Midkine Promotes Selective Expansion of the Nephrogenic Mesenchyme During Kidney Organogenesis

Abstract

During kidney development, the growth and development of the stromal and nephrogenic mesenchyme cell populations and the ureteric bud epithelium is tightly coupled through intricate reciprocal signaling mechanisms between these three tissue compartments. Midkine, a target gene activated by retinoid signaling in the metanephros, encodes a secreted polypeptide with mitogenic and anti-apoptotic activities in a wide variety of cell types. Using immmunohistochemical methods we demonstrated that Midkine is found in the uninduced mesenchyme at the earliest stages of metanephric kidney development and only subsequently concentrated in the ureteric bud epithelium and basement membrane. The biological effects of purified recombinant Midkine were analyzed in metanephric organ culture experiments carried out in serum-free defined media. These studies revealed that Midkine selectively promoted the overgrowth of the Pax-2 and N-CAM positive nephrogenic mesenchymal cells, failed to stimulate expansion of the stromal compartment and suppressed branching morphogenesis of the ureteric bud. Midkine suppressed apoptosis and stimulated cellular proliferation of the nephrogenic mesenchymal cells, and was capable of maintaining the viability of isolated mesenchymes cultured in the absence of the ureteric bud. These results suggest that Midkine may regulate the balance of epithelial and stromal progenitor cell populations of the metanephric mesenchyme during renal organogenesis.

Introduction

The development of the mammalian kidney depends on reciprocal inductive interactions between the ureteric bud and the metanephric mesenchyme that control cellular proliferation, survival, and differentiation in these two tissues.^1–3 The identification of the epithelial-mesenchymal signaling pathways responsible for these complex inductive events has been a major objective in the effort to understand kidney organogenesis in molecular terms. The ureteric bud provides essential survival factors for the metanephric mesenchyme^4–6 and recent studies employing metanephric organ culture have demonstrated that Tissue Inhibitor of Metalloproteinase II (TIMP2),⁷ Fibroblast Growth Factor 2 (FGF2) and Bone Morphogenetic Protein 7 ( BMP7) may play important roles in these processes.^8,9 In turn, retinoid-dependent mesenchymal stromal cell signaling activates ureteric bud branching morphogenesis by maintaining the expression of the c-ret receptor tyrosine kinase in the ureteric bud epithelium¹⁰ which functions as the signal transducer for the mesenchymal cell-secreted factor GDNF.^11,12

In previous studies we reported that serum-free conditioned media prepared from the G401 renal tumor cell line (G401 CM) was required for cellular proliferation of microdissected mesenchymal cells including epithelial progenitor cells obtained from the metanephric kidney.¹³ G401 cell secreted polypeptide factors include FGF2 and BMP7¹⁴ both of which function as mitogens and survival factors for the metanephric mesenchyme in vitro.⁸ Midkine,^15,16 a third secreted polypeptide growth factor identified in G401 CM^17,18 is a retinoic acid inducible gene implicated in the regulation of Wilms' tumor growth and differentiation which is expressed during kidney development.¹⁹ Since retinoic acid plays an essential role in development of the kidney,^10,20–23 we reasoned that Midkine might play an important role during kidney organogenesis. Although genetic inactivation of the Midkine gene did not result in overt developmental abnormalities of the kidney,²⁴ anti-Midkine neutralizing antibodies have recently been shown to inhibit metanephric kidney development in organ culture²⁵ suggesting that this growth factor may have important functions when conditions for organ development are sub-optimal. In the studies reported here we show that Midkine addition to cultured metanephroi results in marked expansion of the nephrogenic mesenchyme without commensurate increases in the stromal mesenchyme cell compartment, and without inhibition of overall organ growth. The biologic activities of Midkine are unique, and suggest that this polypeptide signaling factor may regulate renal progenitor cell number during kidney development.

Materials and Methods

Metanephric Kidney Organ Culture.

