Insertional Mutation of the Collagen Genes Col4a3 and Col4a4 in a Mouse Model of Alport Syndrome

Genomics 61, 113–124 (1999) Article ID geno.1999.5943, available online at http://www.idealibrary.com on

Insertional Mutation of the Collagen Genes Col4a3 and Col4a4 in a Mouse Model of Alport Syndrome Wei Lu,* Carrie L. Phillips,† Paul D. Killen,† Tommy Hlaing,† Wilbur R. Harrison,‡ F. F. B. Elder,‡ Jeffrey H. Miner,§ Paul A. Overbeek, ¶ and Miriam H. Meisler* ,1 *Department of Human Genetics and †Department of Pathology, University of Michigan, Ann Arbor, Michigan 48109; ‡Department of Pathology and Laboratory Medicine, The University of Texas Health Science Center, Houston, Texas 77225; §Renal Division, Washington University School of Medicine, St. Louis, Missouri 63110; and ¶ Department of Cell Biology, Baylor College of Medicine, Houston, Texas 77030 Received May 28, 1999; accepted July 26, 1999

Mice homozygous for the transgenic insertion in line OVE250 exhibit severe progressive glomerulonephritis. Ultrastructural changes in the glomerular basement membrane (GBM) at 2 weeks of age resemble those in Alport syndrome. The transgenic insertion site was mapped by FISH to mouse chromosome 1 close to Pax3. Genetic and molecular analyses identified a deletion of genomic DNA at the transgene insertion site. Exons 1 through 12 of the collagen IV gene Col4a4, exons 1 and 2 of the adjacent Col4a3 gene, and the intergenic promoter region are deleted. Transcripts of Col4a3 and Col4a4 are undetectable in mutant kidney, and both proteins are missing from the GBM. Persistent cellular proliferation in mutant kidneys suggests that interaction with the extracellular matrix may be important for cell maturation. Evolutionarily conserved sequence elements in the promoter regions of human and mouse Col4a3 and Col4a4 include a 19-bp element that was tandemly duplicated in the human lineage and a CTC box element common to several genes encoding extracellular matrix proteins. This new animal model of Alport syndrome, Col4D3-4, lacks both a3 and a4 chains of collagen IV and exhibits an earlier disease onset than mice lacking a3 only. © 1999 Academic Press

INTRODUCTION

During the production of transgenic mice by microinjection of fertilized eggs, the random insertion of the injected DNA sometimes disrupts an endogenous mouse gene, resulting in loss of expression of the endogenous gene and generation of an insertional mutation. Approximately 3% of transgenes are associated Sequence data from this article have been deposited with the EMBL/GenBank Data Libraries under Accession Nnos. AF169387– AF169389. 1 To whom correspondence should be addressed at Department of Human Genetics, 4708 Medical Science II, University of Michigan, Ann Arbor, MI 48109-0618. Telephone: (734) 763-5546; Fax: (734) 763-9691. E-mail: [email protected].

with visible recessive mutant phenotypes, and another 10% result in recessive prenatal lethality (Meisler, 1992). The presence of transgene DNA at the mutant loci facilitates chromosomal mapping, evaluation of candidate genes, and cloning of the mutated genes (Meisler et al., 1997). This strategy has most recently been used to identify and isolate genes responsible for generation of left–right asymmetry (Mochizuki et al., 1998; Morgan et al., 1998), neuronal migration during cerebellar development (Ackerman et al., 1997), and neuronal signaling at the neuromuscular junction (Burgess et al., 1995; Meisler et al., 1997). We now report the phenotypic and molecular characterization of a transgene insertional mutation in a new mouse model of autosomal recessive Alport syndrome. The tyrosinase minigene in this mutant was designed to permit visual identification of transgenic mice and is particularly useful in insertional mutagenesis studies since a gene dosage effect allows heterozygous and homozygous mice to be visibly identified (Overbeek et al., 1991). Alport syndrome is a progressive hereditary disease of the glomerular basement membrane (GBM) that can result from mutations in three different collagen IV genes (Hudson et al., 1993; Kashtan and Michael, 1993). The more common X-linked disorder is caused by mutations in COL4A5 (Barker et al., 1990; Knebelmann et al., 1996). The rare autosomal recessive forms are associated with mutations in COL4A3 or COL4A4 (Lemmink et al., 1994; Mochizuki et al., 1994). Type IV collagen is a major structural component of basement membranes. Each collagen IV molecule is a triple helix of monomeric collagen polypeptides called a chains. The type IV a chains are encoded by six genes that are dispersed as three head-to-head pairs: COL4A1 and COL4A2 on human chromosome 13q34, COL4A3 and COL4A4 on chromosome 2q36, and COL4A5 and COL4A6 on chromosome Xq22 (Heikkila¨ and Soininen, 1996). The proteins encoded by the collagen IV genes are referred to as a1(IV) to a6(IV). Each pair of linked genes is separated by a small promoter

113

0888-7543/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.

114

LU ET AL.

TABLE 1 PCR Primers and Conditions Primer

Gene

Sequence

PCR condition

A B C D E F G H I J K L M N O P Q R S T U

Col4a3 exon 2 F Col4a3 exon 3 R Col4a3 exon 3 F Col4a3 exon 4 R Col4a4 exon 11 F Col4a4 exon 12 R Tyrosinase exon 2 R Tyrosinase exon 1 F Col4a1 cDNA F Col4a1 cDNA R Col4a3 cDNA F Col4a3 cDNA R Col4a4 cDNA F Col4a4 cDNA R Col4a5 cDNA F Col4a5 cDNA R GAPDH F GAPDH R Col4a3 cDNA nt1360 R Col4a4 cDNA nt1045 F Col4a4 cDNA nt1845 R

TCTGCAAAGGCAAAGGACAATG AAGCCTTCAGGACCTGGAAATC AAGGATTTCCAGGTCCTGAAGG ATGCCTTTGGGGCCAGTGAG TGGGTTGAAGGGAAATCCTGC AATCAGGCGGCTGTACCAACAG GATTACGTAATAGTGGTCCCTCAGG CTGTCCAGTGCACCATCTGGACCTC CTTCCTTGTGACCAGGCATA GGCTTCTTGAACATCTCGCT TCACCCGACACAGTCAAACC AAGCCAGCCAGAAACTGTAG TTCCTCCTGGTTCTCCACAG CCTGGCACCTGCTGATTTTC CTCTCCTGTATGTACAAGGA CTCAAGTCTCCTGCTTTCAG TCACCATCTTCCAGGAGCG CTGCTTCACCACCTTCTTGA TTCCCTTAATGCCAGCTTCAC TTGGTACAGCCGCCTGATTTG TTCTCCATGCCAACCAGGGAG

I I II II I I III, annealing at 62°C III, annealing at 62°C IV (pH 8.5, 2 mM MgCl 2) IV IV IV IV, pH 9.0 IV, pH 9.0 IV, 2.5 mM MgCl 2 IV, 2.5 mM MgCl 2 IV, pH 9.0 IV, pH 9.0 III III III

Note. F, forward; R, reverse. PCR conditions are described under Materials and Methods.

