Bone morphogenetic protein 4 (BMP4), a member of the transforming growth factor-
(TGF-) superfamily, has been implicated in several aspects of embryonic development, from establishment of the basic embryonic body plan to morphogenesis of some organs, by regulating cell proliferation, differentiation, apoptosis, and chemotaxis1. We showed earlier2 that mice heterozygous for a null mutation in the Bmp4 gene (Bmp4+/-) exhibit anomalies that anatomically and histologically mimic human congenital anomalies of the kidney and urinary tract (CAKUT)3. Thus, a diverse spectrum of anomalies appears within one litter, including hypo/dysplastic kidneys, hydroureter, ectopic ureterovesical junction, and double collecting system. The study suggested that BMP4 has important functions in controlling the site of budding of the ureter from the epithelium of the Wolffian duct, and the branching of the ureter bud itself. BMP4 appears to accomplish these by locally antagonizing bud-inducing signal(s) from the metanephric mesenchyme that are acting ubiquitously on the duct and the ureter and by promoting the growth and elongation of the ureter bud, respectively. A deficiency in the level of BMP4 causes ectopic and accessory budding and reduced branching of the ureter in Bmp4+/- embryos. These defects lead to ectopic ureterovesical junction and double collecting system, and to hypoplastic kidney, respectively.
Ectopic ureterovesical junction is linked to the development of CAKUT, as it leads to megaureter, ureterovesical junction obstruction, vesicoureteral reflux, and hypo/dysplastic kidney4. In Bmp4+/- mutants, however, some of the kidney and ureter anomalies are seen even in the absence of ectopic ureterovesical junction. In addition, Bmp4 continues to be expressed throughout the embryonic kidney and urinary tract system beyond the stage of initial ureteral budding from the Wolffian duct. Likewise, the BMP type Ia receptor gene (Alk3) is transcribed in the metanephric and periureteral mesenchymes, as well as in the epithelium2,5,6. It is conceivable, therefore, that BMP4 can have diverse regulatory roles in the growth and differentiation of mesenchymes in different regions of the developing kidney.
By performing experiments on Bmp4+/- embryonic tissues in vivo and cultured explants in vitro, we found that BMP4 can have several other specific biological actions in the ontogeny of the excretory system.
METHODS
Mice
F2 male Bmp4 heterozygous null mutant mice (Bmp4lacZneo)7 on the 129/SvEv x Black Swiss genetic background were serially backcrossed with C57BL/6 females (The Jackson Laboratory, Bar Harbor, ME, USA), as described2. Heterozygous (Bmp4lacZneo+/-) and wild-type (+/+) mice between N2 and N5 generations were used in the present study. The Bmp4lacZneo genotype was determined by polymerase chain reaction (PCR) analysis for the neo gene with the following primer sequences: neo1, 5'-TCCTGCCGAGAAAGTATCCATCAT-3', and neo2, 5'-GTAAAGCACGAGGAAGCGGTCAGC-3'.
Explant culture
Kidney rudiments were dissected from E11.5 wild-type embryos. To obtain isolated metanephric mesenchymes, the kidney rudiments were soaked in 0.02% ethylendiamine tetraacetic acid (EDTA, pH 8.0) for 15 minutes, and the mesenchyme was manually separated from the ureteric bud with a 27 G needle. In the experiments on the mesenchyme surrounding the ureter epithelium, the ureter and a part of the Wolffian duct, but not the metanephros, was dissected from E11.5-12.0 embryos. Explants of the whole kidney, the isolated metanephric mesenchyme, or the ureter were cultured as described previously2. In some experiments, the medium was supplemented with human recombinant BMP4 (R&D Systems, Inc., Minneapolis, MN, USA) or fibroblast growth factor 2 (FGF2; R&D Systems, Inc.). Affi-Gel Blue beads (100 to 200 mesh, 75 to 150 
m; Bio-Rad Laboratories, Inc., Hercules, CA, USA) soaked with BMP4 or BMP7 (R&D Systems, Inc.), or heparin acrylic beads (Sigma Chemical Company, St. Louis, MO, USA) soaked with FGF2 were prepared as described2.
Histology
Samples were fixed in 4% buffered paraformaldehyde, routinely processed, embedded in paraffin and sectioned at 4 m. The sections were routinely stained with hematoxylin for qualitative and quantitative analyses.
In situ hybridization
Whole-mount in situ hybridization with digoxigenin (DIG) uridine triphosphate (UTP)-labeled antisense riboprobes and histologic section in situ hybridization using 35S-labeled antisense probes were performed as described previously2. The following mouse cDNA fragments were used as templates: 0.3 kb Pax2 cDNA2, 0.9 kb Foxd1 (previously Bf2) cDNA (provided by Dr. Tsutomu Kume, Vanderbilt University Medical Center, Nashville, TN, USA), and 0.9 kb Bmp7 cDNA8.
