GENOMICS 48, 178–185 (1998)ARTICLE NO. GE975170

Cloning and Characterization of a Novel Gene (TM7SF1)Encoding a Putative Seven-Pass Transmembrane Protein

That Is Upregulated during Kidney Development

Christian Spangenberg,1 Andreas Winterpacht, Bernhard U. Zabel, and Ralf W. Lobbert

Children’s Hospital University of Mainz, Langenbeckstrasse 1, 55101 Mainz, Germany

Received October 27, 1997; accepted December 10, 1997

compare two RNA populations, the differential displayWe have used the cDNA differential display of mRNA approach allows comparative analysis of multiple sam-

technique to isolate genes differentially regulated dur- ples at a time. Additionally, since the products areing kidney development. Here we report the identifi- tagged at both ends with known primers, they can becation of a novel gene, TM7SF1, which is upregulated easily reamplified and used for subsequent analysis.in the course of kidney development. The full-length In the light of rapidly growing sources of expressedcDNA of TM7SF1 is about 2.4 kb and contains an open sequences generated in the course of EST projects (Ad-reading frame of 1197 nucleotides. The predicted sec- ams et al., 1991), sequence information initially ob-ondary structure of the corresponding protein dis- tained by differential display can be easily expanded byplays seven putative helical transmembrane domains,

rapid access to public databases and resource centers.a structural feature shared by all members of the G-A complex series of differential gene expression takesprotein-coupled receptor class of transmembrane pro-

place during kidney development, a process that culmi-teins. Two minor alternatively spliced versions ofÇ2.3nates in the formation of mature nephrons. Startingand Ç2.2 kb could be detected, one of which containsfrom the ureteric bud and the metanephric mesen-a nearly identical open reading frame with a truncatedchyme, a series of reciprocal interactions between thesecarboxy-terminus of the deduced protein. The second

alternatively spliced version harbors a completely two tissues, which controls proliferation, differentia-shifted open reading frame with a potential new ATG tion, and death of the cells involved, occurs, therebystart codon. By the use of single-chromosome hybrid determining kidney morphogenesis. These interactionscells and fluorescence in situ hybridization experi- have been shown to be mediated by direct cellular con-ments, TM7SF1 could be localized to chromosome tacts, extracellular matrix components, and diffusible1q42–q43. Human multiple tissue Northern blot analy- peptide growth hormones (for reviews, see Schofieldsis revealed TM7SF1 transcripts in human kidney, and Boulter, 1996; Vainio and Muller, 1997). Severalheart, brain, and placenta tissue. Studies on Wilms tu- candidate molecules that seem to be responsible for atmor samples showed variable TM7SF1 expression, least some of the signaling processes during kidneyranging from nearly undetectable levels to an abun-

development have been described (Herzlinger et al.,dant level of expression comparable to that of adult1994; Stark et al., 1994; Sariola et al., 1991; Vukicevickidney tissue. q 1998 Academic Presset al., 1996; Woolf et al., 1995; Rogers et al., 1991).Current data suggest that ureter development is initi-ated by a mesenchyme-derived signal. Ureteric tissueINTRODUCTIONreversely induces the cells of the metanephric mesen-

The RT-PCR-based method of differential display is chyme to condense and subsequently patterns thema powerful tool for the identification of genes that show into at least two different populations, a stromal cellaltered expression in different tissues and cell lines or population and a tubular cell population. Cells in thethat are developmentally regulated (Liang and Pardee, tubular zone are then believed to undergo the mesen-1992; Liang et al., 1994; Ozaki et al., 1996; Lin et al., chyme-to-epithelial transition, thereby forming the re-1997; Debernardi et al., 1997). In contrast to sub- nal tubules.tractive hybridization techniques which are designed to An example of aberrant kidney development is

Wilms tumor (WT), or nephroblastoma. Tumor tissueSequence data from this article have been deposited with the Gen- consists of persisting, undifferentiated mesenchymal

Bank Data Library under Accession No. AF027826. cells, which are believed to have not received and/or1 To whom correspondence should be addressed. Telephone://49-processed the adequate inducing signal(s) to differenti-6131-173334.Fax: //49-6131-175528 or -176610. E-mail: spange@

wserv.kinder.klinik.uni-mainz.de. ate into the well-defined structures constituting ma-

1780888-7543/98 $25.00Copyright q 1998 by Academic PressAll rights of reproduction in any form reserved.

AID GENO 5170 / 6r60$$$$$1 02-25-98 14:50:49 gnma

NOVEL GENE TM7SF1 UPREGULATED IN KIDNEY DEVELOPMENT 179

200 mM NaHPO4/1% SDS buffer, 20 min in 100 mM NaHPO4/1%ture kidney tissue (Park et al., 1993). Several candidateSDS, and 10 min in 80 mM NaHPO4/1% SDS at 657C prior to autora-genes and chromosomal regions have been implicateddiography. Hybridization and washing procedures utilizing humanin Wilms tumorigenesis (for a review, see Grundy and Multiple Tissue Northern Blot I and human RNA Master Blot (Clon-

Coppes, 1996) but, due to the genetic heterogeneity of tech) were carried out according to the manufacturer’s instructions.Reverse transcription for RT-PCR analysis was performed usingthis tumor, the major pathways that lead to the devel-