region of less than 500 bp (Burbelo et al., 1988; Kaytes et al., 1988; Momota et al., 1998; Poschl et al., 1988; Sugimoto et al., 1994). The a1(IV) and a2(IV) chains are widely distributed in basement membranes of many tissues. The a3, a4, and a5 chains are most abundant in lung and kidney (Zhou and Reeders, 1996). In kidney, these chains are thought to form a molecular network that does not include the a1 and a2 chains and that is restricted to glomeruli and a subset of tubules (Kashtan and Kim, 1992). Targeted disruption of the a3(IV) gene in mice results in glomerulonephritis (Miner and Sanes, 1996; Cosgrove et al., 1996). In the dog, both X-linked and autosomal recessive forms of Alport syndrome have been described (Zheng et al., 1994; Lees et al., 1998). The renal glomerulus has a thick compound basement membrane that arises from the developmental fusion of two discrete basal laminae associated with vascular endothelial and visceral epithelial cells (podocytes). This basement membrane is postulated to play an important role in resisting the extraordinary transmural hydrostatic pressure gradient necessary to produce a voluminous ultrafiltrate. Alport patients have a thin, multilaminated basement membrane that is thought to be more fragile than normal. The resulting disruption of the basement membrane is presumably the cause of microscopic and episodic gross hematuria. With time, the glomerular basement membrane becomes

structurally disorganized and thickens with multiple interwoven layers. Similar structural abnormalities are observed in the mouse model described below. MATERIALS AND METHODS Animals. We generated the transgenic founder mouse for line OVE250 by methods described previously (Yokoyama et al., 1990). The 4.1-kb tyrosinase minigene construct TyBS, containing the mouse tyrosinase cDNA under the control of a 2.25-kb endogenous promoter fragment, was microinjected into one-cell embryos of the inbred albino strain FVB/N using standard techniques (Hogan et al., 1986). Transgenic offspring from this line can be recognized by their grayish coat color, with homozygotes darker than heterozygotes. The line is maintained by crossing with strain FVB/N. Isolation of the transgene insertion site. Tg/1 genomic DNA prepared from spleen was partially digested with MboI and cloned into the cosmid vector ScosSal, as previously described (Ting et al., 1994; Burgess et al., 1995; Meisler et al., 1997). Approximately 4 3 10 5 colonies were plated on nitrocellulose filters and screened by hybridization with a radiolabeled tyrosinase minigene probe (Overbeek et al., 1991). Positive hybridizing colonies were confirmed by PCR using transgene-specific primers. Cosmid DNA was digested with Sau3A and subcloned into pUC18 using the Ready-To-Go kit (Pharmacia). Fluorescence in situ hybridization (FISH). Spleen lymphocytes from transgenic mice were cultured and metaphase slides were prepared as described (Robinson and Elder, 1987). Slides were aged at 37°C for 3 days and GTG banded, and representative metaphases were photographed. The slides were then destained in 70% ethanol and treated at room temperature for 2.5 min in 3% buffered formalin, 4.5 min in 0.2 N HCl, and 30 min in 23 SSC at 37°C. FISH was then

FIG. 1. Renal histopathology. Light micrographs of renal cortex from wildtype and mutant mice. A and B were stained with PAS; original magnification 4253. C and D were stained with trichrome; original magnification 1703. (A, C) 1/1 controls; (B, D) tg/tg homozygous mutant mice. FIG. 2. Ultrastructure of glomerular capillaries. Representative segments of the glomerular basement membrane (GBM) at 2 and 5 weeks of age. The fenestrated endothelium is to the left of the GBM, and the podocytes are to the right. A schematic representation is shown below the electron micrographs. Original magnification: 2 weeks 36,8003; 5 weeks, 18,4003.

DELETION OF Col4a3 AND Col4a4 CAUSES KIDNEY DISEASE

115

116

LU ET AL.

carried out using digoxigenin-labeled TyBS minigene DNA as a probe for the transgenic insertion site. For accurate assignment of the transgenic insertion site, previously photographed G-banded metaphases were relocated and rephotographed on an Olympus BX60 microscope equipped with epifluorescence. In the two-color FISH experiments, cosmid DNA for the villin, Pax3, and Acrd genes was labeled with biotin (Gibco BRL, Gaithersburg, MD) and cohybridized to slides with the digoxigenin-labeled TyBS probe. Biotinlabeled probes were detected with avidin CY-3 (Amersham), and the digoxigenin label was detected with anti-digoxigenin FITC (Boehringer Mannheim). Cosmid DNA containing the villin gene was obtained from Deborah Gumucio (University of Michigan, Ann Arbor, MI), the Pax3 gene was obtained from Philippe Gros (McGill University, Montreal, Quebec), and the Acrd gene was obtained from James Patrick (Baylor College of Medicine). Northern blots and RT-PCR. Kidneys from 5-week-old mice were sonicated in Trizol Reagent (Gibco BRL). Total RNA was purified according to the manufacturer’s instructions and dissolved in RNasefree water. Aliquots of RNA (10 mg) were electrophoresed through 1.0% agarose/formaldehyde/MOPS gels and transferred to a Zetaprobe GT membrane (Bio-Rad). Filters were hybridized with murine Col4a3 cDNA (GenBank Accession No. AF169387) or Col4a4 cDNA cDNA (GenBank Accession No. AF169388) that was radiolabeled with [a- 32P]dCTP by random priming. For RT-PCR, oligo(dT)-primed cDNA was reverse transcribed from 20 mg of total RNA and diluted to bring the amplification products within the linear range of the assay (Lee et al., 1997). RT-PCRs were carried out as previously described (Lee et al., 1997) except that the time at 94°C in each cycle was 1.5 min. Primer sequences are given in Table 1. The amplification products were resolved by PAGE, and the radioactivity was quantitated using a phosphorimager (Molecular Dynamics, Sunnyvale, CA). Genomic Southern blots. Genomic DNA was prepared from spleens of homozygous transgenic mice and normal FVB/N mice. A 10-mg aliquot of DNA was digested with HindIII, electrophoresed through 1.0% agarose gel, and transferred to Zetaprobe GT membrane. Blots were hybridized at 65°C overnight in 0.25 M Na 2HPO 4, pH 7.2, containing 7% SDS and 1 mM EDTA, washed under stringent conditions at 65°C in 0.23 SSC, and exposed to Kodak Biomax film. Isolation of the promoter region for mouse Col4a3 and Col4a4. Recombinant phage (1.25 3 10 6) from a library of mouse genomic DNA cloned in Lambda FIX were lifted and hybridized according to Church and Gilbert (1984). The [a- 32P]dCTP-labeled mouse cDNA probes were prepared from the Col4a3 cDNA (nt 1–773, GenBank Accession No. AF169387) or the 59UTR of the Col4a4 cDNA (nucleotides 1–303, GenBank Accession No. AF169388). A 5.5-kb HindIII subclone was sequenced. PCR of genomic DNA. Four PCR conditions were used. Under condition I, the Expand High Fidelity PCR System (Boehringer Mannheim) was used with the following cycling conditions: 93°C for 3 min followed by 10 cycles of 93°C for 30 s; 60°C for 30 s; 68°C for 2 min 30 s and 20 cycles of 93°C for 30 s; 60°C for 30 s; 68°C for 2 min 30 s 1 20 s/cycle and a final extension at 68°C for 7 min. Under condition II, the Expand Long Template PCR System (Boehringer Mannheim) was used with the following cycling conditions: 93°C for 2 min followed by 10 cycles of 93°C for 10 s; 60°C for 30 s; 68°C for 10 min and 20 cycles of 93°C for 10 s; 60°C for 30 s; 68°C for 10 min 1 20 s/cycle and a final extension at 68°C for 7 min. Under condition III, standard PCR was performed at 94°C for 4 min followed by 30 cycles of 94°C for 30 s; 60°C for 40 s; 72°C for 3 min and a final extension at 72°C for 8 min. For condition IV, See Lee et al. (1997). Transgene-positive cosmids were identified with the TyBS-specific primers G and H (Table 1) using PCR condition III, with the annealing temperature at 62°C. For routine identification of transgenic mice, genomic DNA was amplified with a mixture of primers A, B, and G or a mixture of primers E, F, and G, using PCR condition II. To generate cDNA probes 1 and 3 for the genomic Southern hybridizations, full-length mouse Col4a4 cDNA was amplified with primers T and U (Table 1), and full-length Col4a3 cDNA was amplified with primers C and S using condition III.

