Role of Hyaluronan and CD44 in in Vitro Branching Morphogenesis of Ureteric Bud Cells

Developmental Biology 224, 312–325 (2000) doi:10.1006/dbio.2000.9783, available online at http://www.idealibrary.com on

Role of Hyaluronan and CD44 in in Vitro Branching Morphogenesis of Ureteric Bud Cells Martin Pohl, Hiroyuki Sakurai, Robert O. Stuart, and Sanjay K. Nigam 1 Division of Nephrology and Hypertension, Department of Pediatrics and Department of Medicine, University of California at San Diego, La Jolla, California 92093

Mutual interaction between the metanephric mesenchyme (MM) and the ureteric bud (UB) in the developing kidney leads to branching morphogenesis and the formation of the ureteric tree. A UB-derived cell line, stimulated by conditioned medium derived from an embryonic MM cell line (or, similarly, by 10% fetal calf serum), forms branching tubules under three-dimensional culture conditions (H. Sakurai et al., 1997, Proc. Natl. Acad. Sci. USA 94, 6279 – 6284). The formation of branching tubules in this simple in vitro system for early nephrogenesis is highly sensitive to the matrix environment, a key component of which is the glycosaminoglycan hyaluronan (HA). Consistent with this, we found that HA in the extracellular environment markedly stimulated the formation of cellular processes and multicellular cords (early steps in branching morphogenesis) and also acted as a cell survival factor. Inhibition of HA binding to the cells by addition of blocking antibodies to CD44, the principal cell surface receptor for HA, or degradation of HA by the addition of Streptomyces hyaluronidase resulted in decreased cell survival and diminished morphogenesis, indicating that the HA–CD44 axis plays a central role in in vitro branching morphogenesis. Analysis of the expression of a large number of genes displayed on a cDNA array revealed that significant changes in gene expression in cells undergoing morphogenesis in the presence of HA were limited to a small subset of genes regulating apoptosis, proliferation, and morphogenesis. This included upregulation by HA of its receptor, CD44, which was found to largely localize to the tips of branching cellular processes. In the embryonic kidney, HA was found near the developing ureteric tree and CD44 was expressed basolaterally in UB-derived structures. In addition, both UB and MM appear to express HA synthase, suggesting their ability to secrete HA. We propose that HA promotes branching morphogenesis by creating a positive feedback loop that results in (1) enhanced interaction of HA–CD44 at branching tips (possibly leading to localization of HA binding morphoregulatory factors at the tips) and (2) an activated transcriptional program favoring cell survival/proliferation and migration/morphogenesis of cells through matrix by the expression of key morphoregulatory molecules. Furthermore, since HA, hyaluronidase, and CD44 have been functionally implicated in branching morphogenesis in this model, and since HA, CD44, and HA synthase are all expressed in an appropriate spatiotemporal fashion in the developing kidney, we propose that these molecules may, together, constitute a morphoregulatory pathway that plays a key role in sequential cycles of branching morphogenesis in the UB. © 2000 Academic Press

Key Words: kidney development; ureteric bud; tubulogenesis; branching morphogenesis; extracellular matrix; threedimensional cell culture; cell survival; CD44; hyaluronan; hyaluronic acid; nephrogenesis.

INTRODUCTION A variety of organs arise through the interaction of mesenchymal and embryonic epithelial tissues. In organs 1 To whom correspondence should be addressed at the Division of Nephrology/Hypertension, Departments of Pediatrics/Medicine, University of California at San Diego, 9500 Gilman Drive (0693), La Jolla, CA 92093-0693. Fax: (858) 822-3281. E-mail: [email protected].

312

such as kidney, lung, liver, or pancreas, branching morphogenesis of the epithelial component is a central feature of the developmental program. Mediators of branching morphogenesis include growth factors, proteinases, proteinase inhibitors, and the extracellular matrix (ECM) (Kanwar et al., 1997; Sakurai and Nigam, 1998; Pohl et al., 2000). One common theme in the function of many of these mediators is an interaction with proteoglycans and their component glycosaminoglycans. Hyaluronan (HA) is the largest of the glycosaminoglycans 0012-1606/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.

313

Hyaluronan and CD44 in in Vitro Branching Morphogenesis

and consists of disaccharide repeats of up to several million Dalton molecular weight. It is produced by most cell types in variable amounts and detected in every tissue and body fluid studied in mammals (Knudson, 1998; Laurent and Fraser, 1992). Experimental evidence in a number of systems supports the importance of HA in embryogenesis. For example, depletion of HA interferes with the fusion of the neural folds (Morriss-Kay et al., 1986) and leads to severe growth reduction in the whole embryo (Schoenwolf and Fisher, 1983). In the kidney, HA content is maximal around gestational day 11 (when branching begins) and decreases toward birth (Belsky and Toole, 1983). More recently, the principal cell surface receptor for HA, CD44, has been implicated in developmental processes (Gakunga et al., 1997; Tucker et al., 1999). The tissue and developmental regulation of CD44 expression is complex. The gene contains at least 19 exons, which are differentially spliced into a variety of CD44 isoforms (Screaton et al., 1992). Isoforms of CD44 have been found to be important in limb and mandibular development, where they act through low-affinity binding of growth factors (Sherman et al., 1998; Tucker et al., 1999). This occurs mainly in regions of epithelial–mesenchymal contact, suggesting that CD44 is important in these interactions during development. In addition, kidney tubule development is delayed in the offspring of pregnant mice that have been injected with antibodies against CD44, though it is not known how this occurs (Zoller et al., 1997). During kidney development, the collecting system arises by branching morphogenesis of the ureteric bud (UB) and reciprocal interactions between it and the metanephric mesenchyme (MM) (al-Awqati and Goldberg, 1998; Lechner and Dressler, 1997; Stuart and Nigam, 2000; Pohl et al., 2000). We have recently established an in vitro model for UB cell outgrowth, proliferation, and formation of branching tubular structures (Sakurai et al., 1997), which, in contrast to organ culture, can be used to explore roles of specific molecules in branching morphogenesis at the cellular level. In this system, branching morphogenesis of immortalized UB-derived cells seeded in a matrix gel is stimulated by conditioned medium elaborated by cells thought to be derived from the MM (Sakurai et al., 1997) or, similarly, by 10% fetal calf serum (FCS); this is perhaps the most elemental system currently available to analyze early branching events pertaining to early kidney development. Using this in vitro model system, we now demonstrate the involvement of the HA–CD44 axis in epithelial branching morphogenesis. HA-mediated branching morphogenesis appeared to occur as a result of enhanced cell survival as well as through a direct effect on the ability of the cells to extend unicellular processes and to form branching multicellular cords; consistent with this, CD44 was found to localize to the tips of cellular processes. In addition, the presence of HA in the ECM appeared to alter the UB cell transcriptional state to favor cell survival/proliferation and branching. Consistent with these results, HA was found to accumulate around the structures of the developing ureteric

tree and CD44 was expressed basolaterally in the UBderived epithelium.