4 mg of total RNA as template in a total reaction volume of 40 ml.opment of Wilms tumors are far from being understood.After initial heat denaturation of RNA (10 min at 657C), the samplesTo elucidate the changes in gene expression that oc-were chilled on ice for 1 min, and RT mix containing the following

cur during nephrogenesis, we carried out differential components was added: 5.0 ml of dT16-oligonucleotide (50 mM), 8.0 mldisplay experiments on human adult and fetal kidney of 51 first-strand buffer (Gibco BRL), 4.0 ml of DTT (0.1 M; Gibco

BRL), 3.6 ml of dNTP mix (10 mM each; Boehringer Mannheim), 2.0tissue. In this paper we report the identification andml of RNasin (200 U/ml; Perkin–Elmer), 1.5 ml of M-MLV reversecharacterization of a novel human gene, designatedtransciptase (200 U/ml; Gibco BRL, omitted in negative controls).TM7SF1 (transmembrane 7 superfamily member 1), RT reaction was carried out for 90 min at 377C, followed by heat

that is transcriptionally upregulated in the course of inactivation of the enzyme at 947C for 10 min. Subsequent PCRkidney development. A prominent structural feature of amplifications (50 ml total volume) contained 2.0 ml of RT reaction

mixture as template and 40 pmol of each gene-specific primer andthe predicted protein consists of seven helical trans-were done using standard PCR conditions. Specific PCR productsmembrane domains (HTMDs), making it a candidatewere gel-purified using the Gene Clean purification kit (Bio 101) andmember of the superfamily of G-protein-coupled recep- subcloned into the T-vector.

tors (GCRs). Nucleotide sequencing and analysis. Cycle sequencing using theABI Prism dye-terminator sequencing kit was performed accordingto ABI standard protocol with either 1 mg of plasmid DNA preparedMATERIALS AND METHODSwith the RPM kit (Bio 101) or 100 ng of isopropanol-precipitatedPCR products generated using T3 or T7 sequencing primers or gene-

Differential mRNA display. Differential display was performed specific primers. All sequence information was obtained by sequenc-essentially as previously described (Liang and Pardee, 1992; Liang ing both strands of the cloned or amplified cDNA fragments (theet al., 1994). DNA-free total RNA (0.5 mg) from adult and fetal kidney sequence was submitted to GenBank under Accession No.tissue in a total volume of 7.5 ml was heated at 657C for 10 min. Four AF027826). Editing and analysis of nucleotide sequences were per-microliters of first-strand buffer (Gibco BRL), 2 ml of 0.1 M DTT formed using the Sequencher 3.0 (Gene Codes Corporation) software(Gibco BRL), 2 ml of 200 mM dNTPs (Boehringer Mannheim), 1 ml of package. Database searches using the BLAST algorithm (AltschulRNasin (200 U/ml; Perkin–Elmer), and 50 pmol of 3* anchored oli- et al., 1990) were carried out at NCBI (http://www.ncbi.nlm.nih.gov/go(dT) primer were added. Duplicate positive samples contained 300 blast). Software tools used for secondary structure and pattern pre-U of M-MLV reverse transcriptase (RT) (Gibco BRL) in a total volume diction are available at http://expasy.hcuge.ch.of 20 ml, whereas negative controls contained no enzyme. Reverse Chromosome mapping analysis. Screening of human 1 rodenttranscription was carried out at 377C for 1 h with heat-inactivation somatic cell hybrids was done by PCR using the oligonucleotidesat 947C for 10 min. Two microliters of RT reaction mixture were Local-5 and Local-3 as primers. The same primer set was used forPCR-amplified in a final volume of 20 ml in the presence of 2 ml of the identification of TM7SF1-positive PAC clones in a human PAC101 PCR buffer (Boehringer Mannheim), 2 ml of 20 mM dNTPs, 1 ml of library (Ioannou et al., 1994). Probe DNA for fluorescence in situ[a-32P]dCTP (Amersham; 1 mCi/ml), 1.25 U of Taq DNA polymerase hybridization was isolated from PAC 43-C3 utilizing the Nucleobond(Boehringer Mannheim; 5 U/ml), 50 pmol of 3 *-anchored oligo(dT) kit (Macherey & Nagel) according to the manufacturer’s instructions.primer (T11A), and 10 pmol of arbitrary 5*-decamer (10-1). Amplifica- The conditions applied during in situ hybridization were identical totion was carried out in a Perkin–Elmer thermal cycler (Cetus) under those described by Lobbert et al. (1996).the following conditions: 947C for 4 min followed by 40 cycles at 947C

Oligonucleotide sequences. Oligonucleotide sequences were as fol-for 45 s, 407C for 2 min, and 727C for 1 min. 32P-labeled PCR productslows: T11A, 5-TTTTTTTTTTTA-3; 10-1, 5-AATCGTCCAT-3; Phage-were separated on a 6% polyacrylamide DNA sequencing gel, excised,R, 5-CTTCTGGCTGCTCTACTGC-3; 5* For, 5-AAGCTGTTAGTT-processed, reamplified, and then subcloned into a T-tailed EcoRV-TGTTGTCCCC-3; Local-5, 5-CGATTCTGAGTGCCACATTG-3; anddigested pBluescript SK(/) cloning vector (T-vector) (Marchuk et al.,Local-3, 5-TAGCGCCCTACTGAAAGGAA-3.1991) using standard molecular cloning techniques.

cDNA library screening. Approximately 2.5 1 105 clones from ahuman adult kidney cDNA library (Bell et al., 1986) in lgt10 (kindly RESULTSprovided by Dr. D. Plachow) were screened with the radiolabeled1AA1 fragment of TM7SF1 by plaque hybridization using Church

cDNA Isolation and Assemblyhybridization buffer (Church and Gilbert, 1984) and Hybond-N/ filterunder standard conditions.

