The cis-Regulatory Element Gsl5 Is Indispensable for Proximal Straight Tubule Cell-specific Transcription of Core 2 β-1,6-N-Acetylglucosaminyltransferase in the Mouse Kidney
2005; Elsevier BV; Volume: 281; Issue: 2 Linguagem: Inglês
10.1074/jbc.m509307200
ISSN1083-351X
AutoresMichiko Sekine, Chouji Taya, Hiroshi Shitara, Yoshiaki Kikkawa, Noriko Akamatsu, Masaharu Kotani, Masao Miyazaki, Akemi Suzuki, Hiromichi Yonekawa,
Tópico(s)RNA Interference and Gene Delivery
ResumoGsl5 regulates the expression of a glycolipid and glycoproteins that contain the LeX epitope in the mouse kidney through tissue-specific transcriptional regulation of the core 2 β-1,6-N-acetylglucosaminyltransferase (core 2 GnT) gene. The core 2 GnT gene has six exons and produces three alternatively spliced transcripts. Gsl5 regulates only the expression of the kidney-type mRNA, which is transcribed from the most 5′-upstream exon. By introducing a 159-kb bacterial artificial chromosome (BAC) clone that carries the mouse core 2 GnT gene and its 5′-upstream region into DBA/2 mice that carry a defective Gsl5 allele, we were able to rescue the deficient phenotype. The BAC clone was subsequently engineered to replace the core 2 GnT gene with the sequence of enhanced green fluorescent protein (EGFP) as a reporter by an inducible homologous recombination system in Escherichia coli. The transgenic mice derived from the modified BAC clone expressed EGFP in the kidney, which suggests that the candidate Gsl5 is in the 5′-upstream region of the core 2 GnT gene. Sequence analysis of the 5′-upstream regions of the BAC clone and DBA/2 genomic DNA revealed a candidate sequence for Gsl5 at about 5.5 kb upstream of exon 1. This sequence consisted of eight repeats of two GT-rich units in the wild-type mice, whereas it consisted of only one pair of GT-rich units with a minor modification in the DBA/2 mice. Transgenic mice produced with the EGFP reporter gene construct that included this candidate sequence expressed EGFP exclusively in the proximal straight tubular cells of the kidney. These results indicated that this unique repeat is indeed the Gsl5, and it is a cis-regulatory element responsible for proximal straight tubule cell-specific transcriptional regulation. Gsl5 regulates the expression of a glycolipid and glycoproteins that contain the LeX epitope in the mouse kidney through tissue-specific transcriptional regulation of the core 2 β-1,6-N-acetylglucosaminyltransferase (core 2 GnT) gene. The core 2 GnT gene has six exons and produces three alternatively spliced transcripts. Gsl5 regulates only the expression of the kidney-type mRNA, which is transcribed from the most 5′-upstream exon. By introducing a 159-kb bacterial artificial chromosome (BAC) clone that carries the mouse core 2 GnT gene and its 5′-upstream region into DBA/2 mice that carry a defective Gsl5 allele, we were able to rescue the deficient phenotype. The BAC clone was subsequently engineered to replace the core 2 GnT gene with the sequence of enhanced green fluorescent protein (EGFP) as a reporter by an inducible homologous recombination system in Escherichia coli. The transgenic mice derived from the modified BAC clone expressed EGFP in the kidney, which suggests that the candidate Gsl5 is in the 5′-upstream region of the core 2 GnT gene. Sequence analysis of the 5′-upstream regions of the BAC clone and DBA/2 genomic DNA revealed a candidate sequence for Gsl5 at about 5.5 kb upstream of exon 1. This sequence consisted of eight repeats of two GT-rich units in the wild-type mice, whereas it consisted of only one pair of GT-rich units with a minor modification in the DBA/2 mice. Transgenic mice produced with the EGFP reporter gene construct that included this candidate sequence expressed EGFP exclusively in the proximal straight tubular cells of the kidney. These results indicated that this unique repeat is indeed the Gsl5, and it is a cis-regulatory element responsible for proximal straight tubule cell-specific transcriptional regulation. Glycoconjugates are constituents of the plasma membrane and play important roles in cell-cell and cell-matrix interactions (1Lowe J. Marth J. Annu. Rev. Biochem. 