METHODS Cell Culture Ureteric bud cells, derived from the isolated UB of SV40 large T transgenic mouse embryos at E11.5 as described elsewhere (Barasch et al., 1996), were cultured in Minimum Essential Medium (MEM; Gibco) supplemented with 10% (v/v) FCS (BioWhittakar) at 32°C and 5% CO 2. Cells were passaged biweekly and were used for assays 3–7 days after passage, when they had reached confluence. Cells derived from isolated metanephric mesenchyme of SV40 large T transgenic mice at E11.5 (BSN cells) (Sakurai et al., 1997), were maintained in culture at 37°C and 5% CO 2 in DMEM/F12 containing 10% (v/v) FCS and passaged once or twice weekly. BSN cell conditioned medium (BSN-CM) was produced by changing the medium to serum-free DMEM/F12 (Gibco) after three thorough washes with Hanks’ balanced salt solution (HBSS; Gibco) when confluence was reached. BSN-CM was harvested after 3– 4 days of serum-free culture and stored at 4°C after low-speed centrifugation and filtration through 0.22-␮m membranes for removal of cell debris. All media were supplemented with antibiotics/ antimycotics (Gibco).

Antibody Purification Rat anti-mouse CD44 antibodies (KM201) were produced by a rat hybridoma (ATCC, Tin 240) cultured in RPMI 1640 medium (Gibco) supplemented with 15% FCS, 50 ␮M mercaptoethanol, and 50 U/ml penicillin/50 ␮g/ml streptomycin (Sigma) (Miyake et al., 1990). The supernatant was collected after 3 weeks incubation at 37°C and stored under sterile conditions without further addition of antimicrobial agents. For purification, 50 –150 ml of supernatant was passed through a protein G–Sepharose column (Hitrap Protein G, 1 ml; Pharmacia Biotech) and extensively washed, and the adsorbed protein was eluted with 0.1 M Glycine–HCl, pH 2.7. The eluted solution was passed through a PD-10 desalting column (Pharmacia) and the buffer changed to HBSS. The protein content of the resulting solution was measured spectroscopically (Beckman BU-65). After concentration by centrifugation at 2000g with Centricon spin columns (cutoff 30 kDa), the protein concentration was approximately 2 mg/dl. The final antibody solution was sterilized by filtering through 0.22-␮m syringe filters (Millipore) and stored at 4°C. As a negative control in the three-dimensional cell culture, identically processed hybridoma growth medium (RPMI 1640 with 15% FCS, 50 ␮M mercaptoethanol and penicillin/streptomycin) and commercially available polyclonal rat IgG (Sigma) were used.

Three-Dimensional Cell Culture For evaluation of the effect of HA on UB cells in threedimensional culture, gels with or without HA were prepared by mixing 1 part of 10⫻ DMEM (Sigma), 1 part of 200 mM Hepes (Gibco), 0.5 parts of 74 mg/ml sodium bicarbonate with 3.75 parts of 3.95 mg/ml rat tail collagen I (Collaborative Biomedical) and 3.75 parts of either 5 mg/ml HA (derived from umbilical cord; Sigma) in water or sterile water. The pH was titrated with 1 M NaOH to 7.4 and all solutions were kept on ice until the addition of the cells. Cells were added in a concentration of approximately 100,000

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.

314

Pohl et al.

cells/ml and 100-␮l aliquots of the cell suspension were pipetted into 96-well plates after thorough mixing. After 15 min at 37°C to allow for gelation, 100 ␮l medium was added to each well. For culture in serum-free medium only, DMEM/F12 alone was used. In experiments with a single growth factor, HGF (Sigma; 50 ng/ml) was added to DMEM/F12. Anti-CD44 antibodies (KM201, 200 ␮g/ml) or Streptomyces hyaluronidase (Sigma; 100 U/ml) was added to BSN-CM for disruption of the HA–CD44 axis. Other experiments were performed utilizing DMEM/F12 containing 10% FCS. Results were recorded with a phase-contrast microscope (Nikon) by measuring 10 consecutive maximal cell/colony lengths in each well with the aid of an ocular micrometer in randomly selected fields within the gel after 24 – 48 h and by counting the cells/colonies throughout the entire depth of the gel (omitting the flattened cells on the top and bottom surface) in a 1-mm 2-sized field on alternate days, beginning after 24 h. Error bars indicate the standard error of the mean, and P values were calculated using the Student t test. Differential staining of living and dead cells was performed after 5 days of culture. Living cells enzymatically cleaved calcein-AM and released a green-fluorescent substrate, whereas the nuclei of dead cells were stained by an ethidium homodimer not penetrating intact cell membranes (Molecular Probes, Eugene, OR). The unfixed gels were incubated for 30 min at room temperature in a 96-well plate and then photographed with a fluorescence microscope (Nikon). Areas shown with fluorescently stained living and dead cells are identical for allowing comparison of the relative density of living or dead cells.

Dolichus biflorus Staining of Three-Dimensional Cell Structures UB cells were grown for several days in pure collagen type I or collagen type I with the addition of 20% (v/v) growth factorreduced Matrigel (MG; Collaborative Biomedical) in BSN-CM. After two washes with phosphate-buffered saline (PBS) the gels were fixed with 4% paraformaldehyde (Sigma) for 1 h at room temperature (RT) and subsequently washed with PBS for 6 h. The gels were then exposed to PBS with 0.1% (w/v) saponin (Sigma) and 3% bovine serum albumin (BSA; Sigma) for 15 min at RT. After a 1-h incubation with fluorescein-coupled lectin from D. biflorus (50 ␮g/ml; Vector Laboratories) the gels were washed again in PBS overnight. The gels were then counterstained with propidium iodine (Sigma) in PBS for 10 min in RT and visualized by confocal microscopy after additional washes in PBS.

Immunoblotting UB and BSN cells were cultured until confluence in six-well plates, washed once with ice-cold PBS, and then harvested on ice in 300 ␮l ice-cold RIPA buffer (100 mM NaCl, 1% TX-100, 0.5% deoxycholate, 0.2% SDS, 2 mM EDTA, 10 mM Hepes, 1 mM NaV) supplemented with protease inhibitors (10 ␮M PMSF, 20 ng/ml aprotinin, 20 ng/ml leupeptin; all from Sigma). For immunoblotting, equal amounts of the samples were mixed with a 4⫻ nonreducing sample buffer, boiled for 5 min, and applied to a 10% polyacrylamide gel. After electrophoresis, proteins were blotted onto nitrocellulose membranes (Micron Separation Inc.), blocked with 5% (w/v) nonfat dry milk in PBS with 0.1% (v/v) Tween 20 (Sigma), and incubated for 2 h at RT with the undiluted supernatant from the hybridoma containing the anti-CD44 antibody KM201. After extensive washing, the membranes were incubated with

peroxidase-conjugated anti-rat IgG and the signals detected using a chemiluminescence kit (Pierce) following the manufacturer’s instructions.