Histology. Formalin-fixed tissues were paraffin embedded, sectioned at 3 mm, and stained with hematoxylin and eosin, periodic acid Schiff, or Masson’s Trichrome. Tissue for electron microscopy was fixed in a solution of 3% glutaraldehyde and 3% formaldehyde buffered with cacodylate, pH 7.3 (Tousimis Research Corp., Rockville, MD). Immunohistochemistry. Kidneys were frozen in liquid nitrogencooled isopentane and sectioned at 7 mm on a cryostat. Sections were fixed and denatured with acid urea as previously described (Miner and Sanes, 1994). Rabbit anti-mouse collagen a1 and a2 (IV) antibodies were purchased from Collaborative Biomedical Products/Becton Dickinson Labware (Franklin Lakes, NJ). Rabbit anti-mouse collagen a3-a5(IV) antisera have been described (Miner and Sanes, 1994). Primary antibodies were diluted 1:400 in PBS containing 1% BSA and applied for 1 to 2 h. FITC-conjugated goat anti-rabbit second antibody (ICN Pharmaceuticals, Inc., Costa Mesa, CA) was diluted 1:200 in PBS/BSA and applied for 1 h. Sections were viewed and photographed on an Olympus Vanox S microscope. Incorporation of BrdU. Adult mice were injected intraperitoneally with 5-bromo-29-deoxyuridine (BrdU) (Sigma) and 5-fluoro-29deoxyuridine (Sigma), an inhibitor of thymidine synthesis, in phosphate-buffered saline (PBS) [10 mM NaPO4 (pH 7.5), 0.9% NaCl] at 100 mg of BrdU per gram of body weight and 6.7 mg of 5-fluoro-29deoxyuridine per gram of body weight and sacrificed 1 h later. Kidneys were fixed in 10% formalin, embedded in paraffin, and sectioned at 5 mm. Immunohistochemistry was performed as previously described (Fromm et al., 1994).

RESULTS

Renal Failure in Homozygous Transgenic Mice We generated the transgenic line OVE250 by microinjection of fertilized eggs. During breeding of this line, unexpected lethality was observed in the homozygous transgenic mice at 10 –14 weeks of age. Renal disease was immediately suggested by the pale appearance of the kidneys. Routine histopathology of all of the major organs demonstrated morphological abnormalities in kidney only (see below). Analysis of urine revealed persistant hematuria and proteinuria that were detectable as early as 2 weeks of age. Urinary protein was 10-fold elevated at 1 month of age. Blood urea nitrogen (BUN) was normal at 6 weeks of age (24 –28 mg/dl), but was 10-fold elevated at 12 weeks. Light microscopy detected intratubular red blood cells and protein casts in the kidney at 2 weeks of age (not shown). At 5 weeks, the glomeruli were enlarged with hyperplasia of parietal epithelial cells (Figs. 1A and 1B). Glomerular lesions were characterized by protein-rich exudates in the urinary space and crescentic glomerulonephritis (Figs. 1C and 1D). Occasional glomeruli demonstrated increased mesangial or endocapillary cells, and rare segments contained karyorrhectic cellular debris. There was evidence of tubular injury that progressed by 8 weeks to tubular atrophy that was accompanied by glomerular obsolescence and interstitial fibrosis (data not shown). Ultrastructural analysis demonstrated that the GBM was thin and focally duplicated at 2 weeks of age (Fig. 2). In older animals, the GBM became thicker and more highly disorganized with the characteristic “basketweaving” of the lamina densa characteristic of Alport syndrome (Fig. 2). No immune deposits were observed. The

DELETION OF Col4a3 AND Col4a4 CAUSES KIDNEY DISEASE

TABLE 2

117

and Col4a4 were tested as candidate genes for the mutation (see below).

Cell Proliferation within the Kidney Glomerulus of Mutant Mice

Expression of Collagen IV a3 and a4 in Transgenic Kidney

Number of labeled cells Genotype

Glomeruli

Parietal epithelial

Endocapillary

Total

tg/tg tg/1

100 100

74 1

98 5

172 6

Note. Proliferating cells were labeled by in vivo incorporation of BrdU for 1 h.

kidneys of transgenic heterozygotes were ultrastructurally indistinguishable from wildtype (1/1) littermates up to 8 weeks of age, the latest point examined. Proliferating Cells in Mutant Kidney To evaluate the origin of the hyperplasia, we measured cellular proliferation by administration of BrdU to control (tg/1) and mutant (tg/tg) mice. In the control mice at 4 weeks of age, a low level of BrdU incorporation into tubular epithelial cells was observed (Fig. 3A). In mutant mice, BrdU-positive cells were visible both in the tubular epithelium (Fig. 3B) and in the glomeruli (Fig. 3C). Approximately half of the proliferating glomerular cells appeared to be parietal epithelial cells lining Bowman’s capsule (Table 2). The remaining BrdU stained cells were located within the glomerular tufts (endocapillary cells, Table 2). These cells may be endothelial cells, mesangial cells, or infiltrating inflammatory cells, but are unlikely to be podocytes based on their location. Proliferation continued in older mice, indicating that the mutant kidneys are undergoing persistent remodeling. Chromosomal Mapping of the Insertion Site and Tissue-Specific Expression of the Transgene The tyrosinase promoter of the transgene construct is normally expressed in melanocytes and the retinal pigment epithelium (Beermann et al., 1990; Yokoyama et al., 1990). To determine whether ectopic expression of tyrosinase in the transgenic mice might be responsible for the renal pathology, kidney RNA was analyzed. The predicted 800-bp product of the transgenic transcript was amplified by RT-PCR from skin RNA, but not from kidney, liver, or heart (data not shown). To localize the transgenic insertion site, the tyrosinase minigene was used as a probe for FISH (Fig. 4). The transgene hybridized to mouse chromosome 1, distal to villin (Fig. 4D) and proximal to Acrd (Fig. 4E). The hybridization signal from Pax 3 could not be resolved from the transgene signal, indicating that the two are closely linked (Fig. 4F). The orthologous human genes VIL, PAX3, and ACRD map to a conserved linkage group on chromosome 2q36, close to the genes COL4A3 [encoding a3(IV) collagen] and COL4A4 [encoding a4(IV) collagen] (Mariyama et al., 1992; Lu-Kuo et al., 1993). Because of their role in human Alport syndrome, mouse Col4a3

