UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The molecular biology of calbindin-D₉ Jeung, Eui-Bae 1993

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1993_fall_phd_jeung_eui-bae.pdf [ 8.05MB ]
JSON: 831-1.0086436.json
JSON-LD: 831-1.0086436-ld.json
RDF/XML (Pretty): 831-1.0086436-rdf.xml
RDF/JSON: 831-1.0086436-rdf.json
Turtle: 831-1.0086436-turtle.txt
N-Triples: 831-1.0086436-rdf-ntriples.txt
Original Record: 831-1.0086436-source.json
Full Text

Full Text

THE MOLECULAR BIOLOGY OF CALBINDIN-D 9kByEui-Bae JeungD.V.M., Seoul National University, Seoul, Korea, 1984M.Sc., Seoul National University, Seoul, Korea, 1986A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIES(DEPARTMENT OF OBSTETRICS AND GYNAECOLOGY)(REPRODUCTIVE AND DEVELOPMENTAL SCIENCES PROGRAM)We accept this thesis as conformingto the required^dardTHE UNIVERSITY OF BRIT SH COLUMBIA1993© Eui-Bae JeungIn presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature) Department of ^\;.-S' 1 11- % c_S co. a CipiA€ coicThe University of British ColumbiaVancouver, CanadaDate ^2- 7?3DE-6 (2/88)11ABSTRACTCalbindin-D9k is a cytosolic calcium binding protein expressed in themammalian intestine, placenta, and uterus. The protein is probably involved incalcium transport across the intestinal and placental epithelia. The objective ofthis thesis was to study the structure of the human and porcine calbindin-D9k atthe cDNA and genomic level. The cDNAs for the bovine, murine, and ratcalbindin-D9k as well as the gene for rat calbindin-D9k had been cloned previously.The full length cDNA encoding the human and porcine calbindin-D 9k were clonedusing a modified PCR (polymerase chain reaction) technique called RACE (rapidamplification of cDNAs ends) with rat and bovine cDNA sequence-derived primersfor amplification. The full length of these sequences was confirmed by primerextension assay. The human cDNA includes a coding region for 79 amino acids,57 nucleotides 5'- and 159 nucleotides 3'-non-coding region, and a poly(A+) tail.Northern analysis showed that the calbindin-D 9k mRNA is expressed in humanduodenum (600 nucleotides in length) but not in reproductive tissues such asplacenta and uterus. The porcine clone revealed a full-length cDNA encodingcalbindin-D9k , 79 amino acids, 57 nucleotides 5'- and 149 nucleotides 3'- non-codingregion, and a poly(A+) tail. The inferred amino acid sequence of the porcinecalbindin-D9k is identical to the published amino acid sequence, except for oneresidue. Northern analysis of porcine tissues showed a 600 nucleotide transcriptin intestine, kidney, and uterus.iiiThe gene was found to be located on the human X-chromosome by meansof PCR of hybrid DNAs from human and hamster somatic cells. The humancalbindin-D9k gene was isolated from a human X-chromosome library. Thegenomic clone contained the complete gene as well as 5' and 3' flanking regions.The structural gene, approximately 1.3 kb of 5' flanking, and 0.5 kb of 3' flankingregion were sequenced. The gene spans about 4.5 kb and consists of three exonsseparated by two introns. The first exon represents only non-coding sequences,while the second and third exons encode the two calcium binding domains of theprotein. A partial genomic sequence from pig DNA was also cloned by the use ofPCR techniques. The genomic sequences were analyzed by computer to identifyconsensus sequences for transcription factor binding sites. The human andporcine calbindin-D9k gene sequences revealed the presence of a putative estrogen-responsive element located at the boundary of exon I and intron A. This type ofsequence in the analogous location of the rat gene has been shown to mediateestrogen regulation in the uterus. The ability of these elements to bind theestrogen receptor was investigated by gel retardation assay. No binding wasdetected when the human and porcine sequences were used. The inability of thehuman sequence to bind the estrogen receptor may explain why the gene is notexpressed in human uterus (and placenta). In the case of the porcine calbindin-D9k an estrogen receptor binding site could be located in the gene's 5' regulatoryregion. In the human gene there was no such putative region found, neither inthe 5' nor 3' flanking region.ivThis study is the first report on the genomic structure and expression of thehuman calbindin-D9k. The human gene appears to be unique in that it is the onlymammalian calbindin-D 9k gene which is not expressed in the uterus and placenta.This has potential significance for fetal/maternal calcium transfer and uterinefunction in human.TABLE OF CONTENTSABSTRACT ^  iiLIST OF TABLES  ixLIST OF FIGURES ^  xLIST OF ABBREVIATIONS ^  xiiACKNOWLEDGEMENTS  xivI. INTRODUCTION^  11.1 Physiochemical properties ^  21.2 Tissue and species distribution of calbindins (Table 1) ^ 81.2.1 Calbindin-Dsk ^  81.2.2 Calbindin-D28k  101.3 Evolutionary Studies  101.4 Function ^  131.4.1 Intestine  131.4.2 Kidney  201.4.3 Placenta and yolk sac ^  221.4.4 Uterus ^  251.4.5 Bone  291.4.6 Nervous system  311.5 Molecular Biology ^  321.5.1 cDNA cloning  321.5.2 Molecular structure of the calbindin gene ^ 341.5.3 Chromosomal localization ^  351.5.4 Calbindin gene regulation  361.6 Objective ^  43viII. MATERIALS AND METHODS ^  452.1 General molecular biological techniques ^  452.1.1 Preparation of plasmid vector and bacteriophage DNA ^ 452.1.1.1 Small scale DNA preparation  452.1.1.2 Large scale DNA preparation  462.1.1.3 Bacteriophage DNA preparation ^ 472. Preparation of plating cells  472. Preparation of lysate  482. DNA preparation from lysate ^ 482.1.2 DNA/RNA preparation ^  492.1.2.1 Genomic DNA preparation ^  492.1.2.2 Total RNA preparation  502.1.2.3 mRNA isolation  512.1.3 Preparation of hybridization probes  522.1.3.1 Random primers DNA labelling system ^ 522.1.3.2 Riboprobe system^  522.1.3.3 End labelling of oligonucleotides  532.1.4 DNA / RNA analysis  542.1.4.1 Southern blot analysis  542.1.4.2 Northern blot analysis ^  552.1.4.3 Hybridization/Washing  562.1.4.4 S1 nuclease protection assay  562. Single stranded DNA preparation ^ 572. Hybridization ^  582. S1 nuclease digestion ^ 582.1.5 Cloning techniques ^  592.1.5.1 Preparation of competent cells  592.1.5.2 Transformation  592.1.5.3 DNA sequence analysis ^  602. Preparation of template DNA ^ 602. Annealing reaction  612. Labelling reaction  612. Termination reaction ^  622. Gel electrophoresis  622.2 Cloning of the human/porcine Calbindin-9k cDNAs ^ 632.2.1 Generation of primers for human/porcine Calbindin-D9k PCR ^ 632.2.2 cDNA synthesis ^  632.2.3 Core product  652.2.4 3'RACE (3' Rapid Amplification of cDNA Ends) ^ 652.2.5 5'RACE (5' Rapid Amplification of cDNA Ends)  66vu2.2.6 Cloning of PCR products ^  662.2.7 Northern analysis using species specific probes ^ 672.2.8 PCR detection of human/porcine calbindin-D 9k mRNAexpression ^  682.3 Cloning of the human/porcine calbindin-D 9k genes ^ 692.3.1 Amplification of human/porcine calbindin-D 9k intron Aregions ^  692.3.2 Primer Extension analysis ^  693.3.3 Genomic DNA Southern blot analysis ^  702.3.4 Chromosomal localization  712.3.5 Screening of cDNA and genomic libraries  712.3.6 Subcloning/Gene sequence analysis  753.4 Gel retardation assay ^  763.4.1 DNA probes for receptor binding ^  763.4.2 Preparation of rat uterine cytosol  793.4.3 Gel retardation assay ^  80III. RESULTS ^  813.1 Cloning of the human calbindin-D9k cDNA ^  813.2 Cloning of the porcine calbindin-D9k cDNA  953.3 Expression of human calbindin-D 9k mRNA ^  1003.4 Expression of porcine calbindin-D 9k mRNA  1033.5 Identification of the transcription initiation site of the human andporcine calbindin-D9k .^  1053.6 Genomic Southern blot analysis. ^  1093.7 Chromosome assignment^  1113.8 Cloning of the human calbindin-D9k intron A region ^ 1153.9 Cloning of the porcine calbindin-D 9k intron A region  1183.10 Isolation of human calbindin-D 9k gene ^  1223.11 Analysis of the human calbindin-D 9k gene sequence ^ 126viii3.12 Estrogen receptor binding assay to an ERE-like sequence in thehuman and porcine calbindin-D 9k genes ^  139W. DISCUSSION^  1454.1 Cloning of the human calbindin-D9k cDNA ^  1454.2 Cloning of the porcine calbindin-D 9k cDNA  1504.3 Calbindin-D9k gene expression in human tissues ^ 1524.4 Calbindin-D9k gene expression in porcine tissues  1554.5 Cloning of the human calbindin-D 9k gene and its analysis ^ 1574.6 Analysis of the porcine intron A sequence ^  1734.7 Estrogen receptor binding assay with an ERE-like sequence in thehuman and porcine calbindin-D9k genes  174V. SUMMARY AND CONCLUSIONS ^  180VI. REFERENCES ^  184ixLIST OF TABLESTable 1. Tissue and species distribution of calbindins. ^ 9Table 2. Chromosome contents of a human x hamster hybrid andassignment of the human calbindin-D 9k gene.  113Table 3. Chromosome localization of calcium binding proteins^ 160xLIST OF FIGURESFig. 1. Diagram of the global fold of calbindin-D 9k^  4Fig. 2. Schematic model for the action of 1,25(OH)2D3 in the intestine^ 16Fig. 3. General mechanisms of action for steroid hormones. ^ 37Fig. 4. Location and sequence of PCR primers  64Fig. 5. Screening procedure^  73Fig. 6. Diagrammatic comparison of the sequences of the seven DNAelements used in the binding studies with estrogen receptor^ 77Fig. 7. PCR strategy for the amplification of human/porcine calbindin-D 9kcDNA. ^  83Fig. 8. Southern blot analysis of core-PCR amplified calbindin-D9k cDNA. ^ 86Fig. 9. 3'RACE and 5'RACE of the human calbindin-D9k cDNA. ^ 88Fig. 10. Nucleotide and derived amino acid sequence of the humancalbindin-D9k cDNA. ^  89Fig. 11. DNA sequence alignment of calbindin-D 9k mRNAs. ^ 90Fig. 12. Interspecies comparison of calbindin-D 9k amino acid sequences.^94Fig. 13. 3'RACE and 5'RACE of the porcine calbindin-D 9k cDNA. ^ 96Fig. 14. Nucleotide sequence of the porcine calbindin-D9k cDNA.  98Fig. 15. Northern blot analysis of human and rat intestinal RNA. ^ 101Fig. 16. PCR detection of human and porcine calbindin-D9k mRNAexpression. ^  102Fig. 17. Northern analysis of porcine RNA. ^  104Fig. 18. Determination of the transcription initiation site of the human andporcine calbindin-D9k. ^  107Fig. 19. Southern blot analysis with the human calbindin-D 9k 3'RACEriboprobe. ^  110Fig. 20. Chromosome assignment of the human calbindin-D 9k gene. ^ 112Fig. 21. Southern blot analysis of human and porcine genomic PCRproducts.  116Fig. 22. Nucleotide sequence of the human calbindin-D 9k intron A region. ^ 117Fig. 23. Nucleotide sequence of the porcine calbindin-D9k intron A region. ^ 120Fig. 24. Southern blot analysis of the human calbindin-D 9k clone. ^ 124Fig. 25. Organization and sequencing strategy of the human calbindin-D 9kgene. ^  125Fig. 26. Organization of the human calbindin-D 9k promoter region. ^ 127Fig. 27. Schematic comparison of the genomic sequences of porcine, humanand rat calbindin-D9k gene. ^  128Fig. 28. Nucleotide sequence of the human calbindin-D 9k gene. ^ 129Fig. 29. Gel retardation assay using the porcine element, vitellogenin ERE,rat calbindin-D9k ERE, a mutant ERE, human calbindin-D 9k ERE,m ERE dl and m ERE dr. ^  140Fig. 30. Gel retardation in the presence of estrogen receptor antiserum. ^ 143xiLIST OF ABBREVIATIONSATP^adenosine-5'-monophosphate°C degree CelsiusCi Curie(s)CAT^chloramphenicol acetyl transferasecpm counts per minutecDNA^complementary deoxyribonucleic acidDEP diethylpyrocarbonatedNTP^deoxyribonucleoside triphosphates (dATP, &I'll', dGTP, dCTP)dATP deoxyadenosine-5'-triphosphatedCTP^deoxycytidine-5'-triphosphatedGTP deoxyguanosine-5'-triphosphatedTTP^deoxythymidine-5'-triphosphateddATP dideoxyadenosine-5'-triphosphateddCTP^dideoxycytidine-5'-triphosphateddGTP dideoxyguanosine-5'-triphosphateddTTP^dideoxythymidine-5'-triphosphateDNA deoxyribonucleic acidDNase^deoxyribonucleaseDTT dithiothreitolE.coli^Escherichia coliEDTA ethylene diaminetetraacetic acidERE^estrogen responsive elementg acceleration of gravityh hourIPTG^Isopropyl-fl-D-thiogalactopyranosideKb kilobaseKd kilodaltonsLB^Luria-BertanimCi milli-Curiesmin minute(s)Mr^apparent molecualr weightmRNA messenger RNAnt nucleotide(s)OD^optical densityPAGE polyacrylamide gel electrophoresisPCR polymerase chain reactionPEG^polyethylene glycolPEI polyethyleneiminePIPES^Piperazine-N,N' -bis(2-ethanesulfonic acid)pfu plaque forming unitpoly(A+)^polyadenylatedpoly(dI.dC)^homopolymer polydeoxyinosinic acid and polydeoxycytidylicacidrpm^revolutions per minuteRIA radioimmunoassayRNA^ribonucleic acidRNase ribonucleaserRNA^ribosomal RNART/PCR reverse transcriptase/polymerase chain reactionS Svedberg unit of sedimentationSDS^sodium dodecyl sulphatesec secondTaq Thermus acluaticus TBE^100 mM Tris-HC1, boric acid, 1 mM EDTA (pH 8.4)TE 10 mM Tris-HC1 (pH 7.4), 1 mM EDTATEMED^N,N,N',N'-TetramethyletheylenediamineTLC thin-layer chromatographyTris Tris (hydroxy methyl) aminomethanetRNA^transfer RNAUV ultravioletU unit(s)vol^volumeX-Gal 5-Bromo-4-Chloro-3-Indoy1-0-D-Galactopyranoside1,25(OH)2D3^1,25-dihydroxyvitamin D324,25(OH)2D3^24,25-dihydroxyvitamin D325(OH)D3^25-hydroxyvitamin D3xiiACKNOWLEDGEMENTSI would like to express my sincerest appreciation to my supervisors DrsJohn Krisinger and Peter C.K. Leung who have consistently provided invaluableadvice, support, guidance and friendship throughout the duration of this thesis.I again would like to thank Dr. John Krisinger for enlightened discussions andfellowship during immense beer consumption. I would like to thank mysupervisory committee, Dr. D.Rurak, Dr. S.Vincent and Dr. H.Pritchard for theirhelpful comments at committee meetings; Drs. H.F.DeLuca and J.Gorski forproviding constructs and antiserum for gel retardation assay; Cêline Conti for herskilful drawings.I would like to thank all members of Drs Krisinger and Leung's laboratoriesfor their support and for tolerating my improbable excuses. Finally, I would liketo express my deepest appreciation to my wife, Hyun-Hi, my son and daughter,Su-Young and Mi-Jin and my family their patience, understanding, love andencouragement throughout this entire thesis.1L INTRODUCTIONThe calbindins, vitamin D dependent calcium binding proteins were firstidentified in chick (Wasserman and Taylor, 1966) and rat intestine (Kallfelz et al.,1967; Drescher and DeLuca, 1971), and shown to be inducible in those tissues by1,25 dihydroxyvitamin D3 (1,25(OH)2D3). Two major subclasses of calbindins havebeen described biochemically (Fullmer and Wasserman, 1981; Fullmer andWasserman, 1987). The calbindin from chick intestine and that from mammalianintestine have been isolated and their properties studied by a number of groups(Hitchman et al., 1973; Arnol et al., 1975; Kallfelz and Wasserman, 1972; Fullmerand Wasserman, 1975; Champman et al.,1972; Tarylor et al., 1968; Wassermanand Taylor, 1971; Wasserman and Taylor, 1971; Alpers et al., 1972; Rhoten et al.,1984; Gona et al., 1986; Rhoten et al., 1986; Christakos et al., 1987). Two classesof calbindins, which differ in their calcium binding sites, have been identified inmammals The calbindin-D9k, which is characteristic for the mammalianduodenum, has two calcium binding domains, whereas the calbindin-D28k molecule,found in kidney and brain, has four calcium binding sites. The chicken expressesonly the single calbindin-D 28k with four calcium binding sites which has beenfound in high concentration in intestine, egg shell gland, kidney and cerebellum(Rhoten et al., 1984; Gona et al., 1986; Rhoten et al., 1986; Christakos et al., 1987;Parmentier et al., 1987). Chick calbindin-D 28k and mammalian calbindin-D28k havesome immunological features in common. In contrast, antibodies raised against2mammalian calbindin-D 9k are species specific. Although the intestinal and renalcalbindins have been reported to be vitamin D-dependent, calbindin-D 28k in ratand chick brain are unresponsive to vitamin D or 1,25(OH) 2D3 (Christakos et al.,1979; Sonnenberg et al., 1984; Sonnenberg et al., 1986; Schneeberger et al., 1985).To date three other vitamin D-dependent calcium binding proteins, not formallyclassified as calbindins, have also been isolated and