To identify new genes implicated in kidney develop-RNA preparation and analysis. Total RNA from human kidney

ment, we compared differential display patterns usingtissue [adult, fetal (22nd week of gestation), and WT] was preparedmRNAs isolated from human adult and fetal kidneyusing the RNeasy Total RNA kit (Qiagen) according to the manufac-

turer’s instructions. To generate DNA-free RNA, samples were di- tissue.gested with DNase I (Boehringer Mannheim) as described elsewhere PCR amplification with the oligonucleotides T11A(Sambrook et al., 1989) and subsequently isolated according to the and 10-1 as primers and denaturing polyacrylamide gelQiagen RNA-cleanup protocol.

electrophoresis resulted in a prominent band in adultElectrophoresis of 10 mg of RNA and Northern transfer on Hybond-kidney RNA (data not shown). Reamplification yieldedN/ filters (Amersham) were carried out using techniques described

by Mason et al. (1993). Slot blots were generated using a slot blot a 436-bp fragment, initially named 1AA1, which wasvacuum device (Hybri-Slot manifold; BRL) according to the protocol cloned and subsequently sequenced. Database searchessupplied with Hybond-N/ filters (Amersham) with 2 mg of total RNA. revealed sequence identity to several overlapping ESTHybridization was done using radiolabeled probes generated by ran-

clones. One of them (EST32 in Fig. 1A; clone namedom hexamer labeling of gel-purified PCR fragments with 1–2 1 106

cpm/ml hybridization buffer. The blots were washed for 20 min in IMAGp998P011414) was ordered from the IMAGE

AID GENO 5170 / 6r60$$$$$1 02-25-98 14:50:49 gnma

SPANGENBERG ET AL.180

1776-bp TM7SF1 cDNAs were examined. Open readingframes (ORFs) for all three splice variants of TM7SF1[(L)ong, (I)ntermediate, and (S)hort] could be detected.While TM7SF1-L and TM7SF1-I share most parts oftheir ORFs, the alternative splicing event in version Sresults in a frameshift that opens a different ORF thatcurrently lacks an ATG start codon (Fig. 1B). To obtaininformation about the secondary structure of thededuced proteins, tools available at the EXPASYhomepage (http://expasy.hcuge.ch) were applied to theTM7SF1 sequence and resulted in the features givenin Fig. 2. While L and I versions of TM7SF1 were eachpredicted to contain seven helical transmembrane do-mains, with TM7SF1-I missing 71 amino acids at itsC-terminus (Fig. 2A), secondary structure prediction ofTM7SF1-S displayed a strongly different feature pat-

FIG. 1. (A) Cloning strategy of the human TM7SF1 coding se- tern (data not shown). A scan of TM7SF1-L and -Squence. Clones were derived from the initial differential display ex- against consensus sequences present in the PROSITEperiment (1AA1), from a human adult kidney cDNA library

database revealed several putative sites for posttrans-(c1AA1A), and from the IMAGE Consortium (EST32). RT-PCR prod-lational modifications of TM7SF1, as indicated in Fig.uct AS3 * was generated using total adult kidney RNA as template

with oligonucleotides Phage-R and 5* For as primers. PCR amplifica- 2B. These include an Asn-glycosylation site in the po-tion with this primer set revealed two additional fragments. One of tential extracellular N-terminal region, several consen-these was shown to contain both alternatively spliced exons (thereby sus phosphorylation sequences for serine/threonine ki-representing TM7SF1-L), whereas the second one was missing the

nases in the putative C-terminal tail region of5* alternatively spliced exon but contained the 125 bp of the alterna-TM7SF1-L only, and a tyrosine residue in a potentialtively spliced 3 * region (thereby representing TM7SF1-I). Both frag-

ments are omitted for clarity. (B) Schematic representation of the tyrosine kinase consensus sequence.predicted ORFs of TM7SF1-I (top), TM7SF1-L (middle), andTM7SF1-S (bottom). Rectangles represent the cDNAs of the three

Northern Analysis of TM7SF1 ExpressionTM7SF1 species with the white part indicating the positions of theORFs. The alternative splicing event in the 3 * region of TM7SF1-L

To determine whether TM7SF1 is developmentallygives rise to the short form, TM7SF1-I, which contains a truncatedcarboxy-terminus when compared to the long version. TM7SF1-S regulated during nephrogenesis, we examined its ex-results from an alternative splicing event in the more 5* region of pression level in total RNA isolated from human adultTM7SF1. The predicted amino acid sequence of this isoform is com- and fetal kidney tissue. Using the initially identifiedpletely different from the TM7SF1-L and -I translations. Nucleotide

