Secreted Frizzled-related proteins can regulate metanephric development

Kiyoshi Yoshinoa,*, Jeffrey S. Rubinb, Kathleen G. Higinbothama, Aykut UÈ renb,Vasiliki Anestb, Sergei Y. Plisova, Alan O. Perantonia

aLaboratory of Comparative Carcinogenesis, Building 538/Room 205E, National Cancer Institute, Frederick, MD 21702, USAbLaboratory of Cellular and Molecular Biology, Building 37/Room 1D28B, National Cancer Institute, Bethesda, MD 20892, USA

Received 26 September 2000; received in revised form 20 December 2000; accepted 23 December 2000

Abstract

Wnt-4 signaling plays a critical role in kidney development and is associated with the epithelial conversion of the metanephric mesench-

yme. Furthermore, secreted Frizzled-related proteins (sFRPs) that can bind Wnts are normally expressed in the developing metanephros, and

function in other systems as modulators of Wnt signaling. sfrp-1 is distributed throughout the medullary and cortical stroma in the

metanephros, but is absent from condensed mesenchyme and primitive tubular epithelia of the developing nephron where wnt-4 is highly

expressed. In contrast, sfrp-2 is expressed in primitive tubules. To determine their role in kidney development, recombinant sFRP-1, sFRP-2

or combinations of both were applied to cultures of 13-dpc rat metanephroi. Both tubule formation and bud branching were markedly

inhibited by sFRP-1, but concurrent sFRP-2 treatment restored some tubular differentiation and bud branching. sFRP-2 itself showed no

effect on cultures of metanephroi. In cultures of isolated, induced rat metanephric mesenchymes, sFRP-1 blocked events associated with

epithelial conversion (tubulogenesis and expression of lim-1, sfrp-2 and E-cadherin); however, it had no demonstrable effect on early events

(compaction of mesenchyme and expression of wt1). As shown herein, sFRP-1 binds Wnt-4 with considerable avidity and inhibits the DNA-

binding activity of TCF, an effector of Wnt signaling, while sFRP-2 had no effect on TCF activation. These observations suggest that sFRP-1

and sFRP-2 compete locally to regulate Wnt signaling during renal organogenesis. The antagonistic effect of sFRP-1 may be important either

in preventing inappropriate development within differentiated areas of the medulla or in maintaining a population of cortical blastemal cells

to facilitate further renal expansion. On the other hand, sFRP-2 might promote tubule formation by permitting Wnt-4 signaling in the

presence of sFRP-1. Published by Elsevier Science Ireland Ltd.

Keywords: Embryonic induction; Secreted Frizzled-related proteins; Kidney; Tubulogenesis; wnt

1. Introduction

The morphogenesis of kidney requires reciprocal interac-

tions between the metanephric mesenchyme and ureteric

bud (SaxeÂn, 1987). It is a relatively unique developmental

process involving the epithelial conversion of blastemal

stem cells within the metanephric mesenchyme to form

the tubular epithelia of the nephron. There appear to be at

least two distinct stages in this development based upon

growth factor involvement: ®rst, compaction or condensa-

tion of the mesenchyme, which can be induced with ®bro-

blast growth factor (FGF2; Perantoni et al., 1995); and then,

epithelial conversion of the mesenchymal condensates,

which requires additional soluble factors (Karavanova et

al., 1996; Barasch et al., 1999; Plisov et al., 2001). The

complete process has been characterized by speci®c

morphogenetic structures that arise following induction,

i.e. condensed mesenchyme, pretubular aggregate of polar-

ized cells, comma-shaped body, S-shaped body, and

segmented tubule/nephron. While these events are induced

by soluble signaling factors released from an apposing

tissue, the ureteric bud, (Perantoni et al., 1995; Karavanova

et al., 1996; Barasch et al., 1997), epithelial conversion is

thought to be mediated in the induced mesenchyme by a

secreted patterning molecule derived from the mesenchyme

itself, namely wnt-4 (Stark et al., 1994; Kispert et al., 1998).

The wnt gene family consists of at least 16 different

conserved glycoproteins that regulate pattern formation

during embryogenesis (reviewed by Dale, 1998). Results

from expression and gene targeting studies indicate that

Wnt-4 plays a critical role in renal tubule formation (Stark

et al., 1994). During the preliminary phases of morphogen-

esis, it is localized to mesenchymal condensates and primi-

tive epithelia derived from the metanephric mesenchyme,

while other kidney-expressed wnt molecules, e.g. wnt-7b or

wnt-11, are detected only in the ureteric bud or interstitial

stroma (Kispert et al., 1996; Karavanova et al., 1996). Gene

Mechanisms of Development 102 (2001) 45±55

0925-4773/01/$ - see front matter Published by Elsevier Science Ireland Ltd.

PII: S0925-4773(01)00282-9

www.elsevier.com/locate/modo

* Corresponding author. Tel.: 11-301-846-1242; fax: 11-301-846-5946.

E-mail address: yoshino@mail.ncifcrf.gov (K. Yoshino).

targeting of wnt-4 results in the formation of small undiffer-

entiated renal rudiments, which show no evidence of tubu-

logenesis. In addition, the co-cultivation of metanephric

mesenchyme with cells expressing wnt-4 induces tubule

formation, although other ectopically-expressed Wnt mole-

cules are equally ef®cient (Kispert et al., 1998).

Wnt signaling is initiated in mammals by interaction of

the Wnt protein with a homolog of the Drosophila gene

family of frizzled receptors (Chan et al., 1992; Wang et

al., 1996). These membrane-bound proteins consist of a

lengthy extracellular portion with a cysteine-rich domain

(CRD), seven putative transmembrane segments, and an

intracellular tail. Signaling is thought to be dependent

upon the interaction of the Wnt ligand with the CRD

(Hsieh et al., 1999). In addition to the membrane-bound

forms of Frizzled, several naturally occurring genes that

encode soluble homologous proteins have also been identi-

®ed. These secreted Frizzled-related proteins (sFRPs) lack

the seven transmembrane segments, but have a Frizzled-like

CRD. Many of them have been shown to bind Wnt proteins,

and can function as antagonists of Wnt activity (Moon et al.,

1997; Rattner et al., 1997; Finch et al., 1997; Leyns et al.,

1997; Wang et al., 1997). The role of the sFRPs in organo-

genesis remains unknown, although there is evidence that

they can modulate Wnt signaling during embryogenesis.

Notably, Frzb-1 or sFRP-3 and sFRP-1 can inhibit Wnt-

dependent axis duplication in Xenopus embryos (Finch et

al., 1997; Leyns et al., 1997; Wang et al., 1997). More

recently, the inhibitory effect of sFRP-1 on Wnt-dependent

axonal remodeling has been reported (Hall et al., 2000).

At least three different sFRP family members thus far

described have been implicated in renal organogenesis,

and each family member shows a characteristically distinct,

and often complementary, pattern of expression, involving

stroma or epithelia (Leimeister et al., 1998; Lescher et al.,

1998). To determine a possible function for these proteins in

renal organogenesis, we have evaluated the ability of sFRP-

1 and sFRP-2 to modulate development in cultures of meta-

nephric tissues. In these studies, sFRP-1 blocked tubule

formation in metanephroi and isolated induced metanephric

mesenchymes. Furthermore, sFRP-2 partially rescued tubu-

logenesis from inhibition by sFRP-1. Finally, sFRP-1 was

found to interact directly with Wnt-4 and to inhibit the

DNA-binding activity of T cell factor (TCF) in explanted

metanephric mesenchymes.