purified: 1) an integralmembrane calcium binding protein(IM-CAL) with a Mr 200,000 which was firstidentified in intestine and subsequently has been reported in a number of othertissues (Kowarski and Schachter, 1975; Kowarski et al.,1987) 2) a 6,000 Mrcalcium binding protein found in bone which is also dependent on vitamin K andis termed the "bone gla protein" (Price and Baukol, 1981; Pan and Price, 1984).This is reflective of the vitamin K-dependent posttranslational carboxylation ofglutamic acid residues to give gamma-carboxyglutamyl or gla residues.; 3) an11,500 Mr calcium binding protein found in skin (Laouari et al., 1980; Pavlovitchet al., 1983; MacManus et al., 1985; Risk et al., 1986).1.1 Physiochemical propertiesThe vitamin D-dependent calcium binding proteins, calbindin-D 9k andcalbindin-D28k, belong to a family of intracellular proteins which have highaffinities for calcium [dissociation constant (Kd) = 10 .8 - 10 -6M]. The two calcium3binding sites in calbindin-D 9k show positive cooperativity. The primary structuresof bovine, porcine, and rat intestinal calbindin-D 9k have been obtained by aminoacid sequencing techniques (Hofmann et al., 1979; Fullmer and Wasserman, 1981;MacManus et al., 1986). Amino acid sequences have also been deduced from ratand bovine cDNA clones (Darwish et al., 1987; Desplan et al., 1983; Kumar et al.,1989). The best studied of these proteins is the bovine intestinal calbindin-D 9k(Mr, 11,000). The high-resolution (2.8 A) X-ray crystallographic structure of atruncated form of the bovine protein has been determined (Szebebyi et al., 1981;Szebenyi and Moffat, 1986). Other members of a class of proteins with "tight"calcium binding (which is not dependent upon vitamin D) include regulatory lightchain, essential light chain, calmodulin, parvalbumin, troponin C as well as thebrain S-100 protein.A common structural feature of all these proteins is the presence of anoctahedral calcium binding structure formed by a helix-loop-helix conformation ofthe polypeptide chain termed an EF hand (Fig. 1). This structural unit is madeup of two helices separated by a calcium binding loop which is typically 12 aminoacids long and is wrapped around the calcium ion in such a way that the ion isapproximately octahedrally coordinated with oxygen atoms. Pairs of EF handsappear to be the functional unit. Linker sequences connect multiple EF handswithin one protein. It has been suggested that differences in function are due todifferences in the conformation of the calcium binding loops, in the position of theNH I4Fig. 1. Diagram of the global fold of calbindin-D9k. Helices are indicated bycylinders, other sections of the protein by thick lines. Calcium ions are bound atthe positions indicated by the filled circles. Helices II and IV are held in closeproximity at their N termini by the hydrogen bonds between the calcium-bindingloops (indicated by the thin parallel lines)5helices with respect to one another and variations in the linker regions. Thestructures of parvalbumin, troponin C, calmodulin, calbindin-D 9k and calbindin-D28k have been determined crystallographically (Kretsinger and Nockolds, 1973;Moewa and Kretsinger, 1975; Sundarlinggam et al., 1985; Herzberg and James,1985; Babu et al., 1985). Determining the structure is an important first step indetermining the mode of action of these calcium binding proteins. Skeletaltroponin C and calmodulin contain 4 EF hands occupied by calcium ions whereasparvalbumin and calbindin-D 9k contain 2 EF hands occupied by calcium ions.Calbindin-D9k is a single polypeptide chain consisting of 78 amino acid residuesand an acetylated amino terminus (Fullmer and Wasserman, 1981; Hofmann etal., 1979; MacManus et al., 1986). The structure of calbindin-D 9k is shownschematically in Fig 1.. The calcium-induced changes in structure and dynamicsform a basis for their regulatory function and constitute a central issue inbiophysical studies of EF-hand proteins (Szebenyi et al., 1981; Szebenyi andMoffat, 1986). The calcium-binding site in the C-terminal half (site II) of themolecule has the same fold as that in the archetypal EF-hand, whereas the sitein the N-terminal half (site I) constitutes a variant hand (sometimes known as a'pseudo EF-hand' with two extra amino acids in the loop, one of which is a proline.The first and third loops provide the calcium ligands in sites I and II, respectively.A short antiparallel beta-sheet is formed between these loops. Essentially thereis no difference between solution and crystal structures in regards to helicalsegments and beta-sheet. Between helices I and II is a slightly longer loop whose6calcium ligands are mostly peptide carbonyls. Substitutions and / or deletions inthe N-terminal calcium-binding loop of calbindin-D 9k affect the dynamic propertiesof this site with only minor effects on the C-terminal calcium binding site (Linseet al., 1987; Johansson et al., 1990).The binding process involves dynamic changes in the proteins structure,and protein surface charges may therefore exert a profound influence on thebinding affinity through effects on the kinetics of ion association. It has beensuggested that the variant hand may be a structural site, saturated with calciumin all physiological conditions, whereas the EF hand in the C-terminal domainmay be a regulatory site which binds or releases calcium as intracellular calciumlevels vary. The powerful molecular genetic technique of site-specific mutagenesishas had a major impact on structure-function studies of proteins (Brodin et al.,1990). This technique has shown marked effects on calbindin of the contributionof specific amino acid side chains to enzyme mechanisms and of hydrophobicinteractions to protein stability, isolated helices, and ion transport, but thecalcium induced conformational changes are relatively small and do not affect thegeneral fold.Bovine, porcine, and rat intestinal calbindin-D 9k share 80-87% amino acidsequence identity (Fullmer et al., 1981; Hofmann et al., 1979; MacManus et al.,1986). However, these proteins do not cross-react immunologically (Arnold et al.,71975; Thomasset et al., 1982; Wasserman et al., 1977). The 2 loop areas arehighly conserved, but differences in the amino acid sequence are found in thelinker regions and in the helices. These regions are more exposed and presentmore immunoreactive sites. Thus the differences in these regions may explain theabsence of immunological cross-reactivity between porcine, rat, and bovinecalbindin-D9k. Calbindin-D9k proteins lack cysteine, histidine, tryptophan, and,with the exception of porcine calbindin-D 9k, arginine. Calbindin-D28k proteinscontain all these amino acid residues. The complete amino acid sequence of theMr 28,000 type was first reported for chicken intestinal calbindin-D28k (Wilson etal., 1985; Fullmer and Wasserman , 1985). Amino acid sequences havesubsequently been determined either directly or from studies with nucleic acidsfor both avian (Hunziker et al., 1986; Fullmer and Wasserman, 1987) andmammalian (Parmentier et al., 1987; Wood et al., 1988; Hunziker and Schrickel,1988; Takagi et al., 1986) calbindin-D 28k. Sequence analysis of calbindin-D 9kindicated that it is a single polypeptide chain consisting of 78 amino acid residueswith a blocked amino terminus. Calbindin-D 9k proteins from various mammalianspecies are very sequence identical to each other. Hydrophilicity plots of theproteins are similar (Hopp and Woods, 1981; Kumar et al., 1989). However, itshould be noted that there is no amino acid sequence identity between calbindin-D9k and calbindin-D 28k. In fact it has been reported that calbindin-D 9k is moreclosely related to S100 than to calbindin-D 28k (Desplan et al., 1983).81.2 Tissue and species distribution of calbindins (Table 1)Although calbindin-D9k was originally thought to be restricted tomammalian intestine and calbindin-D 28k to the avian intestine and avian andmammalian kidney, brain, and pancreas, the specific cellular distribution of thesetwo calbindins is now extended to other tissues. Calbindins in many of theseother tissues have been detected only by immunological criteria.1.2.1 Calbindin-D9kIn addition to mammalian intestine, calbindin-D9k has been found in mousekidney (Delorme et al., 1983; Rhoten et al., 1985; Schreiner et al., 1983), inneonatal rat kidney (Rhoten et al., 1985), in mammalian placenta (Bruns et al.,1978; Bruns et al., 1982; MacManus et al., 1986; Marche et al., 1978; Bruns et al.,1981; Bruns et al., 1985; Mathieu et al., 1988, Krisinger et al., 1992a,b) and yolksac (Bruns et al., 1985; Mathieu et al., 1988; Delorme et al., 1983; Bruns et at,1986), in rat uterus (Bruns et al., 1988b; Delorme et al., 1983; Mathieu et al.,1988, Krisinger et al., 1992a,b), in rat growth cartilage (Balmain et al., 1986b), inrat osteoblastic cells (Moue et al., 1988), in ameloblasts of rodent teeth (Taylor etal., 1984b; Berdal et at, 1988), in mouse embryonal carcinoma cell lines (Brunset at, 1988a), and in rat lung (Riggle et al., 1988).Table 1. Tissue and species distribution of calbindins.Calbindin-D 9k C,albindin-D 28kMammalian intestine a ' 1' Avian intestine a,bRat and mouse placenta a,b Avian,' reptilian,' amphibian,' mammalianincluding rat," mouse,' bovine,' human' kidneyRat and mouse yolk sac' Aviana and mammalian ' pancreasRat uterus a Egg shell gland (uterus) of the laying hen'Mouse kidney' Avian,' reptilian,' amphibian,' molluslcan, a fish,'and mammalian a,b brain including cells of thesensory pathwayNeonatal rat kidney' a. MamMalian cochlear and vestibular cells'Rat growth cartilage`Rat osteoblastic cells' b. Avian basilar papilla'c. Avian and mammalian retina'd. Pineal transducers 'Ameloblasts of rodent teeth'Mouse embryonal carcinoma cells` Avian and mammalian bone'Rat lung' Calcitonin cells of the chick ultimobranchial gland''By both biochemical and immunological criteria.b Sequence data available.`Principally by immunological criteria.9101.2.2 Calbindin-D28kIn addition to avian intestine and avian and mammalian kidney andpancreas, calbindin-D28k has also been localized in the egg shell gland (uterus) ofthe egg laying hen (Corradino et al., 1968; Bar and Hurwitx, 1973; Lippiello andWasserman, 1975; Fullmer et al., 1976; Jande et al., 1981b) and in brain(calbindin-D28k is widely distributed in brain) (Jande et al., 1981a; Roth et al.,1981; Baimbridge et al., 1982; Feldman and Christakos, 1983). More recentevidence localized calbindin-D 28k in specific cells of sensory pathways includingmammalian cochlear and vestibular hair cells in the inner ear (Rabie et al., 1983;San et al., 1986; Dechesne et al., 1987; Legrand et al., 1988), the basilar papillaof the chick (Oberholtzer et al., 1988), avian and mammalian retina (Rabie et al.,1985; Schrener et al., 1985; Verstappen et al., 1986), in pineal transducers (Romanet al., 1988), in the calcitonin-containing cells of the chick ultimobranchial gland(Taylor et al., 1987), and in avian and mammalian bone (Christakos and Norman,1978; Taylor, 1984b; Elms and Taylor, 1987; Balmain et al., 1986a; Zhou et al,1986).1.3 Evolutionary StudiesAlthough calbindin-D 9k appears to be confined to mammalian species,calbindin-D28k has been described in a variety of species, suggesting that calbindin-11D28k may have more widespread functional significance. Early studies(Wasserman et al., 1977) indicated that chick intestinal calbindin-D 28k antiserumcross-reacted with intestinal homogenates from a variety of species, such asamphibia, reptiles, and mammals. Further immunochemical studies (Parmentieret al., 1987) also using chick calbindin-D28k antiserum found that only a singleprotein was present in fish, but two crossreacting proteins were detected in thebrain of amphibia, birds and mammals One protein corresponded to the Mr27,000 calbindin (calbindin-D28k) and the additional one was a Mr 29,000 protein(now termed calretinin). More interestingly, the affinity of the antibody for thecalbindin-D28k was decreasing from birds (source of the antigen) to amphibia, whilethe affinity for the calretinin was almost not affected. These results indicate thatcalretinin could have evolved independently from the calbindin-D 28k, andconsequently be encoded by a different gene. The evolutionary rates werecalculated for different calcium-binding proteins of the troponin C superfamily andreference proteins (Kimura, 1983). Interestingly, calbindin-D 28k and calretininappear, with evolutionary rates of 0.27 and 0.30 x 10 -9/amino acid/yearrespectively, strongly conserved in evolution. Proteins with low evolutionary ratesare generally considered as essential proteins having extensive interactions withother macromolecules. The biological function of calbindins, usually described ascalcium buffers (Jande et al., 1981a), could therefore be underestimated andregulatory actions should be considered and more actively searched for.12The position of calbindins in the evolutionary tree of EF hand domaincontaining proteins was analyzed by comparing the homology between theindividual domains of the various calcium-binding proteins (Perret et al., 1990).It appeared clearly that both calbindins were derived from a common ancestorhaving 6 domains. An evolutionary tree was constructed and it was suggestedthat all members of the troponin C superfamily derive from a common ancestorhaving 2 domains, but that duplications leading to calbindins and to the calcium-binding proteins having 4 domains took place independently on different branchesof the evolutionary tree.In the calbindin-D9k gene each calcium-binding site is encoded by one exon(Krisinger et al., 1988). Therefore, the position of the intron correlates with thismodel of duplication and is in agreement with the suggestion of Gilbert (Gilbert,1978) that exons correspond to functional domains of proteins that can bereshuffled by recombination within the introns to create new proteins. Thegreater homology seen between the rat calbindin-D 9k (Perret et al., 1988) and ratS100 (Szebenyi and Moffat, 1986) protein than between rat calbindin-D 9k and ratcalbindin-D28k gene is not found at the nucleotide level suggesting that thestructural organization of the rat 5100 gene could be dissimilar to that of ratcalbindin-D9k gene.131.4 FunctionThe definitive role of cytosolic calbindins has been difficult to assess.Although the exact function of the calbindins has not been elucidated, hypothesesbased on experimental data and immunocytochemical localization studies havebeen presented.1.4.1 IntestineThe major factor influencing calcium absorption is vitamin D and, morespecifically, the hormonal form of this steroid, 1,25(OH)2D3 . Vitamin D deficiencyin the young, growing individual results in rickets, a disease characterized byundermineralized bone and a decreased absorption of calcium (Hess, 1929).Vitamin D deficiency in the adult, termed osteomalacia, is characterized by a lackof remineralization of the skeleton during the course of bone turnover (Sebrell andHarris, 1954). The absorption of calcium is also impaired.The source of vitamin D is from the diet or from the conversion of 7-dehydrocholesterol in the skin to vitamin D by ultraviolet light. Vitamin D istransported in blood primarily in association with the vitamin D binding protein.In the liver, vitamin D is converted to the 25-hydroxylated derivative (DeLuca and14Schnoes, 1984) and, in kidney, 25(OH)D3 is converted to 1,25(OH) 2D3 (Yamamotoet at, 1984). During periods of calcium need, 1,25(OH) 2D3 is preferentiallysynthesized and, during periods of calcium adequacy, 24,25(OH) 2D3 ispreferentially produced, the latter having a considerably lower bioactivity than1,25(OH)2D3 on calcium absorption. A high correlation exists between theconcentration of calbindin in the intestine and the efficiency or rate of calciumabsorption (Bruns et al., 1986), and this relationship seems to pertain more to theactive transport of calcium than to the non-saturable passive component.Correlations between calcium absorption and intestinal calbindin concentrationwere observed