Metanephric organ culture was performed as previously described²⁶ with some modifications. Briefly, CD-1 mice were mated and females sacrificed by CO₂ asphyxiation 11 or 12 days after plug formation. Their embryos were aseptically removed and metanephric kidneys dissected out in serum-free medium. To obtain isolated mesenchyme, kidneys were treated with 100 mM EDTA for 30 minutes on ice. The mesenchyme was teased away from the ureter. Explants of mesenchyme or whole kidneys were cultured in serum-free conditions at 37°C, 5% CO₂ on 0.4 µm pore size Transwells (Costar, cat# 3450) in DMEM/F12 medium supplemented with 45 mM sodium bicarbonate, 1X ITS (Sigma), 2 nM tri-iodothyronine and 70 nM prostaglandin E1 in the presence or absence of either 7 nM Midkine (7 nM) or 100 ng/ml FGF2. Recombinant murine Midkine was purified to homogeneity from Midkine-expressing CHO cells using heparin-affinity chromatography and G401 cell proliferation assays established that this 7 nM Midkine concentration produced the maximal mitogenic effect.²⁷ The culture medium was changed daily and no antibiotic or fungicide was present throughout the culture period.

Quantification of Ureteric Bud Branching Growth and Glomerulogenesis in Organ Culture.

Filter-grown kidney rudiments were fixed in 4% paraformaldehyde for 2 hours at room temperature and rinsed twice with ice-cold PBS and were then carefully detached from the transwell membrane. After overnight incubation in 50 mM NH₄Cl at 4°C, kidney rudiments were permeabilized with 0.075% saponin for 1 hour and incubated with 0.2% gelatin for 1 hour at room temperature. Explants were washed twice with neuraminidase buffer (50 mM sodium acetate, 150 mM NaCl, 9 mM CaCl₂, pH 5.5) followed by 3 hours of incubation at 37°C with Vibrio cholera neuraminidase (1 Unit/ml, Sigma). The rudiments were washed twice with PBS and stained for 1 hour at room temperature with 50 µg/ml FITC-coupled peanut agglutinin (PNA), which labels glomeruli, or FITC-coupled dolichos biflorus agglutinin (DBA), which labels the ureteric bud epithelium (Vector). After several washes with ice-cold PBS, kidneys were mounted in glycerol with Vectashield (Vector). The numbers of glomeruli or ureteric bud branches and tips were counted under a fluorescence microscope.

Kidney Growth Measurements.

To assay overall growth of kidney rudiments in culture, metanephric kidneys were separated from 12 day mouse embryos and cultured in the condition described above for 5 days in the presence or absence of 7 nM Midkine. Uncultured E12 kidneys were used as controls and all kidneys selected for use in these assays were matched for the number of branch tips in each group to ensure that comparisons after culture would be valid. For measurements,^9–12 rudiments were solubilised in 0.5 ml of 0.25 M NaOH for 30 minutes at 100°C. The kidney lysates were assayed for protein concentration using the BCA protein assay system (Pierce) and total protein estimated from a calibration curve of bovine serum albumen similarly treated with NaOH.

Immunohistochemistry.

Metanephric kidneys were dissected from CD-1 mouse embryos day 12 and cultured for 5 days in serum-free medium as described above in the absence or presence of 7 nM Midkine. Fetal kidneys and cultured kidney rudiments were fixed in 4% paraformaldehyde in PBS for 2 hours at room temperature and then processed for paraffin sectioning using Histogel (Richard Allan Scientific Cat# HG-400-012). Sections were cut at five microns and slides were air-dried overnight at 37°C. Immunohistochemistical staining was carried out using an avidin-biotin indirect immunoperoxidase technique (Vectastain, Vector). Briefly, endogenous peroxidase activity was blocked using 0.3% H₂O₂ in methanol for 40 minutes and in some cases (e.g., for Pax-2 staining), tissues were permeabilized in 0.1% Triton X-100 for 5 minutes. After 30 minutes of block with 10% normal goat or horse serum, sections were incubated with a primary antibody for 45 minutes, washed 3 times in PBS/Tween for 5 minutes each and then incubated with an appropriate biotinylated secondary antibody for 45 minutes. After 3 washes of 5 minutes each in PBS/Tween, sections were then incubated with peroxidase-coupled ABC reagent for 30 minutes and washed. Development was carried out for 10–30 minutes with aminoethylcarbazole substrate until suitable staining developed. The primary antibodies used were Pax-2 (1:100, Berkeley Antibody Company, cat# LN118519001), N-CAM (1:30, Sigma, cat# C9672) and Midkine (10 µg/ ml, Santa Cruz Biotechnologies, SC# 1398). Western immunoblotting has demonstrated that this anti-C-terminal peptide polyclonal antibody detects only Midkine¹⁸ and not pleiotrophin.