The distribution of Col4a3 and Col4a4 proteins was examined by immunocytochemistry (Fig. 5). In normal controls, collagen IV a3, a4, and a5 chains are localized to the GBM and a subset of tubular basement membrane (TBM); the a1 and a2 chains are distributed throughout the TBM but in mature glomeruli they are largely restricted to the mesangial matrix. In kidneys from tg/tg mice, the a3 and a4 proteins were absent from the GBM (Fig. 5). In addition, a5(IV) protein was missing (Fig. 5), as predicted by the previously observed dependence of a5(IV) protein stability on association with the a3(IV) and a4(IV) proteins (Gubler et al., 1995; Cosgrove et al., 1996; Miner and Sanes, 1996). Abnormally high levels of a1 and a2 chains were detected in the capillary loops of the GBM in tg/tg mice (Fig. 5). Hybridization of Northern blots with a full-length Col4a3 cDNA probe detected an 8-kb transcript in RNA from tg/1 kidney, but no transcript was detected in RNA from tg/tg homozygous kidney (Fig. 6A). Analysis of mRNA levels with a more sensitive and semiquantitative RT-PCR assay (Lee et al., 1997) indicated that both Col4a3 and Col4a4 transcripts were reduced to ,1% of normal levels in tg/tg homozygotes (Fig. 6B). In contrast, transcripts of Col4a1 and Col4a5 were present in the mutant kidney (Fig. 6B), suggesting that the loss of the a5 protein results from posttranscriptional events. Interruption of the Col4a3 and Col4a4 Genes by the Transgenic Insertion Genomic DNA from tg/tg and 1/1 mice was hybridized with Col4a3 and Col4a4 cDNA probes (Fig. 7). Probes 1 and 3 contained the 39 portions of the cDNAs and hybridized to identical fragments in transgenic and wildtype DNA. Probe 2, containing the shared promoter and the first exon of Col4a3 and Col4a4, hybridized with a 4.7-kb fragment in 1/1 DNA that was absent in the tg/tg homozygote. This result indicates that the 59 portions of Col4a3 and Col4a4 and the shared intergenic promoter region are deleted in the transgenic mice. PCR primers corresponding to predicted exons of Col4a4 and Col4a3 were synthesized, based on the published structures of human collagen IV genes (Heikkila¨ and Soininen, 1996) (Fig. 8A). Interexon PCR was carried out to detect specific exons. Amplification from exon 3 to exon 4 of Col4a3 produced a normal 6-kb product from tg/tg DNA (Fig. 8B, lanes 3 and 4), but amplification from exon 2 to exon 3 produced no product from tg/tg DNA (lanes 1 and 2), indicating that exon 2 is deleted. Amplification from exon 12 to exon 13 of Col4a4 (lane 6) produced a 2.2-kb product from wildtype DNA and no product from transgenic DNA (Fig.

118

LU ET AL.

FIG. 3. Increased cell proliferation in mutant kidney. Incorporation of BrdU was assayed at 4 weeks of age. Cells in S-phase are stained brown by DAB. (A) tg/1 heterozygote (1703); (B) tg/tg homozygote (1703); (C) tg/tg (4253).

8B, lanes 6 and 7). Exon 13 was intact in a cosmid clone containing transgene DNA (not shown), indicating that the deletion terminates near the end of exon 12 or within intron 12 of Col4a4 (Fig. 8A). PCR primers were designed to amplify the transgene junctions from genomic DNA. Using a mixture of primers A, B, and G for the Col4a3-transgene junction, a 2-kb product (A and B) was obtained from 1/1 DNA and a 6-kb product (B to G) from tg/tg DNA. Both products are amplified from tg/1 DNA (Fig. 8C, lanes 1–3). For the Col4a4 junction, similar results were obtained (Fig. 8C, lanes 5–7). The amplification of the junction fragments confirms the extent of the collagen gene deletion and the orientation of transgene copies shown in Fig. 8A. Based on the intron sizes of human COL4A3 and COL4A4 genes (Boye et al., 1998; Momota et al., 1998), we estimate that the length of the

deletion is 50 kb. This deletion, with the systematic designation Del(1) TgN250OVE, is referred to in this report as Col4D3-4. Characterization of the Col4a3–Col4a4 Promoter The intergene region including the first exons of Col4a3 and Col4a4 genes was sequenced (GenBank Accession No. AF169389). Alignment with the corresponding human sequence (Momota et al., 1998) demonstrated 72% nucleotide identity for the 186 bp between the termination of the cDNA clones for each gene that extend furthest in the 59 direction (Fig. 9). Conserved consensus sequence motifs include an E-box in the first exon of Col4a3, two CACCC motifs (one included in Fig. 9), and two CTC boxes in the first exon of Col4a3 (not shown). Only one copy of the 19-nt tandem repeat in the human sequence was present

FIG. 4. Chromosomal localization of the transgene insertion site. (A) Partial G-banded metaphase. Arrow, transgene insertion site. (B) Same partial metaphase after hybridization with the transgene probe (arrow). (C) Standard idiogram of mouse chromosome 1 (Lyon and Searle, 1989), indicating the region of transgene insertion. (D–F) Double-color FISH using probes for the transgene detected by FITC (green) and for marker genes detected with Avidin CY-3 (red). (D) The hybridization signal for the transgene probe (small arrow) is distal to the signal for villin (large arrow). (E) Transgene hybridization (small arrow) is proximal to the probe for Acrd (large arrow). (F) Hybridization signals for transgene probe and a cosmid containing Pax3 are superimposed. On the consensus genetic map of mouse chromosome 1, the marker genes are located at 41 cM (villin), 44 cM (Pax3), and 52 cM (Acrd) from the centromere (www.informatics.jax.org).

DELETION OF Col4a3 AND Col4a4 CAUSES KIDNEY DISEASE

119

FIG. 5. Immunohistochemical analysis of collagen IV chains in mutant kidney. Frozen sections were stained with specific antisera to each of the collagen IV a chains, followed by FITC-conjugated secondary antibody, as described under Materials and Methods.

in mouse, indicating that the tandem duplication occurred subsequent to divergence of the two species. One nucleotide within these 19 is altered in the mouse sequence (Fig. 9). DISCUSSION

The Col4D3-4 deletion at the transgene insertion site in line OVE250 removed the 59 portions of two collagen IV genes and their shared promoter region. Col4D3-4 kidneys exhibit morphological and ultrastructural features characteristic of human Alport syndrome, including

disorganization and multilamellar structure of the glomerular basement membrane and delayed onset glomerulonephritis. The age of onset of hematuria and endstage renal disease varies among Alport syndrome kindreds. Persistent hematuria may be detected in males within the first year of life. Renal failure may develop in early childhood, adolescence, or middle age. The early onset of hematuria and proteinuria (2 weeks of age) and rapid progression to renal failure (8 to 12 weeks) in the COL4D3-4 mice resembles the juvenile form of human Alport syndrome (Gregory et al., 1996).