2. Results

2.1. Expression of sfrp-1 and sfrp-2 in rat metanephros

To determine expression patterns of sfrp-1 and sfrp-2 in

rat metanephroi, in situ hybridization was performed using

thin sections from 16- and 19-dpc embryos, that correspond

developmentally to 14- and 17-dpc mouse embryos. sfrp-1

expression in the developing rat kidney is similar, but not

identical to that described for the mouse (Lescher et al.,

1998). In the 16-dpc rat kidney, sfrp-1 transcripts were

most pronounced around the ureter, but high levels were

also noted in all stromal elements, including the interstitial

populations of the medulla and the outer layer of the cortex

(Fig. 1A±C). Epithelial structures of nephron or duct consis-

tently failed to express this molecule. Neither was it

expressed in condensed induced mesenchyme. In 19-dpc

kidneys, a similar pattern was maintained, except that

medullary expression was further increased (data not

shown), and stromal elements remained the principal site

of sfrp-1 expression. Epithelial structures showed no detect-

able levels of expression. Sections probed with sense tran-

scripts of sfrp-1 or sfrp-2 yielded no signal (data not shown).

In 16-dpc kidneys, sfrp-2 expression occurred most

prominently in the mesenchymal tissue surrounding the

developing ureter (Fig. 1D,E), which overlaps with one

area noted for sfrp-1 expression. In other regions, the

expression was complementary to that for sfrp-1. It was

found in pretubular condensates (polarized epithelia)

beneath the tip of the ureteric bud and in newly formed

epithelia of the nephron (comma- and S-shaped bodies),

but not in the bud itself or the condensed mesenchyme at

the bud tip (Fig. 1F). In 19-dpc rudiments (data not shown),

expression occurred predominantly in the cortical nephro-

genic zone in the same structures just described; however, it

was diminished in more mature structures positioned near

the medullary region. Glomeruli did not express sfrp-2.

Therefore, sfrp-2 serves as an appropriate marker for

newly formed epithelia in the metanephros.

2.2. Effects of sFRPs in cultures of metanephroi

To evaluate the effects of sFRPs in the metanephros,

isolated rat metanephroi (13-dpc) were cultured for 5 days

in medium containing 10% serum in the presence or absence

of sFRPs. After ®xation, epithelial components (both the

ureteric bud and newly formed tubule) were labeled with

anti-E-cadherin (Rhodamine). The speci®city of this bind-

ing has been shown previously (Vestweber et al., 1985). The

ureteric bud was also stained with Dolichos bi¯orus (DB)-

lectin (FITC). The speci®city of DB-lectin binding has been

demonstrated previously (Laitinen et al., 1987). The merged

images distinguish newly formed tubules (red) from ureteric

bud (yellow). In each treatment, three metanephroi were

placed on the ®lter, and the experiment was repeated at

least twice with similar results. Representative cultures are

shown in Fig. 2. In control cultures, the bud had branched

repeatedly and extensive tubule formation occurred at the

bud tips (Fig. 2A,B). sFRP-2 treatment did not change these

phenotypes (data not shown). In the presence of 5 mg/ml of

sFRP-1, although tubules were detected, bud growth and

tubule formation were markedly inhibited (Fig. 2C,D).

This inhibitory effect of sFRP-1 was partially opposed by

the addition of sFRP-2 (10 mg/ml; Fig. 2E,F). A higher

concentration of sFRP-1 (10 mg/ml) completely blocked

K. Yoshino et al. / Mechanisms of Development 102 (2001) 45±5546

both bud growth and tubulogenesis of the mesenchyme (Fig.

2G); however, the addition of sFRP-2 (10 mg/ml) rescued

some branching and tubular morphogenesis, suggesting that

sFRP-2 opposes the effect of sFRP-1 on Wnt-4 signaling.

2.3. Effect of sFRP-1 on morphogenesis of metanephric

mesenchyme

To further de®ne the ability of the sFRP-1 protein to

regulate molecular events in tubulogenesis, our well-char-

acterized model system of explanted non-induced meta-

nephric mesenchymes was used (Perantoni et al., 1995;

Karavanova et al., 1996; Karavanov et al., 1998). Mesench-

ymes were separated from the ureteric bud prior to induction

and then placed on porous Nuclepore ®lters. When explants

are incubated with an inductive medium (IM), which

includes an aliquot of cell culture medium conditioned by

a cell line derived from rat ureteric bud in combination with

basic FGF2, they can be induced to differentiate completely,

forming tubules and avascular glomeruli (Karavanova et al.,

1996). To test the effects of sFRP-1 on this differentiation,

explant cultures were exposed to IM with or without various

amounts of sFRP-1 (1.25±10 mg/ml). In cultures treated

with IM alone, explants were replete with tubules after 72

h (Fig. 3A). Under similar conditions, but with the addition

of 10 mg/ml of sFRP-1, tubule formation was completely

blocked without any apparent effect on the morphological

condensation of the mesenchyme (Fig. 3B). Epithelial struc-

tures were totally absent from these cultures. This could not

be attributed to toxicity, since there was no evidence of

necrosis in explanted tissues and numerous mitotic ®gures

were apparent. Of 20 explants exposed to IM, all 20 showed

extensive tubular development, while 20 additional explants

exposed to IM plus sFRP-1 contained no demonstrable

tubules. At 5 mg/ml of sFRP-1, an inhibitory effect on tubu-

logenesis was apparent but diminished.

2.4. Expression of early marker wt1 in sFRP-1-treated

metanephric mesenchymes

Since the condensation of the mesenchyme, an early

morphogenetic event in nephrogenesis, did not seem to be

affected by sFRP-1 treatment, we evaluated the explants for

the expression of wt1, a transcription factor that is up-regu-

lated in the condensed mesenchyme. Gene targeting studies

have shown that the expression of this gene is critical to the

maintenance of the metanephric mesenchyme and its

condensation (Kreidberg et al., 1993). In explants, high

levels of wt1 expression were observed in both sFRP-1

untreated (Fig. 3C,E) and treated (Fig. 3D,F) cultures. In

Fig. 3C,E, sections of explants included both condensed

and tubular areas, and expression was detected in the

condensed, but not mature, tubular regions as occurs in

vivo. In sFRP-1-treated explants, high levels of wt1 were

K. Yoshino et al. / Mechanisms of Development 102 (2001) 45±55 47

Fig. 1. Expression of sfrp-1 and sfrp-2 in 16-dpc rat kidney by in situ hybridization. (A) Bright®eld; (B), dark®eld; and (C), enlarged bright®eld images of sfrp-

1 expression. sfrp-1 is distributed throughout the stromal elements, both in the medulla and the outermost region of the cortex. Especially high levels of

expression are observed surrounding the ureter, while all epithelial elements show no obvious expression. (D) Bright®eld; and (E), dark®eld images of sfrp-2

expression; (F), enlarged area from (D). Expression is prominent in the mesenchymal tissue surrounding the ureter and in newly formed epithelia. (C) Enlarged

region shows expression in a pretubular aggregate (arrow) beneath the tip of the ureteric bud (ub), but not in adjacent condensed mesenchyme. Bars, 50 mm.

noted throughout the large condensate, suggesting that treat-

ment with this protein did not in¯uence the early events of

differentiation.