2003; 72: 643-691Crossref PubMed Scopus (529) Google Scholar, 2Haltiwanger R. Lowe J. Annu. Rev. Biochem. 2004; 73: 491-537Crossref PubMed Scopus (649) Google Scholar). The glycan structures of these complex glycoconjugates change during the processes of embryogenesis, differentiation, and malignant transformation (3Varki A. Glycobiology. 1993; 3: 97-130Crossref PubMed Scopus (5004) Google Scholar). The synthesis of glycan structures begins in the endoplasmic reticulum and is completed in the Golgi apparatus, where most of the structural variations are produced. The biosynthetic pathways that produce this level of diversity are considered to be strictly regulated in tissue- and stage-specific procedures as the result of the ordered expression of glycosyltransferase activities within the Golgi apparatus. However, the molecular mechanisms remain to be revealed and constitute an important issue in glycobiology research. We focused on the polymorphic expression of kidney glycolipids among inbred strains of mice, and we found a single autosomal gene that controls the expression of a glycolipid (Galβ1-4(Fucα1-3)GlcNAcβ1-6(Galβ1-3)GalNAcβ1-3Galα1-4Galβ1-4Glcβ1-ceramide (GL-Y) 2The abbreviations used are: GL-YGalβ1-4(Fucα1-3)GlcNAcβ1-6(Galβ1-3)GalNAcβ1-3Galα1-4Galβ1-4Glcβ1-ceramidecore 2 GnTcore 2 β-1,6-N-acetylglucosaminyltransferaseBACbacterial artificial chromosomeEGFPenhanced green fluorescent proteinPFGEpulse field gel electrophoresisRTreverse transcriptionGL-XGalβ1-3GalNAcβ1-3Galα1-4Galβ1-4Glcβ1-ceramideLeXLewis X antigenntnucleotide. ) through the regulation of β-1,6-N-acetylglucosaminyltransferase (β6GlcNAc-T) activity (4Sekine M. Hashimoto Y. Inagaki F. Yamakawa T. Suzuki A. J. Biochem. (Tokyo). 1990; 108: 103-108Crossref PubMed Scopus (15) Google Scholar). We named this gene Gsl5 and mapped the Gsl5 locus to chromosome 19 (5Sekine M. Sakaizumi M. Moriwaki K. Yamakawa T. Suzuki A. J. Biochem. (Tokyo). 1989; 105: 680-683Crossref PubMed Scopus (8) Google Scholar). DBA/2 and other several strains of mice carry a recessive allele for Gsl5, which originated from Mus musculus, Asian subspecies (6Sekine M. Kikkawa Y. Takahama S. Tsuda K. Yonekawa H. Suzuki A. J. Biochem. (Tokyo). 2002; 132: 387-393Crossref PubMed Scopus (7) Google Scholar). We have made the following observations regarding Gsl5. 1) The amino acid sequence deduced from the cDNA sequence of Gsl5-controlled β6GlcNAc-T is identical to that of the mouse core 2 β6GlcNAc-T (core 2 GnT). 2) The mouse core 2 GnT gene produces three alternatively spliced transcripts, and Gsl5 controls only one of these, transcribed from exon 1, which is located closest to the 5′ end. 3) Gsl5 controls the activity of β6GlcNAc-T, whose substrates are the glycolipid Galβ1-3GalNAcβ1-3Galα1-4Galβ1-4Glcβ1-ceramide (GL-X) and the oligosaccharide Galβ1-3GalNAcα- and β-p-nitrophenyl derivatives. 4) In addition to the glycolipid, glycoproteins that bear the core 2-LeX epitope (Galβ1-4(Fucα1-3)GlcNAcβ1-6(Galβ1-3)GalNAc-) are regulated by Gsl5. 5) Gsl5 controls the level of mRNA that encodes β6GlcNAc-T in a kidney proximal tubule cell-specific manner (7Sekine M. Hashimoto Y. Suzuki M. Inagaki F. Takio K. Suzuki A. J. Biol. Chem. 1994; 269: 31143-31148Abstract Full Text PDF PubMed Google Scholar, 8Sekine M. Nara K. Suzuki A. J. Biol. Chem. 1997; 272: 27246-27252Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar, 9Sekine M. Taya C. Kikkawa Y. Yonekawa H. Takenaka M. Matsuoka Y. Imai E. Izawa M. Kannagi R. Suzuki A. Eur. J. Biochem. 