120

LU ET AL.

FIG. 6. Absence of Col4a3 and Col4a4 transcripts in homozygous transgenic mice. (A) Northern blot. A filter containing 10 mg of each RNA was hybridized with radiolabeled full-length a3 cDNA. (B) RT-PCR amplification. Primers sequences are given in Table 1 (I to R). The GAPDH mRNA level was used as a normalizing control for quantitation.

Renal disease in patients with Alport syndrome usually progresses slowly but inexorably to renal failure. This slow progression may be associated with the development of secondary glomerular changes as a consequence of reduced numbers of functional nephrons. The light microscopic findings in patients include hypertrophy of epithelial cells in the glomerulus, capillary wall thickening, tubular dilatation and atrophy, and interstitial fibrosis. All of these features are observed in the COL4D3-4 mice. Although fetal-like glomeruli have been reported in patients with Alport syndrome, this was not observed in the mice. The glomerular hyperplasia (crescents) observed in COL4D3-4 mice (Fig. 1) or mice with targeted inactivation of Col4a3 (Miner and Sanes, 1996) resemble those in human immune mediated crescentic glomerulonephritis. However, the COL4D3-4 kidneys do not contain immune deposits, and the serum does not contain anti-dsDNA or anti-ssDNA activity (unpublished observations), both common features of murine and human immune mediated glomerulonephritis. Glomerular crescents are only occasionally observed in patients with human Alport syndrome (Gregory et al., 1996). Cresents are most commonly noted when there is severe glomerular inflammation with proteolysis and oxidant-mediated disruption of the GBM and subsequent hemorrhage into the urinary space. Tissue procoagulants are necessary to activate the coagulation cascade leading to fibrin polymerization and generating a variety of proinflammatory mediators, thereby perpetuating and amplifying the inflammatory response (Bergstein, 1990). In the absence of procoagulant activity, crescents fail to develop in models of anti-GBM antibody-induced crescentic glomerulone-

phritis. There appears to be a species difference in these secondary responses to the disruption of the GBM. An interesting and novel finding in the COL4D3-4 mutant was persistent proliferation of glomerular cells, detected by BrdU incorporation (Fig. 3). The proliferating cells include parietal epithelial cells of Bowman’s capsule and other cells located within the glomerular capillary tufts. Proliferation of parietal epithelial cells, endothelial cells, and mesangial cells accompanies many forms of glomerular disease (Johnson, 1994; Kim et al., 1999). The parietal epithelial cells are bathed by the urine filtrate and can respond to the altered factors present in the urinary space after disruption of the GBM (Bergstein, 1990). Another cellular abnormality is the effacement and flattening of the podocyte foot processes detected as early as 2 weeks of age (Fig. 2). Podocytes undergo little if any proliferation in response to injury, perhaps because their terminal differentiation limits proliferative capacity (Kriz et al., 1998). The rear-

FIG. 7. Disruption of the Col4a3 and Col4a4 genes by the transgene insertion. Ten micrograms of genomic DNA from wildtype mice of strain FVB (1/1) and homozygous transgenic (tg/tg) mice was digested with HindIII. The Southern blot was successively hybridized with probes 1, 2, and 3; before reprobing, the filter was washed to remove the previous probe. Probe 1 is a Col4a4 cDNA fragment (nt 1045 to 1845); probe 2 is a 470-bp genomic fragment containing the common promoter of the two collagen genes; probe 3 is a Col4a3 cDNA fragment Probe 3 (nt 557 to 1360). The probes are described in the text, and their positions in the locus are shown in Fig. 8A.

DELETION OF Col4a3 AND Col4a4 CAUSES KIDNEY DISEASE

121

FIG. 8. Characterization of the transgene-induced deletion. (A) Organization of the Col4a3 and Col4a4 genes and their shared intergenic promoter. The extent of the transgene induced deletion in line OVE250 is indicated. The sizes of most introns have not been determined. The locations of PCR primers and hybridization probes, and the orientation of the outermost copies of the transgene, are indicated. There are approximately six tandem copies of the transgene at the insertion site, based on intensity of hybridization on Southern blots (not shown). (B) Interexon PCR. Genomic DNA was amplified with the indicated primers (Table 1). (C) Exon-to-transgene PCR. Genomic DNA was amplified with a mixture of primers A, B, and G or primers E, F, and G.

rangement of podocyte foot processes may lead to decreased filtration capacity because of diminshed filtration area. COL4D3-4 mutant mice resemble the previously described mice with targeted mutation of Col4a3 with respect to the absence of a5 protein, the elevated levels and altered distribution of a1 and a2 proteins, and the overall clinical phenotype (Cosgrove et al., 1996; Miner and Sanes, 1996). The kidneys of COL4D3-4 mice were approximately 15% larger than kidneys from control littermates (data not shown), unlike the shrunken kid-

neys described for the Col4a3 knockout (Miner and Sanes, 1996). The earlier onset of hematuria and proteinuria (2 weeks versus 5 weeks) and the more rapid disease progression in COL4D3-4 mice may be related to the difference in genetic background, since this mutant is congenic on strain FVB, and the targeted mutations are maintained on a mixture of strains C57BL/6J and 129. Patients with mutations in different collagen genes have overlapping clinical symptoms, and within each group there is a wide spectrum of phenotypes, consistent with an effect of human genetic

122

LU ET AL.

FIG. 9. Evolutionary conservation of the Col4a3–Col4a4 intergenic promoter region. The sequence of the mouse promoter region (GenBank Accession No. AF169389) was aligned with the corresponding human promoter region (Momota et al., 1998). The depicted orientation of the collagen genes is opposite to that shown in Fig. 8. Pairwise sequence alignment was carried out using ALIGN through the BCM Search Launcher (http://kiwi.imgen.bcm.tmc.edu:8088/search-launcher/launcher.html). Transcription factor recognition sites were identified using the BCM Search Launcher Promoter and Transcription Factor Binding Site Prediction site TESS (http://dot.imgen.bcm. tmc.edu:9331/seq-search/gene-search.html). Dots indicate nucleotide identity; dashes represent missing nucleotides.