2.5. Expression of late markers, lim-1, sfrp-2 and E-

cadherin in sFRP-1-treated metanephric mesenchymes

To characterize the effect of sFRP-1 on late morphoge-

netic events, which include epithelial conversion and tubule

formation, we analyzed the expressions of lim-1 and sfrp-2.

The homeobox gene, lim-1, plays a signi®cant role in the

development of the urogenital tract, as observed in null

mutants which fail to form either kidneys or gonads (Shaw-

lot and Behringer, 1995). It is expressed in the developing

metanephric mesenchyme, ®rst in pretubular condensates

which consist of polarized epithelia, and subsequently in

primitive tubules, but is absent from non-induced mesench-

yme or mesenchymal condensates (Karavanov et al., 1998).

Thus, its expression is associated only with developing

epithelia. In explants treated with IM, extremely high levels

of expression, which are prominent even in bright®eld

photomicrographs (Fig. 3G) due to the heavy distribution

of emulsion granules, were detected throughout the tissue

(Fig. 3I). In explants treated with IM plus sFRP-1, a punc-

tate pattern was apparent in bright®eld (Fig. 3H), and the

overall level of expression appeared to be markedly reduced

(Fig. 3J) relative to IM-treated explants.

Using sfrp-2 as a marker for tubulogenesis in explants,

intense expression was demonstrable throughout the form-

ing epithelia (Fig. 3K,M). Furthermore, the expression

occurred in pretubular aggregates, but not in areas of

condensed mesenchyme. In cultures treated with sFRP-1

protein, the expression of sfrp-2 was weak and approached

background levels (Fig. 3L,N). No foci of expression were

apparent in these cultures.

To provide further evidence for the inhibition of epithe-

lial conversion by sFRP-1, E-cadherin expression, which is

widely used as an epithelial marker, was analyzed by immu-

nohistochemistry. E-cadherin is expressed in newly formed

tubules of the metanephros and its expression pattern has

been described in detail (Vestweber et al., 1985; Cho et al.,

1998). In IM-treated explants, E-cadherin expression was

clearly demonstrated in tubular structures (Fig. 3O), while

its expression was absent in explants treated with IM plus

sFRP-1 (Fig. 3P).

Thus, in light of our ®ndings for lim-1, sfrp-2 and E-

cadherin expression, it appears that sFRP-1 blocks not

only the morphogenesis associated with tubulogenesis, but

also molecular events associated with epithelial differentia-

tion.

2.6. Binding of sFRP-1 to Wnt family members

Investigations by others have demonstrated a comple-

mentarity of expression patterns for wnt and sfrp molecules

in tissues for which sFRP antagonism has been described

(Leyns et al., 1997; Wang et al., 1997). Additionally, wnt-4

expression has been reported in induced cultures of meta-

nephric mesenchyme (Karavanova et al., 1996), and thus

far, it is the only Wnt family member associated with the

differentiation of this mesenchyme. To determine if the

sFRP-1 protein binds Wnt-4 and antagonizes its activity,

we ®rst tested the ability of the protein to interact with

Wnt-4 in an ELISA. For this, cell lysates were prepared

from RatB1a ®broblasts transfected with HA-tagged Wnt-

4 expression vector, Wnt-3A expression vector, or an empty

control vector (Shimizu et al., 1997). Using increasing

K. Yoshino et al. / Mechanisms of Development 102 (2001) 45±5548

Fig. 2. Effects of sFRP-1 and combinations of sFRP-1 and sFRP-2 in

cultures of metanephroi. Rat metanephroi (13-dpc) were isolated and

cultured in medium containing 10% fetal bovine serum for 5 days in the:

(A,B), absence; or (C±I), presence of sFRPs. Epithelial structures (ureteric

bud and newly formed tubules) are labeled with anti-E-cadherin (red). The

ureteric bud is also stained with DB-lectin (green). After fusion of these two

colors, the yellowish portion represents ureteric bud and red structures

indicate tubule formation (arrows). (A,B) In control metanephroi, repeated

bud branching and extensive tubule formation near the bud tips are

observed. (C,D) In the presence of 5 mg/ml of sFRP-1, small numbers of

tubules are detected, but bud growth and tubule formation are both inhib-

ited. (E,F) This inhibitory effect of sFRP-1 is opposed by sFRP-2 (10 mg/

ml). (G) At 10 mg/ml, sFRP-1 completely blocks both bud growth and

tubule formation. (H,I) With the addition of sFRP-2 (10 mg/ml), bud

branching and tubulogenesis are restored to a limited extent. Images (B),

(D), (F) and (I) show enlarged views of tubular areas for (A), (C), (E) and

(H), respectively. Bars, 100 mm.

concentrations of lysate, Wnt-4 showed saturable binding

with an avidity greater than that for Wnt-3A (Fig. 4). This

result indicated that sFRP-1 can bind to Wnt-4, supporting

the notion that it could modulate Wnt-4 activity in vivo.

2.7. TCF activity is reduced by sFRP-1 but not by sFRP-2 in

metanephric mesenchyme

Signaling by Wnt leads to an accumulation of cytoplas-

mic beta-catenin, allowing its nuclear translocation as a

complex with a member of the TCF transcription family.

This active complex recognizes speci®c DNA-binding sites

to regulate the transcription of target genes. Although there

is no direct evidence for the accumulation of cytoplasmic

beta-catenin as a result of Wnt-4 signaling, the misexpres-

sion of Wnt-4 or beta-catenin in chicken limb development

caused similar histological and molecular phenotypes,

suggesting that Wnt-4 signal is mediated by a beta-cate-

nin/TCF complex (Hartmann and Tabin, 2000). To deter-

mine if the inhibitory effect of sFRP-1 on tubule formation

is associated with an inhibition of Wnt signaling and TCF

activity, an electrophoretic mobility shift assay (EMSA)

was employed. In mesenchymes cultured in IM, one predo-

minant band was detected using a labeled TCF-binding

oligonucleotide, indicating the formation of a TCF/DNA

complex (Fig. 5, lane 4). Without nuclear extracts, the

band was not observed (lane 1). Pre-incubation of the

nuclear extracts with an excess amount of unlabeled wild-

type TCF-binding probe signi®cantly reduced the band

intensity (lane 2), while an unlabeled mutant-type motif

K. Yoshino et al. / Mechanisms of Development 102 (2001) 45±55 49

Fig. 3. Explant culture of metanephric mesenchyme in IM. Cultures were treated with IM or IM plus sFRP-1 for 3 days. (A) Without sFRP-1, cultures contain

multiple tubular structures showing a typical epithelial morphology. (B) When sFRP-1 was included, the cultures exhibited a condensed mesenchymal