2001; 268: 1129-1135Crossref PubMed Scopus (11) Google Scholar). Galβ1-4(Fucα1-3)GlcNAcβ1-6(Galβ1-3)GalNAcβ1-3Galα1-4Galβ1-4Glcβ1-ceramide core 2 β-1,6-N-acetylglucosaminyltransferase bacterial artificial chromosome enhanced green fluorescent protein pulse field gel electrophoresis reverse transcription Galβ1-3GalNAcβ1-3Galα1-4Galβ1-4Glcβ1-ceramide Lewis X antigen nucleotide. We postulated that Gsl5 might be a kidney-specific element in transcriptional regulation. It appeared to be an excellent candidate for the analysis of tissue-specific regulation mechanisms. However, the molecular basis of Gsl5 could not be addressed easily, as cultured cells that maintain the expression of Gsl5-regulated core 2 GnT are not available at present. In addition, in our preliminary experiments, even primary cultures of mouse proximal tubule cells lost the expression of kidney-specific core 2 GnT mRNA within 48 h. These results compelled us to use an in vivo genetic approach. We introduced a clone of a bacterial artificial chromosome (BAC) containing the entire core 2 GnT gene derived from a dominant mouse strain into Gsl5-deficient DBA/2 mice, and we tested whether the wild-type BAC clone was able to rescue the defective phenotype of the DBA/2 mice. One clone rescued the defective phenotype and produced GL-Y glycolipid and glycoproteins with the core 2-LeX epitope in the kidneys of transgenic mice. In addition, we were able to modify the BAC clone by homologous recombination and to define the region of the candidate sequence. Here we present for the first time evidence that Gsl5 is a unique cis-regulatory element that controls kidney proximal tubular cell-specific transcription. BAC Rescue Experiments—A mouse CITB BAC library (Research Genetics/Invitrogen) was screened by PCR, and one BAC clone carrying the core 2 GnT gene was isolated. The closed circular form of BAC DNA was purified using an alkaline lysis and cesium chloride gradient ultracentrifugation protocol (10Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY1989: 1.119-1.122Google Scholar). After overnight dialysis, the DNA was diluted to a concentration of 3 ng/μl in TE buffer (10 mm Tris-HCl (pH 7.5), 0.1 mm EDTA) and microinjected into pronucleus stage oocytes isolated from DBA/2 mice. The microinjected oocytes were transplanted into pseudo-pregnant ICR mice. The resulting mice were screened by PCR analysis with PCR primers, BAC-1 and BAC-2, for the BAC vector using genomic DNA purified from tail biopsies. The primer sequences are shown in Table 1. One transgenic line was established by mating the founder mouse and DBA/2 mice.TABLE 1Primers used for PCR and GET recombinationBAC-15′-AGGGATTGGCTGAGACGAAA-3′BAC-25′-TGGCAATGAAAGACGGTGAG-3′GNT-3′5′-GCAAAAATGAAGACAGAGCC-3′GNT-205′-GAGATCAGAAGCCACAGACT-3′GNT-225′-AGTCCCAGGTTTTCCGTGTT-3′GNT-265′-CAGGACACTGGAGGAAACAT-3′GNT-285′-CTGGAACCCAATGAAACCTT-3′GNT-295′-AGCCGATCCAGCAGTGTTTC-3′GNT-385′-GTCACCCTGTAAAATCTTGG-3′GNT-945′-ACTTGGGAATCCCGAAACTAAAGATGTAAT-3′GNT-1035′-AAAGCATTACTGTTCCCCGAGGTTC-3′GNT-170Xho5′-GATCTCGAGAGTTCAAATCCCAGCACCCA-3′GNT-22Xho5′-ATCCTCGAGAGTCCCAGGTTTTCCGTGTT-3′GNT-23Hd5′-GCTAAAGCTTTTGGGTTCCAGGTTCTTTCC-3′EGFP-R5′-TGTGGCCGTTTACGTCGCCG-3′EGFP-15′-TGGTGAGCAAGGGCGAGGAG-3′EGFP-35′-ACTGGAACAACACTCAACCCTATCT-3′EGFP-65′-CTCGTCCATGCCGAGAGTGA-3′kana-s5′-TGACCGCTTCCTCGTGCTTT-3′GNT-egfpaLowercase letters correspond to the sequences in pEGFP-N35′-GTGGAGGGAAAGAACCTGGAACCCAATGAAACCTTGATTTTAATGGGTTTatggtgagcaagggcgaggagc-3′GNT-kanaaLowercase letters correspond to the sequences in pEGFP-N35′-CTGATTAGTTCTCGAGGCTGACATTTTTCACTGGCCTCTTTCCCAATCACccagagtcccgctcagaag-3′a