using several experimental models treated with vitamin Dmetabolites (Wasserman and Fullmer, 1989). Factors important to consider in thecalcium absorptive process are as follows: (a) the concentration of the calbindinsin the chick and rat intestinal mucosa is rather high and about 1-3% of totalcytosolic protein (Shimura and Wasserman, 1984) (b) the disassociation constantof these proteins is about 5 x 10' M; (c) the disassociation constant of thebasolateral adenosine-5'-monophosphate (ATP)-dependent calcium pump is about2 x 10' M and (d) most of the protein is apparently soluble in the cytosoliccompartment or loosely bound to components of the cell (Thorens et al., 1982;Taylor and Inpanbutr, 1988). The intestinal calbindins are detected in thecytoplasm and nuclei of absorptive cells, in goblet cells, on the surface ofabsorptive cells, in the intercellular spaces of absorptive cells, and at the apicaland basal membranes (Arnold et al., 1976; Morrisey et al., 1978; Taylor 1981;15Taylor et al., 1984a). Correlations between the induction and concentration ofcalbindin-D28k or calbindin-D9k in intestinal tissue and the rate and time courseof calcium transport provide strong support for a role of both proteins in thevitamin D-dependent transport. It has been suggested that 1,25(OH) 2D3 plays arole when calcium, after entering the cell, binds to calbindin and the diffusion ofabsorbed calcium from the microvillar region to the basolateral membrane occursin the forms of "free" calcium ion and calbindin-bound calcium ion (Fig. 2). At thebasolateral membrane, the "free" Ca' concentration is replenished by calciumdissociating from calbindin. As the calcium ion concentration in the micro-regionadjacent to the basolateral calcium pump is decreased, bound calcium ions arereleased from calbindin. Thereby the calcium pump, considered to be "starved"for calcium during non-absorptive periods, is provided sufficient calcium ions fromcalbindin to account for the accelerated absorption of calcium in the vitamin D-replete animal (Wasserman and Fullmer, 1989).Bronner et al. (1986) estimated that the rate of free diffusion of calciumthrough the cytosol of the epithelial cell was much too low to account for themaximal rate of active transport. Calbindin is taken into account as a diffusionalfacilitator using an in vitro diffusion system (Feher, 1983; Wasserman andFullmer, 1983). This system is comprised of three chambers, separated by a semi-permeable membrane. The transfer of 'Ca from the precursor compartment(intestinal lumen), through the center compartment (enterocyte) containingBRUSHBORDER MUCOSAL CELL•0 Carl P1,25(OH)2D3"SLOW"Na/Ca-ATPaseNuclearReceptorMa=DNAmRNA•I 0 V Ca.' ATPaSeI C 2,a11I •BLOOD•• 016Fig. 2. Schematic model for the action of 1,25(OH) 2D3 in the intestine. Asa consequence of the presence of the steroid-receptor complex in the nucleus, thereensues a stimulation of template activity, including the biosynthesis of a mRNAfor the calbindin. Calcium, after entering the cell, binds to calbindin and diffusionof absorbed calcium from the microvillar region to the basolateral membraneoccurs in the form of free calcium and calbindin-bound calcium. At the basolateralmembrane, free calcium is rapidly extruded from the cell and replenished bycalcium dissociating from calbindin. As the calcium ion concentration in themicro-region adjacent to the basolateral calcium pump is decreased, bound calciumion is released from calbindin.17calbindin or buffer to the blood compartment is determined. After steady state isachieved, the rate of entry of 45Ca into the blood compartment from the calbindincontaining compartment is greater than when either buffer alone or when buffercontaining bovine serum albumin is in the center compartment. Thus, theexperiments of Feher support the theoretical analysis of Kretsinger et al. (1982)and provide a basis for a mechanism by which calbindin increases calciumabsorption in the vitamin D-replete animal However, more data are required forcalcium transfer into the cell as well as translocation through the cell interior andextrusion from the intestinal cell. The first phase responds more rapidly to1,25(OH)2D3 than does the second phase and correlates with changes inphospholipid metabolism (Fontaine et al., 1981; Matsumoto et al., 1981). The firstphase may not be dependent on de novo protein synthesis. The second phasecorrelates with calbindin synthesis. It has been suggested that calbindin has arole in this second phase of the overall calcium absorptive process. The changesin calbindin synthesis in the intestine of vitamin D-deficient chicks in responseto a primary stimulus of 1,25(OH) 2D3 are well known, but the resulting calbindinconcentration is only about 3-4% of that occurring following a single physiologicaldose of vitamin D (5-10 jig) (Shimura and Wasserman, 1984).18Secondary stimulation of vitamin D-deficient chicks and rats with 1,25(OH) 2D3causes a marked increase in calbindin concentration.Similar effects on target tissue responses have been reported for othersteroid hormones. 1,25(OH)2D3 can therefore have two effects on calbindin geneexpression; enhanced transcription and/or decreased mRNA turnover. Thechanges following secondary stimulation are confined to biochemical events andare not reflected in a further increase in the physiological response after primarystimulation. This suggests that there are two pools of calbindin and that only asmall portion of the total protein available is involved in calcium absorption.Alternatively, calbindin is not an essential component of the transport systeminteracting directly with the other components of the machinery. Thus, calbindincould act partly or solely as a reservoir or buffer for calcium and fine tune thecalcium absorption process permitting faster rates of membrane translocation inthe presence of this protein than in its absence.A key controversial issue has been to determine the subcellular localizationof calbindin in the cells with which it is associated. Immunocytochemicallycalbindin has been localized in membrane-bound cytoplasmic compartments or inpreferential association with cellular membranes (Davis and Jones, 1981; Nemereet al., 1986; Freedman et al., 1981; Kubinoff and Nellans, 1985; Sampon et al.,1970). It should be noted, however, that in a recent report using biochemical19techniques a membrane-associated form of calbindin-D28k was identified in chickintestinal mucosa. The localization of calbindin primarily in the absorptive cellcytoplasm of mammalian and avian duodenum is consistent with the kinetic modelof calbindin as an intracellular facilitator of intestinal calcium diffusion (Feher,1983). The quantitative subcellular localization of calbindin was evaluated in thechick intestine using the protein A-gold technique. The gold particle label in theintestinal absorptive cells was localized in the cytosol and in the nucleareuchromatin but not in membrane-bound cytoplasmic compartments such asmitochondria, lysosomes, cisternal spaces of rough endoplasmic reticulum andGolgi, or in preferential association with cellular membranes (Taylor andInapanbutr, 1988). This suggests that calbindin may be also involved in thetranscellular transport of calcium. In very recent studies, using the colloidal goldtechnique, high concentrations of calbindin-D 28k in the nucleolus of the intestinalcells were reported shortly after 1,25(OH) 2D3 treatment. These results suggestedthat intestinal calbindin, in addition to being involved in transcellular calciumtransport, may possibly also function as a nuclear regulator. Significant nuclearlocalization of calbindin was also previously noted by Thorens et al. (1982).201.4.2 KidneyAlthough the kidney has been identified as a target organ for 1,25(OH) 2D3and although relatively high levels (2-7 pg/mg protein) of calbindin-D28k have beenreported in the kidney (Christakos et al., 1979; Sonnenberg et al., 1984), theintrinsic effects of vitamin D and calbindin on renal function remain unclear. Ithas been suggested that vitamin D is involved in the tubular reabsorption ofcalcium (Puschett et al., 1972a; Puschett et al., 1972b). In a micropuncture studyWinaver et al. (1980) indicated that vitamin D metabolites can enhance calciumtransport beyond the last accessible portion of the proximal convoluted tubules.Consistent with the micropuncture data for the site of action of vitamin D,autoradiographic data demonstrated that the nuclear uptake of [ 3H] 1,25(OH)2D3is localized predominantly in the distal nephron in chick (Stumpf et al., 1979) andrat (Stumpf et al., 1980; Narbaitz et al., 1982). The distribution of calbindin-D 28kin the chick kidney has been determined by immunoperoxidase andimmunofluorescence techniques. Calbindin-D28k was localized exclusively in thedistal convoluted tubule, the initial collecting tubule, and the early part of thecollecting tubules. The intercalated (mitochondria-rich) cells in these tubularsegments were devoid of calbindin-D 28k. The quantitative intensity of staining forcalbindin-D28k was evaluated by the protein-A gold technique in the section of thedistal convoluted tubule of the chick kidney. Calbindin-D 28k was found to beexclusively present in the cytosol and in the nuclear euchromatin but not over21membrane-bound organdies or in preferential association with cell membranes(Christakos et al., 1981; Roth et al., 1981; Taylor et al., 1982; Rhoten andChristakos, 1981; Roth et al., 1982; Christakos et al., 1987). Labelling was absentfrom the cisternae of the rough endoplasmic reticulum and the Golgi apparatus,from mitochondria, lysosomes, and membrane-bound vesicles in the apicalcytoplasm. These data indicate that in the avian kidney, calbindin is present onlyin the distal convoluted tubule, in the initial collecting tubule, and in the earlypart of the collecting tubule. In the kidney of all species studied (chick, rat, andhuman), the calbindin was present only in cells in the tubular regions of the distalnephron, in which a selective calcium reabsorption is known to occur (Delorme etal., 1983). The cytosolic and euchromatin intracellular distribution of calbindinsuggests an involvement in processes related to the regulation of the intracellulartranslocation of calcium ions rather than in regulation of calcium reabsorption perse. These results demonstrate that calbindin is not associated with membranesand, therefore, is probably not involved directly in trans-membrane calciumtransport.The mouse kidney and perinatal rat kidney are unique when compared toother tissues since it has been reported that these tissues contain both the 9 Kdand the 28 Kd calbindins Immunocytochemical studies using dual colour stainingindicated that the localization of the 9 Kd protein in the mouse kidney was foundto be strikingly similar to that of the 28 Kd calbindin (immunoreactivity was22found only in the distal nephron) (Rhoten et al., 1985). Most positive cells hadmixed colour staining, indicating the presence of both calbindins in the same cells.Both proteins were reported to be vitamin D-dependent in the mouse (Delorme etal, 1983). It is possible that the calbindin-D 9k and the calbindin-D 28k may havedifferent cellular actions in mouse kidney. One calbindin may be important as anintracellular calcium buffer whereas the other calbindin may be involved in theallocation of calcium to intracellular calcium storage sites. Why the adult mouseexpresses both calbindins while other adult species need only one type of calbindinin the kidney is not apparent at this time. In the neonatal rat immunoreactivityfor the 9 Kd protein was also confined to the distal nephron. Sinceimmunocytochemical studies indicate that the 9 Kd protein is not present insignificant quantities in the adult, these findings suggest that calbindin-D 9k hasa specialized function in rat kidney early in development (Christakos et al., 1989).1.4.3 Placenta and yolk sacBiochemical and immunological studies have provided evidence forthe presence of calbindin-D 9k in both placenta and yolk sac of rats and mice;concentrations between 1 and 7 pg/mg protein have been reported (Bruns et al.,1978; Bruns et al., 1982; Bruns et al., 1981; Mathieu et al., 1988). These resultshave raised questions concerning the regulation and functional significance of this23protein. The effect of maternal parathyroid hormone and vitamin D metaboliteson calcium transport by the placenta is difficult to interpret in animalexperimentation because of the concomitant changes in plasma calciumconcentrations (Pike et al., 1980). Serum 1,25(OH) 2D3 levels are increased inpregnant women (Steichen et al., 1977; Kumar et al., 1979; Gray et al., 1981).Because of the increase in serum vitamin D binding protein, the free level of thishormone is not elevated until late gestation (Bouillon et al., 1981). Biochemicalevidence has indicated that both placenta and yolk sac contain 1,25(OH) 2D3receptors (Pike et al., 1980). However, mouse placenta and yolk sac calbindin donot increase upon 1,25(OH) 2D3 administration (Delorme et al., 1983). In addition,Halloran and DeLuca(1981) reported that pregnant rats maintained on a longterm vitamin D-deficient diet (concentrations of 1,25(OH) 2D3 were undetectablein the maternal serum) produce grossly normal offspring. Fetuses from thesepregnant rats were able to maintain normal blood calcium and phosphorus levels.In vitro studies by Bruns et al. (1986) have indicated that 1,25(OH) 2D3 canincrease the synthesis of calbindin-D 9k in the cultured yolk sac. These resultssuggest that fetal production of 1,25(OH)2D3 may provide an alternative source ofhormone, at least for yolk sac calbindin-D9k.24Transplacental movement of calcium increases dramatically during the lasttrimester of gestation when fetal skeletal mineralization is occurring. Passivemechanisms are involved in the majority of placental calcium movement. In therat these fluxes are bidirectional with an additional active component whichresults in the net accumulation of calcium in the fetal compartment (Garel, 1983).The levels of both placental and yolk sac calbindin-D 9k rise in late gestation,suggesting a role for the protein in maternal-fetal calcium transport.Immunocytochemical studies within the 16-day placenta and yolk sac indicatedthat yolk sac calbindin-D 9k is localized to the endodermal cells of the villous yolksac and that placental calbindin-D9k is concentrated in the intraplacental yolk sac(Bruns et al., 1985). In the intraplacental yolk sac, calbindin-D9k is localized tocolumnar yolk sac cells lining the sinuses of Duval in close proximity to fetal bloodvessels. It has been suggested that these cells play an important role in thetransport of antibodies and nutrients from mother to fetus. Although the exactrole of calbindin-D9k in these cells is not known, these studies suggest thatcalbindin-D9k does act to facilitate placental and yolk sac transcellular calciumtransport.251.4.4 UterusThe calbindin-D 28k (Corradino et al., 1968; Bar and Hurwitz, 1973) ispresent in the egg shell gland of the laying hen and calbindin-D 9k (Mathieu et al.,1988; Delorme et at, 1983; Bruns et al., 1988) has been demonstrated in ratuterus. Both in chick and rat uterus, calbindins were found not to be vitamin Ddependent (Mathieu et at, 1988; Bar et at, 1984) but they increased after 1713-estradiol administration (Mathieu et at, 1988; Navickis et al., 1979). However,chick egg shell gland and rat uterus were both reported to possess receptors for1,25(OH)2D3 whose number is regulated by 17f3-estradiol (Coty, 1980; Coty et al.,1982; Waters, 1981). The calbindin-D9k in rat uterus and calbindin-D 28k in chickuterus are not stimulated by 1,25(OH)2D3 and its receptor.In the chick shell gland calbindin-D28k was found only in the tubular glandcells and not in the surface epithelium (Lippiello and Wasserman, 1975),indicating that calbindin-D 28k is present within the segment of the oviductprimarily concerned with egg shell formation. The mechanism of calcium transferacross the intestine and egg-shell gland of birds is generally assumed to be verysimilar. The evidence includes high 1,25(OH) 2D3 production and raised plasma1,25(OH)2D3 levels during egg production, the presence of 1,25(OH) 2D3 receptorsin both tissues. Furthermore, calcium absorption and intestinal and egg shellgland calbindin-D 28k also increase during egg formation. Although it is clear that26intestinal calbindin-D28k is 1,25(OH)2D3 dependent, the situation in the egg shellgland is problematical as egg laying birds on a vitamin D deficient diet stop layingand any arrest of egg production causes a marked decline in egg shell glandcalbindin-D28k concentration. 