phenotype, but contained no evidence of tubulogenesis. (C±F) wt1 expression in cultured metanephric mesenchymes. (C) Bright®eld; and (E), corresponding

dark®eld images of explants treated with IM. (D) Bright®eld; and (F), corresponding dark®eld images of explants treated with IM plus sFRP-1. High levels of

wt1 expression, a marker for early nephrogenesis, occur in the presence and absence of sFRP-1. (G±J) Expression of the homeobox gene, lim-1, in explanted

metanephric mesenchymes. (G) Bright®eld; and (I), dark®eld images of an explant culture treated with IM. (H) Bright®eld; and (J), dark®eld images of a culture

treated with IM plus sFRP-1. Intense expression of this marker for primitive epithelia is noted only in cultures treated with IM. A punctate pattern with weak

overall expression is apparent in cultures treated additionally with sFRP-1. (K±N) Expression of sfrp-2 in explanted mesenchymes. (K) Bright®eld; and (M),

dark®eld images of explants treated with IM. (L) Bright®eld; and (N), dark®eld images of explants treated with IM plus sFRP-1. In the absence of sFRP-1, newly

formed epithelia express marked levels of sfrp-2 transcripts, but weak or no expression is observed in the presence of sFRP-1. (O,P) E-cadherin immunohis-

tochemistry. (O) E-cadherin is expressed in tubular structures in IM-treated explants; (P), but is absent in IM 1 sFRP-1-treated explants. Bars, 100 mm.

did not alter the intensity (lane 3), showing the speci®city of

this binding. In the presence of sFRP-1, the band intensity

was markedly reduced (lane 5). This provides additional

evidence supporting the hypothesis that sFRP-1 can act as

an antagonist of Wnt-4 signaling in the metanephric

mesenchyme. On the other hand, sFRP-2 did not reduce

the band intensity (lane 6) despite its reported interaction

with Wnt-4 (Lescher et al., 1998).

3. Discussion

Multiple sfrps are expressed at high levels during renal

development in tissue-speci®c patterns, which complement

or overlap the expression of the Wnt family members. In an

effort to better understand how these putative modulators of

Wnt signaling might function in vivo, we evaluated the

ability of sFRP-1 to in¯uence the morphogenesis of the

metanephric mesenchyme, a process shown by gene target-

ing to be dependent upon Wnt-4 signaling (Stark et al.,

1994). In these studies, we have found that sFRP-1 is

capable of interacting with Wnt-4 protein and blocking

morphogenetic and molecular events associated with Wnt-

4 signaling. We also found that sFRP-2 competes with

sFRP-1 to facilitate tubulogenesis and bud branching.

3.1. Comparison of sfrp expression in rat and mouse

Our studies show that sfrp-1 is expressed in the develop-

ing rat kidney in a manner similar to that described for the

mouse (Leimeister et al., 1998). During organogenesis, it

appears early and in a tissue-speci®c pattern, i.e. it is

detected only in stromal elements, including those that

surround the induced tubule-generating blastemal tissues.

With our probe, it is speci®cally absent from any epithelial

structure in the metanephros. This differs somewhat from

the report for the mouse, in which expression was observed

to shift from the stroma to the epithelium of the nephron as

development progressed, since we detected no expression in

the epithelia even at later stages. This may simply indicate

differing sensitivities of probes or experimental conditions,

although it may also re¯ect bona ®de species distinctions.

Conversely, sfrp-2 expression occurs primarily in newly

formed epithelia, with the exception of the tissue immedi-

ately around the ureter, in a pattern that is complementary to

sfrp-1 expression, suggesting distinct functions in renal

development. The pattern of sfrp-2 expression also differs

slightly between species. In the rat, it is clearly demon-

strable in pretubular aggregates, which consist of polarized

epithelia beneath the bud tip, but not in condensed mesench-

yme at the bud tip as observed in mice.

3.2. sFRP-1 inhibits tubule formation and bud growth in

metanephroi

The metanephros in Wnt-4 null homozygotes is de®cient

not only in tubule formation, but also in the extent of

ureteric bud branching in comparison with normal metane-

phroi (Kispert et al., 1996). In our experiments, signi®cant

inhibition of tubule formation and bud growth by sFRP-1

was demonstrated. Although we have not elucidated the

mechanism for its inhibitory effects on the bud, there are

some possible interpretations. Mesenchymal differentiation

and bud growth are driven by reciprocal inductive factors.

Soluble signals from the metanephric mesenchyme, such as

the glial cell line-derived neurotrophic factor (GDNF) and

other(s), are required for bud growth and branching (Qiao et

al., 1999). It is possible that treatment with sFRP-1 reduces

the production of these necessary soluble signaling factors

in the metanephric mesenchyme. Alternatively, sFRP-1 may

inhibit bud growth directly by regulating Wnt signaling in

the ureteric bud. Wnt-7B, which is expressed in collecting

ducts, and Wnt-11, which is expressed at bud tips, would be

likely candidate molecules, although the functional role of

these Wnts in kidney development has not been established.

K. Yoshino et al. / Mechanisms of Development 102 (2001) 45±5550

Fig. 4. Differential binding of sFRP-1 to Wnt family members. (A) Immunoblot analysis of HA-tagged Wnt-3A and Wnt-4 expressed in Rat B1a cells.

Background signal is observed in RatB1a cells infected with retrovirus lacking a cDNA insert. (B) sFRP-1/Wnt ELISA. Wells were coated with sFRP-1 and

incubated with serial dilutions of cell lysates (squares for Wnt-4, triangles for Wnt-3A, and circles for vector control). Bound protein was detected with HA

monoclonal antibody as described in Section 4. The results are the means ^ SD of triplicate measurements from a representative experiment.

3.3. sFRP-1 activity is speci®c to late events in

morphogenesis

Of particular interest are the noted expression pro®les for

markers of kidney development in our explant cultures.

While wt1 seems to be appropriately expressed in the

presence and absence of sFRP-1, lim-1 expression is

decreased in the presence of sFRP-1 and occurs in a punc-

tate manner. Based on its distribution in vivo in pretubular

aggregates and primitive epithelial structures, this observa-

tion suggests that some polarization or epithelial conversion

of mesenchymal cells has occurred in sFRP-1-treated

cultures, but that these cells have failed to expand in

number. While we have not ruled out the possibility that a

higher concentration of sFRP-1 might completely eliminate

the punctate pattern, an alternative interpretation might be

that Wnt-4 does not initiate the epithelial conversion of

metanephric mesenchyme, but rather stimulates the burst

of epithelial cell proliferation that accompanies the conver-

sion process. This is supported by the fact that FGF2 causes

explanted metanephric mesenchyme to condense and

express early markers of development, including wnt-4

(unpublished observation); however, FGF2 is only weakly

inductive for tubule formation (Karavanov et al., 1998).

