proteoglycans in the developing kidney

Glycosaminoglycans in the form of heparan sulfate proteoglycans (HSPG) and chondroitin sulfate proteoglycans (CSPG) are required for normal kidney organogenesis. The specific roles of HSPGs and CSPGs on ureteric bud (UB) branching morphogenesis are unclear, and past reports have obtained differing results. Here we employ in vitro systems, including isolated UB culture, to clarify the roles of HSPGs and CSPGs on this process. Microarray analysis revealed that many proteoglycan core proteins change during kidney development (syndecan-1,2,4, glypican-1,2,3, versican, decorin, biglycan). Moreover, syndecan-1, syndecan-4, glypican-3, and versican are differentially expressed during isolated UB culture, while decorin is dynamically regulated in cultured isolated metanephric mesenchyme (MM). Biochemical analysis indicated that while both heparan sulfate (HS) and chondroitin sulfate (CS) are present, CS accounts for approximately 75% of the glycosaminoglycans (GAG) in the embryonic kidney. Selective perturbation of HS in whole kidney rudiments and in the isolated UB resulted in a significant reduction in the number of UB branch tips, while CS perturbation has much less impressive effects on branching morphogenesis. Disruption of endogenous HS sulfation with chlorate resulted in diminished FGF2 binding and proliferation, which markedly altered kidney area but did not have a statistically significant effect on patterning of the ureteric tree. Furthermore, perturbation of GAGs did not have a detectable effect on FGFR2 expression or epithelial marker localization, suggesting the expression of these molecules is largely independent of HS function. Taken together, the data suggests that nonselective perturbation of HSPG function results in a general proliferation defect; selective perturbation of specific core proteins and/or GAG microstructure may result in branching pattern defects. Despite CS being the major GAG synthesized in the whole developing kidney, it appears to play a lesser role in UB branching; however, CS is likely to be integral to other developmental processes during nephrogenesis, possibly involving the MM. A model is presented of how, together with growth factors, heterogeneity of proteoglycan core proteins and glycosaminoglycan sulfation act as a switching mechanism to regulate different stages of the branching process. In this model, specific growth factor-HSPG combinations play key roles in the transitioning between stages and their maintenance.

Branching morphogenesis and kidney disease

Branching morphogenesis in the kidney is a tightly regulated, complex process and its disruption potentially can lead to a broad spectrum of diseases, ranging from rare hereditary syndromes to common conditions such as hypertension and chronic kidney failure. This review synthesizes data on branching during kidney development derived from in vitro and in vivo rodent studies and to apply them to human diseases. It discusses how the broad organization of molecular interactions during kidney development might provide a mechanistic framework for understanding disorders related to aberrant branching.

TGF-β superfamily members modulate growth, branching, shaping, and patterning of the ureteric bud

Protein-rich fractions inhibitory for isolated ureteric bud (UB) growth were separated from a conditioned medium secreted by cells derived from the metanephric mesenchyme (MM). Elution profiles and immunoblotting indicated the presence of members of the transforming growth factor-beta (TGF-beta) superfamily. Treatment of cultured whole embryonic kidney with BMP2, BMP4, activin, or TGF-beta1 leads to statistically significant differences in the overall size of the kidney, the number of UB branches, the length and angle of the branches, as well as in the thickness of the UB stalks. Thus, the pattern of the ureteric tree is altered. LIF, however, appeared to have only minimal effect on growth and development of the whole embryonic kidney in organ culture. The factors all directly inhibited, in a concentration-dependent fashion, the growth and branching of the isolated UB, albeit to different extents. Antagonists of some of these factors reduced their inhibitory effect. Detailed examination of TGF-beta1-treated UBs revealed only a slight increase in the amount of apoptosis in tips by TUNEL staining, but diminished proliferation throughout by Ki67 staining. These data suggest an important direct modulatory role for BMP2, BMP4, LIF, TGF-beta1, and activin (as well as their antagonists) on growth and branching of the UB, possibly in shaping the growing UB by playing a role in determining the number of branches, as well as where and how the branches occur. In support of this notion, UBs cultured in the presence of fibroblast growth factor 7 (FGF7), which induces the formation of globular structures with little distinction between the stalk and ampullae [Mech. Dev. 109 (2001) 123], and TGF-beta superfamily members lead to the formation of UBs with clear stalks and ampullae. This indicates that positive (i.e., growth and branch promoting) and negative (i.e., growth and branch inhibiting) modulators of UB morphogenesis can cooperate in the formation of slender arborized UB structures similar to those observed in the intact developing kidney or in whole embryonic kidney organ culture. Finally, purification data also indicate the presence of an as yet unidentified soluble non-heparin-binding activity modulating UB growth and branching. The data suggest how contributions of positive and negative growth factors can together (perhaps as local bipolar morphogenetic gradients existing within the mesenchyme) modulate the vectoral arborization pattern of the UB and shape branches as they develop, thereby regulating both nephron number and tubule/duct caliber. We suggest that TGF-beta-like molecules and other non-heparin-binding inhibitory factors can, in the appropriate matrix context, facilitate "braking" of the branching program as the UB shifts from a rapid branching stage (governed by a feed-forward mechanism) to a stage where branching slows down (negative feedback) and eventually stops.