1,25(OH)2D3 increased the calbindin-D 28kconcentration in the intestine of vitamin D-deficient immature chicks but thesteroid had no effect on egg shell gland calbindin-D 28k. The egg shell glands ofthese immature birds were shown to contain 1,25(OH) 2D3 receptors. In normallaying birds, increased plasma 1,25(OH)2D3 concentrations are associated with anincrease in intestinal calbindin-D 28k concentration but egg shell gland calbindin-D28k is unaffected. Similar relationships were observed with the measurement ofcalbindin-D28k mRNA, strongly suggesting that 1,25(OH)2D3 only has an effect oncalbindin-D28k gene transcription in the intestine and not in the egg shell gland.In the rat uterus, immunocytochemical studies (Delorme et al., 1983)showed that the endometrium and myometrium both contain calbindin-D9k. In theendometrium of the nonpregnant rat calbindin-D 9k was found to be specificallylocalized in the cytoplasm of the stromal cells and not in the epithelium orglandular cells. Calbindin-D9k staining was more intense in the myometrium thanin the endometrium. In pregnant rats calbindin-D9k was localized in the uterineepithelium as well as in uterine myometrium and endometrium. The epitheliallocalization of calbindin-D 9k, was not estradiol induced in non-pregnant rats; this27suggests that its induction depends on an as yet unidentified factor. It has beensuggested that calbindin-D 9k in the epithelium may play a role in the transport ofcalcium to the fetus. The finding of calbindin-D 9k in myometrial fibers is also ofinterest since the protein has not been localized in other smooth or striatedmuscles. Mathieu et al. (1988) reported, in studies concerning the gestationalappearance of calbindin-D 9k, that uterine myometrial calbindin-D 9k was presenton day 0, became decreased to minimal levels in mid-pregnancy and increasedfrom day 15 onwards.Recent observations by Krisinger et al. (1992a, 1992b) on uterine calbindin-D9k mRNA expression demonstrated a tight regulation during the estrous cycleand profound changes in pregnant and lactating rats. The gene is under strictregulation of estradiol and its receptor during the reproductive life of the rat.Uterine gene expression starts with sexual maturity of the female and canbe provoked with estrogen treatment. Upon ovariectomy expression is eliminatedand can be restored by hormone treatment (L'Horset et al., 1990). During theestrous cycle calbindin-D 9k mRNA expression varies from very high levels atproestrus, followed by a 10 fold decrease at estrus to near undetectable amountsin diestrus (Krisinger et al., 1992a). At early pregnancy the gene is downregulated from day 2 to day 10 and mRNA levels increase from day 10 onwardsto reach a maximum at day 21 (Krisinger et al., unpublished observations).28Maximal expression levels are maintained through parturition (day 23) and thegene is abruptly down regulated between lactation day 1 and 2.The expression pattern described above may be an indication of a specificand important function of calbindin-D 9k in uterus, both in the cycling andpregnant rat. At present, there is no evidence pointing towards such a function.The hypothesis for a function of calbindin-D9k in uterus put forward by Bruns etal. (1988b) is based on the calcium binding properties of this cytosolic protein. Aneffect of calbindin-D 9k on the myometrial calcium concentration appears very likelygiven the high concentration (up to 4 pg/mg protein) and affinity for calcium.Calcium plays a key role in the control of myometrial activity. Uterine calbindinmay be involved in the regulation of myometrial cytoplasmic free calcium, andthereby exhibit an effect on uterine activity.The finding of calbindin-D9k in the epithelial cells of the fallopian tubesuggests that the calbindin-D 9k protein may be involved in transcellular movementof calcium across the fallopian tube into the luminal fluid (Bruns et al, 1985).291.4.5 BonePrevious studies using RIA have identified the presence of immunoreactivecalbindin-D9k in rat and chick bone tissue, with a range of 4-460 ng/mg proteinreported. Studies using immunocytochemical methods have indicated thatcalbindin-D9k and/or calbindin-D28k is present in chondrocytes of growth platecartilage in rats (Balmain et al., 1986a; Balmain et al, 1986b; Zhou et al., 1986)and chick (Zhou et al., 1986), in ameloblasts of rodent teeth (Taylor et al., 1984b;Berda et al., 1988; Elms and Taylor, 1984), and in rat osteoblastic cells in culture(Moue et al., 1988). RIA results suggest vitamin D dependency of calbindin-D 28kin chick bone. Vitamin D-inducible calbindin-D 9k has been reported in rat boneand calbindin-D28k in human osteoblasts (Moue et al., 1988). Bone is a secondtissue at which vitamin D acts to mobilize calcium for the circulation. The processof calcium resorption from bone has been investigated in vivo and in organ culture(Raisz et al., 1972; Reynolds et al., 1974; Stern et al., 1976; Gharabedian et al.,1974). It is not clear from immunohistochemical studies whether calbindin-D 9k orD28k in chondrocytes or in ameloblasts is vitamin D dependent. However, Sudaet al. (1985) showed that specific 1,25(OH)2D3 receptors in developing chick boneare present in dividing chondrocytes. On the other hand, autoradiographic studiesin the rat have indicated that [3H] 1,25(OH)2D3 is localized specifically not inameloblasts but rather in rat pulpal cells of the fetal rat incisor (Kim et al., 1983;Kim et al., 1985; Clark et al, 1985). The significance of these observations and the30determination of whether increased synthesis of these proteins occurs in responseto vitamin D needs further investigation.Calbindin-D9k has been localized in the cytoplasm of maturing ratchondrocytes (Balmain et al., 1986). The protein was found in the extracellularlateral edges of longitudinal septa, i.e. where mineralization of cartilage isinitiated and where matrix vesicles are preferentially localized. These findingssuggest that calbindin-D 9k may be involved in the matrix vesicle-associated processof cartilage calcification. Calbindin-D 28k has also been localized in rat growthcartilage chondrocytes (Balmain et al., 1986; Zhou et al., 1986). However, thereis some discrepancy between the reported immunocytochemical studies. Accordingto Balmain et al. (1986) calbindin-D28k appears earlier, in the resting andproliferative zone of the growth cartilage, but not in the mature hypertrophicchondrocytes. The major localization was reported to be nuclear. Zhou et al.(1986) indicated that calbindin-D28k was located mainly in the cytoplasm. It ispossible that differences in calbindin-D 28k localization may be due to differencesin antisera or in immunocytochemical procedures. Taken together, however, theimmunocytochemical studies do suggest that calbindins play a role in themovement of intracellular calcium in the chondrocyte and that they may beinvolved in the movement of calcium toward extracellular sites of calcification inthe growth plate.31In rodent teeth both the calbindin-D 9k and the calbindin-D28k appear to beconfined to a single cell type, the ameloblast (Taylor et al., 1984b; Berdal et al.,1988; Elms and Taylor, 1987). Interestingly, calbindin-D 9k and calbindin-D28k arealternatively present and absent in the same areas of the inner dental epitheliumduring enamel maturation. 'Ca has been reported to be located preferentiallyover ameloblasts in the initial enamel surface (Belanger, 1957), thus suggestinga role for ameloblasts and calbindins in the process of calcium transport into toothenamel and therefore in tooth formation.In summary, these results demonstrate the presence of calbindins in certaincells of bone and teeth. Calbindins may be vitamin D dependent in some of thesecells and not others. It should be noted that, since calbindins have not as yet beenpurified from bone, it is not yet known whether the immunologically cross-reactingproteins in bone are structurally identical to the well defined chick or ratintestinal calbindins.1.4.6 Nervous systemCalbindin-D28k was found to have a widespread distribution throughout thecentral nervous system. The cross reactivity between calbindin-D 28k and calretinindoes not allow one to exclude that both proteins were codetected in some of these32studies. For this reason, the distribution of both proteins in the rat brain werestudied using in situ hybridization techniques and two cRNA probes that do notcross hybridize with each other (Pasteels et al., 1986; Lee et al., 1987; Wassef etal., 1985). However, calbindin-D9k has not been detected in the nervous systemusing immunological techniques.1.5 Molecular Biology1.5.1 cDNA cloningA cDNA clone for the rat intestinal calbindin-D 9k was first reported byDesplan et al. (1983) using a library constructed from size-selected mRNA fromthe intestine of vitamin D-treated rats. The library was screened by a differentialcolony in situ hybridization technique by using a vitamin D-deficient(-D) and a +D[32P]cDNA probe. Soon after the molecular cloning of calbindin-D 9k (Desplan et al.,1983), the calbindin-D 28k cDNA from chick intestine was cloned with a modifiedtechnique using a probe prepared from poly(A+)-RNA enriched for calbindin-D 28kmRNA by an immunochemical technique (Hunziker et al., 1983; Wilson et al.,1985). The complete sequence of the rat intestinal calbindin-D 9k derived from thecDNA was reported (Darwish et al., 1987). Recently, the cDNA for bovineintestinal calbindin-D 9k has been isolated by screening a cDNA library prepared33from bovine intestinal poly(A+) RNA in a AZAP vector (Kumar et al., 1989). Thecloned DNA contained a 207 base pair (bp) open reading frame followed by a stopcodon and a 104 by 3' untranslated region. The deduced amino acid sequence is87 and 81% sequence identity with that of the porcine and rat calbindin-D 9k ,respectively.The amino acid sequence of the chick calbindin-D28k protein deduced fromthe cDNA sequence was reported by Hunziker et al. (1983) and Wilson et al.(1985). The molecular cloning of the cDNA for mammalian calbindin-D 28k hasbeen reported (Parmentier et al., 1987; Wood et al., 1988; Hunzinker andSchrickel, 1988). cDNA clones were isolated from either mouse (Wood et al.,1988), rat (Hunziker and Schrickel, 1988), or human (Parmentier et al., 1987)Xgtll brain libraries using antibody screening or screening with a fragment of thechicken intestinal calbindin-D28k cDNA. The coding sequence of mammaliancalbindin-D28k showed 79% homology with that of chicken (Parmentier et al., 1987;Wood et al., 1988; Hunzinker and Schrickel, 1988). The 5'nontranslated regionof rat calbindin-D28k is GC rich (72.5%); however, there is little homology to the5' nontranslated region of chicken calbindin-D 28k (Parmentier et al., 1987). The3'nontranslated region of mammalian calbindin-D 28k shares only 39% identity withthe 3'untranslated region of the chicken cDNA.341.5.2 Molecular structure of the calbindin geneThe molecular structure of the chicken calbindin-D 28k gene and the ratcalbindin-D9k gene have recently been characterized (Minghetti et al., 1988;Krisinger et al., 1988; Perret et at, 1988; Wilson et al., 1988). The calbindin-D 9kgene is 2.5 kb long and contains three exons interrupted by two introns. The firstexon represents only noncoding sequences, while the second and third encode thetwo calcium binding domains of the protein. The exon-intron boundary is locatedin between the two calcium binding domains. The CCAAT sequence has beenreported to be a crucial component of several eukaryotic promoters (Benoist et al.,1980). The promoter region of the rat calbindin-D 9k gene has a TATA box at -30,and three CAAT-like boxes at -91, -167, and -174, respectively. At the boundaryof exon I and the first intron, an imperfect palindromic sequence is detected thatoverlaps the first exon by 2 nt at position +51 and has high homology to theconsensus sequences of the estrogen- and glucocorticoid-responsive elements(Darwish et al., 1991). Within the indicated 13 nt, it matches all but one of the10 specific nucleotides of the Vitellogenin A2 derived consensusGGTCANNNTGACC, constituting an estrogen responsive element (ERE) and isresponsible for estrogen induction of calbindin-D9k in the rat uterus. A vitaminD response element has been identified in the upstream region of the rat genemediating the tissue specific induction in the rat intestine by 1,25(OH) 2D3 and itsreceptor. Repetitive elements, which may be components of putative regulatory35signals, are present 5 kb upstream from the cap site and in the 3' flanking region.The consensus sequence for splicing proposed for the 5' splice site is 5'-AG/GTAAG-3' where only GT is always present. In the calbindin-D 9k gene the two5' splice sites are 5'-AG/GTCAG-3' and 5'AG/GTGAG-3', and so are in agreementwith this model. The consensus sequence preceding the 3' splice site consists ofa stretch of 11 or more pyrimidines followed by CAG where only AG is strictlyconserved in all functional introns.1.5.3 Chromosomal localizationThe human calbindin-D 28k (Parmentier et al., 1989) and calretinin genes,two closely related calcium binding proteins, were localized by in situhybridization with the calbindin-D 28k and calretinin probes. They mapped to the8q21.3-q22.1 and 16q-q23 regions of the human genome, respectively. Theselocalizations match the chromosomal regions for the carbonic anhydrase isozymegene cluster (CA1, CA2, CA3) and the related gene CA7, respectively. Thissuggests a common duplication of the calbindin-D 28k/calretinin and the carbonicanhydrase ancestral genes. It is of interest to note that the calbindin-D 28k genedoes not colocalize with any other calcium binding protein genes mapped to date.The genes for S100 (Allore et al., 1988) and parvalbumin (Berchtold et al., 1987)have been localized to chromosomes 21 and 22, respectively. Two calmodulin36genes in the rat genome have been reported, and preliminary evidence hasindicated that calmodulin-related sequences are on many different humanchromosomes. Perhaps the separation of genes encoding the family of calciumbinding proteins has favoured the evolution of different developmental and/orregulatory functions.1.5.4 Calbindin gene regulationThe mechanism of action of 1,25(OH) 2D3 is similar to that of other nuclear-acting hormones; after binding to its specific receptor, it induces specific geneexpression (Fig. 3). The general mechanism of 1,25(OH) 2D3-mediated geneexpression is believed to involve the binding of the vitamin D receptor to1,25(OH)2D3 and the localization of the ligand-bound receptor to the cell nucleus.The hormone-receptor complex then recognizes and binds to cis-acting enhancer-like DNA sequences, located in the vicinity of 1,25(OH) 2D3-regulated genes. Thesespecific protein-DNA interactions are proposed to lead to conformational changesin the chromatin structure, which in turn may increase or decrease the rate oftranscription by RNA polymerase II. The DNA-bound vitamin D receptor mightalso interact with other transcription factors that may have positive or negativeeffects.Biological Response37Fig. 3. General mechanisms of action for steroid hormones. (A) homodimermodel: Steroid(s) passively enters the nucleus where it encounters and binds tothe steroid receptor protein (R). The receptor-ligand complex transforms to aDNA-binding form (homodimer) and interacts with specific genomic sequences(regulatory elements), resulting in