Cell Proliferation and Apoptosis Assays.

The bromodeoxyuridine (BrdU) cell proliferation assay was applied to detect cell growth in kidney organ culture after treatment with Midkine. Mouse E12.5 metanephric kidneys were cultured for 3 days in the presence or absence of 7 nM Midkine in serum-free medium (n = 5 kidneys in each group). The kidneys were washed twice with PBS and labeled at 37°C, 5% CO₂ for 1 hour with 1:1,000 BrdU labeling reagent (Amersham Pharmacia, cat# RPN 201). Following two washes of 3 minutes with PBS, kidney rudiments were fixed in 4% paraformaldehyde in PBS for 3 hours at room temperature. The rudiments were then processed for paraffin sectioning. Sections were cut at five microns and slides were air-dried overnight at 37°C. Sections were incubated with anti-BrdU antibody containing nuclease (Amersham Pharmacia, cat# RPN 202) for 1 hour followed by 1:100 peroxidase-coupled anti mouse IgG (KPL, cat # 074-1806) for 40 minutes at room temperature. After washing, color development was carried out with diaminobenzidine (Sigma, cat# D 4293) as substrate for 30–45 minutes until suitable staining developed.

The TUNEL assay detection (Roche) of apoptotic cells and TOPRO-1 (Molecular Probes) fluorescence for detection of cell death in metanephric mesenchymes⁸ were used to analyze the activity of Midkine as a survival factor. For the TUNEL assay, mouse E12.5 metanephric kidneys were cultured for 5 days in the presence or absence of 7 nM Midkine in serum-free medium (n = 5 kidneys in each group). The kidney rudiments were fixed in 4% paraformaldehyde in PBS for 2 hours at room temperature and then processed for paraffin sectioning using Histogel. After dewaxing and rehydration of the tissue sections, they were incubated with 20 µg/ml proteinase K for 20 minutes at 37°C and rinsed with PBS. Endogenous peroxidase activity was blocked using 3% H₂O₂ in methanol for 10 minutes and tissues were permeabilized with 0.1% Triton X-100 in 0.1% sodium citrate for 2 minutes on ice. DNA fragmentation of apoptotic cells was then labeled and detected by the TUNEL assay according to the manusfacturer's instructions (Roche, cat# 1 684 817). For TOPRO-1 staining, mouse E11.5 mesenchymes (n = 5 for each culture group) were isolated from ureteric buds and grown in culture for 48 hours in the presence or absence of 7 nM MK or 100 ng/ml FGF2. Following culture, mesenchymes were stained with 1µM TOPRO-1 (Molecular Probes, cat# T3602) diluted in Dulbecco's modified Eagle medium (DMEM) medium for 1 hour at 37°C and washed with large volumes of PBS for 30 minutes. Samples were mounted on slides in glycerol, and labeled cells were detected using a fluorescence microscope.

X-gal Staining for LacZ Activity.

BF-2/Lac-Z knock-in C57BL6 male mice (+/−) were mated with normal female mice (+/+) and pregnant mice were sacrificed at E12.5 by CO₂ asphyxiation. Their embryos were aseptically removed and tails were cut from each embryo and placed in a 96-well plate. X-gal staining (described below) was performed for each tail to screen for positive embryos. Metanephric kidneys were dissected from LacZ positive embryos and grown for 6 days in the presence or absence of 7nM MK, at 37°C, 5% CO₂ on 0.4 µm pore size transwells as described above. Kidneys were fixed in 4% paraformaldehyde for 1.5 hours and washed with tissue rinse solution A (0.1 M PO₄ buffer, 2 mM MgCl₂ and 5 mM EDTA, pH 7.3) for 30 minutes and tissue rinse solution B (0.1 M PO₄ buffer, 2 mM MgCl₂, 0.01% Na Deoxycholate and 0.02% NP 40, pH 7.3) for 5 minutes at room temperature. Kidneys were then incubated with 1 mg/ml X-Gal in staining solution (0.1 M PO₄, 2 mM MgCl₂, 0.01% Na deoxycholate, 0.02% NP-40 and 5 mM K₃Fe(CN)₆/K₄Fe(CN)₆) for 1 hour at 37°C. The tissues were then processed for paraffin sectioning using Histogel. Sections were cut at 5 microns and slides were air-dried overnight at 37°C in room air. X-Gal staining (blue color) was observed and analyzed under a microscope.