Northern Blotting Total RNA was isolated from UB cells grown in 5 ml collagen I gels with or without the addition of HA (as described above) in the presence of an equal volume of serum-free DMEM/F12, BSN-CM, or DMEM/F12 with 10% FCS (in the presence of HA only with DMEM/F12 with 10% FCS). After 20 –22 h incubation the gels were compacted by brief centrifugation, immediately dissolved in Tri-Reagent LS (Molecular Research Center, Inc.), and homogenized in several steps using 19- to 23 gauge needles in 3-ml syringes. After an additional spin at 12,000g to clear insoluble debris, the procedure was continued according to the manufacturer’s directions. The resulting total RNA was dissolved in water and stored at ⫺80°C. Poly(A) RNA was then isolated from 200 ␮g total RNA from each condition using the Micro Poly(A) Pure Kit (Ambion, Inc.). An amount of 1 ␮g of poly(A) RNA was electrophoresed in 1% agarose-formamide and blotted overnight on Magnagraph nylon membranes (Micron Separations, Inc.). A 573-bp cDNA probe for CD44 was generated by reverse transcriptase-PCR from isolated UB cell RNA using primers spanning almost the entire 5⬘-end constant region (exons 1–5) of the gene (GenBank Accession No. X66084). The following primers were used: forward 5⬘-CTTTGCCTCTTGCAGTTGAGC-3⬘, reverse 5⬘-GCTGTTCCAGTAGGAAGGTAGG-3⬘. The resulting cDNA-probe thus recognized all known isoforms, which differ only in exons 6 –15 (Screaton et al., 1992). The probe was fully sequenced and matched the expected sequence of CD44. After radioactive labeling (Ready-to-Go labeling kit; Pharmacia) the cDNA probe was added to the prehybridized membranes and allowed to bind at 42°C for 16 –20 h in the presence of a hybridization buffer (40% formamide, 4⫻ SSC, 2.5⫻ Denhardt’s solution, 10% Dextran sulfate, 50 mg/L salmon sperm DNA). The membranes were then washed twice at RT for 20 min in 2⫻ SSC and 0.1% SDS and exposed to Kodak Biomax MS films at ⫺80°C.

Reverse Transcriptase-PCR For the determination of hyaluronate synthase (Has) expression in embryonic kidney and the cell lines, isolated poly(A) RNA was DNase treated (DNase I; Life Technologies) for 15 min at 37°C and then reverse transcribed using the Superscript Preamplification System (Life Technologies). Poly(A) RNA was isolated (Micro Poly(A) Pure; Ambion, Inc.) from separated UBs and MMs of embryonic day 13 rat kidneys, from UB cells (grown in collagen I gels with BSN-CM), and from BSN cells grown as monolayer in serum-free DMEM/F12 for 1 day. PCR on the generated cDNAs was performed with an annealing temperature of 66 (mouse and rat Has2, WT1) or 70°C (mouse Has1 and 3) with HotStarTaq DNA polymerase (Qiagen) in 35 cycles using the following primer pairs: mouse Has1, forward 5-gagacaggacatgccaaagccctca-3, reverse 5-cacgcacctgcgtgttctcaccag-3, predicted product size 640 bp (Itano et al., 1999); mouse Has2, forward 5-ctgccaaagcaaattgatacatcag-3, reverse 5-agcatagccagtggctttccaac-3, predicted product size 315 bp (Spicer and McDonald, 1998); mouse Has3, forward 5-accggtgcagctgactacagccct-3, reverse 5-tcacaccgcaaaagccaggc-3, predicted product size 1659 bp (Itano et al., 1999); rat Has2, forward 5-ctgccaaagcaaattgatacatcag-3, reverse 5-agcatagccagtggcttttcg-3, predicted product size 341 bp, derived from coding sequence (GenBank Accession No. AF008201); and rat Wilms tumor suppressor (WT-1), forward

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.

315

Hyaluronan and CD44 in in Vitro Branching Morphogenesis

5-gaagcgtcctttcatgtgtgc-3, reverse 5-cctgggagttaatcaagagtgg-3, predicted product size 605 bp, derived from coding sequence (GenBank Accession No. X69716).

Immunocytochemistry UB cells were grown in 60 ␮l collagen I in 96-well plates in DMEM/F12 with 10% FCS. After 24 h, the gels were washed three times with PBS and fixed with 4% paraformaldehyde for 1 h at RT. The gels were then taken out of the wells after an additional wash with PBS and washed for 2 h at RT in 20 ml PBS. After blocking with 2% (w/v) BSA (Sigma) in PBS for 1 h, anti-CD44 antibody (KM201) was applied as a 1:10 dilution of the purified preparation in 2% BSA and incubated for 1 h at 37°C. After three washes in PBS and an additional wash in 20 ml PBS for 2 h rhodamine, (TRITC)labeled secondary antibody (Jackson ImmunoResearch) was added and incubated for 1 h in 37°C. Then the gels were washed overnight in PBS. The next day, they were placed onto microscope slides and immediately evaluated under a fluorescence microscope (Nikon). For monolayer staining, UB cells were seeded onto glass coverslips and grown in MEM with 10% FCS for several days. The monolayers were then fixed with methanol for 20 min at ⫺20°C, washed, blocked with 2% BSA for 30 min, and stained with a 1:10 dilution of the concentrated antibody in 2% BSA for 1 h in RT. After washing, the TRITC-labeled secondary antibody was incubated for 1 h in RT. The slips were then washed and mounted for observation with a fluorescence microscope (Nikon).

Immunohistochemistry Murine embryonic kidneys were dissected at E13, dehydrated in 40% sucrose for 3 h, embedded in OCT compound (Tissue Tek; Sakura Finetek U.S.A.), frozen at ⫺20°C, and cut in 6-␮m sections. For HA staining, the sections were fixed in 4% paraformaldehyde for 2 min, blocked for 1 h in 3% BSA (Sigma) containing 0.5% hydrogen peroxide, and then incubated with 7.3 ␮g/ml biotinylated proteoglycan (b-PG, extracted from bovine nasal cartilage, a gift from Dr. C. Underhill) in PBS with 3% BSA for 1 h at RT. Nonspecific staining was controlled for by adding 0.1 mg/ml HA to 7.3 ␮g/ml b-PG before the application to the section. For CD44 staining, the frozen sections were fixed for 2 min in methanol at ⫺20°C, blocked for 90 min with 3% BSA in PBS containing 0.5% hydrogen peroxide, and exposed to either KM201 hybridoma supernatant or a 1:1 dilution of the concentrated anti-CD44 antibody in PBS with 3% BSA for 90 min at 37°C. Control sections were exposed to the hybridoma growth medium or the blocking solution alone. After a wash with PBS, a biotin-coupled goat anti-rat antibody (1:100 in PBS with 3% BSA) was added for 90 min at RT. All sections (HA and CD44 stained) were exposed to a horseradish peroxidase-coupled streptavidin– biotin complex (Vectastain; Vector Laboratories) for 45 min at RT and developed by the addition of the substrate 3,3-diaminobenzidine (Sigma). Washes in between the steps were performed with PBS.

cDNA Array Analysis UB cells were suspended in either collagen I alone or collagen I with HA as described above and cultured with DMEM/F12 containing 10% FCS. After 18 h incubation at 32°C, total RNA was extracted as described for the Northern blotting procedure. Poly(A) RNA was isolated from 150 ␮g of total RNA with the Oligotex kit (Qiagen). The poly(A) RNA was then used for

synthesis of complex 32 P-labeled cDNA strands using a master mix of gene-specific primers provided by the manufacturer (Clontech). After separation from unincorporated 32 P, the probe was hybridized with membranes displaying 588 gene-specific murine cDNAs. After six washes at 68°C first in 1⫻ SSC with 1% SDS and then in 0.1⫻ SSC with 0.5% SDS, the membranes were exposed to Kodak Biomax MS for autoradiographic evaluation. The individual autoradiographic signals were quantified with the aid of a computer program (VectorArray; Pavlova et al., 1999), which provided area- and background-corrected signal intensities (integrated pixel density, IPD) corresponding to each gene present in the arrays. The IPD of visually detectable signals exceeding a threshold of 4000 in at least one condition and showing concordant differences in two different sets of membranes were converted to logarithmic values and plotted as a scatter graph. The distance of the individual points from a regression line accounting for the expression of housekeeping genes was taken as a measure of differential expression. An increase greater than log 0.25 (a roughly 75% increase) was regarded as a significant increase in gene expression (Table 1).