the enhanced transcription of several targetgenes. The transcribed mRNA is translated in the cytosol, resulting in thesynthesis of several proteins and a multitude of other biological responses. (B)heterodimer model: A single steroid receptor (R) bound to specific genomicsequences (SRE) even when steroid is not available, may bind othertranscriptional factors (F) instead of the homodimer binding as a response toligand binding.A. Homodimer ModelTARGET CELL^TARGET CELLB. Heterodimer ModelTARGET CELL^TARGET CELL^TARGET CELLO O DNA38In the intestine, the 1,25(OH)2D3 receptor complex induces the geneencoding calbindin-D9k. The cloning of rat intestinal calbindin-D9k has allowed thedirect quantitation of calbindin-D 9k mRNA. Complementary DNA to rat intestinalcalbindin-D9k hybridizes to single 500- to 600-nucleotide long mRNA and does notcross-hybridize to mRNA for intestinal calbindin-D 28k (Perret et al., 1985).However, three species of calbindin-D28k mRNA at approximately 2.0, 2.6 and 3.1kb are present in chick intestine and in chick and rat brain and kidney (Hunziker,1986; King and Norman, 1986). It has been suggested that the three species ofchick calbindin-D28k mRNA are generated as a result of polyadenylation from threedifferent sites at the 3' end. Southern analysis suggests that the three species ofcalbindin-D28k mRNA are transcripts of a single gene.The kinetics of mRNA induction of chicken (King and Norman, 1986;Theofan et al., 1986) and rat intestinal (Perret et al., 1983; Kessler et al., 1986)and renal (Varghese et al, 1988) calbindins after 1,25(OH) 2D3 administration areconsistent with a genomic action of the hormone. Calbindin-D9k mRNA is inducedwithin lh following the administration of 1,25(OH) 2D3. In vitro induction ofchicken intestinal calbindin-D28k and calcium transport in an organ culturepreparation of embryonic chick intestine is blocked by inhibitors of RNA synthesissuch as actinomycin D, or a-amanitin. Emtage et al. (1974) isolated polysomesfrom chicken intestinal tissue capable of synthesizing calbindin-D28k anddemonstrated vitamin D dependence of calbindin-D 28k mRNA synthesis.39A good correlation is found between tissue levels of calbindin-D 28k and the levelsof calbindin-D 28k mRNA in total polysomes.In addition, administration of 1,25(OH) 2D3 to vitamin D-deficient animalshas been shown to induce the transcription of the calbindin gene in chick and ratintestine and in rat kidney.There is evidence that vitamin D and its activemetabolites may not be the exclusive regulators of calbindin-D28k or calbindin-D9kconcentrations in the intestinal and renal cells. 1,25(OH)2D3 has been shown tomodulate calbindin gene expression by a rapid transcriptional stimulation (thepeak of calbindin gene transcription occurs at 1-3 h) (Theofan et al., 1986; Dupretet al., 1987; Varghese and Christakos, 1988) which precedes the peak ofaccumulation of calbindin mRNA (12h) and calbindin protein (48h). The synthesisof calbindin-D28k from nuclear RNA is not seen until nuclear 1,25(OH)2D3accumulation is at a maximum. These results suggest the presence of additionalnuclear and/or cytosolic events in the regulation of calbindin-D 28k .The kidney of some mammals is a unique tissue because both calbindin-D 9kand calbindin-D28k are localized in the same cells of the distal convoluted tubule.The time course of calbindin-D9k and -D28k responses to 1,25(OH) 2D3 differedmarkedly using the mouse kidney as a model (Li and Christakos, 1991). The peakinduction of renal calbindin-D 28k mRNA occurred at 12h after a single injection of1,25(OH)2D3 to vitamin D-deficient mice, and a decrease was observed at 24h40(similar to the time course of the calbindin-D 9k gene in the intestine and othersteroid regulated genes). Unlike calbindin-D28k, there was a delayed response ofcalbindin-D9k in mouse kidney to 1,25(OH)2D3 (the peak of induction wasconsistently observed at 24h after 1,25(OH) 2D3 adminstration). Differentialregulation may be due to a requirement by different genes for differenttranscription factors and/or for different amounts of receptor. In addition, bindingaffinities may differ between the DNA of different genes and the hormonereceptor. Alternatively, differences may exist in posttranscriptional regulation.Differential regulation of the two calbindins by 1,25(OH) 2D3 may also be relatedto their different functions. Because a delayed response to 1,25(OH) 2D3 is notobserved for calbindin-D9k in intestine, this suggests that tissue-specific factorsmay also be involved in the regulation of gene expression by 1,25(OH) 2D3 .The data from several laboratories provide evidence that calbindins, similarto other steroid-induced proteins, are regulated at the transcriptional andposttranscriptional level and that stabilization of transcribed mRNA may be animportant mechanism in the regulation of calbindin-D9k gene expression by1,25(OH)2D3 . Most recently an element in the rat calbindin-D 9k gene has beenidentified to mediate responsiveness to 1,25(OH)2D3 and its receptor (Darwish andDeLuca, 1992). The approach has been to ligate 5' flanking regions of thecalbindin gene to the gene for chloramphenicol acetyl transferase (CAT) and a tk-promoter. Cotransfection of CV-1 cells (African green monkey kidney cell) with41vitamin D receptor expressing constructs was necessary to elicit a vitamin Dregulation in these cells. It was suggested that the endogenous number ofreceptors does not suffice to allow a vitamin D regulation of the calbindin-D 9kgene. This may also be the reason for unresponsiveness of the calbindin-D 9k geneto 1,25(OH)2D3 regulation in certain tissues despite the presence of the vitaminD receptor. A vitamin D-responsive element in the osteocalcin gene has also beenidentified recently. The calbindin-D 9k D-response element (GGGTGT CGGAAGCCC) differs by 4 nucleotides from the element of the rat osteocalcin gene(Bortell et al., 1992; Demay et al., 1990; Markose et al., 1990). This element wasfound to form complexes with 1,25(OH) 2D3 receptor in vitro, and such protein DNAbinding was shifted to a higher molecular weight form during gel retardation uponthe addition of a monoclonal antibody specific to the 1,25(OH) 2D3 receptor.Further studies have provided evidence that the calbindin genes are notalways regulated by exogenous 1,25(OH)2D3 adminstration and that factors otherthan 1,25(OH)2D3 can modulate their expression. Unlike calbindin-D9k in intestineand -28k in kidney, brain calbindin-D 28k and its mRNA are not affected by vitaminD-depletion. Also brain calbindin-D 28k does not respond to vitamin Dadministration. These observations suggest tissue specificity in the regulation ofcalbindin gene expression. Additional evidence of 1,25(OH)2D3 independent factorsthat can regulate calbindin gene expression was recently presented by Hall et al.(1987) who noted that glucocorticoid treatment could alter calbindin-D 28k gene42expression in the intestine of vitamin D-deficient chicks; this observation suggestsa possible regulation of the calbindin-D28k gene by glucocorticoids.431.6 ObjectiveThe objective of this thesis was to study the structure and regulation of thehuman and porcine calbindin-D9k at the cDNA and genomic level. Especially inhuman, the information on calbindin-D 9k is very scarce and this lack of dataprevents any projections of findings in animal models to humanphysiology/pathophysiology.Recent evidence from work in the rat model suggests a crucial role ofintestinal and placental calbindin-D 9k in calcium transport. At least in theintestine, compromised calcium absorption is associated with decreased calbindin-D9k expression and hypocalcemia. Although data on human calbindin-D 9k are verylimited, the protein has been identified in small intestine in analogy to animalmodels. Experiments to study the regulation of human calbindin-D 9k in theintestine in a similar way as has been done in the rat model are not feasible.However molecular cloning of the corresponding gene would allow to address thequestion whether the human calbindin-D 9k gene is under vitamin D control as inthe rat and serves as an essential part of the calcium absorptive apparatus inhuman.44A function of calbindin-D9k in uterus is unknown. The current hypothesisis that mediated through estrogen control, the protein may alter myometrialcalcium and thereby influence smooth muscle activity. Therefore, a participationof uterine calbindin-D9k in maintenance of pregnancy and parturition is possible.No data are available on calbindin-D 9k in the human uterus. In human pregnancycontrol myometrial activity is also in part subject to steroid hormone control.Idiopathic, preterm labor (abnormal myometrial activity) is a disorder unique tohuman. The increasing body of information on the rat uterine calbindin-D 9k hasprompted the investigation into the human calbindin-D 9k. The hypothesis wasthat: If uterine calbindin-D 9k is directly associated with uterine activity in humanpregnancy, (a) the gene will be expressed in this tissue in a similar, steroid-dependent manner as in the rat and (b) there should be a difference in calbindin-D9k expression correlated with clinical signs of uterine activity (labor).Finally, placental calbindin-D 9k in human has not been investigated at alland this project will address the question whether the gene is expressed in thistissue.Overall, the aim is to clone the human cDNA and gene encoding calbindin-D9k which will provide the basis for the specific studies mentioned above and thetools for analysis of the tissue specific expression in vitro using moleculartechniques.45II. MATERIALS AND METHODS2.1 General molecular biological techniques2.1.1 Preparation of plasmid vector and bacteriophage DNA2.1.1.1 Small scale DNA preparationPlasmid vectors were propagated in E.coli strain DH5a. Five ml of LB-Broth with the appropriate antibiotic inoculated with cells were grown overnightat 37°C and then collected by centrifugation. Cells were resuspended in 150 illsolution I (50 mM glucose, 25 mM Tris-HC1 (pH 8.0), and 10 mM ethylenediaminetetraacetic acid (EDTA) (pH 8.0)) and chilled on ice. After 5 min, 200 Alof freshly prepared solution II (0.2 N NaOH and 1 % SDS) was added and mixedby inversion. Another 300 Sul of ice-cold solution III (3 M potassium acetate, 5 Macetic acid) was added and mixed for 5 min on ice. The mixture was extractedwith an equal volume of phenol/chloroform. After brief centrifugation thesupernatant was taken into a fresh tube and precipitation followed with 2 vol ofethanol at -20°C for 5 min. After centrifugation at 14,000 rpm for 10 min, thesupernatant was discarded and the DNA pellet washed with 70 % ethanol. Thepellet was dissolved in 10 Al of TE buffer. This plasmid "mini-prep" method wasused for a rapid analytical restriction digest analysis of plasmids before large scale46preparation and sequencing reactions. Large scale DNA preparationFor large scale preparation of plasmid DNA a single step alkali preparationmethod was performed as follows: From a culture of bacteria containing therecombinant plasmid, overnight grown cells were harvested at 5,000 x g at 4°C for15 min. Cells were resuspended in 6 ml of freshly prepared ice-cold lysis buffer(25 mM Tris-HC1 (pH 8.0), 10 mM EDTA, 15 % sucrose, and 2 mg/ml lysozyme),and placed in ice water for 20 min. Twelve ml of freshly prepared 0.2 N NaOH,1 % SDS were added, mixed by inversion, and kept on ice for 10 min. Another 7.5ml of 3 M sodium acetate (pH 4.6) was added, and placed on ice for 20 min. Forseparation of the DNA containing solution from the precipitate, the sample wascentrifuged at 10,000 x g for 15 min. The supernatant was removed andincubated with 50 pl of RNAse A (stock 1 mg/m1) for 20 min at 37°C. One volumeof phenol:chloroform (1:1) saturated with TE buffer was added and vortexed for5 min and centrifuged at 10,000 x g for 10 min. This step was performed twice.The upper aqueous phase was transferred into a fresh tube and a third extractionwith an equal volume of chloroform:isoamyl alcohol (24:1) followed. The aqueousphase was transferred to a fresh tube containing 2 vol of 95% ethanol, mixed andleft at -20 °C. After 30 min, the precipitated DNA was centrifuged at 10,000 x g47for 20 min. To dissolve the pellet, 1.6 ml water was added. For differentialplasmid precipitation, 0.4 ml of 4 M NaCl and 2 ml of 13 % polyethylene glycol(PEG, Mr 6,000) were added and mixed. After 60 min in ice water, the pellet wasspun down at 10,000 x g for 10 min, washed with 70 % ethanol, and dissolvedwith appropriate amount of water. Bacteriophage DNA preparationThis method was used for rapid initial characterization of multiple samplesof cloned bacteriophage DNA. A high-titer phage lysate (10 nil) was made for eachclone of interest. Several micrograms of DNA can be prepared form a portion ofthe 10 ml lysate for further restriction enzyme analysis. Preparation of plating cellsCells were grown to an OD of 0.5 at 600 nm with 50 ml of LB- broth at37°C. The cells were harvested at 3,000 x g for 10 min and resuspended with 25ml 10 mM MgSO 4. These plating cells were kept at 4°C and used for 1 week.482. Preparation of lysateUsing a pasteur pipette an agar plug containing a single phage plaque wasadded to 10 ml of LB-broth supplemented with MgSO 4 (10 mM final concentration)in a 50 ml tube. Fifty ml of plating cells were added and shaken vigorously withgood aeration until the culture became clear indicating cell lysis. After lysis, 100pl of chloroform was added to tube and shaken for 2 min at 37°C. The mixturewas spun down for 10 min at 3,000 x g at room temperature to remove bacterialdebris. The supernatant was transferred to a new tube and 100 pl of a sterile 1M MgSO4 solution was added to 10 ml of lysate. DNA preparation from lysateTen ml of TM buffer (50 mM Tris-HC1 (pH 7.4) and 10 mM MgSO 4) and 320pl of freshly made deoxyribonuclease (DNase) I solution were added to 10 ml oflysate, mixed gently by inversion and incubated for 15 min at room temperature.Two ml of 5 M NaCl and 2.2 g of solid PEG-6000 were added and incubated for15 min on ice. The mixture was then spun down for 10 min at 12,000 x g at 4°C.The supernatant was poured off. The precipitated phage pellet was resuspendedwith 300 pl of TM buffer and transferred to a microfuge tube. After twochloroform extractions, 15 p1 of 0.5 M EDTA (pH 8.0) and 30 pl of 5 M NaCl were49added to the aqueous phase. After another phenol/chloroform extraction, the DNAwas precipitated with ethanol.2.1.2 DNA/RNA preparation2.1.2.1 Genomic DNA preparationHigh molecular-weight cellular DNA from tissues was extracted by usingthe phenol/chloroform method. After a tissue sample of 1 - 2 g was frozen inliquid nitrogen, the tissue was crushed with a mortar and pestle. The powderedtissue was suspended in 1.2 ml digestion buffer (100 mM NaC1, 10 mM Tris-HC1(pH 8.0), 25 mM EDTA (pH 8.0), 0.5% SDS, and 0.1 mg/ml proteinase K) per 100mg tissue. The mixture was incubated with shaking at 50°C overnight in a tightlycapped tube. After incubation, the digested homogenate was extracted three timeswith an equal volume of phenol/chloroform/isoamylalcohol and once withchloroform. The extracted DNA was then precipitated by adding 1/5 vol of 10 Mammonium acetate and 2 vol of 95 % ethanol. After centrifugation, the DNA wasdissolved in TE buffer. Residual RNA was removed by adding 0.1 % sodiumdodecyl sulfate (SDS) and 1 pg/m1 DNase-free RNase and incubated lh at 37°C.The concentration of DNA was determined spectrophotometrically by absorbanceat 260 nm. The quality of DNA was evaluated by agarose gel electrophoresis.502.1.2.2 Total RNA preparationTotal RNA from porcine and human tissues was prepared by theguanidinium isothiocyanate-cesium chloride (CsC1) technique (Glisin et al., 1974).Tissue samples were homogenized in 8 - 10 ml of 4 M guanidinium isothiocyanatebuffer containing 5 mM sodium citrate (pH 7.0), 0.5 % sarcosyl and 1% 6-mercaptoethanol with a high speed polytron homogenizer for 30 seconds. Thisstep was performed in a sterile 50 ml conical culture tube. The