1AA1 fragment as a probe, a strong transcript of aboutpositions of start and stop codons are given on top of each sequencebar. The light gray shaded parts of TM7SF1-L indicate the positions 2.4 kb could be detected in adult kidney RNA (Fig. 3A).of the alternatively spliced exons. The sizes (in amino acids) of the Even after long exposures, only a faint band of thecorresponding TM7SF1 proteins are indicated on the left. same size was visible in fetal kidney RNA, confirming

the differential expression pattern of TM7SF1 duringkidney development. To examine tissue-specific expres-Consortium, distributed by the Resource Center/Pri-sion of TM7SF1, the PCR-based Local-5/-3 product (lo-mary Database of the German Human Genome Project,cated in the putative 3 * untranslated region) was radio-Berlin (Lehrach et al., 1990), and its sequence was ana-labeled and used as a probe on a human multiple tissuelyzed. To obtain information about the 5* region, weNorthern blot containing 2 mg poly(A)/ RNA in eachused the 1AA1 cDNA as a probe to screen a lgt10-basedlane. Overnight exposure revealed strong 2.4-kb sig-adult kidney cDNA library. A positive clone, designatednals in human kidney and heart RNA as well as weakerc1AA1A, could be identified. Sequence analysis and fur-signals in brain and placenta RNA (Fig. 3B). Hybridiza-ther rounds of sequence similarity searches yieldedtion experiments on a human RNA master blot showednew overlapping EST clones. Comparison with thestrongest signals in adult kidney, nucleus caudatus,phage cDNA revealed an additional 223 bp in the ex-putamen, and spinal cord RNAs, whereas weaker hy-pressed sequence tags. Through RT-PCR, the existencebridization was observed on RNA from heart and totalof both RNA species as well as a new transcript ofbrain (data not shown).intermediate length could be confirmed (Figs. 1A and

WTs are believed to arise from primitive, mesenchy-1B), and all three cDNA fragments were subcloned.mal kidney cells that are not properly induced duringmetanephrogenesis to develop into mature kidneySequence Analysisstructures (Park et al., 1993). Therefore, we screeneda subset of our Wilms tumor collection to look for ex-As initial comparison of the nucleotide sequence of

TM7SF1 did not show striking homology to any se- pression of TM7SF1 transcripts in these tumors. Forthis purpose an RNA slot blot filter harboring RNAquence reported in the nonredundant database of Gen-

Bank, the putative translations of the 1998-, 1874-, and from normal adult and fetal kidney tissue as well as

AID GENO 5170 / 6r60$$$$$1 02-25-98 14:50:49 gnma

NOVEL GENE TM7SF1 UPREGULATED IN KIDNEY DEVELOPMENT 181

FIG. 2. Predicted structural features of TM7SF1-L and -I. (A) Diagram of the results obtained from the DAS transmembrane predictionserver for the ORFs of TM7SF1-L and TM7SF1-I. The X axes indicate the linear amino acid sequences of the two protein isoforms. Horizontallines represent strict (solid) and loose (dashed) cutoff levels for transmembrane-segment predictions. Predictions obtained from the programsTopPred2, SOSUI, TMPRED, PSA, and PSORT were essentially the same as the ones represented. (B) Schematic model derived from thepredicted features of TM7SF1-L shown in (A). Symbols mark consensus sites for posttranslational modifications as indicated by the Prositedatabase (black square, Asn glycosylation/Asn26; open circle, Ser/Thr residues representing CK2 substrates/Ser348/Thr385/Ser389; blackcircle, PKC phosphorylation site/Ser334; gray circle, phosphorylation by Tyr kinase/Tyr154). The predicted TM domains are indicated byvertical rectangles.

RNA from five WTs was generated. Hybridization of ing the results obtained by PCR screening of the so-matic cell hybrid DNAs.this blot with the Local-5/-3 probe showed a varying

abundance of the TM7SF1 mRNA, ranging from nearlyundetectable levels to expression levels comparable to DISCUSSIONthose of adult kidney tissue (data not shown).

Differential display has been widely used to identifyChromosomal Assignment of TM7SF1 todifferentially expressed genes in eukaryotic cells or tis-Chromosomal Region 1q42–q43sues (Liang and Pardee, 1992; Liang et al., 1994; Ozakiet al., 1996; Lin et al., 1997; Debernardi et al., 1997).Oligonucleotides Local-5 and Local-3 were used to

screen DNA from human x rodent somatic cell hybrids. To identify genes involved in kidney development, thismethodology was employed to analyze differences inAmplification products of the correct size could be de-

tected with human chromosome 1 DNA as a template. gene expression between adult and fetal (22nd weekof gestation) kidney tissue. This report describes theFluorescence in situ hybridization localized the

TM7SF1-containing PAC clone 43/C3 to chromosomal identification and characterization of a novel gene,TM7SF1, which is expressed as a 2.4-kb transcript pre-region 1q42–q43 (Fig. 4) without further cross-hybrid-

ization to other chromosomal locations, thus confirm- dominantly in adult kidney tissue. The amino acid se-

AID GENO 5170 / 6r60$$$$$1 02-25-98 14:50:49 gnma

SPANGENBERG ET AL.182

new ATG start codon. Whether this version of TM7SF1contains an upstream start codon in the new ORF and,therefore, may give rise to a largely different proteincould not be answered as the most 5* region of theTM7SF1 mRNA is still unknown.