Detection of programmed cell death
Cell death was examined in whole tissues and sections. Explants were incubated in the medium with 2.5 g/mL propidium iodide (Boehringer Mannheim, Indianapolis, IN, USA) for 30 minutes at 37°C. Tissues were then washed in large volume of the medium without propidium iodide for 30 minutes. Labeled cells were detected using a Zeiss Universal fluorescence microscope (Carl Zeiss, Oberkochen, Germany). To analyze apoptotic cells in sections, the TUNEL assay was performed using the ApoTag detection kit (Intergen, Purchase, NY, USA) and Apoptosis detection kit (Promega, Madison, WI, USA), according to the manufacturers' protocol. Six nonadjacent coronal sections from embryonic kidneys were examined and the percentage of terminal deoxy transferase uridine triphosphate nick end-labeling (TUNEL)–positive nuclei against total nuclei was determined for the metanephric mesenchymal area.
Immunostaining for 
-smooth muscle actin
The sections were stained for 
-smooth muscle actin (-SMA), with an antihuman -SMA antibody (DAKO, Carpinteria, CA, USA), following the manufacturer's protocol. The density of -SMA expressing cells was determined as a percentage of the -SMA positive cells against total mesenchymal cells surrounding the epithelium at the most cranial portion of the ureter.
Cell migration assay
Ureters from E12.0 embryos were treated with 0.02% ethylenediaminetetraacetic acid (EDTA), and the epithelium was then manually removed. The remaining periureteral mesenchyme was dissociated into single cells with 0.1% tripsin-EDTA (GIBCO BRL, Gaithersburg, MD, USA) and 0.2% collagenase A (Boehringer Mannheim) for 60 minutes at 37°C followed by pipetting. After resuspended in Dulbecco's modified Eagle's medium (DMEM) containing 20% fetal bovine serum (FBS), cells were subjected to a modified Boyden chamber assay. Recombinant BMP4 or platelet-derived growth factor (PDGF)–BB (R&D Systems, Inc.) was added to the bottom chamber or both the bottom and the top chamber at a concentration ranging from 0 to 30 ng/mL. In a pilot study, BMP4 at the above concentrations does not change the number of viable cells (more than 90% of cells excluded trypan blue) after 24 hours of culture compared to vehicle alone. A polycarbonate membrane filter (8 m pore size) (Neuro Probe, Inc., Cabin Jone, MD, USA) was placed between the bottom and top chambers. The mesenchymal cell suspension was added to the top chamber at a cell density of 2 
103/mm2. After an 18-hour incubation, cells on the top of the filters were scraped, the remaining cells were fixed with methanol, and stained with hematoxylin. The number of cells was determined under 200 magnification at five random fields.
RESULTS
Enhanced apoptosis in the metanephric mesenchyme of Bmp4+/- embryos
We found earlier that the kidneys of Bmp4+/- mice at birth contain multicystic dysplastic regions, which are completely devoid of nephrogenic components, but filled with cysts and stromal mesenchymal cells. This suggests that BMP4 is also essential for later processes of kidney development. In the present study, we examined apoptotic activity in the mutant kidneys at E14.5 and E16.5, as increased cell death has been reported prior to the development of hypoplastic cystic kidneys in some mutant mice9,10. In wild-type embryos at E14.5, TUNEL-positive cells are present in the developing nephron epithelium and loose stromal cells located in the center of metanephros, but hardly seen in the metanephric mesenchyme located in peripheral areas Figure 1a. By contrast, in +/- mutants, a large number of TUNEL-positive cells is seen in the superficial metanephric mesenchymal area Figure 1b. Quantitation of the results showed that, although there is variation among mutant embryos, the number of apoptotic cells is significantly higher in +/- mutants than in wild-type embryos (% total metanephric mesenchymal cells, Bmp4+/+, 0.59 
0.06, N = 10; Bmp4+/-, 1.60 0.28, N = 12, mean SEM, P < 0.01 Student t test) Figure 1g. Moreover, analyses of molecular markers in an adjacent section revealed that TUNEL-positive cells of the mutant kidneys are mainly located in the area that expresses Foxd1Figure 1d, but not Pax2Figure 1c. Thus, Bmp4+/- embryos have an abnormally high apoptotic activity in the stromal components of metanephric mesenchyme at this stage. Note that Bmp4+/- kidneys already contain region that lacks superficial nephrogenic components at E16.5 Figure 1e, f, while the apoptotic activity in the nephrogenic zone is comparable between Bmp4+/+ and +/- embryos at this stage (% total metanephric mesenchymal cells, Bmp4+/+, 0.47 0.05, N = 11; Bmp4+/-, 0.52 0.13, N = 10, mean SEM). This suggests that an abnormal cell death in the metanephric mesenchyme at mid-stages can contribute to the development of hypodysplastic kidney in Bmp4 heterozygous mice.
Figure 1.