Results

Midkine Expression During Kidney Development.

Midkine expression was analyzed by immunohistochemical staining at different stages of metanephric kidney development. During the earliest phase of kidney development, Midkine was most abundant in mesenchymal cells of the peripheral subcapsular nephrogenic zone in E11 metanephroi (Fig. 1.A). Midkine was not present in the condensed mesenchyme surrounding the ureteric bud at this stage of development, and only modest staining for Midkine was observed in the ureteric bud epithelial cell plasma membranes. By E12 (Fig. 1B), Midkine staining was most intense in ureteric bud basement membranes (see arrow in (Fig. 1B) as previously demonstrated in E13 mouse kidneys.¹⁹ At later developmental stages, Midkine continued to be most abundant in ureteric bud and collecting duct epithelial cell basement membranes (E14, Fig. 1C and E17 metanephroi, Fig. 1D) as previously reported.¹⁹

Figure 1.

Immunohistochemical analysis of Midkine expression in the developing murine metanephric kidney from E11 to E17. Midkine was mainly detected in the outer rim of uncondensed mesenchyme of the E11 metanephros. (A) Arrow marks the ureteric bud). From E12 until birth (B), E12; (C), E14; (D), E17), Midkine was most intensely localized to ureteric bud and collecting duct basement membranes (B–D, arrows). Scale Bars: (A and B), 100 µm; (C and D), 50 µm.

Midkine Stimulates the Expansion of the Non-Stromal Nephrogenic Mesenchyme Cell Compartment in Cultured Metanephric Kidneys.

Midkine was purified to homogeneity by two rounds of heparin-affinity chromatography from a Midkine-producing recombinant CHO cell line as previously described¹⁸ and titration experiments established that concentrations of 7 nM produced maximal proliferative effects in the G401 cell line.¹⁷ Explanted E12 metanephroi cultured for five days in the presence of 7 nM Midkine in serum-free conditions developed subcapsular lucent regions (arrows, Fig. 2) not seen in control cultures. Subsequent microscopic study of these clear regions in hematoxylin-eosin stained tissue sections showed them to consist of large masses of undifferentiated blastemal cells surrounding ureteric bud ampullae (Fig. 2L and M). In these regions, nephron Anlagen including vesicles, comma-shaped and S-shaped bodies were absent. However, many regions of the kidney revealed normal maturation of renal vesicles with no histological differences compared to control cultures (data not shown).

Figure 2.

Midkine induces specific expansion of the blastemal component of the nephrogenic zone. E12.5 kidneys were isolated and cultured in serum-free medium in the absence (A–E or presence F–J) of 7 Nm Midkine for 5 days. Photomicrographs of kidney rudiments are shown for each day of culture; day 1, A and F; day 2, B and G; day 3, C and H; day 4, D and I; day 5, E and J. Midkine did not inhibit overall organ growth, but induced the formation of large lucent mesenchymal regions in the subcapsular nephrogenic zone (arrowheads, lower panels I and J). (K–M) Midkine stimulates expansion of the undifferentiated mesenchymal cell compartment of the outer nephrogenic zone in cultured metanephric kidneys. E 12.5 mouse kidneys were dissected and cultured with Midkine (L, M and N) and compared to control cultures without Midkine (K) for 5 days. Midkine induced the development of nodular masses of undifferentiated mesenchymal cells (marked by asterisks in L–N), not seen in control metanephroi cultured in serum-free control cultures (K). Scale bars: K, 50 µm; L, 100 µm; M, 50 µm; N, 25 µm.

Since the nephrogenic zone includes BF-2 positive stromal cells and Pax-2 positive nephrogenic mesenchyme cells, we used these markers to distinguish which cell population was affected by addition of Midkine to metanephric kidney organ cultures. As shown in Figure 3, the large blastemal cell aggregates produced by Midkine treatment are composed of BF-2 negative cells (Fig. 3B, C and D) and Pax-2 (Fig. 3F) and N-CAM positive cells (Figure 3H); control metanephroi cultured in the absence of Midkine are shown in Fig. 3A, E and G). Although occasional BF-2 positive stromal cells are seen infiltrating the mesenchymal aggregates (Fig. 3D), the preponderance of cells have the BF-2 negative, Pax-2 positive, Wnt-4 positive and N-CAM-positive cell phenotype of nephrogenic mesenchymal cells in the earliest stage of mesenchymal to epithelial conversion.