Lowercase letters correspond to the sequences in pEGFP-N3 Open table in a new tab To examine the expression of mRNA for core 2 GnT, total RNA samples from the kidneys or other tissues of the transgenic mice were subjected to RT-PCR. Omniscript™ reverse transcriptase (Qiagen) was used, and the PCR was performed with the following primers: GNT-28, for the kidney-type cDNA; GNT-29, for the ubiquitous-type cDNA; and GNT-38, the common antisense primer. Kidney neutral glycolipids were isolated and identified as described previously (6Sekine M. Kikkawa Y. Takahama S. Tsuda K. Yonekawa H. Suzuki A. J. Biochem. (Tokyo). 2002; 132: 387-393Crossref PubMed Scopus (7) Google Scholar). Briefly, total glycolipids were extracted from kidneys in a chloroform/methanol mixture. The neutral glycolipid fraction was eluted from a DEAE-Sephadex A-25 column and subjected to mild alkaline treatment. After neutralization, the neutral glycolipid fractions were desalted on a Sephadex LH-20 column and analyzed by TLC. Glycolipids were visualized by spraying the TLC plates with an orcinolsulfuric acid reagent and heating at 110 °C. Microsomal fractions of mouse kidneys were prepared as described previously (9Sekine M. Taya C. Kikkawa Y. Yonekawa H. Takenaka M. Matsuoka Y. Imai E. Izawa M. Kannagi R. Suzuki A. Eur. J. Biochem. 2001; 268: 1129-1135Crossref PubMed Scopus (11) Google Scholar). Protein content was determined using the BCA reagent (Pierce). The protein samples (20 μg) were applied to an SDS-polyacrylamide 3-15% gradient gel, electrophoresed, and blotted to an Immobilon-P membrane (Millipore Corp.). The IgM monoclonal antibody SA024, which recognizes the carbohydrate epitope Galβ1-4(Fucα1-3)GlcNAcβ1-6(Galβ1-3)GalNAc, and horseradish peroxidase-conjugated anti-mouse IgM antibody were used. Peroxidase activity was detected with the ECL kit (Amersham Biosciences). Modification of the BAC Clone by Homologous Recombination in Escherichia coli—The BAC clone that contained Gsl5 was modified to replace the core 2 GnT gene with the EGFP reporter gene using an inducible homologous recombination system, GET recombination, in E. coli (11Narayanan K. Williamson R. Zhang Y. Stewart A.F. Ioannou P.A. Gene Ther. 1999; 6: 442-447Crossref PubMed Scopus (139) Google Scholar). The 6578-bp pGETrec plasmid, which contains the E. coli recE and recT genes and the bacteriophage λ gam gene in a polycistronic operon (donated by Dr. P. A. Ioannou, The Royal Children's Hospital, Melbourne, Australia), was electroporated into E. coli DH10B that carried the BAC clone. The DH10B cells that carried both the BAC clone and pGETrec were prepared for recombination. The expression levels of the recE, recT, and gam genes were induced by the addition of l-arabinose to a final concentration of 0.2% (w/v) for an additional 40-min incubation. The cells were harvested and made electrocompetent (10Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY1989: 1.119-1.122Google Scholar). The PCR product that contained the EGFP/kanr cassette was amplified using the GNT-egfp and GNT-kana primers (Table 1). In Table 1, the capital letters (50 nt) correspond to the homology-targeting arms of the genomic sequences, and the lowercase letters refer to those used for primer amplification of the EGFP/kanr cassette. The pEGFP-N3 vector (Clontech) was used as the template. The purified PCR product (300-500 ng) was then electroporated into the DH10B cells that carried both the BAC clone and pGETrec (12Orford M. Nefedov M. Vadolas J. Zaibak F. Williamson R. Ioannou P.A. Nucleic Acids Res. 2000; 28: 84Crossref PubMed Scopus (46) Google Scholar). PCR screening was performed to identify recombinant clones using the combination of primer sets designed for the