Apoptosis in the metanephric mesenchyme ofBmp4+/- embryos. Sections from wild-type (A) and mutant (B) kidneys at E14.5 were subjected to terminal deoxy transferase uridine triphosphate nick end-labeling (TUNEL) staining. Many TUNEL-positive cells (stained in brown) are seen in the peripheral mesenchymal area of Bmp4+/- mutants (arrows). Bars are 200 m. An adjacent section of (B) was subjected to in situ hybridization with Pax2 (C) or Foxd1 probes (D). TUNEL positive cells in (B) were present in Pax2-negative and Foxd1-positive area of the metanephric mesenchyme (arrows). Sections from wild-type (E) and mutant (F) kidneys at E16.5 are shown. Condensed and noninduced mesenchymal cells are lower in number in Bmp4+/- mutants. The percentage of apoptotic nuclei against total nuclei in the metanephric mesenchyme (G) is shown in kidneys from wild-type (N = 10) and mutant (N = 12) embryos. Mean values and standard errors from each group are also shown. The difference is statistically significant by Student t test (P < 0.01).
BMP4 promotes expansion of the stromal mesenchyme and inhibits the condensation of the metenephric mesenchyme
Using an explant culture system, we next examined whether BMP4 has direct actions on the metanephric mesenchyme. In whole metanephric explants, BMP4 uniformly inhibits the branching of the ureter (N = 20) Figure 2b, an observation consistent with earlier reports2,11. In addition, exogenous BMP4 down-regulated the expression of Pax2 and Bmp7 (N = 14), markers for the condensed mesenchyme, which surround the tip of the branching ureters Figure 2b, h. Histologic examinations of tissue sections reveal that the condensation of metanephric mesenchyme is inhibited around the ureteric buds when explants were treated with BMP4 (data not shown). Moreover, when the explants are treated with both BMP4 and FGF2, the latter being a known antiapoptotic factor for the metanephric mesenchyme12, the peripheral portion of the mesenchyme increases in size compared to explants with FGF2 alone (Figures 2f, j, and 5 in reference2). The increase in size of this portion is further confirmed by histologic analyses of tissue sections (data not shown). Whole mount in situ hybridization reveals that this mesenchymal component is negative for Bmp7 (N = 14) Figure 2j and Pax2 (Figure 5 in reference2) and weakly and heterogeneously positive for Foxd1 (N = 14) Figure 2f, a winged-helix transcription factor that marks a peripheral undifferentiated stromal cell population13. In this setting, the expression area of Bmp7, which is known to have survival and growth-promoting effects on the metanephric mesenchyme14, is decreased in explants treated with BMP4 Figure 2j.
Figure 2.
Bone morphogenetic protein 4 (BMP4) inhibits induction of metanephric mesenchyme around the ureteric buds, and promotes expansion of the peripheral stromal zone in the presence of fibroblast growth factor 2 (FGF2). Kidney rudiments dissected from E11.5 wild-type embryos were cultured for 48 hours in the presence of vehicle alone (A, C, G), 100 ng/mL of BMP4 (B, D, H), 100 ng/mL of FGF2 (E, I), or both BMP4 and FGF2 (F, J). The explants were subjected to whole mount in situ hybridization with a probe encoding Pax2 (A, B), Foxd1 (C to F), or Bmp7 (G to J). Each panel represents a typical result of more than 14 explants obtained from three independent experiments. BMP4 inhibits branching of the ureters (B, H) and down regulates Pax2 (B) and Bmp7 (H) expression surrounding the ureteric buds (arrowheads). Treatment with both BMP4 and FGF2 promotes expansion of Bmp7-negative (J) and Foxd1-weakly positive (F) stromal compartment of the metanephric mesenchyme (arrows). Arrowheads indicate an edge of the explants. Bars are 300 m.
Figure 5.
Bone morphgenetic protein 4 (BMP4) induces an attractive chemotaxis for the periureteral mesenchymal cells. Bovine serum albumin (BSA) control beads (A, D, G), 3 g/mL (B, E, H), or 10 g/mL (C, F, I) BMP4 beads were placed in the mesenchyme that surrounds the epithelium of the ureter explants dissected from E12.0 embryos. After incubation for 24 hours, the explants were routinely processed, sectioned, and stained with hematoxylin. BMP4 beads at both 3 g/mL (N = 10) and 10 g/mL (N = 15) concentrations, but not control beads, form several mesenchymal cell layers around the beads. (A to C), (D to F) and (G to I) are from three independent experiments, respectively. Ut is ureter epithelium. Bar is 100 m. Sections were subjected to terminal deoxy transferase uridine triphosphate nick end-labeling TUNEL staining. Compared to the BSA bead (J), a larger number of apoptotic cells are present in the metanephric mesenchyme (arrowheads) around the BMP4 bead (K). Note that apoptotic activity is comparable in the periureteral mesenchyme between these two beads. The density of periureteral mesenchymal cells is shown (L) as the cell number within the area as far as 25 m in distance from the bead per square micrometer. The density is significantly higher for 10 g/mL BMP4 beads than control beads. The number of migrating cells is shown (M) in a modified Boyden chamber assay. Dissociated periureteral mesenchymal cells are placed on a membrane with 8 m pores and incubated for 18 hours. The migrating cells through the membrane are fixed, stained, and counted under 200 magnification at random five fields. Addition of BMP4 to the bottom chamber increases the number of migrating cells. Addition of BMP4 to both chambers inhibits the migration. Addition of platelet-derived growth factor (PDGF)-BB has no effect. Values are expressed as mean SEM. *P < 0.001 vs. control; #P < 0.01 vs. BMP4 in the bottom chamber alone by analysis of variance (ANOVA) followed by Bonferroni test.