disrupted tissuemixture was centrifuged 5 min at 15,000 rpm. The supernatant was transferredto a fresh tube and 1 g CsC1/2.5 ml supernatant was added. The solution waslayered onto 3 ml 5.7 M CsC1 containing 0.1 M EDTA (pH 7.5) in 13 ml BeckmanUltra Clear polyallomer tubes (Beckman Instruments, Palo Alto, CA). Thesamples were centrifuged at 28,000 rpm in a Beckman Ultracentrifuge (modelLM8) using an SW 40 rotor for 18 h at 18°C. The supernatant was removed andthe pellet was washed with 1 ml of diethylpyrocarbonate (DEP) water. The RNApellet was dissolved in 10 mM Tris-HC1 (pH 7.4), 5 mM EDTA, and 1% SDS andtransferred to a 15 ml corex tube. The dissolved sample was extracted withchloroform/n-butanol (4:1) and precipitated with 1/10 vol 3 M sodium acetate (pH5.2) and 2.2 vol ethanol at -20°C overnight. The pellet was dissolved withappropriate amount of DEP water. The purity and integrity of RNA wasdetermined by A260/A280 and denaturing agarose gel electrophoresis. RNAproperly isolated with this method was used directly for oligo(dT) selection of poly51A(+) RNA, Northern blots, and cDNA synthesis. mRNA isolationAffinity chromatography using oligo(dT) cellulose was performed at roomtemperature. The column was prepared by pouring oligo(dT) cellulose slurry intoa sterile disposable plastic column. Total RNA samples were heated to 70°C for10 min. This mixture was cooled down to room temperature for 2 min and appliedat a flow rate of 0.5 ml/min to the oligo(dT) cellulose packed column. The eluatewas collected and reapplied to the column. The column was washed andribosomal RNA containing washes discarded. The column was eluted 5 times withelution buffer (100 mM Tris-HC1 (pH 8.0) and 1 mM EDTA) and the RNAconcentration of each fraction was measured by IN spectrometry (OD at 260nm).All positive fractions were pooled and precipitated with ethanol. The pellet wasresuspended with DEP water and used for either northern blot analysis or PCRexperiments.522.1.3 Preparation of hybridization probes2.1.3.1 Random primers DNA labelling systemRecombinant cDNA inserts were isolated by restriction enzyme digestion,followed by electrophoresis on 1% agarose and electro-eluted. cDNA (25 ng) insertwas labelled with a[32P]dCTP by a random primers DNA labelling kit using theprocedure described by the manufacturer (BRL, Burington, Ontario, Canada). Todenature the DNA, the sample was boiled for 5 min followed by rapid chilling onice. To the boiled cDNA solution were added the following: 2 p.1 of each of 0.5 mMdATP, dTTP and dGTP, 15 pl of random primers buffer, and 50 pCi a[32P]dCTP.One pl of DNA polymerase I, Klenow fragment (3 U) was added and the mixturewas incubated at 25°C for lh. Prior to use, the probe was denatured at 100°C, for3 mM. The probe was labelled to a specific activity of 10 9 counts per minute(cpm)/pg DNA. Riboprobe systemA DNA fragment was inserted into the polylinker cloning site (EcoRI) of thevector pSPT 19 containing the DNA-dependent RNA polymerase promoter. Thetemplate was prepared by growing and amplifying the plasmid in E.coli DH5a.53To linearlize the recombinant plasmid template, the DNA was digested with anappropriate restriction enzyme leaving the insert connected to the promoter.Following digestion, the plasmid was extracted with phenol/chloroform twice andprecipitated with ethanol. The pellet was dissolved with water at 200 ng/pl. Thefollowing reagents were added in a total volume of 20p1: 4 pl reaction buffer, 0.5p.1 1 M dithiothreitol (DTT) solution, 10 pl a[ 32P]UTP, 14 U of the DNA-dependentRNA polymerase. The mixture was incubated for 60 min at 37°C. To stop thereaction, 20 pl of H2O, 25 p.1 2 % SDS containing 2 mM EDTA and 100 pl 0.1 Msodium acetate (pH 5.0) were added. The solution was extracted with 100 pl eachof phenol and chloroform, and precipitated with 500 pl of ethanol. The probe wasdissolved in 1 ml of DEP water. The specific activity of the synthesized RNA wasmeasured by TCA (Trichloroacetic acid) precipitation. The specific activity of theRNA probes was 10 9 cpm/pg. End labelling of oligonucleotidesHigh specific activity 7[32P]dATP was used with polynucleotide kinase toend-label DNA by phosphorylation of 5'-hydroxyl. To 10 pmole oligonucleotidewere added 1.5 pl of kinase buffer (500 mM Tris-HC1 (pH 8.3), 100 mM MgC1 2, 50mM DTT, 1 mM spermidine, and 1 mM EDTA), 50 pCi A 3211dATP, 10 U T4Kinase (BRL). The mixture was incubated at 25°C for 30 min. To test the54incorporation of 7{3213, a TLC procedure was used. A small aliquot (<1 pl) oflabelled sample along with free ..43213]clATP were spotted on PEI-cellulose TLCplates and chromatographed in 0.5 M sodium dihydrogen phosphate (pH 3.5). Thesamples were chromatographed until the solvent front reached the top (10 cm).The plate was dried and exposed to X-ray film for 5 sec. An autoradiographicsignal at the origin indicated labelled DNA whereas free yrPRIATP migrated withRf (retention factor) of 0.5. Prior to use the 32P-labelled probe was boiled for 10min then cooled in ice-water for five min.2.1.4 DNA / RNA analysis2.1.4.1 Southern blot analysisDNA samples were size-fractionated by electrophoresis on 1.0% agarosecontaining lx TBE buffer. After electrophoresis the DNA was visualized with along-wave length UV light source by staining with liig/m1 ethidium bromidesolution. The gel was soaked in 0.2 N NaOH, 0.6 M NaC1 for 45 min at roomtemperature in order to denature the DNA fragments. The alkali solution wasremoved and the gel was neutralized by soaking in 1 M Tris-HC1 (pH 7.4) and 0.6M NaCl for 45 min. The gel was then placed on a Whatman 3MM filter paperresting on a glass plate set up in a pyrex dish. The bottom of the pyrex dish was55covered with 3x SSC (stock 20x SSC; 3.0 M NaCl and 0.3 M sodium citrate (pH7.0)) but only to a level lower than the Whatman 3MM filter. The filter waslonger than the glass plates so that the edges were soaked in the 3x SSC solution,and the entire filter is kept wet by capillary action. A nylon membrane filter, cutto the exact dimensions of the agarose gel, was then prewet with 3x SSC solutionand laid on top of the gel. A second Whatman 3MM filter, identical in size to thegel was layered on top of the nylon membrane. A stack of paper towels, about 10cm in height, was placed on top of the second Whatman 3MM filter. This set upwas left overnight at room temperature for capillary transfer of the DNA from thegel onto the nylon membrane. The nylon membrane was subsequently rinsed oncewith lx SSC, air dried and irradiated by UV light for 5 min in order to immobilizethe DNA. Northern blot analysisTotal RNA samples were denatured in 50% formamide and 2.2 Mformaldehyde at 65°C for 15 min. The samples were electrophoresed through1.0% agarose, containing 2.2 M formaldehyde, 20 mM 3-(N-morpholino)propanesulfonic acid (MOPS), 8 mM sodium acetate and 1 mM EDTA (pH 8.0).Ribosomal RNAs (28S and 18S) were visualized under UV light to ensure theintegrity of RNA samples. RNA was transferred to nylon membranes by capillary56transfer using 3x SSC. After transfer, the RNA was cross linked to the membraneby -UV light for 5 min. Hybridization/WashingThe filter was soaked in a prehybridization solution (50% formamide, 5xDenhardt's solution (lx Denhardt's solution is 0.02% ficoll, 0.02% bovine serumalbumin, 0.02% polyvinylpyrrolidone), 5x SSPE (lx SSPE is 3 M NaC1, 200 mMsodium phosphate, 20 mM EDTA (pH 7.4)), 0.1% SDS and 0.1 mg/ml denaturedsalmon sperm DNA) at 42°C for at least 12 hours. The filter was subsequentlyhybridized with a radioactive labelled probe overnight at 42°C. The filter wasthen washed 3 times at 42°C for 20 min with about 100 ml of lx SSC containing,0.1% SDS and once with 0.1X SSC, 0.1% SDS at 50°C for 20 min. The blot wasthen dried briefly and autoradiographed by using intensifying screens at -70°C. S1 nuclease protection assayThe restriction enzyme digested DNA was labelled by the filling-in reactionmethod using DNA polymerase I, Klenow fragment. The reaction (total volume. 201x1) was as follows: 2.0 ill 10x Klenow buffer (0.1 M Tris-HC1 (pH 8.0), 50 mM57MgC12, 300 mM NaCl), 50 iaCi a[32P]dATP, 2.0 Ill deoxyribonucloside triphosphates(dNTPs) (C, G, T) and 10 U Klenow fragment for 2 hr at 25°C. After incubationthe reaction mixture was precipitated with ethanol including glycogen ascoprecipitator. Single stranded DNA preparationThe pellet was resuspended with 20 pl strand separation dye. To denaturethe DNA, samples were boiled for 5 min, and chilled on ice before running on 6%PAGE (acrylamide : bisacrylamide . 50 : 1). After overnight electrophoresis at 60V, the gel sandwich was taken apart, covered with saran wrap and exposed for 5min onto X-ray film. To mark the orientation, fluorescent dye strips were placedbeside the gel. To recover the sense and anti-sense DNA, the bands on the X-rayimage were cut out and used as windows to excise the radioactive gel slices. Thegel slices were placed into a pipette tip sealed by melting and plugged withglasswool. After 20 min at 60°C, the solution was eluted through glasswool, theDNA was ethanol precipitated and counted.582. HybridizationAfter the pellets were redissolved with hybridization buffer (100 mMPIPES, 2 M NaC1, 5 mM EDTA) at 10,000 cpm/30111, 24111 formamide and 20 jigtotal RNA were added. To denature the samples, the reactions were incubatedat 75°C for 15 min and transferred directly to hybridization temperature (51°C)without letting the tube cool. S1 nuclease digestionTo digest unhybridized single stranded DNA, Si nuclease digestion followedwith S1 nuclease buffer (280 mM NaCl, 30 mM sodium acetate (pH 4.4), 4.5 mMZnC12), 1111 glycogen (20)4111), and 200 - 300 units/ml S1 nuclease (1,000 U/ml) atroom temperature for lh. The reaction was stopped with 70)11 stop buffer (4 Mammonium acetate, 50 mM EDTA (pH 8.0 )) and precipitated with ethanol. Afterresuspended with TE buffer,the sample was run on 8% PAGE.592.1.5 Cloning techniques2.1.5.1 Preparation of competent cellsTo prepare competent cells for transformation, a single colony of E.coliDH5a was incubated with 10 ml of LB-broth at 37°C overnight with constantshaking. One ml of the cell culture was added to 50 ml of LB and incubated for2 to 3 h until the cell density reached an OD of 0.6 at 600nm. The culture waschilled on ice for 5 min and harvested by centrifugation for 5 min at 2,500 x g at4°C. The cell pellet was resuspended in 25 ml of solution A (10 mM MOPS (pH7.0), 10 mM RbC1). After additional centrifugation, the cell pellet wasresuspended in 25 ml of solution B (100 mM MOPS (pH 6.5), 50 mM CaC1 2, 10mM RbC1) and kept for 15 min on ice. After the above preparation, the cells wereready for DNA transformation. Aliquots of competent cells were snap frozen inethanol/dry ice and kept in -70°C. TransformationIn a sterile Eppendorf tube, 10 Al (1-50 ng DNA) of ligated recombinantplasmid was mixed with 200 Al of competent cells and kept for 30 min on ice. Thetube was transferred to a 42°C water bath for 1 min to heat-shock the cells and60then transferred to a 37°C incubator for 1 h in the presence of 0.5 ml LB-broth toallow expression of the antibiotic resistance gene of the plasmid. 0.2 ml of the cellmixture was spread over the surface of an LB-agar plate supplemented withampicillin. Following overnight culture, positive colonies were selected by smallscale plasmid preparation. DNA sequence analysisThe dideoxy sequencing method from denatured, double stranded plasmidtemplates was used for the DNA sequence analysis (Sanger et al., 1977). Preparation of template DNAThe DNA of interest was cloned into the pUC19 and the recombinantplasmid was prepared by the small or large scale procedure. Eight pl DNA (2 pg)was mixed with 2 pl of 2 N NaOH and incubated at room temperature for 10 min.Then, 7 pl water, 3 pl of 3 M sodium acetate (pH 4.8) was added and thedenatured DNA was precipitated by the addition 60 pl of ethanol at -70°C for 5min. The precipitate was harvested by centrifugation at 14,000 rpm for 5 min,rinsed once with 70% ethanol, and dried under vacuum. The denatured DNA wasdissolved in 10 pl of water prior to the annealing reaction.612. Annealing reactionEnzyme reactions for sequencing including template-primer annealing andlabelling-extension reactions were described in the instruction manual of the T7DNA polymerase sequencing system (Pharmacia, Vancouver, B.C, Canada). Tothe 10 ill of denatured template DNA from above, 2111 of annealing buffer (40 mMTris-HC1(pH 7.4), 20 mM MgC12, 50 mM NaC1) and 2 Al of primer were added andincubated at 37°C for 20 min. The mixture was allowed to cool slowly to roomtemperature (about 10 min). Labelling reactionTo the tube containing the annealed template and primer, 3 pi of labellingMix-dATP, 1 ill of water, 10 liCi of a[ 35S]dATP, and diluted T7 DNA Polymerase(3 U) were added and incubated at room temperature for 5 min. While thisincubation was in progress, the four sequencing mixes (each termination mixcontained 80 pM dGTP, dATP, dTTP, dCTP and 50 mM NaCl - in addition, the'G'mix contained 8 pM ddGTP, the 'A'mix contained 8 pM ddATP, the 'T'mixcontained 8 pM ddTTP, and the 'C'mix contained 8 pM ddCTP) were dispensedand warmed by placing the microcentrifuge tubes at 37°C for at least 1 min. Afterthe 5 min incubation of the labelling reaction, the samples were immediatelysubjected to the termination Reaction.622. Termination reactionAfter the labelling reaction 4.5 ul aliquots were transferred into each of thefour pre-warmed sequencing mixes. After incubation at 37°C for 5 min, 5 ul ofstop solution (98% formamide, 10 mM EDTA, 0.1 % xylene cyanol, and 0.1 %bromophenol blue) was added into each tube. Just prior to loading the gel, thereactions were heated to 95°C for 5 min, and 3 p1 of sample was immediatelyloaded onto a sequencing gel. Gel electrophoresisThe sequencing gel (38 x 50 cm, 0.4 mm) was prepared according to themethod described in the manufacturer's instruction manual (BRL, Burlington,Ont., Canada) Three Ill of each reaction sample was applied to one lane of thesequencing gel. After electrophoresis at 2,000 V for 2.5 h, the gel was dried at80°C for 1 h on a gel dryer under vacuum. Finally, the dried gel was placed in acassette with x-ray film for autoradiography. The result of the sequence wasanalyzed using the GCG program (Genetics Computer Group, Inc., Madison, Wi).632.2 Cloning of the human/porcine Calbindin-9k cDNAs2.2.1 Generation of primers for human/porcine Calbindin-D 9k PCRA series of oligonucleotides (synthesized by J. Hewitt, Dept. of Biochemistry,U.B.C.) was used as primers for PCR amplification of calbindin-D 9k cDNAs.Names, locations, lengths and sequences of the primers are shown in Fig. cDNA synthesiscDNA was synthesized from total RNA (20 pg) using Avian MyeloblastosisVirus (AMV) Reverse Transcriptase (RT, 20 U/ pg total RNA). cDNA synthesiswas primed with either 2.0 pg oligo d(T) 1248 for core PCR, or 0.2 pg NotI-d(T) 19for 3' RACE (3' Rapid Amplification of cDNA Ends), or 1.0 pg B1 for 5' RACE (seeFig. 