sequences of the BAC clone and the inserted PCR product. We obtained one recombinant clone in which the EGFP/kanr cassette replaced the core 2 GnT gene. The modified BAC DNA was digested with SacII or SalI and analyzed by pulse field gel electrophoresis (PFGE) (10Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd Ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY1989: 1.119-1.122Google Scholar). The closed circular form of the recombinant BAC DNA was purified and microinjected into pronucleus stage oocytes obtained from C57BL/6 mice, as described above for the original BAC clone. To identify transgenic mice, PCR analysis was performed with the GNT-22 and EGFP-R primers. The 707-bp fragment of the EGFP gene was amplified by PCR using the EGFP-1 and EGFP-6 primers with kidney total RNA as the template. To ensure that the PCR products were not derived from some contamination of the genomic DNA, a negative control reaction without reverse transcriptase was performed. Three transgenic mouse lines were established by mating each founder animal with C57BL/6 mice. An MZ FLIII fluorescence stereomicroscope (Leica) with a filter for green fluorescent protein was used for the histological observations. Sequencing of the BAC Clone and the Corresponding DBA/2 Genomic DNA—The BAC clone was digested with BssHII and SalI. A 40-kb fragment from the 5′ end of the clone, which included exon 1 and intron 1 of the core 2 GnT gene, was subcloned and sequenced using the BigDye Terminator cycle sequencing kit (Applied Biosystems) and the ABI PRISM 377 sequencer (Applied Biosystems). PCR products that corresponded to those of the BAC clone were obtained from the DBA/2 genomic DNA and were analyzed. Gsl5-EGFP Transgenesis—Using the BAC clone that included the core 2 GnT gene as the template, a PCR product was amplified with the GNT-103 and GNT-170Xho primers and was digested with XhoI. Another PCR product, which contained exon 1 of the core 2 GnT gene and its immediate 5′-flanking region, was amplified with the GNT-22Xho and GNT-23Hd primers and was digested with XhoI and HindIII. These fragments were purified using the QIAquick gel extraction kit (Qiagen). The pEGFP-N3 plasmid was digested with HindIII and DraIII. The fragment that included the EGFP gene was ligated with the two PCR fragments described above. The Gsl5-EGFP construct was then amplified by PCR using the GNT-103 and EGFP-3 primers with the ligated DNA as template, purified using the QIAquick gel extraction kit, and microinjected at a concentration of 3 ng/μl in TE buffer into pronucleus stage oocytes from C57BL/6 mice. A Gsl5-EGFP transgenic mouse was identified and established by mating the founder animal with C57BL/6 mice. Cryostat sections (10 μm thick) from unfixed frozen kidney samples of the transgenic mouse were placed on poly-l-lysine-coated slides, dried for 5 min at room temperature, and then exposed for 16 h at -20 °C to the vapor of formaldehyde produced from a filter paper soaked in 37% formaldehyde solution in a tightly closed plastic box (13Jockusch H. Voigt S. Eberhard D. J. Histochem. Cytochem. 2003; 51: 401-404Crossref PubMed Scopus (53) Google Scholar). After fixation, the sections were washed with TBS-Ca buffer (50 mm Tris-HCl (pH 7.6), 150 mm NaCl, 1 mm CaCl2). An MZ FLIII fluorescence stereomicroscope (Leica) with a filter for green fluorescent protein was used. Then the sections were incubated, after preincubation in TBS-Ca buffer containing 10% goat serum, with rabbit polyclonal antibody against keyhole limpet hemocyanin-conjugated C-terminal peptide, (C)TATEDTFKDTANLVKEDSDV, of mouse megalin which is a member of low density lipoprotein receptor and is localized at the proximal tubule cells (14Zou Z. Chung B. Nguyen T. Mentone