Figure 3.

Midkine promotes the expansion of BF-2 negative, Pax-2 and NCAM positive nephrogenic mesenchyme progenitor cells in cultured metanephroi. (A–D) E12.5 metanephric kidneys microdissected from BF-2/lacZ knock-in mice embryos were cultured for 5 days. (A) control culture; (B–D) Midkine (7 Nm) addition. Kidney rudiments were stained with X-Gal to identify BF-2 positive stromal cells without (A and B) or with (C and D) eosin counterstaining. Midkine-induced nodular masses of mesenchymal cells were composed of BF-2 negative cells (asterisk). Scale bars: (A), 50 µm; (B), 50 µm; (C), 25 µm; (D), 25 µm. (E–H) Explanted E12.5 metanephroi were cultured for 5 days. (E and G), control cultures; (F and H), Midkine (7 Nm) treated cultures. Immunohistochemical detection of Pax-2 (E and F) showed Pax-2 positive cells in the ureteric bud and condensing nephrogenic mesenchyme in control kidneys and in large mesenchymal cell masses induced by Midkine treatment (marked by asterisk in F). NCAM immunostaining (G and H) revealed NCAM positive cells in mesenchymal cell aggregates around ureteric bud tips of untreated kidneys (G) and in large masses of mesenchymal cells induced by Midkine treatment (marked by asterisk in H). Scale bars: (E), 50 µm; (F), 50 µm; (G), 25 µm; (H), 25 µm.

Midkine Inhibits Nephrogenesis in Metanephric Kidney Organ Culture.

E11.5–12.5 kidney rudiments containing both the ureteric bud and the surrounding metanephric mesenchyme were cultured in serum-free medium in the absence or presence of 7 nM Midkine for 2 to 6 days. Whole-mounts of kidney rudiments were stained with FITC-coupled peanut agglutinin (PNA), which labels glomeruli, or FITC-coupled dolichos biflorus agglutinin (DBA), which labels the ureteric bud epithelium to allow quantitative analysis of the number of glomeruli and ureteric bud branches formed in culture. As shown in Figure 4, ureteric bud branching morphogenesis resulted in an arborized tubular structure in both control (Fig. 4A) and Midkine-treated (Fig. 4B) cultured metanephroi although Midkine appeared to have an inhibitory effect on this process. By quantitative analysis, Midkine significantly inhibited ureteric bud branching growth in both cultured E 11.5 metanephroi (see Fig. 4C, control culture ureteric branch points 3.8 ± 0.31 versus Midkine treated branch points 1.9 ± 0.23, p < 0.0002; control culture ureteric bud tips 5.3 ± 0.49 versus Midkine-treated branch tips 2.9 ± 0.23, p < 0.0003 (data shown are mean ± SEM)) and in cultured E 12.5 metanephroi (see Fig. 4D, control culture ureteric branch points 28 ± 1.1 versus Midkine treated branch points 16 ± 0.07, p < 0.001; control culture ureteric bud tips 33 ± 1.4 versus Midkine treated branch tips 19 ± 1.1, p < 0.001 (data shown are mean ± SEM)). This suppressive effect of Midkine on branching growth of the ureteric bud in cultured metanephric kidneys contrasts to the stimulatory effect of the Midkine-related heparin-binding growth factor pleiotrophin on isolated ureteric buds grown in defined culture media.²⁸

Figure 4.