RESULTS Hyaluronan Prevents Cell Death and Promotes Morphogenesis in an in Vitro Model of Ureteric Bud Branching Morphogenesis Embryonic murine immortalized UB cells grown in three-dimensional collagen type I gels form structures reminiscent of the early steps of the branching embryonic collecting system of the kidney, when stimulated with BSN-CM (a conditioned medium derived from a putative metanephric mesenchyme cell line) or serum. When 20% (v/v) growth factor-reduced MG is added to the collagen I, the morphogenetic process is more sustained and leads to larger structures consisting of more cells and, over time, branching multicellular cords and tubules with lumens (Fig. 1). These morphogenetically distinguishable stages can be correlated with expression of specific, though overlapping, subsets of genes (Pavlova et al., 1999). Gene products involved in cell–matrix interactions were found to be upregulated in this in vitro model system under conditions favoring branching morphogenesis (Pavlova et al., 1999). Since the ECM composition is known to influence in vitro branching morphogenesis of cultured cells (Sakurai et al., 1997; Santos and Nigam, 1993), albeit in poorly defined ways, as a first step toward further dissecting the potential role of HA–CD44 interactions in this process, we added HA to the collagen matrix and compared the resulting UB cell morphogenesis with growth in collagen I alone. When cultured in the presence of HA, UB cells/colonies developed significantly longer processes (P ⬍ 0.05) and were more likely to form multicellular cords (Figs. 2A and 2B). This effect was clear in serum-free medium, under stimulation with a single pure growth factor (HGF 50 ng/ml), in BSNCM, or in medium containing 10% FCS. In addition, the number of cells/colonies with processes and branches was increased in the cells exposed to HA, when cultured under

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.

316

Pohl et al.

FIG. 1. UB cells were grown in collagen type I gels (A, B) or MG/collagen I gels (C, D, E) for 7 days (A, B, C, D) or 10 days (E) in BSN-CM. Structures were more elaborate and contained more cells in the MG/collagen I gel (C, D) after 7 days than in the collagen I gel (A, B), suggesting an influence of noncollagen ECM components on branching morphogenesis. In addition, structures in the MG/collagen I gels proceeded to form multicellular cords (D, E) and tubules (E) with lumens (*), the latter of which could not be readily observed in collagen I gels alone. The structures were stained with fluorescein-coupled D. biflorus agglutinin (green) and propidium iodine (red). Scale bars correspond to 40 ␮m.

serum-free conditions or stimulated with HGF or 10% FCS (P ⬍ 0.005) (Fig. 3A). In principle, the larger size and increased complexity of the multicellular structures resulting from culture in the presence of HA may have been a function of an increase

in cell survival. Even in the absence of a specific morphogenetic effect, a HA-induced lengthening of cell life might be expected to directly result in larger structures and perhaps to result in more complex structures as well, if a default pathway for branching exists in epithelial

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.

317

Hyaluronan and CD44 in in Vitro Branching Morphogenesis

FIG. 2. UB cells grown for 48 –72 h in collagen I gels without (A) or with (B) the addition of HA in 10% FCS. UB cells/colonies develop longer processes in HA-containing gels and form more multicellular cords. Scale bars correspond to 50 ␮m. Quantification of cell/colony length revealed a significant increase in maximal length in HA-containing collagen I compared to culture in collagen I only, regardless of whether cultured in 10% FCS (FCS), BSN-CM (BSN), 50 ng/ml HGF (HGF), or serum-free medium (SF) alone (C; P ⬍ 0.05 for ⫹HA versus ⫺HA under all conditions). Single UB cell/colony length was recorded after 48 –72 h by measuring the longest distance between the tips from opposite cell/colony extensions in 10 consecutive structures starting with a randomly chosen structure. Lengths were determined in 10 cells/colonies per well and three or four wells were counted for each condition. Assays were performed at least in triplicate. Error bars show the standard errors of the mean, P values were calculated with the Student t test.

cells. Such an intrinsic program for branching morphogenesis may exist in certain clonal cell lines when cultured under three-dimensional conditions (Bennet, 1980) and the explanted isolated ureteric bud (Qiao et al., 1999). We therefore analyzed cell death in HA-treated and control cells grown in collagen I gels. HA inclusion resulted in markedly fewer dead cells after 5 days compared to the UB cells dispersed in collagen I alone, as shown by staining with an ethidium homodimer, which binds to nucleic acids but cannot penetrate intact cell membranes (Fig. 3B). The cell densities in the evaluated areas were similar (Fig. 3B).

Addition of Streptomyces Hyaluronidase or Blocking Anti-CD44 Antibodies Markedly Inhibits UB Cell Morphogenesis The potential role of HA was further investigated by addition of Streptomyces hyaluronidase, an enzyme that

specifically degrades HA, to the three-dimensional UB cell culture in collagen I. This resulted in significantly decreased numbers of cells/colonies with processes after 5 days of culture (P ⬍ 0.00001, Fig. 4A). Addition of heat-inactivated Streptomyces hyaluronidase had no significant effect compared to culture medium alone. In keeping with the hyaluronidase data, the addition of purified blocking anti-CD44 antibody (KM201) to the three-dimensional cultures in collagen I had the same effect on process formation as did active hyaluronidase (P ⬍ 0.00002, Fig. 5A). Identically purified hybridoma growth medium or polyclonal rat IgG in the same concentration had no effect on cell behavior. Both responses were dose-dependent (Figs. 4B and 5B) and point toward decreased morphogenesis when HA–CD44 signaling is impaired. Taken together with data presented in Figs. 2 and 3, these observations suggest that, under threedimensional conditions, HA acts as a cell survival factor

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.

318

Pohl et al.

observed effects of HA are at least partially the result of receptor-mediated signaling via the principal cell surface HA receptor, CD44. Therefore, we investigated the expression of CD44 in further detail. Immunoblotting of UB cell monolayers grown in normal growth medium with the anti-CD44 antibody KM201 showed mainly the 80kDa standard form of the HA receptor (Fig. 6A). The antibody employed recognizes the extracellular constant HA-binding region, which is preserved in all CD44 isoforms (Zheng et al., 1995). When grown in threedimensional collagen I culture, expression of CD44mRNA was upregulated under conditions favoring branching in comparison with culture in serum-free DMEM/F12 (Fig. 6B), suggesting a role for the CD44 receptor in branching morphogenesis. Addition of HA into the collagen I matrix stimulated CD44 expression even further (Fig. 6B). Because both UB and BSN cells

FIG. 3. After 2–5 days, UB cells cultured in HA-containing collagen I gels and grown in serum-free medium (SF), HGF (HGF, 50 ng/ml), or 10% FCS (FCS) showed higher cell density and a higher number of cells/colonies with processes (an early step in branching tubulogenesis) in HA-containing matrix than UB cells cultured in collagen I alone (A; P ⬍ 0.005 for ⫹HA versus ⫺HA under all conditions). All cells/colonies with processes were counted through the entire width of the gel in a randomly chosen high-power field and three or four wells for both conditions were evaluated. Assays were performed at least threefold. Error bars show the standard errors of the mean and P values were calculated with the Student t test. Staining of unfixed cells in the gel with fluorescent compounds, specifically labeling living and dead cells, showed a much lower number of dead cells in the HA-containing matrix after 5 days in regions of equal numbers of living cells (B).

and thereby favors structural morphogenesis, but they also raise the possibility that the HA–CD44 axis plays an additional, more direct role in facilitating the formation of branching tubular structures.