S. Thomson B. Biemesderfer D. J. Biol. Chem. 2004; 279: 34302-34310Abstract Full Text Full Text PDF PubMed Scopus (142) Google Scholar), and stained with goat anti-rabbit IgG conjugated with Alexa Fluor 543 (Molecular Probes). The stained sections were observed under an LSM510 laser scanning confocal microscope (Carl Zeiss). For Western blotting of the EGFP protein, kidney homogenates were subjected to SDS-PAGE (10% polyacrylamide gel). The affinity-purified anti-EGFP monoclonal antibody JL-8 was purchased from Clontech. The secondary antibody was peroxidase-conjugated anti-mouse IgG (Santa Cruz Biotechnology), and the ECL Advance system (Amersham Biosciences) was used for visualization. BAC Rescue Experiments—The transgenesis of DBA/2 mice with a 159-kb BAC clone (BAC367O1) that contains the core 2 GnT gene rescued the defective kidney-specific transcription of the enzyme (Fig. 1A). The core 2 GnT gene consists of at least six exons spanning about 45 kb and produces three alternatively spliced transcripts (8Sekine M. Nara K. Suzuki A. J. Biol. Chem. 1997; 272: 27246-27252Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar). The kidney-type transcript is composed of exons 1-3; the ubiquitous-type contains exons 1′, 2, and 3, as well as an additional transcript, which was cloned from MDAY-Dw33 cells established from a DBA/2 mouse and encompasses exons 1″-1, 1″-2, 2, and 3. 3C. E. Warren, GenBank™ accession number MMU19265. Exons 2 and 3 are common to these three transcripts. The coding region is included only in exon 3, which indicates that the same protein is produced from the three different transcripts. BAC367O1 has 24 kb of the 5′-flanking region of exon 1, 45 kb of exons 1-3, and 90 kb of the 3′-flanking region of exon 3, giving a total size of 159 kb. The transgenic mice that carried the BAC vector sequence in their genomic DNA expressed the kidney-type transcript of core 2 GnT as well as the ubiquitous-type transcript in their kidneys (Fig. 1, B and C). Four of six transgenic F1 offspring were transgenic. These mice expressed the GL-Y glycolipid and glycoproteins that incorporated the core 2-LeX, which are expressed in dominant-type mice but not in DBA/2 mice (Fig. 1, D and E). The two nontransgenic F1 mice expressed neither GL-Y nor the glycoproteins. Modification of the BAC Clone by Homologous Recombination—We modified the BAC clone using the homologous recombination system in its host E. coli strain DH10B. The targeting arm of the GNT-egfp primer was designed for exon 1 of the core 2 GnT gene and that of GNT-kana was designed for the 3′-noncoding region of the core 2 GnT gene, so that the core 2 GnT gene (45 kb) was replaced with the EGFP/kanr cassette (2857 bp) (Fig. 2A). Electroporation of the EGFP/kanr cassette followed by selection with chloramphenicol and kanamycin resulted in the isolation of a number of clones that were resistant to both antibiotics. However, we obtained only one clone (rBACegfp) that showed PCR products that spanned the recombination junctions, which were amplified with the primer sets GNT-22/EGFP-R and kana-s/GNT-3′ (Fig. 2B). The recombinant also contained the EGFP gene (EGFP-1/EGFP-6 primer set) but not the core 2 GnT gene (GNT-26/GNT-38). PFGE analysis demonstrated that the SacII 53- and 29-kb fragments and the SalI 9- and 72-kb fragments in the original BAC disappeared, and the new SacII 40-kb and SalI 35-kb fragments were detected, indicating that the SacII and SalI restriction sites in exon 3 of the core 2 GnT gene were deleted (Fig. 2C). The rBACegfp construct was introduced into mice, generating three transgenic lines (Fig. 3A). RT-PCR analysis showed that each rBACegfp transgenic line expressed EGFP mRNA (Fig. 3B) in the kidneys. Fluorescence stereomicroscopy