background on the clinical severity of Alport syndrome (Lemmink et al., 1997). We have assigned the murine Col4a3 and Col4a4 genes to mouse chromosome 1 band C5-D, close to the Pax3 gene at 44 cM from the centromere. The genes are located within the syntenic segments of mouse chromosome 1 and human 2q36 that also contain the unpaired collagen genes COL3A1 and COL6A3 (Schurr et al., 1990). The collagen IV genes are organized in three head-to-head pairs. The transcription start sites of the a1–a2 gene pair are separated by a 130-bp bidirectional promoter that is highly conserved between human and mouse (Heikkila¨ and Soininen, 1996). The bidirectional promoter of the COL4A5 and COL4A6 genes is 442 bp in length (Zhou et al., 1993), but the a6 gene has a second transcription initiation site located 1.3 kb downstream of the intergenic promoter (Sugimoto et al., 1994). The 72% sequence identity of the intergenic region between Col4a3 and Col4a4 in mouse and human suggests that most of the 186-bp region is involved in transcriptional regulation. A 19-bp element that is duplicated in the human a3–a4 promoter (Momata et al., 1998) is present in a single copy in the mouse promoter (Fig. 9), indicating that the duplication occurred after the divergence of human and mouse species. The evolutionary conservation of this novel element supports a functional role. Other elements found in both the human and the mouse promoters include an E-box motif, which is recognized by bHLH transcription factors, and two CACCC motifs that bind Kruppel-like zinc finger transcription factors. Two CTC-box motifs in the 59UTR of the mouse a3 gene are conserved in the 59 UTR of the human gene. A CTC-box motif is common to the promoter regions of several genes encoding extracellular matrix proteins, including the a1–a2 and a5–a6 collagen IV genes (Bruggeman et al., 1992; Sugimoto et al., 1994; Heikkila¨ and Soininen, 1996). The CTC-box in the a1–a2 promoter binds a transcription factor that regulates the expression of both genes (Fischer et al., 1993). It has been suggested that the sharing of bidirectional regulatory elements allows the paired type IV collagen genes to be transcribed coordinately and facil-

itates incorporation into heterotrimers (Heikkila¨ and Soininen, 1996). The a1 and a2 genes are both ubiquitously expressed (Hudson et al., 1993). The a3 and a4 chains are colocalized in the GBM, portions of kidney tubular basement membrane, and several other tissues (Kleppel et al., 1989; Miner and Sanes, 1994). The a5 and a6 genes are expressed in the distal tubules, collecting ducts, and Bowman’s capsule of the kidney, as well as other tissues. The exception to coordinated expression is the presence of a5 without a6 in the GBM (Peissel et al., 1995). In the kidney glomerulus, a3, a4, and a5 form a collagen network that is spatially and temporally distinct from the network formed by the a1 and a2 chains (Kleppel and Michael, 1990; Kleppel et al., 1992; Miner and Sanes, 1994). In early stages of glomerular development, basement membranes contain only a1 and a2. From the capillary stage onward, a3, a4, and a5 chains appear in the GBM and gradually replace a1 and a2. In the mature glomeruli, a1 and a2 are concentrated in the mesangium, and a3, a4, and a5 are the main components of the GBM. Further analysis of the interesting bidirectional regulatory regions of the collagen IV genes could contribute to understanding their coordinated spacial and temporal expression and the factors responsible for high-level expression of a3, a4, and a5 in the glomeruli. The mechanism underlying the progressive nature of the GBM pathology in Alport syndrome is not well understood. The absence of the collagen network formed by the a3, a4, and a5 chains is structurally compensated for by the a1/a2 network, but function is not fully corrected (Nakanishi et al., 1994; Gubler et al., 1995). Three pathological mechanisms have been proposed. One suggests that the collagen network formed by a3, a4, and a5 chains is stronger or more stable than that formed by the a1 and a2 chains, because the high cysteine content of the a4 and a3 chains may enable them to form more interchain disulfide crosslinks (Leinonen et al., 1994; Mariyama et al., 1994). In support of this view, bulk collagen IV isolated from human Alport kidney (containing primarily a1 and a2 chains) was more susceptible to endoproteolysis than a similar isolate from normal kidney (containing a1 to a6 chains), suggesting that Alport GBM is slowly

DELETION OF Col4a3 AND Col4a4 CAUSES KIDNEY DISEASE

damaged by endogenous proteases that have little effect on normal GBM (Kalluri et al., 1997). A second model suggests that abnormal accumulation of matrix proteins such as fibronectin and collagens I, V, and VI at the GBM, which has been described in human Alport syndrome patients and in the a3 mutant mice, plays a key role in the progressive glomerulonephritis phenotype (Kashtan and Kim, 1992; Miner and Sanes, 1996; Muda et al., 1997). The third hypothesis suggests that a possible mismatch between the collagen network formed by the a1 and a2 chains and the b2 chain-containing laminin network may be responsible for glomerulonephritis (Miner and Sanes, 1996). Our observations in the COL4D3-4 mice suggest that the failure to switch to the a3/a4/a5 network may cause stress and proliferation of kidney cells, contributing to disease progression. The Col4D3-4 mutant provides an opportunity for physical and physiological studies of the differences between a1/a2 basement membranes and a3/a4/a5 basement membranes at the same anatomical location. The coat color marker linked to the mutation permits heterozygote transgenic carriers and homozygote mutants to be visually identified after 1 week of age for developmental, physiological, and pathological studies. ACKNOWLEDGMENTS We thank Deborah Gumucio, Philippe Gros, and James Patrick for providing cosmid clones and David Kohrman and Nicholas Plummer for advice in construction of the cosmid library. This work was supported by the Michigan Kidney Foundation and by USPHS Grants GM24872 (M.H.M.), HL49953 and AR45316 (P.A.O.), and DK53196 (J.H.M.). W.L. was supported by the NIH Training Grants in Genetics (T32 GM07544) and Genome Sciences (T32 HG00040). C.L.P. acknowledges support by NIH Training Grant T32 HL07517.

REFERENCES Ackerman, S. L., Kozak, L. P., Przyborski, S. A., Rund, L. A., Boyer, B. B., and Knowles, B. B. (1997). The mouse rostral cerebellar malformation gene encodes an UNC-5-like protein. Nature 386: 838 – 842. Barker, D. F., Hostikka, S. L., Zhou, J., Chow, L. T., Oliphant, A. R., Gerken, S. C., Gregory, M. C., Skolnick, M. H., Atkin, C. L., and Tryggvason, K. (1990). Identification of mutations in the COL4A5 collagen gene in Alport syndrome. Science 248: 1224 –1227. Beermann, F., Ruppert, S., Hummler, E., Bosch, F. X., Mu¨ller, G., Ru¨ther, U., and Schu¨tz, G. (1990). Rescue of the albino phenotype by introduction of a functional tyrosinase gene into mice. EMBO J. 9: 2819 –2826. Bergstein, J. M. (1990). Glomerular fibrin deposition and removal. Pediatr. Nephrol. 4: 78 – 87. Boye, E., Mollet, G., Forestier, L., Cohen-Solal, L., Heidet, L., Cochat, P., Grunfeld, J. P., Palcoux, J. B., Gubler, M. C., and Antignac, C. (1998). Determination of the genomic structure of the COL4A4 gene and of novel mutations causing autosomal recessive Alport syndrome. Am. J. Hum. Genet. 63: 1329 – 40. Bruggeman, L. A., Burbelo, P. D., Yamada, Y., and Klotman, P. E. (1992). A novel sequence in the type IV collagen promoter binds nuclear proteins from Engelbreth–Holm–Swarm tumor. Oncogene 7: 1497–1502. Burbelo, P. D., Martin, G. R., and Yamada, Y. (1988). Alpha 1(IV) and alpha 2(IV) collagen genes are regulated by a bidirectional