The Hyaluronan Receptor, CD44, Is Expressed Mainly in Its Standard Form and Is Upregulated by BSN-CM, 10% FCS-Containing Medium, and the Addition of HA to the ECM; Expression of HA Synthase in the Embryonic Kidney The comparable effect of Streptomyces hyaluronidase and blocking anti-CD44 antibodies suggested that the

FIG. 4. UB cells were cultured in collagen I gels and stimulated with BSN-CM. Cells/colonies with processes and branching structures were counted in randomly selected high-power fields. Addition of Streptomyces hyaluronidase (Hyal.; 100 U/ml) to the cultures led to a significant decrease in UB cells/colonies with processes after 5–7 days (A; P ⬍ 0.00001) compared to addition of heat-inactivated Streptomyces hyaluronidase. These effects were dose-dependent (B). Error bars indicate the standard errors of the mean and P values were calculated using the Student t test.

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.

319

Hyaluronan and CD44 in in Vitro Branching Morphogenesis

FIG. 5. UB cells were grown in collagen I gels and stimulated with BSN-CM. Addition of purified blocking rat anti-mouse CD44 antibody (KM201; 200 ␮g/ml) significantly decreased the number of cells/colonies with processes at 5 days compared to control (A; P ⬍ 0.00002). The decrease was dose-dependent and still significant at lower concentrations (B; P ⬍ 0.001 at 20 ␮g/ml). Polyclonal rat IgG or purified hybridoma medium had no effect. Error bars indicate the standard errors of the mean and P values were calculated using the Student t test.

express the hyaluronate synthases (Has) 1 and 2 (Fig. 6C), these results suggest a positive feedback loop. Has2 also appears to be expressed in dissected E13 rat UB and MM (Fig. 6D). Expression of Has3 could not be found in the cells using previously published primers (Itano et al., 1999). The effect of the HA–CD44 axis on branching morphogenesis could be either direct (by participating in the alteration of cell shape) or indirect (by favoring cell survival) or both.

Immunostaining of UB Cells in Developing Branching Structures with Anti-CD44 Antibody Shows Two Different Pools of the Hyaluronan Receptor, One of Which Localizes to Branch Tips; Both CD44 and HA Are Found in or Near the Developing Ureteric Tree The cellular distribution of the HA receptor was determined by staining UB cell monolayers and UB cells grown in three-dimensional gels with anti-CD44 antibodies. The anti-CD44 antibody (KM201) strongly stained the plasma-

lemma of the cell bodies in a “grainy” pattern (Figs. 6E and 6F). In the gel, the cells formed processes. The stalks of those outstretched cellular processes were only faintly stained, but at the tips a strong immunofluorescent signal was detectable in many developing branching structures (Fig. 6F). The specifically stained tips appear to represent areas of cellular protrusion into the surrounding ECM. Together with the increased process length in cells/colonies cultured in HA-containing ECM, this localization further supports a role for HA and its main cell surface receptor in process formation, which is essential for branching morphogenesis in this model. Another pool of CD44 was distributed on the membrane of the cell body. A direct role of this fraction of CD44 in branching morphogenesis would seem less likely, since these receptors are far from plasma membrane domains actively involved in morphogenetic movements. Nevertheless, these CD44 molecules could play a role in the effect of HA observed on cell survival. When frozen sections of E13 mouse kidneys were stained for HA (Fig. 7A) or CD44 (Fig. 7C), staining of HA was observed around the developing collecting tubules, and a basolateral localization of the CD44 receptor on collecting duct cells was found, further supporting their potential interaction and significance during collecting system development.

Addition of Hyaluronan to UB Cell Cultures in Three-Dimensional Collagen I Matrices Leads to Increased Expression of Proliferation Markers and Antiapoptotic Factors and Leads to Upregulation of Its Receptor (CD44) The HA–CD44 interaction possibly serves several purposes, including creation of membrane domains (i.e., branching tips) with high local concentrations of HA, subsequent creation of “clear spaces” for protrusion of cellular processes, opportunities for localized sequestration, or presentation of secreted products such as growth factors, proteinases, or proteinase inhibitors that facilitate branching morphogenesis. An additional potential function is CD44-mediated signal transduction and cell activation and change in the transcriptional state of the cell. In order to detect transcripts which were differentially expressed in the presence of HA, a membrane carrying an array of 588 different specific mouse cDNA sequences was hybridized with cDNA derived from mRNA of UB cells grown in either collagen or collagen/HA gels for 18 h (Fig. 8A). The signal intensities, measured as integrated pixel density (IPD), were summarized in the form of a two-dimensional scatter graph (Fig. 8B). Of the 588 genes, 133 were found to be expressed, but only 14 increased more than log 0.25, corresponding to an increase in signal strength of more than approximately 75% (Table 1). Remarkably, nearly all these genes (which include CD44) are known to play key roles in cell proliferation, survival, and morphogenesis, consistent with data already presented that support an important role for the HA–CD44 axis in epithelial branching morphogenesis. (The

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.

320

Pohl et al.

FIG. 6. UB cells were examined for the expression of the HA receptor CD44 by immunoblotting (A), Northern hybridization (B), and immunocytochemistry (E, F). Expression of hyaluronate synthases (Has1, 2, 3) was tested by reverse transcriptase-PCR on UB and BSN cells (C) and (Has2) on dissected E13 rat UB (rUB) and MM as an indicator of HA production. Wilms tumor suppresser-1 (WT1) expression was investigated as a measure of contamination of the dissected rUB and only minimally, if at all, expressed (D). Monolayer cell extracts probed

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.

Hyaluronan and CD44 in in Vitro Branching Morphogenesis

321

FIG. 7. Frozen sections of E13 mouse embryonic kidneys were stained for hyaluronan (A) and the HA receptor CD44 (C), utilizing a hyaluronan binding protein from bovine nasal cartilage (gift from Dr. C. Underhill) and the anti-CD44 antibody KM201. HA is found near the basal cell surface of the developing ureteric tree (A) and CD44 stains the basolateral surfaces of ureteric bud-derived structures (C). Negative controls showed no specific signals (B, D). Scale bars correspond to 100 ␮m.

with the anti-mouse antibody KM201 showed abundant protein of the standard form (CD44H) at approximately 85 kDa (A). Hybridization of blotted poly(A) RNA from UB cells in collagen I gels showed increasing expression of the receptor in cultures stimulated with BSN-CM (BSN) or DMEM/F12 with 10% FCS (FCS) compared to serum-free (SF) culture after 18 h (B). Addition of HA to collagen I also increased CD44 expression by the UB cells (cultured in 10% FCS). The increase in expression upon stimulation with soluble factors that induce UB cell branching morphogenesis suggests a role of this receptor in the morphogenetic process. The CD44 cDNA-probe was produced by reverse transcriptase-PCR from UB cells using primers specific for the constant region of this receptor (exons 1–5) and sequenced to ensure correct amplification. Both UB cells and BSN cells express Has1 and 2 (C), whereas Has3 could not be detected with previously published primers in the cell lines. UB tissue appears to express at least Has2 (D), suggesting the production of HA by the UB epithelium. UB cells, grown either as monolayer (E) or in collagen I (F), were stained with the anti-CD44 antibody KM201. The obtained pattern showed cell surface staining of UB cells in a monolayer (E). When cultured in three-dimensional gels, the cells showed much stronger staining at the tips of the cell processes (arrows) compared with the stalks of those processes, while maintaining cell surface staining (F). Together with the data suggesting HA production by the UB itself and increasing expression of CD44 as a reaction to HA in the cell environment, these data are consistent with a positive feedback loop, involved in branching morphogenesis of the developing ureteric tree. Scale bars correspond to 10 ␮m.