detected EGFP fluorescence in the corticomedullary region of the kidney (Fig. 3C). EGFP was not detected in the kidney medulla. Sequencing of the BAC Clone and the Corresponding DBA/2 Genomic DNA—The alignment of the 24-kb sequence, which includes the region 5′-upstream of the core 2 GnT gene in BAC367O1, and the DBA/2 genome revealed five deletions and four insertions of >20 nt as well as a number of single nucleotide polymorphisms in the DBA/2 genome (data not shown). Two exons of the riboflavin kinase (Rfk) gene were detected at the 5′ end of the BAC clone in the reverse direction, which represents the border of the core 2 GnT gene. There was no significant difference in the sequences of exon 1 and its immediate flanking region extending up to -400 bp of the core 2 GnT gene between BAC367O1 and the DBA/2 genome (AB219562 and AB219563). The TATA box was detected in both sequences at about 30 bp upstream of exon 1, where the kidney-type mRNA is transcribed as reported in previous paper (8Sekine M. Nara K. Suzuki A. J. Biol. Chem. 1997; 272: 27246-27252Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar). We found that DBA/2 contained an ∼350-bp deletion at 5.5 kb upstream of exon 1 (AB219561). The corresponding region of BAC367O1 included a unique tandem repeat (AB219560) (Fig. 4A), which consisted of a GT-rich, 16-nt unit (A,TGTGTGTGTGTATGTA) plus a GT-rich, 36-nt unit (B, TGGGTTTGTGTGTATGTGTT-TGTGTGGTTATATGTA). BAC367O1 contained these units repeated eight times: A-B-A′-B-A-B-B-A-B′-A-B′-A-B-A-B-A′, where A′ (TGTGTGTGTATGTA) is 2 nt shorter than A, and B′ has a C substitution for the T of TGGGTTT...in B. In contrast, DBA/2 had only one A″-B″, in which A″ (TGTGTGTGTTAGTGTATGTA) is 4 nt longer than the wild-type A, and B″ has an A substitution for the G of TGGGTTTGTGTGTATG... in B. We performed PCR analysis to determine the correlation between this tandem repeat and the expression of GL-Y in the inbred strains of mice that lacked GL-Y in their kidneys, i.e. MSM, JF1, BLG2, KJR, CHD, and SWN. All six strains had the same deletion. Mus spretus, which is the closest species to M. musculus, which does not express GL-Y, also had this deletion. On the other hand, A/Wy, BALB/c, C57BL/6, and PGN/2 mice that expressed GL-Y in their kidneys had PCR products of the same length as that of BAC367O1 (Fig. 4B). These correlations support the notion that this deletion is a candidate for the recessive Gsl5. Gsl5-EGFP Transgenesis—The Gsl5-EGFP construct shown in Fig. 5, A and B (AB234879), was introduced into C57BL/6 oocytes, and one transgenic line was established. RT-PCR analysis showed that the expression of EGFP was restricted to the kidney (Fig. 5C). Although the submaxillary gland showed slight contamination of the genomic DNA, we could not find EGFP mRNA in the liver, spleen, submaxillary gland, heart, or lung tissues of the transgenic mouse. EGFP fluorescence was detected in the proximal tubule cells in the corticomedullary region of the kidney (Fig. 6A). EGFP was not detected in either the medulla or outer cortex regions. Confocal microscopy at a higher magnification demonstrated that EGFP was localized at the proximal straight tubule cells, and megalin distribution at the luminal side of proximal tubule cells supported the EGFP localization (Fig. 6B). Western blotting with the anti-EGFP antibody confirmed that EGFP protein was expressed in the kidneys of transgenic mice (Fig. 6C). Thus, we were able to establish a transgenic mouse line that expresses EGFP in the kidneys. Taken together, these results demonstrate that the Gsl5-EGFP construct contains a kidney-specific cis-element and that A-B-A″-B-A-B-B-A-B′-A-B′-A-B-A-B-A′ is absolutely required for the transcription of the core 2 GnT gene in a proximal