123

promoter and a shared enhancer. Proc. Natl. Acad. Sci. USA 85: 9679 –9682. Burgess, D. L., Kohrman, D. C., Galt, J., Plummer, N. W., Jones, J. M., Spear, B., and Meisler, M. H. (1995). Mutation of a new sodium channel gene, Scn8a, in the mouse mutant ‘motor endplate disease.’ Nat. Genet. 10: 461– 465. Church, G. M., and Gilbert, W. (1984). Genomic sequencing. Proc. Natl. Acad. Sci. USA 81: 1991–1995. Cosgrove, D., Meehan, D. T., Grunkemeyer, J. A., Kornak, J. M., Sayers, R., Hunter, W. J., and Samuelson, G. C. (1996). Collagen COL4A3 knockout: A mouse model for autosomal Alport syndrome. Genes Dev. 10: 2981–2992. Fischer, G., Schmidt, C., Opitz, J., Cully, Z., Ku¨hn, K., and Po¨schl, E. (1993). Identification of a novel sequence element in the common promoter region of human collagen type IV genes, involved in the regulation of divergent transcription. Biochem. J. 292: 687– 695. Fromm, L., Shawlot, W., Gunning, K., Butel, J. S., and Overbeek, P. A. (1994). The retinoblastoma protein-binding region of simian virus 40 large T antigen alters cell cycle regulation in lenses of transgenic mice. Mol. Cell. Biol. 14: 6743– 6754. Gregory, M. C., Terreros, D. A., Barker, D. F., Fain, P. N., Denison, J. C., and Atkin, C. L. (1996). Alport syndrome—Clinical phenotypes, incidence, and pathology. Contrib. Nephrol. 117: 1–28. Gubler, M., Levy, M., Broyer, M., Naizot, C., Gonzales, G., Perrin, D., and Habib, R. (1981). Alport’s syndrome: A report of 58 cases and a review of the literature. Am. J. Med. 70: 493–504. Gubler, M-C., Knebelmann, B., Beziau, A., Broyer, M., Pirson, Y., Haddoum, F., Kleppel, M. M., and Antignac, C. (1995). Autosomal recessive Alport syndrome: Immunohistochemical study of type IV collagen chain distribution. Kidney Int. 47: 1142–1147. Heikkila¨, P., and Soininen, R. (1996). The type IV collagen gene family. Contrib. Nephrol. 117: 105–129. Hogan, B., Costantini, F., and Lacy, E. (1986). “Manipulating the Mouse Embryo: A Laboratory Manual,” Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Hudson, B. G., Reeders, S. T., and Tryggvason, K. (1993). Type IV collagen: Structure, gene organization, and role in human diseases. J. Biol. Chem. 268: 26033–26036. Johnson, R. J. (1994). The glomerular response to injury: Progression or resolution? Kidney Int. 45: 1769 –1782. Kalluri, R., Shield, C. F., Todd, P., Hudson, B. G., and Neilson, E. G. (1997). Isoform switching of type IV collagen is developmentally arrested in X-linked Alport syndrome leading to increased susceptibility of renal basement membranes to endoproteolysis. J. Clin. Invest. 99: 2470 –2478. Kashtan, C. E., and Kim, Y. (1992). Distribution of the alpha 1 and alpha 2 chains of collagen IV and of collagens V and VI in Alport syndrome. Kidney Int. 42: 115–126. Kashtan, C. E., and Michael, A. F. (1993). Alport syndrome: From bedside to genome to bedside. Am. J. Kidney Dis. 22: 627– 640. Kaytes, P., Wood, L., Theriault, N., Kurkinen, M., and Vogeli, G. (1988). Head-to-head arrangement of mouse type IV collagen genes. J. Biol. Chem. 263: 19274 –19277. Kim, Y. G., Alpers, C. E., Brugarolas, J., Johnson, R. J., Couser, W. G., and Shankland, S. J. (1999). The cyclin kinase inhibitor p21CIP1/WAF1 limits glomerular epithelial cell proliferation in experimental glomerulonephritis. Kidney Int. 55: 2349 –2361. Kleppel, M. M., Santi, P. A., Cameron, J. D., Wieslander, J., and Michael, A. F. (1989). Human tissue distribution of novel basement membrane collagen. Am. J. Pathol. 134: 813– 825. Kleppel, M. M., and Michael, A. F. (1990). Expression of novel basement membrane components in the developing human kidney and eye. Am. J. Anat. 187: 165–174. Kleppel, M. M., Fan, W., Cheong, H. I., and Michael, A. F. (1992). Evidence for separate networks of classical and novel basement membrane collagen. J. Biol. Chem. 267: 4137– 4142. Knebelmann, B., Breillat, C., Forestier, L., Arrondel, C., Jacassier, D., Giatras, I., Drouot, L., Deschenes, G., Gru¨nfeld, J-P., Broyer,

124

LU ET AL.