BMP4 prevents cell death of the isolated metanephric mesenchyme
Apart from the inhibitory effect of BMP4 on the condensed mesenchyme found in the above experiments, we next tested whether BMP4 can directly act as a survival factor for the isolated metanephric mesenchyme. Mesenchyme separated from the E11.5 ureteric bud epithelium was cultured on filters in the presence of bovine serum albumin (BSA)-, BMP4-, BMP7- or FGF2-soaked beads and then incubated in medium with propidium iodide, a fluorescent dye which can penetrate into dead cells, but not living cells. While fluorescent label is uniformly distributed in the control explants, it is distinctively absent in cells adjacent to BMP4- (N = 16), BMP7- (N = 8) or FGF2-soaked beads (N = 7) Figure 3. The rescue of cell death occurs only in close proximity to the BMP4 beads Figure 3f, when compared to the FGF2-beads Figure 3h. However, the degree is comparable between BMP4 and BMP7 beads Figure 3f, g. Similar results were also seen when BMP4 or FGF2 was added directly to the culture medium, although the degree of antiapoptotic effect of BMP4 is lower than that of FGF2 (data not shown). These results indicate that BMP4, acting on the metanephric mesenchyme, prevents death of the mesenchymal cells and promotes increase in cell number of the stromal cell population in the presence of FGF2.
Figure 3.
Bone morphogenetic protein 4 (BMP4) prevents cell death of the metanephric mesenchyme. Isolated metaneprhic mesenchymes obtained from E11.5 embryos were cultured for 24 hours in the presence of (A, E) bovine serum albumin (BSA)-, (B, F) 3 g/mL BMP4-, (C, G) 3 g/mL BMP7-, or (D, H) 3 g/mL fibroblast growth factor 2 (FGF2)-soaked beads, and incubated with 2.5 g/mL propidium iodide for 30 minutes. The explants were observed under the microscope with (E to G) or (A to D) without a fluorescence. In BMP4 beads (N = 16), BMP7 beads (N = 8), and FGF2 beads (N = 7), cells adjacent to the beads do not fluoresce, while, in the control explants (N = 9), mesenchymal cells show fluorescent signals around the beads. The heparin acrylic bead (C and F) shows an autofluorescence. Bars are 200 m.
Defective smooth muscle development of the ureter in Bmp4+/- embryos
During embryonic development of the ureter, periureteral mesenchymal cells condense around the epithelium to form a distinctive smooth muscle layer. In this process, whereas the expression of Bmp4 is uniformly seen in the ureter mesenchyme at early developmental stages, its expression becomes localized to an area where the smooth muscle layer develops at later stages2. This suggests that BMP4 has a regulatory role in the development of smooth muscle of the urinary tract. At E15.5, the periureteral mesenchymal cells begin to express -SMA, an early marker for smooth muscle cells, in wild-type embryos at the cranial portion of the ureter. We therefore examined the expression of -SMA in the periureteral mesenchyme at this stage. Immunostaining with -SMA antibody reveals that the number of -SMA–positive cells is significantly lower in mutants than in wild-type embryos (% total ureter mesenchymal cells, +/+, 28.4 1.2%, N = 16; +/-, 16.4 1.4%, N = 12, mean SEM, P < 0.001 Student t test) Figure 4a, b. At this stage, the size of the ureter lumen and the morphology of the ureter epithelium are comparable between wild-type and mutant embryos Figure 4a, b. Thus, the reduced expression of the smooth muscle marker in mutant ureters is not likely a result of a mechanical injury due to a possible obstructive uropathy as previously speculated2.
Figure 4.
Ureteral smooth muscle development inBmp4+/- embryos. Transverse sections of the most cranial portion of the ureter are immunostained for -smooth muscle actin (-SMA). Compared to wild-type embryos (A) at E15.5, there are a few smooth muscle cells surrounding the ureter epithelium in Bmp4+/- (B). Abbreviations are: Ao, aorta; Ut, ureter epithelium; Kd, kidney. Bars are 50 m. (C) The percentage of -SMA-expressing cells against total mesenchymal cells around the epithelium is shown at the most cranial portion of wild-type (N = 16) and mutant (N = 12) ureters at E15.5. Mean values and standard errors from each group are also shown. The difference is statistically significant by Student t test (P < 0.001).