4). The reaction buffer for cDNA synthesis contained 10 mM Tris-HC1 (pH8.3), 6 mM MgC12, 40 mM KC1, 50 mM DTT, 1mM dNTPs, and 25 U RNaseinhibitor. The reaction mixtures (25 p.1) were incubated at 42°C for lh, followedby 10 min at 95°C, and chilled on ice. A 1 pl aliquot of the reaction was amplifiedby PCR.64Fig. 4. Location and sequence of PCR primers. The schematic calbindin-D 9kmRNA structure and location of primers are shown. The primer name, sequence,and its 5'-3'-direction are shown on the bottom of the figure.A. 5' non-coding^coding-region^3' non-coding Human 5' 1 1 3'3^1Rat^5' 1 I^ I^ t 3'1^2^3Porcine^5' 1^1^ I^ 1 3'-4-2—Bovine^5' ) I^ I^ 1 3'B.Species Name Length Sequence (5-0--3")Human HI 23 GCC AAA GAA GGT GAT CCA GAC CAH2 22 CTG GGG AAT TCA GAC TGA ATC AH3 21 CCA GAC ACC AGA ATG AGT TCAH4 20 CCA GTCTCTC AGG GTT CTA TrH5 21 CTC CTC 'ITT GAT TCT TCT AGCRat RI 20 AAG AGC ATT TTT CAA AAA TAR2 26 CTT CTC CAT CAC CGT TCT TAT CCA GCR3 25 GAT CCA AAC CAG CTG TCC AAG GAG GPorcine P1 20 ACT CCT ATT TGA TTC TTC CAP2 20 CCT TCT TTG GCT GCA TAT TTTBovine B1 26 CTA ACT TCT CCA TCT CCA TTC TTG TC652.2.3 Core productPCR amplification was performed using a rat sequence derived primer (R1)and bovine derived primer (B1). The PCR mixture contained 10 mM Tris-HC1 (pH8.3), 50 mM KC1, 1.5 mM MgC1 2, 0.01% gelatin, 50 pm dNTPs, and 1 U AmpliTaqDNA polymerase (Perkin Elmer, Vancouver, B.C, Canada). The PCR step cycleprofile (DNA thermal cycler, Perkin Elmer) was used for generating a 167 by coreproduct; denaturing for 30 sec at 96°C, annealing for 30 sec at 50°C, and extensionof primers at 72 °C for 1.5 min. After 30 cycles, an extension step at 72°C for 7min was added.2.2.4 3'RACE (3' Rapid Amplification of cDNA Ends)The amplification of the 3'end was carried out in a 25 pl volume using a 1pl aliquot of the cDNA reaction primed by the NotI-d(T) 18 primer. Primers foramplification were R1 and NotI-d(T) 18 (see Fig. 7). The remaining reagents andPCR conditions were as described above.662.2.5 5'RACE (5' Rapid Amplification of cDNA Ends)cDNA synthesis was primed by a bovine derived primer Bl. For removalof unincorporated nucleotides and excess primer the reaction mixture was purifiedby the Geneclean kit (Bio101, LaJolla, CA., USA). The cDNA was homopolymertailed in a 25 ill reaction using terminal deoxynucleotidyl transferase and dATP.A 1 iul aliquot of the tailed product was used for PCR with NotI-d(T) 18 and R1primer (see Fig. 7). The remaining reagents and PCR conditions were asdescribed above.2.2.6 Cloning of PCR productsA protocol is included here because it was found that the products of PCRreactions do not appear to clone as efficiently as predicted for their massabundance. One reason for this is that most PCR products, although in theoryblunt ended, are not, due to the tendency of Taq DNA polymerase to add one ormore extra dATP residues onto available 3' ends. This problem may becircumvented by the incorporation of restriction sites on the 5' end of PCRoligonucleotides; however, one must be certain that comparable internal restrictionsites are not present in the DNA fragment to be amplified.67To facilitate blunt cloning of PCR amplified DNA fragments, amplified DNAwas incubated with Klenow mix (2311 of 20 mM Tris-HC1, 100 mM MgC1 2, 0.03 UKlenow fragment of E.coli DNA polymerase/311) and incubated at 25°C for 30 minto remove terminal overhanging single stranded DNA. The incubated mixturewas run on a 6% polyacrylamide gel and purified using the Mermaid Kit (BioCan,Mississauga, Ont). The DNA was dissolved in an appropriate volume of water forligation with 10 ng of digested vector DNA (Smal-digested pUC19 plasmid forblunt end cloning) and subsequently transformed into competent cells.2.2.7 Northern analysis using species specific probesSamples from human placenta and myometrium were obtained from theGrace Hospital. Patients undergoing caesarian section for various reasons wereasked to participate in the study. After obtaining the patients' consent, a tissuespecimen of about 500 mg was removed at the site of incision. Other humantissues were received from the National Disease Research Interchange (NDRI,Philadelphia, PA). The protocol was approved by the University of BritishColumbia Ethics Committee. For porcine Northern blot analysis, a 6-wk-oldfemale Yorkshire piglet was obtained from a local farm. The animal was killedwith an injection of pentobarbital sodium in accordance with the guidelines of theAnimal Care Committee of the University of British Columbia. Total RNA from68duodenum, jejunum, ileum, kidney, uterus, liver, lung, heart, pancreas, spleen,aorta, ovary, fimbriae, cervix, and brain was collected. For Northern analysis, 10lig of poly(A+) or 20 jig of total RNA was used. Each Northern blot was probedwith the human and porcine 3'RACE clone.2.2.8 PCR detection of human/porcine calbindin-D 9k mRNA expressionTo analyze expression of the calbindin-D 9k gene in human tissues, totalRNA and first strand cDNA were prepared from tissues using the previouslydescribed protocol. For detection of human calbindin-D 9k gene transcription, analiquot of the first strand cDNA reaction containing RNA:cDNA hybrids wasmixed with 100 pmol of the sense and antisense primers (for oligonucleotidessequences see Fig. 4) for PCR (30 cycles). Routinely, 10% of a PCR sample wereseparated by electrophoresis on a 5% PAGE to analyze the amplified DNAproducts.692.3 Cloning of the human/porcine calbindin-D 9k genes2.3.1 Amplification of human/porcine calbindin-D9k in tron A regionsHuman genomic DNA (100 or 200 ng) was boiled for 5 min and brieflycentrifuged. An aliquot of boiled DNA was amplified with 100 pmol of H5 and H2-primers (see Fig. 4). Porcine DNA was amplified with P1 and H2- primers. PCRamplification was performed in 1 x PCR buffer containing 50 pm dNTP's and 1 UTaq DNA polymerase. Following 30 cycles of PCR (96°C for 45 sec, 50°C for 30sec, 72°C for up to 1.5 min depending on expected size), 10% of each reaction wasanalyzed by agarose gel electrophoresis. The products were cloned and thesequence determined.2.3.2 Primer Extension analysisTo map the sites of transcription initiation for the human and porcinecalbindin-D9k genes, an oligonucleotide was synthesized which was complementaryto nucleotides +45 to +65 of the translation initiation site. Approximately100,000-300,000 cpm of end-labelled primer was incubated with 20 jug of totalRNA isolated from human and porcine duodenum as well as pectoral muscle asnegative control. The RNA was dissolved in a 5 iil reaction mixture containing 0.570M NaC1, 40 mM PIPES (pH 6.8) and 1 mM EDTA for 90 min at 50°C. Theannealed primer reaction mixtures were diluted 1:10 into primer extension buffer(50 mM Tris-HC1 (pH 7.6), 60 mM KC1, 10 mM MgC12, 1 mM of each dNTP, 1 mMDTT, 1 Will RNasin (Promega, Madison, Wis., USA)) and was extended with 5 Uof AMV reverse transcriptase per 50 pl reaction for 60 min at 42°C, and theextension products were separated by electrophoresis on 6% denaturingpolyacrylamide gels. A control sequencing reaction of pUC19 plasmid used as sizemarker.3.3.3 Genomic DNA Southern blot analysisThis procedure was employed to determine the molecular structure of thecalbindin-D9k gene in the human genome. The DNA was digested with a varietyof restriction endonucleases using conditions recommended by the manufacturererovernight. The restriction DNA fragments were separated on 0.8% agarose. 10lig of DNA was applied to each lane of the gel. Blotting and hybridization was asdescribed in Chromosomal localizationGenomic DNA from 25 hamster/human hybrid cell lines (BIOS Laboratories,New Haven, Con., USA) was screened for the detection of human calbindin-D 9kusing PCR. The primer pair was designed after the intronic sequence determinedfor the intron A clone (5'TGCAGAGTGCAGATCAGGTACA3', and5'CCATTTCCTTGGCATAAA3'). The reaction mixtures containing 50 ng of hybridcell line genomic DNA were subjected to 30 cycles PCR. The reaction productswere electrophoresed on a 1.2 % agarose gel and stained with ethidium bromide.2.3.5 Screening of cDNA and genomic librariesFor isolation of the cDNA encoding the human calbindin-D 9k three libraries(term placenta, fetal lung, and fetal kidney)) were screened at a density of 20,000plaque forming unit (pfu) per 150 mm plate. Library screening was performedaccording to the instruction of Clontech (Palo Alto, Cal., USA) with somemodifications. The placental library was constructed in the Xgt 11 vector. Thisphage was propagated in E.coli Y1090 host strain grown in LB-brothsupplemented with ampicillin (50 mg/1). The fetal lung cDNA was oligo(dT)-primed and cloned into the Agt 10 vector. This library was plated on E.coliC600hfi strain in LB-broth containing 0.2% maltose. The fetal kidney library wasalso oligo(dT) primed and cloned into A,gt 10 vector.72For the isolation of the human calbindin-D 9k gene, two different genomiclibraries were used. One was prepared from human leukocytes (MboI-partiallydigested) inserted into EMBL-3, and grown in E.coli P2-392 strain on NZY (10 gNZ amine, 5 g yeast, 5 g NaC1, 0.94 g MgC12/1) medium. The second library wasa human chromosome X library in charon 35 hosted in E.coli K802 recA - straingrown in the LB medium containing 0.2 % maltose. The insert DNA frommultiple X-containing 1635 cell line was partially digested with Sau3AI andcloned into the BamHI site (size; 10 - 22 kb) of this vector. One genome wascompletely covered with 1.45 x 10 4 clones of this library.The standard protocol of library screening is presented in Fig. 5. Briefly,a single, isolated colony of host strain cells was picked and grown to saturationin LB broth containing 0.2% maltose at 37°C with good aeration. Approximately0.5 ml of cell culture and 0.1 ml of cDNA or genomic library containing 25,000recombinant phages were mixed and incubated at 37°C for 15 min After additionof 10 ml molten LB top agar, the culture was mixed and poured onto a 150 mmpetri dish with LB bottom agar and then incubated at 37°C for 8h. Afterincubation, the plates were cooled to 4°C for 30 min to allow the top agar toharden. After nylon filters were placed onto the top agar, they were marked in3 asymmetric locations by stabbing through them and into the agar with an 18-gauge needle. After 60 sec, filters were carefully peeled off the plates and floated,plaque side up, on DNA denaturing solution (1.5 M NaCl, 0.5 M NaOH) for 30 sec,73Fig. 5. Screening procedure. The screening procedure of a genomic/cDNAlibrary is depicted schematically. Following plating and growth to an optimaldensity, DNA containing plaques are lifted onto a nylon membrane and hybridizedwith a radio labelled probe. Location of plaques producing DNA of interest isdetermined by autoradiography of filters. Note that it is important to mark thecorresponding positions in the plate, filter and autoradiogram to be able toreidentify the position of positive plaques.748. RemoveMembrane Positive Clones9. Denature, Washand DryPinholeMarks11.Wash Membrane12.Dry13. Autoradiograph14. Identify PositiveClones10. Hybridize withRadio labelled ProbeSealed boxHybridization Buffer1.Pour Infected cells in TopAgar Onto LB Agar Plate5. Plaques Appear inBacterial "Lawn"2. Swirl3.Wait4. IncubateAgarNeedle6. ApplyNylonMembrane7. Mark Membrane/Agarwith Needle Holes75then immersed for 60 sec. The filters were removed and transferred toneutralizing solution (1.5 M NaC1, 0.5 M Tris-HC1 (pH 8.0)) for 5 min, brieflyrinsed with 3X SSC and placed on filter paper to dry. To cross link the DNA,filters were irradiated by UV light for 5 mM, DNA side down. The hybridizationand washing steps of library screening were as described for Southern blotanalysis. The filters were covered in plastic wrap and subjected toautoradiography at -70°C, using an intensifying screen. After developing, the filmwas aligned with the filter to identify positive plaques. An agar plug containingseveral plaques was removed with a pasteur pipette and transferred into 0.5 mlsterile phage dilution buffer (20 mM Tris-HC1 (pH 7.4), 100 mM NaCl, 10 mMMgSO4) and replated at 1,000 pfu/90 mm plate. These plaques were thenrescreened. A single, well isolated positive plaque was picked as a plate stock.The preparation of X DNA from liquid lysates is described in Subcloning/Gene sequence analysisThe purified bacteriophage DNA was digested with EcoRI, separated on anagarose gel and then transferred to a nylon membrane (see Southern analysis).The filter was hybridized with the human 3'RACE clone and intron A clones asdescribed above. Each EcoRI fragment of the genomic clone was subcloned intopUC19 vector. Plasmid DNA was isolated from each of the subclones.76Approximately 2.0 ug DNA was sequenced using a modified T7 DNA polymerasewith both M13 universal and reverse primers (section Specific sequencingprimers were synthesized from the obtained sequences in order to read throughlong (more than 400 bp) DNA segments.3.4 Gel retardation assay3.4.1 DNA probes for receptor bindingEnhancer elements (estrogen response elements, EREs) from the ratcalbindin-D9k and the Xenopus vitellogenin A2 and a mutant ERE were suppliedby Dr. H.F. DeLuca (Dept. of Biochemistry, Univ. of Wisconsin-Madison, Madison,Wis., USA). Four 21-mer oligonucleotides corresponding to related sequenceswithin the human and porcine calbindin-D 9k genes and two mutants weresynthesized with BamHI overhangs. The sequences of these oligomers are givenin Fig. 6. The complementary oligonucleotides were annealed in 10 mM Tris-HC1(pH 8.0), 200 mM NaCl and 1 mM EDTA by heating to 75°C and cooling to roomtemperature over a period of 3 h. The annealed oligomers were cloned into theBamHI site of the pUC 19 vector. The recombinant clones were confirmed bysequence analysis. The clones were then digested with EcoRI and XbaI enzymesto generate 45-nucleotide fragments with the test sequence approximately in the77Fig. 6. Diagrammatic comparison of the sequences of the seven DNAelements used in the binding studies with estrogen receptor. Rat and vitERE are the estrogen responsive elements from the rat calbindin-D 9k andvitellogenin A2 gene. A mutant sequence (mERE) has two base differences fromthe vit ERE. These three oligomers cloned into BamHI site of pUTKAT1 weresupplied by Dr. H.F. DeLuca. The putative EREs from the human (Human) andporcine (Pig) calbindin-D 9ks gene were synthesized. mERE dr and -dl denote thehuman sequence with the right or left part as found in the human calbindin-D 9kgene and the other part as found in the rat ERE. All sequences were synthesizedwith BamHI ends, cloned into the BamHI site of pUC19 and recovered as 45-ntEcoRI-XbaI fragments, labeled with "P and purified on polyacrylamide gels.78^BamHI^ BamHI^I IIVitellogenin^GATCC I AGGTCA CTG TGATCT GRat^GATCC AGGTCA GGG TGATCT GHuman^GATCC AGGTTA GTG TGATTT Gm ERE dr^GATCC AGGTCA GGG TGA'TTT Gm ERE dl^GATCC AGGTTA GGG TGATCT GPig^GATCC AGGTAA GTG LGATTT Gm ERE^GATCC AGATCA CTG TGATCT G EEI..18nt21nt//6ntXXI- -45nt../../.0"79center. The digested DNA was labelled with arPRICTP and a[ 32P]dATP using theKlenow fragment of DNA polymerase I. The corresponding fragments werepurified over 8% polyacrylamide gels, excised and electroeluted. The elutedprobes were counted using 13-scintillation counting.3.4.2 Preparation of rat uterine cytosolRat uteri were taken from immature 19-day-old female Sprague-Dawleyrats killed by cervical dislocation and obtained by a ventral incision. Afterexcision, the uteri were freed of mesentery and adipose tissues, and placedimmediately in ice-cold saline solution. The tissue was minced in homogenizationbuffer (10 mM Tris-HC1 (pH 7.4), 50 mM NaCl, 1 mM DTT and 10% glycerol; 3uteri/ml) and disrupted by a Dounce homogenizer. The crude homogenate wasthen centrifuged at 212,000 x g in a TL-100.2 rotor (Beckman, Palo Alto, Cal.,USA) for 10 min at 4°C. The supernatant was recovered and 1713-estradiol wasadded to a 10 nM final concentration. The cytosol/ligand mixture was thenincubated for 1 h on ice. For heat transformation of the estrogen receptor, theuterine cytosol was incubated at 25°C for 45 min. Protein concentration wasdetermined by the Bradford method using crystallized bovine serum albumin asstandard. Heat transformed and nontransformed cytosols were frozenimmediately in liquid nitrogen and stored at -70°C until use (not more than 4weeks).803.4.3 Gel retardation assayRat uterine cytosol (10 jig of protein) was incubated for 15 min on ice with1 jig poly(di.dC) on ice for 15 min. Approximately 45,000 cpm of the individuallabelled 45-mer DNA fragments was added to the mixture and incubationcontinued on ice for an additional 15 min. The final mixture volume was 16 pl.A 4% PAGE gel (6.7 mM Tris-HC1 (pH 7.4), 33 mM sodium acetate, and 1 mMEDTA) was pre-electrophoresed at 10 mA for 10 min with the gel running buffercontinuously recirculated. The samples were applied and electrophoresed for 1.5h at 25 mA at 4°C. The gel was dried down on a slab gel dryer, and then exposedto Kodak X-OMAT X-ray film overnight at -70°C.In order to confirm the presence of the estrogen receptor in the retardationcomplex