Since neither the nucleotide sequence nor the aminoacid sequence revealed striking homologies in databasesearches, computer-based protein topology predictionswere used to elucidate the structural nature of theTM7SF1-L and -I gene products. All algorithms appliedindicated the existence of seven helical transmembranedomains. Prediction methods based on the average hy-drophobicity (program MOMENT), on the positive in-side rule (program TOPPRED2; von Heijne, 1992), ona neural network method (PHDhtm), and on the so-called dense alignment surface method (Cserzo et al.,1994), as well as on some other programs (PSA, Stultzet al., 1993; White et al., 1994; and PSORT, Nakai andKanehisa, 1992; Klein et al., 1985) all located the sevenHTMDs of TM7SF1-L and -I between amino acids 40and 313. Only minor differences in length or exact posi-tions of the single transmembrane domains were ob-served. Since all algorithms used resulted in very simi-lar predictions, and current methods can identifyaround 90–95% of all true transmembrane segmentswith only a few overpredictions (von Heijne, 1992), weconclude that the TM7SF1 gene products indeed con-tain seven HTMDs, a structural feature that makes it acandidate member of the G-protein-coupled serpentinereceptor superfamily.

GCRs function by transmitting signals across theFIG. 3. Northern analysis of TM7SF1 expression. (A) 15 mg/lane

plasma membrane after ligand binding to the extracel-of total adult (ad) and fetal (fet) kidney RNA was probed with thelular part of the receptors. They influence the activity1AA1 fragment to confirm differential expression of TM7SF1 during

nephrogenesis. The arrow on the right marks the position of the most of membrane-bound proteins (adenylate cyclase, gua-abundant TM7SF1-L transcript. The bottom shows rehybridization nylate cyclase, phospholipase C, and channel proteins)of the same blot with GAPDH to normalize for RNA loading. Posi- via G proteins, thereby altering cellular concentrationstions of an RNA size standard (in kb) are indicated by arrowheads

of second-messenger molecules (for a review, seeon the left. (B) The radiolabeled Local-5/-3 fragment was hybridizedStrader et al., 1994). In addition to seven HTMDs,to a multiple tissue Northern blot containing 2 mg poly(A)/ RNA of

the indicated human tissues in each lane (sk. muscle, skeletal mus- the predicted proteins encoded by TM7SF1-L andcle). The arrow on the left marks the position of the TM7SF1-L tran- TM7SF1-I possess several features that make themscript. Positions of an RNA size standard are indicated by arrow- candidate GCRs. In both proteins an N-linked glycosyl-heads on the right (in kb).

ation consensus site can be found in the extracellularamino-terminal region, a posttranslational modifica-

quence of the deduced protein indicates that the gene tion present in all GCRs characterized to date. More-encodes a protein of 399 amino acids containing seven over, consensus phosphorylation sites for PKC as wellhelical transmembrane domains, a structural feature as serine and threonine residues, which might be ofshared by all members of the superfamily of G-protein- functional importance, can be found in the carboxyl-coupled receptors. terminal tail of TM7SF1-L (for reviews, see Lefkowitz,

RT-PCR analysis showed that three alternatively 1993; Haga et al., 1994; Franci et al., 1996). Further-spliced versions of TM7SF1, TM7SF1-L, TM7SF1-I, more, alternative splicing has been reported for a groupand TM7SF1-S, exist. Quantification of the alternative of GCRs and seems to affect coupling specificities ofmRNAs showed that version L is the most abundant the proteins, among which are, e.g., the MCP-1 receptorform of TM7SF1 mRNAs (data not shown). Compared and the receptors for prostaglandin and neurokinin-1to isoform L, a 125-bp segment is missing in TM7SF1- (Namba et al., 1993; Fong et al., 1992; Charo et al.,I, giving rise to a truncated protein (328 amino acids 1994). Whether different coupling specificities can bein length) that possesses only a very short intracellular assigned to isoforms L and I of TM7SF1, or if the miss-C-terminal tail. TM7SF1-S results from an alternative ing intracellular C-terminal tail of TM7SF1-I makessplicing event that removes 223 bp with respect to the protein unresponsive to negative regulatory modi-

fications, remains to be clarified.TM7SF1-L. It harbors a shifted ORF with a potential

AID GENO 5170 / 6r60$$$$$1 02-25-98 14:50:49 gnma

NOVEL GENE TM7SF1 UPREGULATED IN KIDNEY DEVELOPMENT 183

FIG. 4. Fluorescence in situ hybridization of PAC clone 43-C3 on metaphase chromosomes. The PAC probe was detected with anti-dig-TRITC (red dots) and shows specific signals on both chromatids of chromosomal region 1q42–q43.

Although TM7SF1 shows several characteristics that TM7SF1 proteins at Tyr154 occurs, this site could actas a scaffold for generating a phosphotyrosine-depen-match the serpentine receptor superfamily, we cannot

exclude that the gene may code for a protein(s) that dent assembly of a signaling complex in direct associa-tion with the GCR.spans the plasma membrane seven times but is not

implicated in cellular signaling. Some evolutionary The data presented in this report show that TM7SF1expression is upregulated during kidney development.conserved Cys residues, found in nearly all members of

the GCR superfamily, are not present in the predicted As metanephrogenesis consists of a series of interac-tions between different tissues (ureteric bud and meta-protein, and a scan against the G-protein-coupled re-

ceptor database (GCRDb) (http://www.gcrdb.uthsc nephrogenic mesenchyme), there is a need for a varietyof participating signaling molecules to be present atsa.edu/) revealed only weak amino acid similarities