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.

322

Pohl et al.

significance of changes in the expression of individual genes is discussed in further detail below.)

DISCUSSION HA, by virtue of its very large size, negative charge, and ability to serve as a scaffold for the assembly of multiproteoglycan aggregates, is an ideal space-filling molecule in the ECM. Nevertheless, HA is not an inert bystander in morphogenetic processes though perhaps even in a spacefilling capacity it may be important for allowing morphogenetic movements and creating cell-free zones for easy invasion by cellular processes. Our data indicate a key role for the HA–CD44 axis in the function of complex branching epithelial structures and suggest two mechanisms by which this may occur during embryogenesis: First, HA–CD44 interactions lead to increased cell number and survival, which is reflected by changes in gene expression patterns favoring cell proliferation and regulating apoptosis (Figs. 3 and 8). Second, HA leads to longer cell processes and to upregulation of its receptor (CD44), a fraction of which localizes to the tips of nascent branching epithelial processes, an early step in the evolution of complex branching structures (Figs. 2, 6B, and 6D). HA–CD44 interactions are accompanied by increased expression of classes of genes believed to play key morphogenetic roles (Fig. 8). As we have shown, blocking the HA–CD44 axis either with neutralizing antibodies against CD44 or by degrading HA in the matrix prevents complex morphogenesis (Figs. 4 and 5). With respect to the in vivo situation, the tissue distribution of HA and CD44 in the embryonic kidney strongly supports a role for their interaction in kidney organogenesis (Fig. 7). The localized expression of the HA receptor CD44 at the leading edges of the advancing UB cell processes in a three-dimensional environment is consistent with a direct role in branch formation. The predominant expression of CD44 at the tips in the in vitro model would seem to indicate a function in tip elongation rather than in cleft or stalk regulation. The presence of Has2 in UB and BSN cells as well as the explanted UB and MM suggests that HA in the embryonic kidney can potentially be made by both UB and MM to regulate branching events. A role for the interaction of CD44 and HA during UB branching morphogenesis in vivo is supported by the expression of HA and CD44 near the UB in the embryonic kidney. A variety of possible mechanisms exist by which HA, through interactions with its receptor CD44 at the tip of branching structures, might facilitate branching. HA, and chondroitin sulfate proteoglycans bound to HA, such as versican, may become localized at the leading edges. The local sequestration of HA on the cell surface via CD44 could potentially participate in structuring the ECM environment with respect to those HA-bound proteoglycans. Indeed, versican has been shown to decrease adhesion and enhance proliferation (Yang, 1999), which is consistent with the reaction of the UB cells to the addition of HA.

Another possible mechanism is the local stabilization of CD44 at the leading edges of cell processes secondary to localized synthesis and extrusion of HA through the plasma membrane; this stabilization could conceivably occur via the intracellular connection of CD44 to the actin cytoskeleton through the ERM (ezrin/radixin/ moesin) complex (Tsukita et al., 1994; Yonemura et al., 1998). ERM proteins are thought to be important for the cell surface distribution of cell adhesion molecules and the formation of cellular protrusions (Helander et al., 1996; Turunen et al., 1998). Thus, this suggests a potential mechanism for the generation of branching tips: localized HA deposition could lead to increased localization of CD44 at the tips, which then, together with proteins of the ERM complex, could reshape the subcortical actin cytoskeleton and the plasmalemma to create branching processes; this possibility is consistent with our finding that HA increases the expression of CD44, a large pool of which localizes to the tips of branching processes (Fig. 6D). Strongly supporting this notion, the formation of lamellipodia of single breast epithelial cells (which derive from a tissue also capable of branching morphogenesis) can be induced by local application of HA, an effect which could be blocked by anti-CD44 antibodies (Oliferenko et al., 2000). Interestingly, several recent studies suggested a role for CD44 in morphogenetic processes independent of HA. For example, CD44 has been shown to bind gelatinase B and could therefore serve as cell surface receptor for mesenchymally or epithelially produced gelatinase B (Bourguignon et al., 1998; Yu and Stamenkovic, 1999). In addition, some heparan sulfate chain-carrying isoforms of CD44 appear to bind FGFs during limb formation, which then can be presented to high-affinity receptors (Sherman et al., 1998; Tucker et al., 1999). It is well established that some growth factors such as fibroblast growth factors bind to heparan sulfate-containing proteoglycans (Taipale and Keski-Oja, 1997). In addition, heparan sulfate is thought to bind urokinase, which could result in local activation of hepatocyte growth factor (HGF) (Mignatti and Rifkin, 1993; Naldini et al., 1995). HGF is a well-established regulator of in vitro branching morphogenesis of epithelial cells (Sakurai and Nigam, 1997; Santos and Nigam, 1993). Urokinase could also activate matrix metalloproteinases and promote local ECM digestion (Mignatti and Rifkin, 1993). Indeed, matrix-degrading proteinases elaborated by multiple sources appear to be functionally important in UB cell branching morphogenesis (M. Pohl, H. Sakurai, and S. K. Nigam, unpublished results). However, our results indicate that, in addition to the possible functions of different heparan-sulfate containing CD44 isoforms, HA itself is important in this morphogenetic process, because the addition of HA led to more multicellular cords and longer processes under all conditions, most strikingly even when no growth factors were present, and destruction of HA by hyaluronidase decreased morphogenesis. Thus, in this in vitro model, HA, as a component of the ECM, appears to act

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.