straight tubule cell-specific manner. We cloned the mouse core 2 GnT cDNA, showed that this gene is transcribed in a kidney-specific manner (8Sekine M. Nara K. Suzuki A. J. Biol. Chem. 1997; 272: 27246-27252Abstract Full Text Full Text PDF PubMed Scopus (27) Google Scholar), and investigated the elements involved in the regulation process defined by Gsl5. During a search for the core 2 GnT gene, we obtained a clone containing the 5′ end of the kidney-type cDNA from a λ library of the mouse genome. This clone included about 4.5 kb of the 5′-upstream region of the transcriptional start site of kidney-type cDNA and had a TATA box-like sequence at about -30 nt. However, in our preliminary promoter assays using primary cultures of mouse kidney proximal tubule cells and luciferase as the reporter, we did not find any promoter activity in the 5′-upstream region of exon 1 of the core 2 GnT gene, even up to about -4500 nt (data not shown). As the primary culture rapidly lost the expression of kidney-type mRNA, promoter activity was assayed although mRNA expression was still maintained. The 5′-upstream region of the 4.5-kb fragment seemed to be insufficient. To investigate further the longer 5′- and 3′-flanking regions of the core 2 GnT gene, we used BAC-mediated transgenesis. BAC clones can stably maintain DNA fragments of up to 300 kb (16Shizuya H. Birren B. Kim U.J. Mancino V. Slepak T. Tachiiri Y. Simon M. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 8794-8797Crossref PubMed Scopus (1381) Google Scholar) and are known to be useful in identifying cis-regulatory elements that lie far from the coding region (17Nielsen L.B. McCormick S.P. Pierotti V. Tam C. Gunn M.D. Shizuya H. Young S.G. J. Biol. Chem. 1997; 272: 29752-29758Abstract Full Text Full Text PDF PubMed Scopus (67) Google Scholar, 18Antoch M.P. Song E.J. Chang A.M. Vitaterna M.H. Zhao Y. Wilsbacher L.D. Sangoram A.M. King D.P. Pinto L.H. Takahashi J.S. Cell. 1997; 89: 655-667Abstract Full Text Full Text PDF PubMed Scopus (552) Google Scholar). The BAC clone BAC367O1, which contains the core 2 GnT gene, was able to rescue the defective phenotype of DBA/2 in terms of the kidney-specific transcription of the gene, which suggests that the Gsl5 locus that plays a key role in the transcriptional regulation of core 2 GnT is contained within this clone. Exon 1, which is transcribed in a kidney-specific manner, is located the farthest 5′-upstream, and the Gsl5-deficient DBA/2 mice retain the ubiquitous-type mRNA, which is transcribed from exon 1′, located 17 kb downstream of exon 1. Gsl5 should lie either 5′-upstream of exon 1 on BAC367O1 or in intron 1 between exons 1 and 1′. BAC engineering using the GET recombination system excluded the possibility that Gsl5 is in intron 1. The engineered BAC clone, rBACegfp, has the EGFP gene inserted between exon 1 and the 3′ end of the core 2 GnT gene, and it expresses EGFP in a kidney-specific manner, despite lacking the region between intron 1 and exon 3 of the core 2 GnT gene. Therefore, Gsl5 is located 5′-upstream of the 24-kb fragment. As a reporter, EGFP has the advantage of being directly visualized in vivo with high sensitivity and without any further treatments (19Yu Y.A. Szalay A.A. Wang G. Oberg K. Luminescence. 2003; 18: 1-18Crossref PubMed Scopus (28) Google Scholar). We analyzed the 24-kb upstream region of exon 1 and the corresponding genomic DNA sequence of DBA/2 mice. A candidate for Gsl5 was located about 5.5 kb upstream of exon 1, where the unique sequence was tandemly repeated eight times in BAC367O1. However, a similar sequence was repeated only once in the DBA/2 genome. We found binding sites for several known trans-acting factors, such as CCAAT/enhancer binding protein (C/EBP)
Referência(s)