3.4. sFRP interactions with Wnts

Since Wnt-4 has been shown to mediate events involved

in the epithelial conversion of the metanephric mesenchyme

or in tubule elongation, sFRP-1 might modulate these activ-

ities through a direct interaction with Wnt-4. Indeed, sFRP-

1/Wnt-4 binding was observed in our experiments. sFRP-1

bound Wnt-4 with an avidity greater than that demonstrated

for Wnt-3A, whose dorsal axis-duplicating activity in Xeno-

pus embryos was only marginally affected by concomitant

sFRP-1 expression (Finch et al., 1997). Another possibility

is that our bud cell conditioned medium contains a soluble

Wnt that is responsible for the induction of tubule formation

in our explant system, and that sFRP-1 interaction with

Wnt-4 in the mesenchymes is not the principal mechanism

causing the inhibition of morphogenesis. While Wnt

proteins have been detected in soluble form (Bradley and

Brown, 1995), most often, they associate with extracellular

matrix components upon secretion and are rarely detected in

culture medium from Wnt-expressing cells (Papkoff and

Schryver, 1990). Thus, Wnt-expressing cells themselves

are required for inductive signaling (Herzlinger et al.,

1994). We have attempted to address this issue by depleting

potential Wnts from bud conditioned medium using sFRP-

1-coated plates, but have been unable to detect any loss of

inductive activity in such experiments (unpublished obser-

vation), suggesting that sFRP-1 is not regulating tubulogen-

esis through interaction with a soluble Wnt in conditioned

medium.

Although binding of sFRPs and Wnts is presumably

required to antagonize Wnt signaling, binding per se is

not necessarily suf®cient to prevent this signaling. Frzb-1,

for example, coimmunoprecipitates with Wnt-5A, suggest-

ing that it interacts with the Wnt protein, but it does not

block Wnt-5A signaling in a Xenopus functional assay for

Wnt-mediated axis duplication. However, Frzb-1 both asso-

ciated with and blocked Wnt-1 signaling (Lin et al., 1997).

The sFRPs have been implicated principally as antagonists

of Wnt signaling, perhaps as proteins that can bind and

sequester active Wnt proteins. Frzb and Xwnt-8 are

expressed in complementary patterns during embryogenesis,

and Frzb binds to Xwnt-8. Furthermore, it apparently inacti-

K. Yoshino et al. / Mechanisms of Development 102 (2001) 45±55 51

Fig. 5. Inhibition of TCF DNA-binding activity in sFRP-1-treated induced

metanephric mesenchymes. Isolated metanephric mesenchymes (13-dpc)

were cultured in IM for 3 days with or without sFRPs (10 mg/ml). The

DNA-binding by TCF was analyzed by EMSA as described in Section 4. In

cultures grown in IM, one predominant band (arrow) is detected (lane 4).

Without nuclear extracts, this band is not observed (lane 1). The intensity of

this band is signi®cantly reduced by pre-incubation with the 80-fold excess

amount of unlabeled wild-type oligonucleotide (lane 2), but not with mutant

oligonucleotide (lane 3). In the presence of sFRP-1, the band intensity is

signi®cantly reduced (lane 5). However, sFRP-2 does not diminish the band

intensity (lane 6).

vates Xwnt-8 during gastrulation to promote dorsalization of

the Xenopus embryo (Leyns et al., 1997; Wang et al., 1997)

and prevents the induction of MyoD in a manner similar to

that reported for a dominant-negative form of Xwnt-8

(Hoppler et al., 1996). In the current studies, sFRP-1 was

applied to a well-characterized biological model for tubulo-

genesis of the metanephric mesenchyme, and it completely

blocked tubule formation in isolated mesenchymes and

metanephroi, providing the ®rst evidence for participation

of sFRPs in organogenesis as well. While we cannot prove

that this antagonism results from the direct interaction of

sFRP-1 and Wnt-4, instead of an as yet unidenti®ed Wnt

family member, the circumstances are comparable with

those described above for Frzb and Xwnt-8. Both molecules

are expressed in a complementary manner; they are interac-

tive; a morphogenetic event associated with Wnt-4 activity

(epithelial conversion/tubule formation) is inhibited by

sFRP-1; and sFRP-1 limits or blocks the expression of mole-

cules associated with Wnt-4 signaling, namely sfrp-2

(Lescher et al., 1998), and the activation of TCF.

3.5. Function of sFRPs in vivo

Finally, as mentioned above, direct interaction of sFRP

and Wnt molecules does not necessarily mean that Wnt

signaling will be blocked. Recently, we observed that

sFRP-1 can have a biphasic effect on Wnt-dependent stabi-

lization of Armadillo (Drosophila ortholog of beta-catenin),

potentiating Wnt activity at low concentrations, but inhibit-

ing it at high concentrations, comparable with those used in

the current study (UÈ ren et al., 2000). This raises the possi-

bility that the response to sFRP-1 in vivo could be strictly

dependent on the proximity to its sites of synthesis, i.e. near

abundant sources, high concentrations of sFRP-1 would be

expected to block Wnt signaling, whereas at a distance, low

concentrations might enhance Wnt activity. Such a pattern

has been described for the regulation of Decapentaplegic

(Dpp) by the protein, Short gastrulation (Sog), which

binds Dpp and inhibits its activity on adjacent cells, but

enhances signaling in a distant cell population (Ashe and

Levine, 1999). Thus, sFRP-1 might function to exclude

nephrogenesis from regions of the developing kidney

where it is expressed and promote nephron formation at

distant sites. The role of sFRP-2 in tubulogenesis would

appear to follow a different pattern. Since sfrp-2 is concur-

rently expressed with wnt-4 in the newly formed epithelia, it

is dif®cult to imagine a similar local antagonistic function

such as the one proposed for sFRP-1. It is more likely that a

high local concentration of sFRP-2 would facilitate Wnt-4

activity. This hypothesis is supported by our ®nding that

sFRP-2 competes with sFRP-1 to promote tubulogenesis.

There is a precedent for sFRP-1 and sFRP-2 having oppos-

ing effects on the same cells (Melkonyan et al., 1997),

adding further credence to this hypothesis.

A direct interaction between sFRP-2 and Wnt-4 has been

demonstrated previously (Lescher et al., 1998), and cells

expressing sFRP-2 can reduce the dermomyotome-inducing

activity of Wnt-4 in cultured presomitic mesoderm explants

(Lee et al., 2000). These observations indicate a possible

antagonism of Wnt-4 signaling by sFRP-2. However, as

shown herein, the DNA-binding activity of TCF in Wnt-4

expressing mesenchyme is not reduced by sFRP-2, indicat-

ing that sFRP-2 does not antagonize Wnt-4 signaling in the

metanephric mesenchyme. Thus, the effect of sFRPs on Wnt

signaling may differ in diverse cell populations and may

re¯ect varying af®nities for different frizzled receptor and/

or associated proteoglycan. The latter has an important role

in Wnt-dependent kidney morphogenesis (Davies et al.,

1995) and can modulate sFRP/Wnt interactions (UÈ ren et

al., 2000).