BMP4 promotes migration of ureter mesenchymal cells
We next examined whether smooth muscle cells can be induced by BMP4 in explant culture system. When E12.0 ureter rudiments are cultured on filters for 72 hours in the absence of exogenous BMP4, -SMA expression is seen in the mesenchymal cells that condense around the ureter epithelium. Addition of exogenous BMP4 (30 to 300 ng/mL) to the medium does not cause an increase in the number of -SMA–positive cells nor a change in 5-bromo-2'-deoxyuridine (BrdU) incorporation or TUNEL positivity (data not shown). In the next experiments, BMP4-soaked beads were placed in the mesenchyme surrounding the ureter epithelium. While BMP4 beads neither induce -SMA expression, promote cell proliferation (data not shown), nor prevent cell death Figure 5j, k, in the mesenchymal cells surrounding the beads, the beads promote the accumulation of the mesenchymal cells to form several layers surrounding the beads, which is not seen in control BSA beads Figure 5a to c. Similar results were obtained in three replicate experiments Figure 5d to i. Quantitation of cell number revealed that the density of mesenchymal cells significantly increases around the 10 g/mL BMP4 bead (cell density, 1.48 0.07 10-2/m2, N = 7, mean SEM), when compared to the control BSA bead (0.9 0.04 10-2/m2, N = 7; mean SEM, P < 0.05) Figure 5l. These results suggest that a localized, not diffuse, high concentration of BMP4 is important for the induction of the ureter smooth muscle and that BMP4 can serve as a chemoattractant for the periureteral mesenchymal cells. To examine the chemotactic activity of BMP4, a modified Boyden chamber assay was performed for dissociated periureteral mesenchymal cells. Addition of BMP4 (30 ng/mL) to the bottom chamber shows a significant increase in the cell migration through the membrane (number of migrating cells, control, 23.9 3.2, N = 11; BMP4, 54.0 3.2, N = 12, mean SEM, P < 0.001) Figure 5m. The increase in cell migration is not seen (23.6 4.2, N = 10), when treated with PDGF-BB, which is implicated in the chemoattraction of vascular smooth muscle cells15. Moreover, addition of BMP4 to both the bottom and top chambers attenuates migration of the cells, when compared to the bottom chamber alone (35.3 3.6, N = 10, P < 0.01) Figure 5m, suggesting that the migration of the cells through the membrane is due, at least in part, to a chemotactic effect of BMP4. Thus, BMP4 can induce attractive chemotaxis, thereby promoting condensation of the mesenchymal cells along the ureter epithelium.
DISCUSSION
In this study, we show that the development of the dysplastic regions seen in the kidneys of Bmp4+/- embryos at E16.5 are preceded by enhanced apoptosis in the stromal cell population of metanephric mesenchyme in mutant embryos Figure 1. A similar pathophysiologic process has been reported for the formation of hypo/dysplastic kidney in other mutant mice. For example, in mice carrying null mutations of the Bcl2 or Bmp7 genes, the development of hypo/dysplastic kidney is preceded by fluminant apoptosis of the metanephric mesenchyme during the mid-stage of nephrogenesis9,10. In Bmp4+/- mutants from E12.5 to E14.5, while nephron number per kidney is appreciably reduced, the nephron density per se and the overall morphology are well preserved. However, the dysplastic lesion develops at E16.5 Figure 1f, just 2 days after the enhanced apoptosis occurs in the stromal population of the metanephric mesenchyme Figure 1b. The dysplastic morphorlogy seen in Bmp4 heterozygozes is characterized by two lesions, one of which is local loss of nephrogenic components and the other is cystic dilatation of tubules and glomeruli. We speculate without conclusive evidence that the enhanced apoptosis in the stromal metanephric mesenchyme at E14.5 contributes to the development of the former abnormality. Other pathophysiologic processes can be involved in the formation of cystic region, given the potentially multiple biologic effects of BMP4, however.