(20.7% to the rat secretin receptor) (Kolakowski, 1994). different stages of nephrogenesis. Several candidatesinvolved in this extensive signaling have been identi-A low sequence homology between members of the GCR

superfamily has been reported previously (Strader et fied and characterized to date, but the picture of kidneydevelopment is far from being complete. The inductional., 1994), so TM7SF1 might code for a member of a

new family of GCRs. of TM7SF1 during kidney development indicates thatthis protein is a cellular marker for a differentiatedCellular signaling utilizes a complex network of pro-

teins and second-messenger molecules that enables the developmental stage. This might be explained by itsupregulation during the mesenchyme-to-epithelialcells to respond to external stimuli by means of chang-

ing their repertoire of expressed genes. As TM7SF1-L transition that takes place in the course of metanephro-genesis. Whether TM7SF1 represents a protein partici-and -I both contain a consensus Tyr-phosphorylation

site in their putative third intracytoplasmic loop (de- pating in normal physiologic kidney function orwhether it plays a more active role in reaching and/orposited in the PROSITE database), it is tempting to

speculate about a crosstalk between signal transduc- maintainence of a differentiated phenotype cannot beanswered. Nevertheless, its predicted protein structuretion modules employing tyrosine kinases [like the mito-

gen-activated protein (MAP) kinase pathway] and het- as a serpentine receptor strongly suggests a role in cell-type-specific differentiation-dependent signaling pro-erotrimeric G proteins. Recently, a convergent signal-

ing of these two classical cascades was described, in cesses.This suggestion is in concordance with two additionalwhich the activation of a G protein via its GCR results

in stimulation and activation of the MAP kinase cas- observations presented in this paper. First, the locallyrestricted expression of TM7SF1 mRNA in organs likecade (Sadoshima and Izumo, 1996; Schieffer et al.,

1996; Luttrell et al., 1997). If phosphorylation of kidney, heart, and brain argues for the protein being

AID GENO 5170 / 6r60$$$$$1 02-25-98 14:50:49 gnma

SPANGENBERG ET AL.184

tive splicing of the carboxyl-terminal tails. Proc. Natl. Acad. Sci.a player in organ- and development-specific signalUSA 91: 2752–2756.transduction. Several GCRs have been implicated in

Church, G. M., and Gilbert, W. (1984). Genomic sequencing. Proc.signaling in these tissues (Rosenthal et al., 1993; Rock-Natl. Acad. Sci. USA 81: 1991–1995.man et al., 1996; Dearry et al., 1990; Zhou et al., 1990).

Cserzo, M., Bernassau, J. M., Simon, I., and Maigret, B. (1994). NewSome of them seem to be responsible for severe diseases alignment strategy for transmembrane proteins. J. Mol. Biol. 243:when overexpressed or mutated. Second, TM7SF1 388–396.mRNA levels have been shown to vary in Wilms tumor Dearry, A., Gingrich, J. A., Falardeau, P., Fremeau, R. T., Jr., Bates,tissues, ranging from a very low level of expression, M. D., and Caron, M. (1990). Molecular cloning and expression of

the gene for a human D1 dopamine receptor. Nature 347: 72–76.comparable to that in fetal kidney, up to levels observedDebernardi, S., Fontanella, E., De Gregorio, L., Pierotti, M. A., andin adult kidney tissue. As nephroblastomas often dis-

Delia, D. (1997). Identification of a novel human kinesin-relatedplay a triphasic morphology (blastemal, stromal, andgene (HK2) by the cDNA differential display technique. Genomicsepithelial elements) with one histological subtype being42: 67–73.

the predominant one, the finding that TM7SF1 mRNAFong, T. M., Anderson, S. A., Yu, H., Huang, R. R., and Strader, C.levels vary may reflect histological differences in the (1992). Differential activation of intracellular effector by two iso-

Wilms tumor samples examined. The expression of one forms of human neurokinin-1 receptor. Mol. Pharmacol. 41: 24–30.known key player in Wilms tumorigenesis, the WT1

gene product, has recently been shown to be correlated Franci, C., Gosling, J., Tsou, C. L., Coughlin, S. R., and Charo, I.(1996). Phosphorylation by a G protein-coupled kinase inhibits sig-with a histological subtype of nephroblastomas (Schu-naling and promotes internalization of the monocyte chemoattrac-macher et al., 1997).tant protein-1 receptor: Critical role of carboxyl-tail serines/threo-Together with the presented chromosomal mapping nines in receptor function. J. Immunol. 157: 5606–5612.

studies of TM7SF1, which place the gene in the chromo- Grundy, P., and Coppes, M. (1996). An overview of the clinical andsomal region 1q42–q43, a region that was implicated molecular genetics of Wilms’ tumor. Med. Pediatr. Oncol. 27: 394–in Wilms tumorigenesis previously (Hohenfellner et al., 397.1989; Mertens et al., 1997), our expression data and Haga, T., Haga, K., and Kameyama, K. (1994). G protein-coupled

receptor kinases. J. Neurochem. 63: 400–412.structural analyses of TM7SF1 suggest a closer exami-Herzlinger, D., Qiao, J., Cohen, D., Ramakrishna, N., and Brown,nation of its role during kidney development in future