323

Hyaluronan and CD44 in in Vitro Branching Morphogenesis

FIG. 8. Differential gene expression in collagen I gels with and without the addition of HA was investigated with the aid of mouse cDNA arrays. mRNA was isolated from both conditions, reverse transcribed, radioactively labeled, and hybridized with membranes carrying 588 known mouse gene cDNAs. A representative blot is shown (A). Computerized calculation and comparison of the signal intensity under both conditions led to the identification of differentially expressed genes. Signals were area- and backgroundcorrected and expressed as log values. Points most distant from the regression line vary most between the conditions (B). Autoradiographically visible signals which showed concordant differences in two sets of blots and increased more than log 0.25 in the presence of HA are described in Table 1.

as a facilitator of morphogenetic events and as a factor promoting cell survival. The effect of HA on cell survival and morphogenesis is presumably a function of signal transduction via CD44 and the regulation of gene expression. Investigation of the mRNA expression pattern induced by HA in this model showed increased expression of genes known to regulate proliferation, apoptosis, and cell–matrix interactions (Table 1), further supporting an active role for HA in the morphogenetic process. Among the genes with increased expres-

sion in the presence of HA were C/EBP, a CCAAT-binding transcription factor, which is thought to be important in mammary gland ductal development (Seagroves et al., 1998); factors implicated in cell survival and proliferation (c-Myc-related transcription factor, PCNA, BAG-1, BID, Grp78); and other molecules likely to be involved in morphogenesis (MMP-11, integrin ␣7, CD44, ezrin villin 2 NF-2 (merlin)-related filament). Interestingly, integrin ␣7 is known to mediate cell adhesion and migration on laminin isoforms and is developmentally regulated during muscle development (Echtermeyer et al., 1996; Kaufman et al., 1985; Yao et al., 1996). Thus, the patterns of gene expression seem consistent with the observed morphogenetic changes in the model, in that genes governing cell survival and the cellular communication with the ECM were upregulated under conditions in which the overwhelming majority of genes were not differentially regulated. Obviously, the individual significance of differentially regulated genes in the morphogenetic process cannot be deduced from this investigation, but the overall pattern adds further support to the proposal that HA sustains cell viability and facilitates morphogenesis. It is intriguing that CD44, the main cell surface receptor for HA (Miyake et al., 1990), is upregulated by soluble factors in the media under conditions favoring morphogenesis and by exposure of the cells to its ligand, HA. Together with the finding of Has expression by the UB itself, this suggests the existence of a positive feedback loop, which may drive process formation and iterative branching and ensure cell survival. In another context, disruption of the HA–CD44 interaction by transfection of a malignant cell line with soluble (truncated) CD44 led to apoptosis of the metastatic tumor cells after injection of the cells into mice (Yu et al., 1997). Fibroblasts, treated with blocking antiCD44 antibodies, underwent apoptosis in threedimensional fibrin gels (Henke et al., 1996). Together with our results, these data seem to establish a role for the

TABLE 1 Genes with Transcription Increasing More Than Log 0.25 CCAAT-binding transcription factor (C/EBP) Integrin ␣7 Glucose-regulated protein 78 kDa (Grp78) Ezrin villin 2 NF-2 (merlin)-related filament BAG-1, bcl-2 binding protein with anti-cell death activity BID, apoptic death agonist Transducin ␤-2 subunit GST Pi 1, glutathione S-transferase Pi 1 preadipocyte growth factor HSP84, heat-shock 84-kDa protein Stromelysin-3 matrix metalloproteinase-11 (MMP-11) c-Myc-related transcription factor CD14 antigen CD44 antigen PCNA, proliferating cell nuclear antigen processivity factor

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.

324

Pohl et al.

HA–CD44 interaction in maintaining cell survival in different cell types and conditions, including developmental settings. Thus, we propose that cycles of branching morphogenesis of embryonic UB cells are facilitated by HA through effects on cell survival, cellular process formation, and branching. HA, presumably acting through CD44 at the leading edges of UB cell processes, facilitates branching morphogenesis of UB cells, leading to specific patterns of gene expression (including increases in CD44) favoring cell proliferation, survival, and morphogenesis. In the embryonic kidney, HA, CD44, and HA synthases are expressed in a spatiotemporal fashion consistent with a key role in migration and branching morphogenesis. Through its capacity to spatially coordinate chondroitin sulfate proteoglycans and its interaction with CD44, which could influence local effector molecule concentrations by its heparan sulfate proteoglycan chains and connections to ERM proteins and the actin cytoskeleton, HA could act as a central regulator of epithelial branching morphogenesis. Furthermore, the possibility that HA may exist in different sizes with different morphogenetic activities, especially in the context of multiple CD44 and Has isoforms, suggests the potential for a complex interplay of ligand and receptor during different steps in the evolution of repetitively branching structures.

ACKNOWLEDGMENTS This work was supported by an NIH grant (DK 49517) and an Established Investigator Award from the American Heart Association to S. K. Nigam. M. Pohl was supported by the Deutsche Forschungsgemeinschaft (DFG), H. Sakurai was in part supported by a fellowship grant from Uehara Memorial Life Science Foundation, and R. O. Stuart was supported by a National Institute of Diabetes and Digestive and Kidney Diseases grant.

REFERENCES al-Awqati, Q., and Goldberg, M. R. (1998). Architectural patterns in branching morphogenesis in the kidney. Kidney Int. 54, 1832– 1842. Barasch, J., Pressler, L., Connor, J., and Malik, A. (1996). A ureteric bud cell line induces nephrogenesis in two steps by two distinct signals. Am. J. Physiol. 271, F50 –F61. Belsky, E., and Toole, B. P. (1983). Hyaluronate and hyaluronidase in the developing chick embryo kidney. Cell Differ. 12, 61– 66. Bennet (1980). Morphogenesis of branching tubules in cultures of cloned mammary epithelial cells. Nature 285, 657– 659. Bourguignon, L. Y., Gunja-Smith, Z., Iida, N., Zhu, H. B., Young, L. J., Muller, W. J., and Cardiff, R. D. (1998). CD44v(3,8-10) is involved in cytoskeleton-mediated tumor cell migration and matrix metalloproteinase (MMP-9) association in metastatic breast cancer cells. J. Cell. Physiol. 176, 206 –215. Echtermeyer, F., Schober, S., Poschl, E., von der Mark, H., and von der Mark, K. (1996). Specific induction of cell motility on laminin by alpha 7 integrin. J. Biol. Chem. 271, 2071–2075.