The observed enhanced apoptosis in the mutant metanephric mesenchyme can be a result of a deficiency in antiapoptotic factors derived from the ureteric bud. In support of this possibility, previous studies suggested that BMP4, acting on the ureter epithelium, stimulates the growth and elongation of the ureter bud and promotes overall branching morphogenesis of the ureter2. To determine whether BMP4 has a direct antiapoptotic effect on the metanephric mesenchyme, we performed in vitro experiments on the metenephric mesenchyme. In the isolated metanephric mesenchyme, BMP4-soaked beads uniformly prevented cell death of the metanephric mesenchyme locally Figure 3. In addition, in the presence of exogenous FGF2, treatment with BMP4 led to an increase in cell number of the Pax2- and Bmp7-negative and weakly Foxd1-positive stromal zone of the metanephric mesenchyme in the whole metanephric explants (Figures 2 and Figure 5 in reference2). Moreover, exogenous BMP4 down-regulates Pax2 and Bmp7 expression that surrounds the ureteric bud, and inhibits the condensation of the metanephric mesenchyme Figure 2. This observation is consistent with the previous finding that BMP4 attenuated tubulogenesis in the isolated metanephric mesenchyme induced by the spinal cord or lithium chloride11. Collectively, these in vivo and in vitro observations indicate that BMP4 can act on the metanephric mesenchyme to regulate its growth and differentiation. With regard to the inhibitory effect of BMP4 on the development of the condensed mesenchyme, however, the possibility remains that the loss of Pax2 or Bmp7 expression is related to an increase in cell death, since a large quantity of TUNEL-positive cells are seen in the condensed mesenchyme that surrounds the BMP4 bead, but not the control bead Figure 5j, k. The data from TUNEL-positive cells in the condenced metanephric mesenchyme appear to argue against the antiapoptotic effect of BMP4 shown in Figure 3. However, the condensed metantphric mesenchyme, which is positive for Pax2 or Bmp7, is small fraction of the total isolated mesenchyme shown in Figure 3. In support of the latter, it has been reported earlier that isolated mesenchyme has a uniformly low level of Foxd1, but they completely lacks expression of Pax214. In this study, we cannot determine whether BMP4 directly induces apoptosis of nephrogenic mesenchyme, or inhibits the differentiation of the mesenchyme, which, in turn, undergoes apoptosis.
The observed effects of BMP4 are reminiscent of those of BMP7 on the metanephric mesenchyme. Thus, BMP7 has been shown to prevent cell death, inhibit tubulogenesis, and promote expansion of stromal progenitor cell population in the presence of FGF2, thereby preserving the metenephric mesenchymal cell population at the tip of the branching ureters14. It has been shown that a functional redundancy exists among the members of BMP superfamily during the embryonic development1, and BMP4 and 7 have a binding affinity to the same BMP receptors16,17, for example, Bmpr1a (Alk3) and Bmpr1b (Alk6), the former being intensely expressed in metanephric mesenchyme2,6. It is conceivable, therefore, that BMP4 and 7 share a common receptor-signaling pathway and identical effects on the development of metanephric mesenchyme. In this regard, one may argue that the observed effects of BMP4 on the metanephric mesenchyme Figures 2 and 3 merely mimic those of BMP7, since, unlike BMP7, BMP4 is not expressed in the metanephric mesenchyme2,5, and BMP4 appears to have only a short-range effect as shown Figure 3. However, in addition to the loose stromal mesenchyme that surrounds the branching ureter, Bmp4 is also distinctively expressed in the superficial capsule of the developing kidney (Figure 1 d, f in reference2). Furthermore, in Bmp4+/- embryos that selectively lack Bmp4, but not Bmp7, a large number of apoptotic cells are seen in the most peripheral portion of the metanephric mesenchyme just beneath the kidney capsule Figure 1. In these heterozygotes, the onset and distribution of the mesenchyme apoptosis closely resemble those in Bmp7 null mutant embryos5. The present study, therefore, indicates that full gene and protein dosages of both BMP4 and BMP7 are required for the normal development of the metanephric mesenchyme.
In addition to the effects on the metanephric mesenchyme, it appears that BMP4 has a different effect on the development of the periureteral mesenchyme, namely, induction of the smooth muscle layer at the site where BMP4 is intensely expressed around the ureter epithelium. In wild-type embryos, -SMA is expressed in the periureteral mesenchyme as early as E15.5. At this stage, the -SMA-expressing cells are significantly lower in number in Bmp4+/- embryos than wild type Figure 4, while the gross appearance of the ureter epithelium is not changed. At later stages, the poor development of the muscle layer is more distinctively seen in some of the mutant embryos, but, at the same time, accompanied by a change in the overall morphology of the ureter (e.g., dilated lumen, flattened epithelium, and tortuous structure)2. As reported previously, Bmp4+/- mutants show an ectopic ureterovesical junction, which can cause an obstructive uropathy2. Thus, the possibility remained that the hypoplasia of the ureter smooth muscle is attributed merely to a mechanical injury as a result of obstruction, but not to an intrinsic defect of smooth muscle development. We, therefore, performed in vitro experiments to test whether BMP4 has a local regulatory role on the induction of the ureter smooth muscle. Using the ureteric tissues separated from the metanephros, addition of BMP4 to the culture medium did not promote smooth muscle development, but caused a slight decrease in the thickness of the whole explants without a change in cell proliferation not cell death. By contrast, BMP4-soaked beads, when placed in the mesenchyme surrounding the ureter epithelium, promote accumulation of the mesenchymal cells to form several layers around the beads Figure 5, with no change in cell proliferation or cell death. This suggests that a localized, not diffuse, high concentration of BMP4 is required prior to the induction of the ureter smooth muscle layer. Moreover, in a modified Boyden chamber assay, BMP4 promotes chemotaxis of periuretral mesenchymal cells through the filter, and the migration is inhibited when BMP4 is present in both top and bottom chambers Figure 5. Thus, these in vitro observations indicate that BMP4 acts as a chemoattractant for the periureteral mesenchymal cells. A similar chemotaxis-promoting effect of BMPs has been reported previously. For example, BMP2 has been shown to induce chemotaxis of human monocytes18, vascular aortic smooth muscle cells19, and corneal fibroblasts20.