A. (1994). Induction of kidney epithelial morphogenesis by cellsexperiments. To address this question, we are cur-expressing Wnt-1. Dev. Biol. 166: 815–818.rently generating GST fusions of TM7SF1-L/-I amino-

Hohenfellner, K., Holl, M., Gutjahr, P., and Zabel, B. U. (1989). Cyto-terminal regions to synthesize antibodies specific forgenetic findings in Wilms’ tumor. Klin. Padiatr. 201: 293–298.TM7SF1 that will provide us with a more detailed pic-

Ioannou, P. A., Amemiya, C. T., Garnes, J., Kroisel, P. M., Shizuya,ture of the place and developmental stage at which H., Chen, C., Batzer, M. A., and de Jong, P. (1994). A new bacterio-TM7SF1 expression occurs. phage P1-derived vector for the propagation of large human DNA

fragments. Nature Genet. 6: 84–89.Klein, P., Kanehisa, M., and DeLisi, C. (1985). The detection andACKNOWLEDGMENTS

classification of membrane-spanning proteins. Biochim. Biophys.Acta 815: 468–476.

We thank Gabi Klemm and Thorsten Enklaar, for their help in Kolakowski, L. F. (1994). GCRDb: A G protein-coupled receptor data-FISH experiments, and Jutta Busch, for excellent technical assis- base, receptors and channels. Int. J. Recept. Channels Transporterstance. The human 1 rodent hybrids used for chromosomal localiza- 2: 1–7.tion were kindly provided by Dr. S. Naylor. This work was supported

Lefkowitz, R. J. (1993). G protein-coupled receptor kinases. Cell 74:by the Deutsche Forschungsgemeinschaft. C.S. is supported by a409–412.fellowship of the ‘‘Graduiertenkolleg Molekulare Mechanismen der

Lehrach, H., et al. (1990). In ‘‘Genome Analysis,’’ Vol. 1, ‘‘GeneticPathogenese.’’and Physical Mapping’’ (K. E. Davies and S. M. Tilghman, Eds.),pp. 39–81, Cold Spring Harbor Laboratory Press, Cold Spring Har-bor, NY.REFERENCES

Liang, P., and Pardee, A. (1992). Differential display of eukaryoticmessenger RNA by means of the polymerase chain reaction. Sci-Adams, M. D., Kelley, J. M., Gocayne, J. D., Dubnick, M., Polymero-ence 257: 967–971.poulos, M. H., Xiao, H., Merril, C. R., Wu, A., Olde, B., Moreno,

R. F., Kerlavage, A. R., McCombie, W. R., and Venter, J. C. (1991). Liang, P., Zhu, W., Zhang, X., Guo, Z., O’Connell, R. P., Averboukh,Complementary DNA sequencing: Expressed sequence tags and L., Wang, F., and Pardee, A. (1994). Differential display using one-human genome project. Science 252: 1651–1656. base anchored oligo-dT primers. Nucleic Acids Res. 22: 5763–5764.

Altschul, S. F., Gish, W., Miller, W., Myers, E. W., and Lipman, D. Lin, H. H., Stubbs, L. J., and Mucenski, M. (1997). Identification(1990). Basic local alignment search tool. J. Mol. Biol. 215: 403– and characterization of a seven transmembrane hormone receptor410. using differential display. Genomics 41: 301–308.

Bell, G. I., Fong, N. M., Stempien, M. M., Wormsted, M. A., Caput, Lobbert, R. W., Winterpacht, A., Seipel, B., and Zabel, B. U. (1996).D., Ku, L. L., Urdea, M. S., Rall, L. B., and Sanchez Pescador, R. Molecular cloning and chromosomal assignment of the human ho-(1986). Human epidermal growth factor precursor: cDNA se- mologue of the rat cGMP-inhibited phosphodiesterase 1quence, expression in vitro and gene organization. Nucleic Acids (PDE3A)—A gene involved in fat metabolism located at 11p15.1.Res. 14: 8427–8446. Genomics 37: 211–218.

Luttrell, L. M., Della Rocca, G. J., van Biesen, T., Luttrell, D. K.,Charo, I. F., Myers, S. J., Herman, A., Franci, C., Connolly, A. J., andCoughlin, S. (1994). Molecular cloning and functional expression of and Lefkowitz, R. (1997). Gbg subunits mediate Src-dependent

phosphorylation of the epidermal growth factor receptor. A scaffoldtwo monocyte chemoattractant protein 1 receptors reveals alterna-

AID GENO 5170 / 6r60$$$$$1 02-25-98 14:50:49 gnma

NOVEL GENE TM7SF1 UPREGULATED IN KIDNEY DEVELOPMENT 185

for G protein-coupled receptor-mediated Ras activation. J. Biol. Sariola, H., Saarma, M., Sainio, K., Arumae, U., Palgi, J., Vaahto-kari, A., Thesleff, I., and Karavanov, A. (1991). Dependence ofChem. 272: 4637–4644.kidney morphogenesis on the expression of nerve growth factorMarchuk, D., Drumm, M., Saulino, A., and Collins, F. (1991). Con-receptor. Science 254: 571–573.struction of T-vectors, a rapid and general system for direct cloning

of unmodified PCR products. Nucleic Acids Res. 19: 1154. Schieffer, B., Paxton, W. G., Chai, Q., Marrero, M. B., and Bernstein,K. (1996). Angiotensin II controls p21ras activity via pp60c-src. J.Mason, P. J., Enver, T., Wilkinson, D., and Williams, G. J. (1993).Biol. Chem. 271: 10329–10333.Assay of gene transcription in vivo. In ‘‘Gene Transcription: A

Practical Approach’’ (B. D. Hames and S. J. Higgins, Eds.), Oxford Schofield, P. N., and Boulter, C. (1996). Growth factors and meta-nephrogenesis. Exp. Nephrol. 4: 97–104.Univ. Press, London.