Midkine inhibits ureteric bud branching morphogenesis and nephron formation in cultured metanephric kidneys. (A and B) E12.5 metanephroi were cultured for 2 days without (A) or with (B) MK addition (7 Nm) and stained with FITC-coupled Dolichos biflorus lectin to visualize ureteric buds and the developing collecting duct system. Midkine qualitatively reduced branching morphogenesis of the developing ureteric bud (compare B to A). For quantitative analysis, the number of branch points or tips from each kidney was determined by fluorescence microscopy on Dolichos biflorus lectin stained metanephroi. (C) Midkine significantly reduced the number of ureteric bud branch points and tips in cultured E11.5 metanephroi (p values using Tukey-Kramer multiple comparison test: p < 0.0002 for comparison of branch point number between control and Midkine-treated cultures and p < 0.0003 for comparison of branch tip numbers between control and Midkine-treated cultures (n = 6 no culture control metanephroi; n = 6 untreated and n = 8 Midkine-treated cultured metanephroi), data shown are mean ± SEM). (D) Midkine significantly reduced the number of ureteric bud branch points and tips in cultured E12.5 metanephroi (p values using Tukey-Kramer multiple comparison test: p < 0.001 for comparison of branch point number between control and Midkine-treated cultures and p < 0.001 for comparison of branch tip numbers between control and Midkine-treated cultures (n = 6 no culture control metanephroi; n = 5 for untreated and Midkine treated cultured metanephroi), data shown are mean ± SEM). (E and F) E12.5 metanephroi were cultured and glomerular number determined by peanut lectin staining. (E) control culture, (F) Midkine (7 Nm) treated cultures. (G) Quantitative analyses of glomerular number in E11.5 or E12.5 kidneys after culture. MK addition inhibits the formation of glomeruli in both cultured E11.5 and E12.5 metanephroi. Data shown are mean ± SEM. p values (2 tailed t test): E11.5 comparison Midkine vs control p < 0.0003 (n = 6 control and n = 6 Midkine treated metanephroi). E12.5 comparison Midkine vs control p < 0.001 (n = 19 control and n = 18 Midkine treated metanephroi). (H) E12.5 mouse metanephroi were cultured for 5 days in the presence or absence of MK (7 Nm). The protein content of metanephroi was determined prior to culture, and after culture as shown. Data shown are mean ± SE. p values using Tukey-Kramer multiple correction test: Midkine vs control p = 0.86; Midkine vs no culture p < 0.01; no Midkine control culture vs no culture p < 0.01 (n = 15 metanephroi in no culture group, n = 20 metanephroi in control culture group and n = 15 in Midkine-treated culture group).

Midkine also significantly reduced the formation of glomeruli in cultured E11.5 and E 12.5 kidney rudiments (see Fig. 4E, control versus Fig. 4F, Midkine treated cultures) and this was confirmed by quantitative analysis (see Fig. 4G, E11.5 control culture glomerular number 11 ± 0.82 versus Midkine treated glomerular number 5.5 ± 0.62 [data shown are mean ± SEM, p < 0.0003]; E12.5 control culture glomerular number 61 ± 2.1 versus Midkine treated glomerular number 39.6 ± 1.89 [data shown are mean ± SEM, p < 0.0001]). In contrast to these inhibitory effects on glomerulogenesis and ureteric bud branching growth, Midkine did not inhibit overall organ growth as judged by protein content of metanephroi after 5 days of culture in the presence or absence of this polypeptide factor (see Fig.4H, control culture kidney protein content 30.5 ± 1.1 versus Midkine treated kidney protein content 31.1 ± 2.8, p = 0.86 [data shown are mean µg/g protein/kidney ± SEM]).

Midkine is a Survival and Mitogenic Factor in the Metanephric Kidney.

Since Midkine protein is expressed from the earliest stage of metanephric kidney development in vivo, and is a potent mitogen for many cell types,¹⁷ we reasoned that the Midkine-induced expansion of the condensed nephrogenic mesenchymal cell compartment might result from mitogenic stimulation rather than Midkine inhibition of mesenchymal to epithelial conversion. E12.5 kidney rudiments cultured for 3 days in the presence or absence of 7 Nm midkine were pulsed with BrdU and proliferating cells were detected by in situ BrdU cell staining. Midkine treatment led to increased numbers of proliferating BrdU positive blastemal cells in aggregates found in the nephrogenic zone (Fig. 5C and D compared to control cultures Fig. 5A and B). The proliferative effect of Midkine was confirmed by quantitative analysis (Fig. 5E). BrdU positive cells in control (12.3 ± 0.057 cells/field) versus Midkine-treated cultures (22.8 ± 0.814 cells/field; p < 0.0001; data shown are mean ± SEM). We next examined whether Midkine prevented programmed cell death in the metanephric kidney culture. E12.5 kidney rudiments containing both the ureteric bud and metanephric mesenchyme were cultured for 5 days in the presence or absence of 7 Nm midkine. Midkine addition to the culture significantly reduced the number of apoptotic cells in the nephrogenic zone (see Fig. 5F for comparison between TUNEL positive cells in control) (11.3 ± 1.02 labelled cells/field) versus Midkine treated cultures (7.64 ± 0.821 cells/field); p = 0.0059 (data shown are mean ± SEM). Within blastemal cell aggregates, very few apoptotic cells were identified by TUNEL staining (Fig. 5H).