In conclusion, these studies demonstrate that sFRP-1 can

behave as an effective antagonist of Wnt signaling in meta-

nephroi and explant cultures of induced metanephric

mesenchyme. Additionally, it is expressed in a complemen-

tary pattern to wnt-4 during kidney development and physi-

cally interacts with the Wnt-4 protein in vitro. Due to its

extensive expression throughout the metanephros in stromal

tissues and the known ability of cells that express sfrps to

secrete them (Leyns et al., 1997; Wang et al., 1997; Finch et

al., 1997; Rattner et al., 1997), our ®ndings implicate this

molecule in the regulation of tubule formation during orga-

nogenesis. However, its exact function in this process

remains to be established. Perhaps it is critical to the main-

tenance of blastemal cells in the nephrogenic zone of the

metanephric cortex. The expansion of renal tissue is depen-

dent upon differentiation in a proximal±distal pattern with

the least differentiated tissues existing in the cortical region.

The stem cells for subsequent rounds of tubular formation

therefore populate the outer region of the cortex and are

directly juxtaposed to a capsular sFRP-1-secreting stromal

layer. Thus, sFRP-1 may help to maintain the blastemal stem

cell population by preventing Wnt-dependent differentiation.

Due to the high levels expressed around the ureter and within

the medulla of the metanephros, another possibility is that it

prevents ductal/tubular expansion in mature tissues, and as

such, may function as a suppressor of inappropriate differ-

entiation. Since sFRP-2 is also expressed in tissues surround-

ing the ureter, co-expression of these sFRPs may indicate

both positive and negative regulation of ureter growth, but

further study of such a complex relationship will be required.

4. Experimental procedures

4.1. Probes

A 480-bp fragment (sfrp-1) and a 586-bp fragment

(sfrp-2), based upon mouse sequences, were generated by

RT-PCR from 19-dpc rat kidney RNA. The primers used in

PCR were 5 0-GAGGCCGTGCGCGACTCGTG-3 0 (sfrp-1,

forward), 5 0-CTCCTTGTTTTTCTTGTCCC-3 0 (sfrp-1,

reverse), 5 0-CGCTCTTCGCCCCTGTCTGTCT-3 0 (sfrp-2,

K. Yoshino et al. / Mechanisms of Development 102 (2001) 45±5552

forward) and 5 0-CTTGCGGATGCTGCGGGAGATG-3 0

(sfrp-2, reverse). Ampli®ed fragments were cloned into a

pGEM-T Easy Vector (Promega), and sequenced with SP6

primer. Probes for wt1 and lim-1 were prepared as

previously described (Perantoni et al., 1995; Karavanova

et al., 1996).

4.2. In situ hybridization

Tissue explants and resected rat metanephroi (16- and 19-

dpc) were ®xed with 4% paraformaldehyde (PFA) in PBS.

For explants, tissues grown on ®lters were ¯oated for 2 h on

PFA, washed once in PBS, incubated for 10 min in 50%

ethanol in PBS, and ®nally washed twice for 10 min each in

70% ethanol. Metanephroi were ®xed overnight, then

exposed to a similar regimen, except that tissues were incu-

bated in ethanol washes for 1 h. In situ hybridization

analyses were performed using 35S-labeled riboprobes

according to Wilkinson and Green (1990).

4.3. Recombinant sFRP-1 and sFRP-2

A human sfrp-1 NotI-SmaI cDNA fragment (Finch et al.,

1997) was cloned into an XhoI site in a pcDNA3.1 expression

vector, expressed in Madin-Darby canine kidney (MDCK)

cells, and puri®ed by heparin-af®nity chromatography as

described (UÈ ren et al., 2000). For sFRP-2, a rat cDNA frag-

ment was cloned and the same protocol for sFRP-1 was used

for expression and puri®cation.

4.4. Cultures of isolated metanephroi

Rat metanephroi (13-dpc) were surgically extracted and

cultured on ®lters (three/®lter). The culture medium

contained a basal medium (DMEM/F12, 50:50 mix), 10%

fetal bovine serum and sFRP-1 (5 or 10 mg/ml), or sFRP-2

(10 mg/ml) or combinations of these sFRPs. The culture

medium was replaced every 48 h. After 5 days, tissues

were ®xed for subsequent double-staining. All experiments

were repeated at least twice.

4.5. Double-staining of cultured metanephroi

Tissues were ®xed with 100% methanol at 2208C for 10

min, washed in Tris-buffered saline (TBS; pH 7.4), and

blocked in 10% sheep serum/TBS for 5 h. Incubation with

anti-E-cadherin antibody (1:50, Santa Cruz Biotechnology,

Inc.) was performed at 48C for 48 h. Secondary antibody

reactions were done with Rhodamine-conjugated anti-rabbit

antibody (Santa Cruz Biotechnology, Inc.) at a 1:100 dilu-

tion overnight after an 8-h wash in 1% sheep serum/TBS.

After washing in TBS, tissues were stained with FITC-

conjugated DB-lectin (20 mg/ml; Sigma) for 1 h, and then

washed and ®xed with 4% PFA for 30 min. For confocal

analysis, tissues were mounted in 50% glycerol/TBS and

images were collected with a laser-scanning confocal

microscope (Bio-Rad MRC 1024).

4.6. Culture conditions for metanephric mesenchymes

Metanephroi from 13-dpc rats were surgically extracted

and non-induced metanephric mesenchyme was isolated as

previously described (Perantoni et al., 1995). Mesenchymes

(generally four/®lter) were placed on collagen IV-coated

Nucleopore ®lters (0.1 mm pore size), and the ®lters were

¯oated on 1-ml medium in a 12-well tissue culture plate.

The serum-free medium conditions were as previously

described (Karavanova et al., 1996). Brie¯y, a basal

medium (DMEM/F12, 50:50 mix), containing transferrin,

hydrocortisone, triiodothyronine, insulin, prostaglandin

E1, and selenium, was further supplemented with transform-

ing growth factor-a (10 ng/ml), FGF2 (50 ng/ml), and cell

culture medium conditioned by a cell line derived from rat

ureteric bud (33 ml/ml). The cultures were placed in a humi-

di®ed incubator with continuous ¯ow CO2 (5%) and air. The

explants were monitored for 3 days and then ®xed.

4.7. Immunohistochemistry

Sectioned tissue explants were dewaxed, blocked with

methanol and incubated with a 1:400 dilution of the anti-

E-cadherin antibody (Santa Cruz Biotechnology, Inc.).

Signals were detected using the VECTASTAIN Elite

ABC Kit (Vector Laboratories) according to the manufac-

turer's instructions.

4.8. Immunoblotting

RatB1a ®broblasts transfected with HA-tagged retroviral

Wnt-4, Wnt-3A expression vector, or an empty control

vector were grown in DMEM (Gibco-BRL) containing

7.5% calf serum (Gibco-BRL) and 2.5% fetal calf serum

(Colorado Serum Co.; Shimizu et al., 1997). A clonal line

was used that expressed Wnt-3A at a level comparable with

Wnt-4 expression in the corresponding mass culture.