Mertens, F., Johansson, B., Hoglund, M., and Mitelman, F. (1997). Schumacher, V., Schneider, S., Figge, A., Wildhardt, G., Harms, D.,Schmidt, D., Weirich, A., Ludwig, R., and Royer-Pokora, B. (1997).Chromosomal imbalance maps of malignant solid tumors: A cyto-

genetic survey of 3185 neoplasms. Cancer Res. 57: 2765–2780. Correlation of germ-line mutations and two-hit inactivation of theWT1 gene with Wilms tumors of stromal predominant histology.Nakai, K., and Kanehisa, M. (1992). A knowledge base for predictingProc. Natl. Acad. Sci. USA 94: 3972–3977.protein localization sites in eukaryotic cells. Genomics 14: 897–

911. Stark, K., Vainio, S., Vassileva, G., and McMahon, A. (1994). Epithe-lial transformation of metanephric mesenchyme in the developingNamba, T., Sugimoto, Y., Negishi, M., Irie, A., Ushikubi, F., Kaki-kidney regulated by Wnt-4. Nature 372: 679–683.zuka, A., Ito, S., Ichikawa, A., and Narumiya, S. (1993). Alternative

splicing of C-terminal tail of prostaglandin E receptor subtype EP3 Strader, C. D., Fong, T. M., Tota, M. R., Underwood, D., and Dixon,R. (1994). Structure and function of G protein-coupled receptors.determines G-protein specificity. Nature 365: 166–170.Annu. Rev. Biochem. 63: 101–131.Ozaki, K., Kuroki, T., Hayashi, S., and Nakamura, Y. (1996). Isola-

tion of three testis-specific genes (TSA303, TSA806, TSA903) by a Stultz, C. M., White, J. V., and Smith, T. (1993). Structural analysisbased on state–space modeling. Protein Sci. 2: 305–314.differential mRNA display method. Genomics 36: 316–319.

Park, S., Bernard, A., Bove, K. E., Sens, D. A., Hazen-Martin, D. J., Vainio, S., and Muller, U. (1997). Inductive tissue interactions, cellsignaling, and the control of kidney organogenesis. Cell 90: 975–Garvin, A. J., and Haber, D. A. (1993). Inactivation of WT1 in neph-

rogenic rests, genetic precursors to Wilms’ tumour. Nature Genet. 978.5: 363–367. von Heijne, G. (1992). Membrane protein structure prediction. Hy-

drophobicity analysis and the positive-inside rule. J. Mol. Biol.Rockman, H. A., Koch, W. J., Milano, C. A., and Lefkowitz, R. (1996).Myocardial beta-adrenergic receptor signaling in vivo: Insights 225: 487–494.from transgenic mice. J. Mol. Med. 74: 489–495. Vukicevic, S., Kopp, J. B., Luyten, F. P., and Sampath, T. (1996).

Induction of nephrogenic mesenchyme by osteogenic protein 1Rogers, S. A., Ryan, G., and Hammerman, M. (1991). Insulin-likegrowth factors I and II are produced in the metanephros and are (bone morphogenetic protein 7). Proc. Natl. Acad. Sci. USA 93:

9021–9026.required for growth and development in vitro. J. Cell Biol. 113:1447–1453. White, J. V., Stultz, C. M., and Smith, T. (1994). Protein classification

by stochastic modeling and optimal filtering of amino-acid se-Rosenthal, W., Antaramian, A., Gilbert, S., and Birnbaumer, M.(1993). Nephrogenic diabetes insipidus: A V2 vasopressin receptor quences. Math. Biosci. 119: 35–75.unable to stimulate adenylyl cyclase. J. Biol. Chem. 268: 13030– Woolf, A. S., Kolatsi-Joannou, M., Hardman, P., Andermarcher, E.,13033. Moorby, C., Fine, L. G., Jat, P. S., Noble, M. D., and Gherardi, E.

(1995). Roles of hepatocyte growth factor/scatter factor and theSadoshima, J., and Izumo, S. (1996). The heterotrimeric G q protein-coupled angiotensin II receptor activates p21 ras via the tyrosine met receptor in the early development of the metanephros. J. Cell.

Biol. 128: 171–184.kinase–Shc–Grb2–Sos pathway in cardiac myocytes. EMBO J.15: 775–787. Zhou, Q. Y., Grandy, D. K., Thambi, L., Kushner, J. A., Van Tol,

H. H., Cone, R., Pribnow, D., Salon, J., Bunzow, J. R., and Civelli,Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989). ‘‘MolecularCloning: A Laboratory Manual,’’ 2nd ed., Cold Spring Harbor Labo- O. (1990). Cloning and expression of human and rat D1 dopamine

receptors. Nature 347: 76–80.ratory Press, Cold Spring Harbor, NY.

AID GENO 5170 / 6r60$$$$$1 02-25-98 14:50:49 gnma