It remains uncertain to date whether the formation of the ureter smooth muscle is preceded by the recruitment of the local mesenchymal cells. While BMP4 can induce migration and condensation of the mesenchyme around the ureter epithelium by its attractive chemotactic activity, BMP4 alone is not sufficient for differentiation into smooth muscle cells, since BMP4 beads do not induce -SMA expression in the accumulated mesenchymal cells around the beads. Thus, we cannot conclude in the present study that the chemotactic effect of BMP4 is indeed essential for the development of the ureter smooth muscle. In this regard, it has been shown that the interaction of the mesenchyme with the epithelium is critical for the differentiation into the visceral smooth muscle, as revealed by experiments on tissue recombinants of embryonic uterus21, stomach22, and bladder23. In embryonic ureters, smooth muscle differentiation is induced in the region somewhat distant from the epithelium, where Bmp4 is intensely expressed. Thus, it is likely that BMP4 determines the site of the periureteral mesenchymal condensation, which, in turn, undergoes differentiation into smooth muscle layer through a pathway governed by the epithelium24,25.
CONCLUSION
In summary, BMP4 can have multiple biologic functions on the morphogenesis of individual tissues of the excretory system (i.e., the development of the kidney, the ureter and the ureterovesical junction). Results from the present study suggest that BMP4, acting on the metanephric mesenchyme, prevents cell death and promotes the growth of the stromal component. In addition, BMP4 inhibits condensation of the nephrogenic mesenchyme either via a direct action on the mesenchyme or an indirect action through the ureter. BMP4 also serves as a chemoattractant for periureteral mesenchymal cells and can induce locally the smooth muscle layer of the ureter at Bmp4-expressing sites. Moreover, as shown earlier2, BMP4, acting on the epithelium, regulates the site of initial ureteral budding on the Wolffian duct, and the branching morphogenesis of the ureter by antagonizing mesenchyme-derived inductive signals and promoting the growth and elongation of the ureter, respectively. Therefore, the wide anatomic spectrum of CAKUT seen in Bmp4+/- mice is attributed, at least in part, to the fact that BMP4 governs organogenesis at many different stages and tissues, each through a distinct mechanism.
References
| 1. | Hogan BLM. Bone morphogenetic proteins: Multifunctional regulators of vertebrate development. Genes Dev 1996; 10: 1580–1594. | PubMed | ISI | ChemPort | |
| 2. | Miyazaki Y, Oshima K & Fogo A et al. Bone morphogenetic protein 4 regulates the budding site and elongation of the mouse ureter. J Clin Invest 2000; 105: 863–873. | PubMed | ISI | ChemPort | |
| 3. | Pope JC, IV, Brock JW, III & Adams MC et al. How they begin and how they end: Classic and new theories for the development and deterioration of congenital anomalies of the kidney and urinary tract, CAKUT. J Am Soc Nephrol 1999; 10: 2018–2028. | PubMed | ISI | |
| 4. | Mackie GG & Stephens FD. Duplex kidneys: A correlation of renal dysplasia with position of the ureteral orifice. J Urol 1975; 114: 274–280. | PubMed | ISI | ChemPort | |
| 5. | Dudley AT & Robertson EJ. Overlapping expression domains of bone morphogenetic protein family members potentially account for limited tissue defects in BMP7 deficient embryos. Develop Dynam 1997; 208: 349–362. | ISI | ChemPort | |
| 6. | Dewulf N, Verschueren K & Lonnoy O et al. Distinct spatial and temporal expression patterns of two type I receptors for bone morphogenetic proteins during mouse embryogenesis. Endocrinology 1995; 136: 2652–2663. | Article | PubMed | ISI | ChemPort | |
| 7. | Lawson KA, Dunn NR & Roelen BA et al. Bmp4 is required for the generation of primordial germ cells in the mouse embryo. Genes Dev 1999; 13: 424–436. | PubMed | ISI | ChemPort | |
| 8. | Furuta Y, Piston DW & Hogan BLM. Bone Morphogenetic Proteins (BMPs) as regulators of dorsal forebrain development. Development 1997; 124: 2203–2212. | PubMed | ISI | ChemPort | |