Con¯uent cultures (10-cm dishes) were washed twice with

cold PBS and treated with 0.3 ml lysis buffer (50 mM

HEPES (pH 7.5); 1% Triton X-100; 50 mM NaCl; 5 mM

EDTA; 50 mM EDTA; 50 mM NaF; 6.7 mM sodium phos-

phate; 1 mM sodium vanadate; 10 mg/ml aprotinin; 10 mg/

ml leupeptin; and 1 mM PMSF). Aliquots of lysates (50 mg

total protein) were resolved in 8% SDS PAGE gels and

transferred to Immobilon-P membranes. The membranes

were blocked with 3% non-fat dry milk in 1% Triton X-

100/Tris-buffered saline (TTBS) for 2 h and incubated over-

night with 0.4 mg/ml anti-HA monoclonal antibody (Roche)

in 0.5% BSA/TTBS. Proteins were visualized using an ECL

Western blotting system (Amersham) and X-Omat AR ®lm

(Kodak) according to the manufacturer's instructions.

4.9. ELISA

Wnt binding to sFRP-1-coated wells was performed as

described (UÈ ren et al., 2000), except that cell lysates were

used instead of conditioned medium and HA-tagged Wnt

K. Yoshino et al. / Mechanisms of Development 102 (2001) 45±55 53

proteins were detected with 0.4 mg/ml anti-HA monoclonal

antibody (Roche).

4.10. Nuclear extract preparation

Thirty-six metanephric mesenchymes (13-dpc) were

cultured for 3 days on ®lters in IM with or without 10 mg/

ml of puri®ed recombinant sFRP1 or sFRP-2. Nuclear

extracts were prepared as described previously (Dignam et

al., 1983). To strip cells, the ®lters were combined and

vortexed for 10 s in 1 ml of ice-cold PBS. The cells were

spun down for 1 min at 3000 rev./min, and then re-

suspended in 300 ml of cold buffer A (10 mM HEPES

(pH 7.4); 1.5 mM MgCl2; 10 mM KCl; 0.5 mM DTT)

with protease inhibitors Completee (Roche) and incubated

on ice for 10 min. The nuclei were collected by centrifuga-

tion at 7000 rev./min for 1 min at 48C. The supernatant was

removed, and the nuclei were re-suspended in 50 ml of cold

buffer B (20 mM HEPES (pH 7.4); 1.5 mM MgCl2; 400 mM

NaCl; 0.5 mM DTT; 0.2 mM EDTA; 20% glycerol with

protease inhibitors Completee) and incubated on ice for

30 min, followed by centrifugation at 14 000 rev./min for

30 min. The protein concentrations were determined using

Bio-Rad protein assay dye reagent. Nuclear extracts were

quickly frozen in liquid nitrogen and stored at 2708C. From

36 metanephric explants, 10±15 mg of nuclear proteins was

obtained.

4.11. EMSA

The DNA-binding reaction was performed for 30 min at

room temperature in a volume of 25 ml, containing 0.5 mg of

nuclear protein extract in 5 ml of buffer B, 7.5 mg of acety-

lated bovine serum albumin (Promega), 0.5 mg of sonicated

salmon sperm DNA (Stratagene), and 0.5 ng of 32P-radiola-

beled probe (1 £ 105 cpm) with or without an 80-fold molar

excess of cold competitor DNA. For the optimal TCF/LEF-

binding sites, we used double-stranded oligonucleotides

GGTACCCCCTTTGATCTTACCAAGCTT (TCF-wt) and

GGTACCCCCTTTGGCCTTACCAAGCTT (TCF-mut),

synthesized according to Korinek et al. (1997) and puri®ed

from 10% polyacrylamide gels.

Acknowledgements

The authors wish to thank Susan Gar®eld (manager,

confocal core facility, Laboratory of Experimental Carcino-

genesis, NCI) and Barbara H. Kasprzak for excellent tech-

nical assistance. We also thank Dr Jan Kitajewski

(Columbia University, New York) for providing the

RatB1a ®broblasts expressing HA-tagged Wnt-3A, Wnt-4

and control vectors. K. Yoshino is a JSPS Research Fellow

in Biomedical and Behavioral Research at NIH.

References

Ashe, H.L., Levine, M., 1999. Local inhibition and long-range enhance-

ment of Dpp signal transduction by Sog. Nature 398, 427±431.

Barasch, J., Qiao, J., McWilliams, G., Chen, D., Oliver, J.A., Herzlinger,

D., 1997. Ureteric bud cells secrete multiple factors, including bFGF,

which rescue renal progenitors from apoptosis. Am. J. Physiol. 273,

F757±F767.

Barasch, J., Yang, J., Ware, C.B., Taga, T., Yoshida, K., Erdjument-Brom-

age, H., Tempst, P., Parravicini, E., Malach, S., Aranoff, T., Oliver,

J.A., 1999. Mesenchymal to epithelial conversion in rat metanephros

is induced by LIF. Cell 99, 377±386.

Bradley, R.S., Brown, A.M., 1995. A soluble form of Wnt-1 protein with

mitogenic activity on mammary epithelial cells. Mol. Cell. Biol. 15,

4616±4622.

Chan, S.D., Karpf, D.B., Fowlkes, M.E., Hooks, M., Bradley, M.S., Vuong,

V., Bambino, T., Liu, M.Y., Arnaud, C.D., Strewler, G.J., 1992. Two

homologs of the Drosophila polarity gene frizzled (fz) are widely

expressed in mammalian tissues. J. Biol. Chem. 267, 25202±25207.

Cho, E.A., Patterson, L.T., Brookhiser, W.T., Mah, S., Kinter, C., Dressler,

G.R., 1998. Differential expression and function of cadherin-6 during

renal epithelium development. Development 125, 803±812.

Dale, T.C., 1998. Signal transduction by the Wnt family of ligands.

Biochem. J. 329, 209±223.

Davies, J., Lyon, M., Gallagher, J., Garrod, D., 1995. Sulphated proteoglycan

is required for collecting duct growth and branching but not nephron

formation during kidney development. Development 121, 1507±1517.

Dignam, J.D., Martin, P.L., Shastry, B.S., Roeder, R.G., 1983. Eukaryotic

gene transcription with puri®ed components. Methods Enzymol. 101,

582±598.

Finch, P.W., He, X., Kelley, M.J., UÈ ren, A., Schaudies, R.P., Popescu, N.C.,

Rudikoff, S., Aaronson, S.A., Varmus, H.E., Rubin, J.S., 1997. Puri®-

cation and molecular cloning of a secreted, Frizzled-related antagonist

of Wnt action. Proc. Natl. Acad. Sci. USA 94, 6770±6775.

Hall, A.C., Lucas, F.R., Salinas, P.C., 2000. Axonal remodeling and synap-

tic differentiation in the cerebellum is regulated by WNT-7a signaling.

Cell 100, 525±535.

Hartmann, C., Tabin, C.J., 2000. Dual roles of Wnt signaling during chon-

drogenesis in the chicken limb. Development 127, 3141±3159.

Herzlinger, D., Qiao, J., Cohen, D., Ramakrishna, N., Brown, A.M., 1994.

Induction of kidney epithelial morphogenesis by cells expressing Wnt-

1. Dev. Biol. 166, 815±818.

Hoppler, S., Brown, J.D., Moon, R.T., 1996. Expression of a dominant-

negative Wnt blocks induction of MyoD in Xenopus embryos. Genes

Dev. 10, 2805±2817.

Hsieh, J.C., Rattner, A., Smallwood, P.M., Nathans, J., 1999. Biochemical

characterization of Wnt±frizzled interactions using a soluble, biologi-

cally active vertebrate Wnt protein. Proc. Natl. Acad. Sci. USA 96,

3546±3551.