Gakunga, P., Frost, G., Shuster, S., Cunha, G., Formby, B., and Stern, R. (1997). Hyaluronan is a prerequisite for ductal branching morphogenesis. Development 124, 3987–3997. Helander, T. S., Carpen, O., Turunen, O., Kovanen, P. E., Vaheri, A., and Timonen, T. (1996). ICAM-2 redistributed by ezrin as a target for killer cells. Nature 382, 265–268. Henke, C., Bitterman, P., Roongta, U., Ingbar, D., and Polunovsky, V. (1996). Induction of fibroblast apoptosis by anti-CD44 antibody: Implications for the treatment of fibroproliferative lung disease. Am. J. Pathol. 149, 1639 –1650. Itano, N., Sawai, T., Yoshida, M., Lenas, P., Yamada, Y., Imagawa, M., Shinomura, T., Hamaguchi, M., Yoshida, Y., Ohnuki, Y., Miyauchi, S., Spicer, A. P., McDonald, J. A., and Kimata, K. (1999). Three isoforms of mammalian hyaluronan synthases have distinct enzymatic properties. J. Biol. Chem. 274, 25085–25092. Kanwar, Y. S., Carone, F. A., Kumar, A., Wada, J., Ota, K., and Wallner, E. I. (1997). Role of extracellular matrix, growth factors and proto-oncogenes in metanephric development. Kidney Int. 52, 589 – 606. Kaufman, S. J., Foster, R. F., Haye, K. R., and Faiman, L. E. (1985). Expression of a developmentally regulated antigen on the surface of skeletal and cardiac muscle cells. J. Cell Biol. 100, 1977–1987. Knudson, W. (1998). The role of CD44 as a cell surface hyaluronan receptor during tumor invasion of connective tissue. Front. Biosci. 3, D604 – 615. Laurent, T. C., and Fraser, J. R. (1992). Hyaluronan. FASEB J. 6, 2397–2404. Lechner, M. S., and Dressler, G. R. (1997). The molecular basis of embryonic kidney development. Mech. Dev. 62, 105–120. Mignatti, P., and Rifkin, D. B. (1993). Biology and biochemistry of proteinases in tumor invasion. Physiol. Rev. 73, 161–195. Miyake, K., Underhill, C. B., Lesley, J., and Kincade, P. W. (1990). Hyaluronate can function as a cell adhesion molecule and CD44 participates in hyaluronate recognition. J. Exp. Med. 172, 69 –75. Morriss-Kay, G. M., Tuckett, F., and Solursh, M. (1986). The effects of Streptomyces hyaluronidase on tissue organization and cell cycle time in rat embryos. J. Embryol. Exp. Morphol. 98, 59 –70. Naldini, L., Vigna, E., Bardelli, A., Follenzi, A., Galimi, F., and Comoglio, P. M. (1995). Biological activation of pro-HGF (hepatocyte growth factor) by urokinase is controlled by a stoichiometric reaction. J. Biol. Chem. 270, 603– 611. Oliferenko, S., Kaverina, I., Small, J. V., and Huber, L. A. (2000). Hyaluronic acid (HA) binding to CD44 activates rac1 and induces lamellipodia outgrowth. J. Cell Biol. 148, 1159 –1164. Pavlova, A., Stuart, R. O., Pohl, M., and Nigam, S. K. (1999). Evolution of gene expression patterns in a model of branching morphogenesis. Am. J. Physiol. 277, F650 –F663. Pohl, M., Stuart, R. O., Sakurai, H., and Nigam, S. K. (2000). Branching morphogenesis during kidney development. Annu. Rev. Physiol. 62, 595– 620. Qiao, J., Sakurai, H., and Nigam, S. K. (1999). Branching morphogenesis independent of mesenchymal– epithelial contact in the developing kidney. Proc. Natl. Acad. Sci. USA 96, 7330 –7335. Sakurai, H., Barros, E. J., Tsukamoto, T., Barasch, J., and Nigam, S. K. (1997). An in vitro tubulogenesis system using cell lines derived from the embryonic kidney shows dependence on multiple soluble growth factors. Proc. Natl. Acad. Sci. USA 94, 6279 – 6284. Sakurai, H., and Nigam, S. K. (1997). Transforming growth factorbeta selectively inhibits branching morphogenesis but not tubulogenesis. Am. J. Physiol. 272, F139 –146.

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.

325

Hyaluronan and CD44 in in Vitro Branching Morphogenesis

Sakurai, H., and Nigam, S. K. (1998). In vitro branching tubulogenesis: Implications for developmental and cystic disorders, nephron number, renal repair, and nephron engineering. Kidney Int. 54, 14 –26. Santos, O. F., and Nigam, S. K. (1993). HGF-induced tubulogenesis and branching of epithelial cells is modulated by extracellular matrix and TGF-beta. Dev. Biol. 160, 293–302. Schoenwolf, G. C., and Fisher, M. (1983). Analysis of the effects of Streptomyces hyaluronidase on formation of the neural tube. J. Embryol. Exp. Morphol. 73, 1–15. Screaton, G. R., Bell, M. V., Jackson, D. G., Cornelis, F. B., Gerth, U., and Bell, J. I. (1992). Genomic structure of DNA encoding the lymphocyte homing receptor CD44 reveals at least 12 alternatively spliced exons. Proc. Natl. Acad. Sci. USA 89, 12160 – 12164. Seagroves, T. N., Krnacik, S., Raught, B., Gay, J., Burgess-Beusse, B., Darlington, G. J., and Rosen, J. M. (1998). C/EBPbeta, but not C/EBPalpha, is essential for ductal morphogenesis, lobuloalveolar proliferation, and functional differentiation in the mouse mammary gland. Genes Dev. 12, 1917–1928. Sherman, L., Wainwright, D., Ponta, H., and Herrlich, P. (1998). A splice variant of CD44 expressed in the apical ectodermal ridge presents fibroblast growth factors to limb mesenchyme and is required for limb outgrowth. Genes Dev. 12, 1058 –1071. Spicer, A. P., and McDonald, J. A. (1998). Characterization and molecular evolution of a vertebrate hyaluronan synthase gene family. J. Biol. Chem. 273, 1923–1932. Stuart, R. O., and Nigam, S. K. (2000). Developmental biology of the kidney In “The Kidney” (B. M. Brenner, Ed.), 6th ed., Vol. 1, Chap. 2, pp. 68 –92. Saunders, Philadelphia. Taipale, J., and Keski-Oja, J. (1997). Growth factors in the extracellular matrix. FASEB J. 11, 51–59. Tsukita, S., Oishi, K., Sato, N., Sagara, J., and Kawai, A. (1994). ERM family members as molecular linkers between the cell surface glycoprotein CD44 and actin-based cytoskeletons. J. Cell Biol. 126, 391– 401.

Tucker, A. S., Al Khamis, A., Ferguson, C. A., Bach, I., Rosenfeld, M. G., and Sharpe, P. T. (1999). Conserved regulation of mesenchymal gene expression by Fgf-8 in face and limb development. Development 126, 221–228. Turunen, O., Sainio, M., Jaaskelainen, J., Carpen, O., and Vaheri, A. (1998). Structure–function relationships in the ezrin family and the effect of tumor-associated point mutations in neurofibromatosis 2 protein. Biochim. Biophys. Acta 1387, 1–16. Yao, C. C., Ziober, B. L., Squillace, R. M., and Kramer, R. H. (1996). Alpha7 integrin mediates cell adhesion and migration on specific laminin isoforms. J. Biol. Chem. 271, 25598 –25603. Yonemura, S., Hirao, M., Doi, Y., Takahashi, N., Kondo, T., and Tsukita, S. (1998). Ezrin/radixin/moesin (ERM) proteins bind to a positively charged amino acid cluster in the juxta-membrane cytoplasmic domain of CD44, CD43, and ICAM-2. J. Cell Biol. 140, 885– 895. Yu, Q., and Stamenkovic, I. (1999). Localization of matrix metalloproteinase 9 to the cell surface provides a mechanism for CD44-mediated tumor invasion. Genes Dev. 13, 35– 48. Yu, Q., Toole, B. P., and Stamenkovic, I. (1997). Induction of apoptosis of metastatic mammary carcinoma cells in vivo by disruption of tumor cell surface CD44 function. J. Exp. Med. 186, 1985–1996. Zheng, Z., Katoh, S., He, Q., Oritani, K., Miyake, K., Lesley, J., Hyman, R., Hamik, A., Parkhouse, R. M., Farr, A. G., et al. (1995). Monoclonal antibodies to CD44 and their influence on hyaluronan recognition. J. Cell Biol. 130, 485– 495. Zoller, M., Herrmann, K., Buchner, S., Seiter, S., Claas, C., Underhill, C. B., and Moller, P. (1997). Transient absence of CD44 expression and delay in development by anti-CD44 treatment during ontogeny: A surrogate of an inducible knockout? Cell Growth Differ. 8, 1211–1223. Received for publication November 11, 1999 Revised April 18, 2000 Accepted May 3, 2000

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.