Figure 5.

(Right). Midkine promotes cell proliferation and suppresses cell death in cultured metanephroi. Explanted E12.5 metanephroi cultured for 3 days. (A–E) Midkine stimulates cell proliferation in cultured metanephroi. (A and B), control cultures; (C and D), Midkine (7 Nm) treated cultures. BrdU labeling was carried out for 1 hour of culture and metanephroi were then analyzed by immunohistochemistry for detection of incorporated BrdU. Scale bars: (A), 100 µm; (B), 50 µm; (C), 100 µm; (D), 50 µm. (E) Quantitative analyses performed by counting the positive cell number of the nephrogenic zone per 40X objective microscope field from control or Midkine-treated groups. Data shown are mean ± SEM {n = 92 fields (control, n = 5 metanephroi) or n = 101 fields (Midkine-treated n = 5 metanephroi)} p < 0.001 for difference between Midkine and control values. (F–K) Midkine suppresses apoptosis in the nephrogenic mesenchyme in cultured E 12.5 mouse kidneys (3 days). (F) Quantitative analyses of the anti-apoptotic effect of Midkine; the TUNEL positive cell number per 40X objective microscope field from control or Midkine-treated group. Data shown are mean ± SEM {n = 47 metanephroi in each group} p = 0.0059 for difference between Midkine-treated and untreated control group. G:TUNEL staining in untreated control organ culture. H: TUNEL staining in Midkine (7 Nm) treated cultures. Scale bars: (G), 50 µm; (H), 50 µm. (I–K) Midkine and FGF2 suppress cell death in isolated mesenchymes. Isolated E11.5 metanephric mesenchymes were cultured for 48 hours after removal of the ureteric buds without (I) or with addition of 7 Nm of MK (J) or 100 ng/ml of b-FGF (K) and then stained with TOPRO-1 for 1 hour prior to fluorescence microscopy. Midkine (J) and FGF-2 (K) suppressed cell death to the same extent. Scale bars: (I), 100 µm; (J), 100 µm; (K), 100 µm. (n = 5 mesenchymes analyzed for each set of culture conditions).

Since the ureteric bud is necessary for the survival of the metanephric mesenchyme as established by transfilter experiments,^9,29 we investigated whether Midkine could act as a direct anti-apoptotic factor in isolated metanephric mesenchymes in culture after ureteric bud removal. Isolated metanephric mesenchymes were cultured for 48 hours in serum-free conditions with or without the addition of either 100 ng/ml FGF-2 or 7 Nm Midkine (Fig. 5I–K). Cell death was detected by TOPRO-1 staining (as in ref. 8). After 48 hours of culture, control explants displayed extensive cell death as indicated by TOPRO-1 staining (Fig. 5I). In contrast, both FGF-2 (Fig. 5J) and Midkine (Fig. 5K) efficiently suppressed cell death.

Discussion

The identification of the signaling mechanisms governing the survival and proliferation of the epithelial progenitor lineages during metanephric kidney development has lagged behind recent dramatic progress in defining the molecular regulation of branching morphogenesis of the ureteric bud. It is, however, evident that these processes must be coupled and regulated in an interdependent fashion to make possible the generation of the complex tissue architecture of the kidney, and to ensure coordinate growth control of the ureteric bud, nephrogenic mesenchyme and stroma. In studies reported here, we have shown that Midkine addition to metanephric organ cultures produces overgrowth of the induced nephrogenic mesenchyme due to a combination of anti-apoptotic and mitogenic effects, and suppression of ureteric bud branching growth. These results illustrate for the first time that the growth and survival of the nephrogenic mesenchyme can be independently regulated from the stromal mesenchyme.

BMP7 and FGF family members have previously been shown to stimulate the survival and proliferation of epithelial progenitor cells, and stromal cells by multiple investigators.^8,9,30 EGF uniquely stimulates stromal cell expansion at the expense of the epithelial progenitor cells of the nephrogenic mesenchyme²⁹ and suppression of apoptosis in the metanephros.³¹ In contrast, our studies show that Midkine treatment produces overgrowth of the nephrogenic mesenchyme, without simultaneously leading to expansion of the stromal mesenchyme. Since Midkine is expressed from the very earliest stages of metanephric kidney development in the uninduced outer mesenchyme layer in E11 metanephroi, we speculate that it may play a role in maintaining epithelial progenitor cell populations during organogenesis.