M., Gubler, M-C., and Antignac, C. (1996). Spectrum of mutations in the COL4A5 collagen gene in X-linked Alport syndrome. Am. J. Hum. Genet. 59: 1221–1232. Kriz, W., Gretz, N., and Lemley, K. V. (1998). Progression of glomerular diseases: Is the podocyte the culprit? Kidney Int. 54: 687– 697. Lee, S. K., Goyal, M., de Miguel, M., Thomas, P., Wharram, B., Dysko, R., Phan, S., Killen, P. D., and Wiggins, R. C. (1997). Renal biopsy collagen I mRNA predicts scarring in rabbit anti-GBM disease: Comparison with conventional measures. Kidney Int. 52: 1000 –1015. Lees, G. E., Helman, R. G., Kashtan, C. E., Michael, A. F., Homco, L. D., Millichamp, N. J., Ninomiya, Y., Sado, Y., Naito, I., and Kim, Y. (1998). A model of autosomal recessive Alport syndrome in English cocker spaniel dogs. Kidney Int. 54: 706 –719. Leinonen, A., Mariyama, M., Mochizuki, T., Tryggvason, K. and Reeders, S. T. (1994). Complete primary structure of the human type IV collagen alpha 4(IV) chain. J. Biol. Chem. 269: 26172– 26177. Lemmink, H. H., Mochizuki, T., van den Heuvel, L. P., Schro¨der, C. H., Barrientos, A., Monnens, L. A., van Oost, B. A., Brunner, H. G., Reeders, S. T., and Smeets, H. J. (1994). Mutations in the type IV collagen alpha 3 (COL4A3) gene in autosomal recessive Alport syndrome. Hum. Mol. Genet. 3: 1269 –1273. Lemmink, H. H., Schroder, C. H., Monnens, L. A., and Smeets, H. J. (1997). The clinical spectrum of type IV collagen mutations. Hum. Mutat. 9: 477– 499. Lu-Kuo, J., Ward, D. C., and Spritz, R. A. (1993). Fluorescence in situ hybridization mapping of 25 markers on distal human chromosome 2q surrounding the human Waardenburg syndrome, type I (WS1) locus (PAX3 gene). Genomics 16: 173–179. Lyon, M. F., and Searle, A. G. (1989). “Genetic Variants and Strains of the Laboratory Mouse,” 2nd ed., Oxford Univ. Press, New York. Mariyama, M., Zheng, K., Yang-Feng, T. L., and Reeders, S. T. (1992). Colocalization of the genes for the alpha 3(IV) and alpha 4(IV) chains of type IV collagen to chromosome 2 bands q35– q37. Genomics 13: 809 – 813. Mariyama, M., Leinonen, A., Mochizuki, T., Tryggvason, K., and Reeders, S. T. (1994). Complete primary structure of the human alpha 3(IV) collagen chain. J. Biol. Chem. 269: 23013–23017. Meisler, M. H. (1992). Insertional mutation of ‘classical’ and novel genes in transgenic mice. Trends Genet. 8: 341–344. Meisler, M. H., Galt, J., Weber, J., Jones, J. M., Burgess, D. L., and Kohrman, D. C. (1997). Isolation of mutated genes from transgene insertion sites. In “Microinjection and Transgenesis of Cultured Cells and Embryos” (A. Cid and Garcia-Carranca, Eds.), pp. 505– 520, Springer Verlag, Heidelberg. Miner, J. H., and Sanes, J. R. (1994). Collagen IV alpha 3, alpha 4, and alpha 5 chains in rodent basal laminae: Sequence, distribution, association with laminins, and developmental switches. J. Cell. Biol. 127: 879 – 891. Miner, J. H., and Sanes, J. R. (1996). Molecular and functional defects in kidneys of mice lacking collagen alpha 3(IV): Implications for Alport syndrome. J. Cell. Biol. 135: 1403–1413. Mochizuki, T., Lemmink, H. H., Mariyama, M., Antignac, C., Gubler, M-C., Pirson, Y., Verellen-Dumoulin, C., Chan, B., Schro¨der, C. H., Smeets, H. J., and Reeders, S. T. (1994). Identification of mutations in the alpha 3(IV) and alpha 4(IV) collagen genes in autosomal recessive Alport syndrome. Nat. Genet. 8: 77– 81. Mochizuki, T., Saijoh, Y., Tsuchiya, K., Shirayoshi, Y., Takai, S., Taya, C., Yonekawa, H., Yamada, K., Nihei, H., Nakatsuji, N., Overbeek, P. A., Hamada, H., and Yokoyama, T. (1998). Cloning of inv, a gene that controls left/right asymmetry and kidney development. Nature 395: 177–181. Momota, R., Sugimoto, M., Oohashi, T., Kigasawa, K., Yoshioka, H., and Ninomiya, Y. (1998). Two genes, COL4A3 and COL4A4, coding for the human alpha3(IV) and alpha4(IV) collagen chains are arranged head-to-head on chromosome 2q36. FEBS Lett. 424: 11–16.

Morgan, D., Turnpenny, L., Goodship, J., Dai, W., Majumder, K., Matthew, L., Gardner, A., Schuster, G., Vien, L., Harrison, W., Elder, F. F. B., Penman-Splitt, M., Overbeek, P., and Strachan, T. (1998). Inversin, a novel gene in the vertebrate left–right axis pathway, is partially deleted in the inv mouse. Nat. Genet. 20: 149 –156. Muda, A. O., Rahimi, S., Renieri, A., Rizzoni, G., Massella, L., and Faraggiana, T. (1997). Ultrastructural immunocytochemistry of collagenous and non-collagenous proteins in fast-frozen, freezesubstituted, and low-temperature-embedded renal tissue in Alport syndrome. J. Pathol. 182: 465– 474. Nakanishi. K., Yoshikawa, N., Iijima, K., Kitagawa, K., Nakamura, H., Ito, H., Yoshioka, K., Kagawa, M., and Sado, Y. (1994). Immunohistochemical study of alpha 1–5 chains of type IV collagen in hereditary nephritis. Kidney Int. 46: 1413–1421. Overbeek, P. A., Aguilar-Cordova, E., Hanten, G., Schaffner, D. L., Patel, P., Lebovitz, R. M., and Lieberman, M. W. (1991). Coinjection strategy for visual identification of transgenic mice. Transgenic Res. 1: 31–37. Peissel, B., Geng, L., Kalluri, R., Kashtan, C., Rennke, H. G., Gallo, G. R., Yoshioka, K., Sun, M. J., Hudson, B. G., Neilson, E. G., and Zhou, J. (1995). Comparative distribution of the alpha 1 (IV), alpha 5 (IV), and alpha 6 (IV) collagen chains in normal human adult and fetal tissues and in kidneys from X-linked Alport syndrome patients. J. Clin. Invest. 96: 1948 –1957 Po¨schl, E., Pollner, R., and Ku¨hn, K. (1988). The genes for the alpha 1(IV) and alpha 2(IV) chains of human basement membrane collagen type IV are arranged head-to-head and separated by a bidirectional promoter of unique structure. EMBO J. 7: 2687–2695. Robinson, T. J., and Elder, F. F. (1987). Multiple common fragile sites are expressed in the genome of the laboratory rat. Chromosoma 96: 45– 49. Schurr, E., Skamene, E., Morgan, K., Chu, M-L., and Gros, P. (1990). Mapping of Col3a1 and Col6a3 to proximal murine chromosome 1 identifies conserved linkage of structural protein genes between murine chromosome 1 and human chromosome 2q. Genomics 8: 477– 486. Sugimoto, M., Oohashi, T., and Ninomiya, Y. (1994). The genes COL4A5 and COL4A6, coding for basement membrane collagen chains alpha 5(IV) and alpha 6(IV), are located head-to-head in close proximity on human chromosome Xq22 and COL4A6 is transcribed from two alternative promoters. Proc. Natl. Acad. Sci. USA 91: 11679 –11683. Ting, C-N., Kohrman, D., Burgess, D. L., Boyle, A., Altschuler, R. A., Gholizadeh, G., Samuelson, L. C., Jang, W., and Meisler, M. H. (1994). Insertional mutation on mouse chromosome 18 with vestibular and craniofacial abnormalities. Genetics 136: 247–254. Yokoyama, T., Silversides, D. W., Waymir, K. G., Kwon, B. S., Takeuchi, T., and Overbeek, P. A. (1990). Conserved cysteine to serine mutation in tyrosinase is responsible for the classical albino mutation in laboratory mice. Nucleic Acids Res. 18: 7293–7298. Zheng, K., Thorner, P. S., Marrano, P., Baumal, R., and McInnes, R. R. (1994). Canine X chromosome-linked hereditary nephritis: A genetic model for human X-linked hereditary nephritis resulting from a single base mutation in the gene encoding the alpha 5 chain of collagen type IV. Proc. Natl. Acad. Sci. USA 91: 3989 –3993. Zhou, J., Mochizuki, T., Smeets, H., Antignac, C., Laurila, P., de Paepe, A., Tryggvason, K., and Reeders, S. T. (1993). Deletion of the paired alpha 5(IV) and alpha 6(IV) collagen genes in inherited smooth muscle tumors. Science 261: 1167–1169. Zhou, J., and Reeders, S. T. (1996). The alpha chains of type IV collagen. Contrib. Nephrol. 117: 80 –104.