| 9. | Nagata M, Nakauchi H & Nakayama K et al. Apoptosis during an early stage of nephrogenesis induces renal hypoplasia in bcl-2-deficient mice. Am J Pathol 1996; 148: 1601–1611. | PubMed | ISI | ChemPort | |
| 10. | Luo G, Hofmann C & Bronckers ALJJ et al. BMP-7 is an inducer of nephrogenesis, and is also required for eye development and skeletal patterning. Genes Dev 1995; 9: 2808–2820. | PubMed | ISI | ChemPort | |
| 11. | Raatikainen-Ahokas A, Hytonen M & Tenhunen A et al. BMP-4 affects the differentiation of metanephric mesenchyme and reveals an early anterior-posterior axis of the embryonic kidney. Dev Dyn 2000; 217: 146–158. | Article | PubMed | ISI | ChemPort | |
| 12. | Perantoni AO, Dove L & Karavanova I. Basic fibroblast growth factor can mediate the early inductive events in renal development. Proc Natl Acad Sci USA 1995; 92: 4696–4700. | PubMed | ChemPort | |
| 13. | Hatini V, Huh SO & Herzlinger D et al. Essential role of stromal mesenchyme in kidney morphogenesis revealed by targeted disruption of Winged Helix transcription factor BF-2. Genes Dev 1996; 10: 1467–1478. | PubMed | ISI | ChemPort | |
| 14. | Dudley AT, Godin RE & Robertson EJ. Interaction between FGF and BMP signaling pathways regulates development of metanephric mesenchyme. Genes Dev 1999; 13: 1601–1613. | Article | PubMed | ISI | ChemPort | |
| 15. | Folkman J & D'Amore PA. Blood vessel formation: what is its molecular basis? Cell 1996; 87: 1153–1155. | Article | PubMed | ISI | ChemPort | |
| 16. | ten Dijke P, Yamashita H & Sampath TK et al. Identification of type I receptors for osteogenic protein-1 and bone morphogenetic protein-4. J Biol Chem 1994; 269: 16985–16988. | PubMed | ChemPort | |
| 17. | Rosenzweig BL, Imamura T & Okadome T et al. Cloning and characterization of a human type II receptor for bone morphogenetic proteins. Proc Natl Acad Sci USA 1995; 92: 7632–7636. | PubMed | ChemPort | |
| 18. | Cunningham NS, Paralkar V & Reddi AH. Osteogenin and recombinant bone morphogenetic protein 2B are chemotactic for human monocytes and stimulate transforming growth factor beta 1 mRNA expression. Proc Natl Acad Sci USA 1992; 89: 11740–11744. | PubMed | ChemPort | |
| 19. | Willette RN, Gu JL & Lysko PG et al. BMP-2 gene expression and effects on human vascular smooth muscle cells. J Vasc Res 1999; 36: 120–125. | Article | PubMed | ISI | ChemPort | |
| 20. | Kim WJ, Mohan RR & Mohan RR et al. Effect of PDGF, IL-1alpha, and BMP2/4 on corneal fibroblast chemotaxis: Expression of the platelet-derived growth factor system in cornea. Invest Ophthalmol Vis Sci 1999; 40: 1364–1372. | PubMed | ISI | ChemPort | |
| 21. | Cunha GR, Battle E & Young P et al. Role of epithelial-mesenchymal interactions in the differentiation and spatial organization of visceral smooth muscle. Epi Cell Biol 1992; 1: 76–83. | ChemPort | |
| 22. | Takahashi Y, Imanaka T & Takano T. Spatial pattern of smooth muscle differentiation is specified by the epithelium in the stomach of mouse embryo. Dev Dyn 1999; 212: 448–460. | ISI | |
| 23. | Disandro MJ, Li Y & Baskin LS et al. Mesenchymal-epithelial interactions in bladder smooth muscle development: Epithelial specificity. J Urol 1998; 160: 1040–1046. | PubMed | ISI | ChemPort | |
| 24. | Schuger L, Skubitz APN & Zhang J et al. Laminin  1 chain synthesis in the mouse developmental lung: requirement for epithelium-mesenchymal contact and possible role in bronchial smooth muscle development. J Cell Biol 1997; 139: 553–562. | Article | PubMed | ISI | ChemPort | |
| 25. | Yang Y, Relan NK & Przywara DA et al. Embryonic mesenchymal cells share the potential for smooth muscle differentiation: Myogenesis is controlled by the cell's shape. Development 1999; 126: 3027–3033. | PubMed | ISI | ChemPort | |
Acknowledgments
The authors thank Dr. Brigid L.M. Hogan for critical reading of the manuscript. We also thank Dr. Tsutomu Kume for technical advice, and Ms. Teresa Bills and Ms. Ellen Donnert for technical assistance. This work was supported by National Institute of Health grants DK-44757 and DK-37868. Dr. Agnes Fogo is a recipient of an Established Investigator Award from the American Heart Association.