Midkine's mitogenic effects have been extensively studied in renal tumor cell lines^17,18 and recent studies have begun to define the receptors and signaling pathways involved in its many observed biological functions.^{17,18,32–38} In the developing kidney, further studies will be needed to determine if any of the proposed cellular receptors mediate the biological effects reported here. It could also be that Midkine's actions in the fetal kidney depend on additional growth factors as reported in studies showing a positive interaction between FGF2 and Midkine in craniofacial development.^19,39 Multiple FGF family members have been implicated in renal development including FGF8, FGF7 and FGF28,^14,40,41 and one or more of these factors might modulate the mitogenic effects of Midkine shown in our studies.

Since the ureteric bud has long been recognized to play an essential role in the survival of the metanephric mesenchyme^5,6,29,42 and is the source of many secreted survival and mitogenic factors,^7,9 we analyzed whether the anti-apoptotic effects of Midkine depend on the presence of the ureteric bud. These studies show that Midkine promotes the survival of isolated metanephric mesenchymes as effectively as FGF2. Taken together these results suggest that Midkine localization to the outer mesenchymal cells of the E11 metanephros and to the ureteric bud ampullary basement membranes in E12 and more mature embryos might be an important factor in the regulation of the survival and proliferation of the adjacent nephrogenic mesenchyme.

One possible mechanism for the Midkine-induced expansion of the nephrogenic mesenchyme is inhibition of mesenchymal to epithelial conversion. For instance, targeted inactivation of the BF-2 winged-helix forkhead gene⁴³ blocks mesenchymal to epithelial conversion and produces localized masses of undifferentiated nephrogenic mesenchyme as does the addition of blocking antilaminin antibodies⁴⁴ to isolated mesenchyme in transfilter induction experiments. The addition of Leukemia Inhibitory Factor (LIF) to cultured metanephroi⁴⁵ produces overgrowth of the nephrogenic mesenchyme which was originally attributed to a blockade of mesenchymal to epithelial conversion. More recently however, it has been found that LIF does not block, but rather promotes, mesenchymal to epithelial conversion in cultured rat metanephric mesenchymes.⁴⁶ In recent experiments designed to determine the effects of Midkine on tubulogenesis in treated mouse metanephric mesenchymes in vitro, Midkine has been shown to have no inhibitory effects on mesenchymal to epithelial conversion (Barasch J, personal communication). These results, combined with the observed mitogenic and anti-apoptotic effects of Midkine on the nephrogenic mesenchyme, provide support for the conclusion that the expansion of the nephrogenic mesenchyme produced by Midkine does not result from an inhibition of mesenchymal-epithelial conversion.

Over the five decades since the development of the metanephric organ culture system,⁴ it has become apparent that important differences exist between in vitro and in vivo kidney development. With regard to Midkine's role in kidney organogenesis, targeted inactivation of the Midkine gene does not produce major abnormalities of genitourinary development in vivo²⁴ while the addition of anti-Midkine neutralizing antibodies to metanephric organ cultures has been shown to inhibit metanephric kidney development in vitro.²⁵ In our studies, highly purified recombinant native Midkine demonstrate that this factor promotes overgrowth of the nephrogenic mesenchyme. These observations suggest that Midkine plays an essential role in the metanephric organ culture model system in ensuring the survival and proliferation of the epithelial progenitor cells of the nephrogenic mesenchyme. With the more favorable conditions for organ growth and development in vivo, Midkine's functions may be rendered less critical by other functionally homologous growth factors.

Our studies establish Midkine as a new member of a growing class of signaling factors (including BMP and FGF family members) that can provide survival and mitogenic stimuli to the nephrogenic mesenchyme, although Midkine uniquely produces expansion of nephrogenic mesenchyme progenitor cells without expanding the stromal compartment. Midkine's suppression of ureteric bud branching growth and simultaneous expansion of the nephrogenic mesenchyme suggests the existence of molecular mechanisms that can couple the relative growth rates of these two tissue compartments, a key requirement for normal kidney organogenesis.