Karavanov, A.A., Karavanova, I., Perantoni, A., Dawid, I.B., 1998. Expres-

sion pattern of the rat Lim-1 homeobox gene suggests a dual role during

kidney development. Int. J. Dev. Biol. 42, 61±66.

Karavanova, I.D., Dove, L.F., Resau, J.H., Perantoni, A.O., 1996. Condi-

tioned medium from a rat ureteric bud cell line in combination with

bFGF induces complete differentiation of isolated metanephric

mesenchyme. Development 122, 4159±4167.

Kispert, A., Vainio, S., Shen, L., Rowitch, D.H., McMahon, A.P., 1996.

Proteoglycans are required for maintenance of Wnt-11 expression in the

ureter tips. Development 122, 3627±3637.

Kispert, A., Vainio, S., McMahon, A.P., 1998. Wnt-4 is a mesenchymal

signal for epithelial transformation of metanephric mesenchyme in the

developing kidney. Development 125, 4225±4234.

Korinek, V., Barker, N., Morin, P.J., van Wichen, D., de Weger, R.,

Kinzler, K.W., Vogelstein, B., Clevers, H., 1997. Constitutive transcrip-

tional activation by a b-catenin±Tcf complex in APC2/2 colon carci-

noma. Science 275, 1784±1787.

K. Yoshino et al. / Mechanisms of Development 102 (2001) 45±5554

Kreidberg, J.A., Sariola, H., Loring, J.M., Maeda, M., Pelletier, J., Hous-

man, D., Jaenisch, R., 1993. WT1 is required for early kidney develop-

ment. Cell 74, 679±691.

Laitinen, L., Virtanen, I., SaxeÂn, L., 1987. Changes in the glycosylation

pattern during embryonic development of mouse kidney as revealed

with lectin conjugates. J. Histochem. Cytochem. 35, 55±65.

Lee, C.S., Buttitta, L.A., May, N.R., Kispert, A., Fan, C.-M., 2000. SHH-N

upregulated Sfrp2 to mediate its competitive interaction with WNT1

and WNT4 in the somitic mesoderm. Development 127, 109±118.

Leimeister, C., Bach, A., Gessler, M., 1998. Developmental expression

patterns of mouse sFRP genes encoding members of the secreted

frizzled related protein family. Mech. Dev. 75, 29±42.

Lescher, B., Haenig, B., Kispert, A., 1998. sFRP-2 is a target of the Wnt-4

signaling pathway in the developing metanephric kidney. Dev. Dyn.

213, 440±451.

Leyns, L., Bouwmeester, T., Kim, S.-H., Piccolo, S., De Robertis, E.M.,

1997. Frzb-1 is a secreted antagonist of Wnt signaling expressed in the

Spemann organizer. Cell 88, 747±756.

Lin, K., Wang, S., Julius, M.A., Kitajewski, J., Moos Jr, M., Luyten, F.P.,

1997. The cysteine-rich frizzled domain of Frzb-1 is required and suf®-

cient for modulation of Wnt signaling. Proc. Natl. Acad. Sci. USA 94,

11196±11200.

Melkonyan, H.S., Chang, W.C., Shapiro, J.P., Mahadevappa, M., Fitzpa-

trick, P.A., Kiefer, M.C., Tomei, L.D., Umansky, S.R., 1997. SARPs: a

family of secreted apoptosis-related proteins. Proc. Natl. Acad. Sci.

USA 94, 13636±13641.

Moon, R.T., Brown, J.D., Yang-Snyder, J.A., Miller, J.R., 1997. Structu-

rally related receptors and antagonists compete for secreted Wnt

ligands. Cell 88, 725±728.

Papkoff, J., Schryver, B., 1990. Secreted int-1 protein is associated with the

cell surface. Mol. Cell. Biol. 10, 2723±2730.

Perantoni, A.O., Dove, L.F., Karavanova, I., 1995. Basic ®broblast growth

factor can mediate the early inductive events in renal development.

Proc. Natl. Acad. Sci. USA 92, 4696±4700.

Plisov, S.Y., Yoshino, K., Dove, L.F., Higinbotham, K.G., Rubin, J.S.,

Perantoni, A.O., 2001. TGF-b2, LIF, and FGF2 cooperate to induce

nephrogenesis. Development (in press).

Qiao, J., Sakurai, H., Nigam, S.K., 1999. Branching morphogenesis inde-

pendent of mesenchymal±epithelial contact in the developing kidney.

Proc. Natl. Acad. Sci. USA 96, 7330±7335.

Rattner, A., Hsieh, J.-C., Smallwood, P.M., Gilbert, D.J., Copeland, N.G.,

Jenkins, N.A., Nathans, J., 1997. A family of secreted proteins contains

homology to the cysteine-rich ligand-binding domain of frizzled recep-

tors. Proc. Natl. Acad. Sci. USA 94, 2859±2863.

SaxeÂn, L., 1987. Organogenesis of the Kidney, Cambridge University

Press, Cambridge.Shawlot, W., Behringer, R.K., 1995. Requirement of Lim1 in head-organi-

zer function. Nature 374, 425±430.

Shimizu, H., Julius, M.A., GiarreÂ, M., Zheng, Z., Brown, A.M.C., Kita-

jewski, J., 1997. Transformation by Wnt family proteins correlates with

regulation of b-catenin. Cell Growth Differ. 8, 1349±1358.

Stark, K., Vainio, S., Vassileva, G., McMahon, A.P., 1994. Epithelial trans-

formation of metanephric mesenchyme in the developing kidney regu-

lated by Wnt-4. Nature 372, 679±683.

UÈ ren, A., Reichsman, F., Anest, V., Taylor, W.G., Muraiso, K., Bottaro,

D.P., Cumberledge, S., Rubin, J.S., 2000. Secreted frizzled-related

protein-1 binds directly to wingless and is a biphasic modulator of

Wnt signaling. J. Biol. Chem. 275, 4374±4382.

Vestweber, D., Kemler, R., Ekblom, P., 1985. Cell-adhesion molecule

uvomorulin during kidney development. Dev. Biol. 112, 213±221.

Wang, Y., Macke, J.P., Abella, B.S., Andreasson, K., Worley, P., Gilbert,

D.J., Copeland, N.G., Jenkins, N.A., Nathans, J., 1996. A large family of

putative transmembrane receptors homologous to the product of the

Drosophila tissue polarity gene frizzled. J. Biol. Chem. 271, 4468±4476.

Wang, S., Krinks, M., Lin, K., Luyten, F.P., Moos, J.r.M., 1997. Frzb, a

secreted protein expressed in the Spemann organizer, binds and inhibits

Wnt-8. Cell 88, 757±766.Wilkinson, D., Green, J., 1990. In situ hybridization and the three-diment-

inal reconstruction of series sections. In: Copp, A.J., Cokroft, D.L.

(Eds.). Postimplantation Mammalian Embryos, Oxford University

Press, London, pp. 155±171.

K. Yoshino et al. / Mechanisms of Development 102 (2001) 45±55 55