a rat estrogen receptor antiserum was obtained from the University ofMaryland School of Medicine, National Hormone and Pituitary Program (providedby Dr. J. Gorski, University of Wisconsin, Madison, Wi). The antiserum wasdiluted in homogenization buffer. Antiserum was included in the incubationmixture for 4 h prior to electrophoresis. The amount necessary to shift theretardation complex further was titrated by using 1:5, 1:25, 1:125, 1:625 dilutionsof the antiserum.81III. RESULTS3.1 Cloning of the human calbindin-D9k cDNAInitially, cloning of the human calbindin-D 9k cDNA was attempted using arat cDNA probe for screening of suitable cDNA libraries. The rat probe wasunable to detect a human cDNA clone from several libraries (placental, fetalkidney, and fetal lung) nor did it detect an RNA molecule on a northern blot ofhuman RNA. Efforts on the human project were additionally hampered bydifficulties to obtain fresh human tissues, which proved to be necessary forisolation of non-degraded RNA. This was especially problematical because of therapid decay of duodenal tissue, which is predicted to have the highest expressionlevel of calbindin-D9k. Furthermore, northern analysis of uterine and placentaltissue failed to reveal calbindin-D9k expression, which was unexpected based ondata from the rat model.It was also decided to examine cross hybridization of the rat probe withporcine calbindin-D 9k, to estimate the potential of this probe to hybridize to otherspecies' calbindin-D 9ks. Isolation of a porcine calbindin-D 9k clone was in partaimed to be used as an alternative probe for isolation of a human clone.82Due to this unexpected difficulty of isolating a human calbindin-D9k cDNAclone, it was decided to devise a PCR strategy for cDNA cloning, hopefully onewhich might provide a universal tool. PCR relies on the presence of uniqueoligonucleotide sequences in the gene of interest, or at least the presence ofoligonucleotides which will define members of a set of related genes. Analysis ofthe rat (Darwish et al., 1987) and bovine (Kumar et al., 1989) calbindin-D9k cDNAsequences revealed potential highly identical regions within the linker of twocalcium binding domains of these proteins. Therefore, reverse transcriptionfollowed by PCR was performed to clone the human calbindin-D 9k cDNA fromduodenal tissue using an upstream primer identical to the rat RNA (R1) and adownstream primer complementary to the bovine RNA sequence (B1). Thestrategy used for amplification is schematically shown in Fig. 7.A PCR product of the expected size (167 bp) was obtained. Hybridizationof the Southern blotted PCR product with a radio labelled rat cDNA probeconfirmed the specificity of the human PCR product. Rat and bovine intestinalcDNAs were used as a positive control in the PCR experiments. The rat probeproduced a strong signal after Southern analysis of the rat, bovine, and porcinePCR products (Fig. 8), while the human product showed only after longerexposure. The 167 by product was cloned and sequenced, which confirmed thehuman origin of the product ruling out contamination by rat cDNA.83Fig. 7. PCR strategy for the amplification of human/porcine calbindin-D 9kcDNA. (1) Core PCR: First strand cDNA synthesis was primed with oligo(dT).PCR was performed using internal rat and bovine derived primers. (2) 3'RACE:First strand cDNA synthesis was primed with NotI-d(T) 18. PCR was performedusing the anchor NotI-d(T)18 and species specific primers. (3) 5'RACE: Firststrand cDNA synthesis was primed with species specific oligonucleotides.Homopolymer tailing with dATP provided the annealing site for the anchoredprimer NotI-d(T) 18. A species specific and the NotI-d(T) 18 primer were used forPCR.(1) Core PCR: Heterologous primers (rat and bovine Calbindin-D 9k)Total RNA (duodenum)reverse transcriptionFirst strand cDNA poolPCR3' ds. PCR productNorthern blot confirmation with core product1Klenow fill-in reactionBlunt-end cloning in pUC19/sequencing(2) 3' RACE (3' Rapid Amplification cDNA End): rat specific primer and NotI-d(T) 18adapter primerTotal RNAreverse transcriptionFirst strand cDNA poolPCRrat specific primer (R 1 )5 , 11111 AAA --AA 3'TTT..TT NotI Southern blot confirmation with core productKlenow fill-in reactionBlunt-end cloning in pUC19/sequencing84(3) 5' RACE (5' Rapid Amplification cDNA End): rat/bovine specific primer and Notl-d(T) 18 primerTotal RNAreverse transcription with bovine specific primer (B 1 )First strand cDNA poolHomopolymer (dA) tailing reactionPCRNotI - TT-TTT-->AA..AAAspecies specific primer (H2 or R2)Southern confirmation with core product1Klenow fill-in reactionBlunt-end cloning/sequencing85RAT COW PIG HUMAN10 11M 1 2 3 4 5 6 7 8 9RAT M 1 2 3COW 4 5 6PIG^HUMAN7 8 9 10 11-167bp-163 by-127bp86Fig. 8. Southern blot analysis of core-PCR amplified calbindin-D9k cDNA.Calbindin-D9k was amplified from human duodenum using the method describedin Fig. 7. Forty % of the amplified cDNA was analyzed on 6% PAGE, thentransferred to a nylon membrane for hybridization with a radio labelled rat cDNAprobe. The upper panel shows the ethidium bromide stained gel of the PCRamplified products; M: 123 by DNA molecular weight ladder, 1, 7: primercombinations with R2 and R3 (163 bp), 2, 4, 8, 10: primer combinations with R1and R3 (167 bp), 3, 5, 9, 11: primer combinations with R1 and B1 (127 bp), 6:negative control without cDNA. The species used for amplification is indicated.The lower panel shows the Southern blot. Note that the size of the hybridizationband is that predicted for the human calbindin-D 9k mRNA.167bp163bp127bp87Subsequently the 3'RACE technique was performed using an upstreamprimer R1, and a downstream primer, NotI-d(T)18, in first round PCR. Theproduct was analyzed by Southern hybridization with the human core product asprobe. A band of correct size could be detected by autoradiography but wasundetectable by ethidium bromide staining. A human specific primer H1 wassynthesized according to the sequence of the core product (Fig. 4). After a secondround of amplification with H1 and NotI-d(T) 18 primers, the product could begenerated in sufficient quantities to be visualized by ethidium bromide staining(Fig. 9).Subsequently, the 5'RACE technique was used to amplify a cDNA fragmentspanning part of the core product and the 5' end of the RNA. The 5'RACEamplification product was detectable by Southern analysis using the 3'RACE cloneas probe. The size of the re-amplification product using the human primer H2and NotI-d(T)18 was 232 nucleotides (Fig. 9) and contained a 74 base pair overlapwith the core product.Both, the 3' and 5' RACE products were cloned into pUC19 and sequenced.The combined sequence representing the full-length cDNA was 456 nucleotideslong (Fig. 10). Fig. 11A shows the schematic alignment of cDNAs with thesequence identities for various regions calculated. Fig. 11B shows the actualalignment generated by the GCG "PILEUP" program. The sequence was aligned88Fig. 9. 3'RACE and 5'RACE of the human calbindin-D 9k cDNA. First strandcDNA prepared from human duodenum was subjected to 30 cycles of PCRamplification as schematically described in the upper panel. The lower panelshows a 6% PAGE analysis of the products. The 100 by size markers areindicated at the left.5' ^  AA._.AAA3'R1(32)^167bp B1(198)vo( R1(32)^359bp^Notl-d(T)18 primer^-4(111(54)^336bp Not1-d(T)18 primerNotl-d(T)18 primer^299bp^81(198)Not1.-d(T)18 primer^232bp 112(126)} S .-End 3'RACE 5'RACE MWCore3'-End--336bp—232bp89Fig. 10. Nucleotide and derived amino acid sequence of the humancalbindin-D9k cDNA. The numbers on the left indicate the nucleotide positionsin the sequence. The amino acids are represented by the single letter code. A,alanine; C, cysteine; D, aspartic acid; E, glutamic acid; F, phenylalanine; G,glycine; H, histidine; I, isoleucine; K, lysine; L, Leucine; M, methionine; N,asparagine; P, proline; Q, glutamine; R, arginine; S, Serine; T, threonine; V,valine; W, tryptophan; Y, tyrosine.AAAAAACTCCTarrTGATTCTTCTAGCTG i i^ i CACTATTGGGCAACCAGACACCAGAATG^1 ^ +^ +^ + +^ +^ + 60I I I i 1^ i GAGGAGAAACTAAGAAGATCGACAAAGTGATAACCCGTTGGTCTGTGGTCTTACAGTACTAAAAAGTCTCCTGAGGAACTGAAGAGGATTTTTGAAAAATATGCAGCCAAAGAA^61 ^ +^ +^ +^ +^ +^ + 120TCATGAT1-1-1-2CAGAGGACTCCTTGACTTCTCCTAAAAACri-ria'ATACGTCGGTTTCTTS T K K S P^ RIFE  K Y A A K EGGTGATCCAGACCAGTTGTCAAAGGATGAACTGAAGCTATTGATTCAGGCTGAATTCCCC^121 ^ +^ +^ +^ +^ +^ + 180CCACTAGGTCTGGTCAACAGTTTCCTACTTGACTTCGATAACTAAGTCCGACTTAAGGGGG DPDQL SK DEL K LL I Q A E F PAGTTTACTCAAAGGTCCAAACACCCTAGATGATCTCTTTCAAGAACTGGACAAGAATGGA^181 ^ +^ +^ +^ +^ +^ + 240TCAAATGAGTri`CCAGG'rn:GTGGGATCTACTAGAGAAAGTTCTTGACCTGTTCTTACCTS L L K GPNTL DDL F QE L D K N GGATGGAGAAGTTAGTTTTGAAGAATTCCAAGTATTAGTAAAAAAGATATCCCAGTGAAGG^241 ^ +^ +^ +^ +^ +^ + 300CTACCTCTTCAATCAAAACTTCTTAAGGTTCATAATCATTria-rCTATAGGGTCACTTCCFEE F QVLVKK I S Q*AGAAAACAAAATAGAACCCTGAGCACTGGAGGAAGAGCGCCTGTGCTGTGGTCTTATCCT301 ^ +^ +^ +^ +^ +^ + 360Terri-EGTTTTATCTTGGGACTCGTGACCTCCTTCTCGCGGACACGACACCAGAATAGGAATGTGGAATCCCCCAAAGTCTCTGGTTTAATTCTTTGCAATTATAATAACCTGGCTGTGA361 ^ +^ +^ +^ +^ +^ + 420TACACCTTAGGGGGTTTCAGAGACCAAATTAAGAAACGTTAATATTATTGGACCGACACTGGTTCAGTTATTATTAATAAAGAAATTATTAGACAT p o lyA421 ^ +^ +^ +^ 456CCAAGTCAATAATAATTATTTCTTTAATAATCTGTA90Fig. 11. DNA sequence alignment of calbindin-D 9k mRNAs. (A) Schematiccomparison of the mRNA encoding the human, porcine, bovine, and rat calbindin-D9k. Sequences are aligned according to their coding regions and polyadenylationsites. Gaps in the 3' regions of bovine, porcine and rat sequences are indicated asnegative numbers. One inserted nucleotide is located in the 3' region of the ratmRNA(4-1). Note the 54 nucleotide gap in the 3' non-coding region of the ratmRNA. Nucleotide sequence identities to the human mRNA are indicated for thecoding and non-coding regions. (B) The nucleotide sequences of the bovine,porcine, human and rat calbindin-D 9k cDNAs are compared using the "PILEUP"program of the GCG sequence analysis system. The coding regions are indicatedin bold and the polyadenylation signals underlined.1 ID1 oly A1 87.1%^-2^-2•^68.7% • oly A' p1 90.4%-10^-6^H ^55.5% • 'poly A57.1%69.4%40.4%1 80.4% 1ihumanbovineporcinerat+1-54 I ^57.3%^•^poly A(A)5' Non-coding^Coding Region^ 3' Non-coding1^50 nt(B)BoySowHumRatBoySowHumRatBoySowHumRatBoySowHumRatBoySowHumRatBoySowHumRatBoySowHumRatBoySowHumRatBoySowHumRatBoySowHumRat^ACACA....AAAAAAGAGACCTCAC51CAAGAACCAACAGGACACCACA.GACACCACAGCACAGAA101TTTCGAAAAATTTTGAAAAATTTTGAAAAATTTTCAAAAA151AGGAGCTGAAAGGAGCTGAAATGAACTGAAAGGAGCTGAA201CCAAGCACCCCCGAGAACCCCCAAACACCCTCAAGTACTC251AGAAGTTAGTAGAAGTTAGTAGAAGTTAGTAGAAGTTAGC301GAAGGAAAGAGAAGGAAAAAGAAGGAGAAAGAAGCCAGAA351ATGTGGTCCTATGCGGCCCTCTGTGGTCTT401GACAATTGTAGGTGATTCTATGCAATTATATCCAATCCCA451ATCCTCAAACCTCCTTAGACATTATTAGACATTCTGAGACCTCCTATTTGCTCCTCTTTGCTGTTCCTGTAAATGAGTGCAAATGAGTGCGAATGAGTACAAATGAGCGCTATGCAGCCATATGCAGCCATATGCAGCCATATGCAGCCAGCTACTGCTTGCAACTGATTGCTATTGATTGCTGCTGATTTCGATGAGCTTAGATGACCTTAGATGATCTTAGACAATCTTTCGAAGAATTTTGAAGAATTTTGAAGAATTATGAAGAATAAAAAATATCCACGCA....8ACAAAATAGAGAAGGAGCTCA..CTCTGTAG..CTCTGTAATCCTATGTGATGATCCAACATGATCCAGCATAACCTGGCAAGATCTAGCATTCTTCCAGATTCTTCTAGCTGACTCTGGCAAAAAGTCTCCAAAAGTCTTAAAAAGTCTTAAGAAATCTAAGAAGGTGAAAGAAGGGGAAAGAAGGTGAAAGAAGGCGACAGACGGAATCAGGCTGAATCAGGCTGAATCAGTCAGAGTTTTTGAAGAACTTTCAGGAACTTTCAAGAACTTTAAAGAGTCCAGGTGTTTCCAGGTGTTTCCAAGTATTTCGAAGTTTTATTCTCAGTA....ACAGTCACCCTGAGCACGACA CAAAACCACCCAAAACCACCGAATCCCCCA ACTGAGGTCTAGTGAGGTCTTGTGAGGTTCTGTGAGAGCACTGCCGCACGCTGTTTCACTCAGCACTCACCCAGAAGAACCCTGCAGAACCCTGAGGAACCCCGAAGAAATCCAAACCAATCCAAACCAGTCCAGACCAGTCCAAACCAGTCCCCAGTTTTCCCCAGTTTTCCCCAGTTTTCCCCAGCCTCTAGACAAGACTGGACAAGACTGGACAAGACTGGATAAGAGGTGAAAAAGAGTGAAAAAGAGTAAAAAAGCTTCAAAAAGCTGGAAGAAGCCAYfTAAGGCTGGAGGAAG5AATGTCACCACATGTCTCTGAAGTCTCTGGCACCTACTGAAATTA..CTGAATT..6...AGTTATTATTAGATACTGTTACTGGACGACATTGGGCAACTGACAGCAAG100TGAAGGGCATTGAAGAGCATTGAAGAGGATTGAAGAGCAT150CTGTCCAAGGCTGTCGAAGGTTGTCAAAGGCTGTCCAAGG200GCTGAAGGGTACTGAAAGGTACTCAAAGGTCCTGAAGGCT250ATGGAGATGGATGGAGATGGATGGAGATGGACGGTGATGG300ATATCCCAGTATATCCCAGTATATCCCAGTTTATCACAAT350AGTGCCTGGGACTGGAAGGGAGCGCCTGTG4 400CT1AATTCTTCTGAATTCTCTTTAATTCTTTTGAATCCTA450AATAAAGCAAAATAAAG 1AAAATAAAG.AAAATAAAGCAA921^50TTC465ATG..ACGTCAT...93with the bovine (Kumar et al., 1989), porcine and rat (Darwish et al., 1987) cDNAsequences (Fig. 12). The cDNA contained a start codon at position 58 and apoly(A) tail preceded by the polyadenylation signal AATAAA. Sequence identitywas calculated for the coding, and the 5' and 3' non-coding regions. Comparisonof the human sequence revealed 87.1% identity to bovine, 90.4% to porcine, and80.4% to the rat cDNA coding regions. Sequence identities within the 5' and 3'non-coding regions were substantially lower.The human 3' non-coding region has 4, 16, or 53 additional nucleotidescompared to bovine, porcine and rat, respectively.The derived amino acid sequence of the human clone was compared to theother known mammalian calbindin-D9k sequences (Fig. 12) as well as searched forcalcium-binding sites by comparing it with several sequences of other members ofthe calcium-binding protein gene superfamily Calcium-binding proteins exhibita common structural motif of helix-loop-helix with the central calcium-binding loopcontaining six amino acids as calcium ligands. The human calbindin-D9k revealeda similar amino acid composition as found in the other species (rat, bovine andporcine).94Fig. 12. Interspecies comparison of calbindin-D 9k amino acid sequences.The complete human amino acid sequence is shown on the top. Matching residuesin the porcine, bovine, rat and murine sequences are indicated as a solid line. Thehelix-loop-helix motifs of the two calcium binding sites are marked above thesequence.HUMANPIGCOWRATMOUSEEmmEFEF^Limmr1 ^helix I ^1 loop I MSTKKSPEELKRIFEKYAAKEGDPDQLSimmANQinnimmiAmminims....4^immAgimmmummimMnimem.immAINEmminimAimiMmeamomiQ^^ 1 ^helix I ^1 1 ^helix II HUMAN^KDELKLLIQAEFPSLLKGPNTLDDLFQEPIG imE^Q RMIIIMMNMEMMEMOMMMI'COW^me Lime^ S^E^EmsRAT mie S SmSomommmEN^KimMOUSE^umE^ s,,......As„,,s.N^ims1 ^loop II ^1 ^helix II  1HUMAN^LDKNGDGEVSFEEFQVLVKKISQPIG mmommimmwmN ^COWRATMOUSE953.2 Cloning of the porcine calbindin-D9k cDNAFor cloning of the porcine calbindin-D 9k cDNA, core PCR amplification wasperformed with primers derived from rat (Darwish et al., 1987) and bovine(Kumar et al., 1989) cDNAs (Fig. 7). Based on sequence identity between rat andbovine, the two calcium binding domains were chosen to synthesize an upstreambovine (B1) and a downstream rat (R1) primer for amplification of a core product.A PCR product of the expected size (167 bp) was obtained. The hybridization ofthe Southern blotted product with a radio-labelled rat probe confirmed thespecificity of the porcine PCR product (Fig. 8). The porcine product was also usedas probe in a Northern blot experiment. Total RNA from rat, bovine, ovine, andporcine duodenum was probed and showed a strong hybridization signal withporcine and only weak signals with the other species (data not shown). The 167-by product was cloned and sequenced. The derived amino acid sequence of thisfragment was identical with the previously published porcine protein sequence(Hormann et al., 1979) with the exception of residue 25. This residue had beendetermined as aspartic acid by amino acid sequencing techniques. The cDNAderived amino acid for this position is asparagine.For 3'RACE, PCR amplification was performed using an upstream primeridentical to the rat RNA sequence (R1) and a nonspecific primer NotI-d(T) 18targeting the poly(A) tail in first-round amplification (Fig. 13). The correct-sized96Fig. 13. 3'RACE and 5'RACE of the porcine calbindin-D 9k cDNA. Firststrand cDNA prepared from porcine duodenum was subjected to 30 cycles of PCRamplification. The upper panel shows the location of the primers. The lowerpanel represents the ethidium bromide stained PAGE of the amplified DNA. The100 by size markers are indicated at the right.5' ^ AA-AAA 3'R1(32)^167 by B1498)R1(32))10■^359 by Notl-d(Iii18 primerCore3' EndNot1-d(T)18 primer^299 by^B1(198)Notl-d(T)18 primer^294 by^R2(193) 5' EndMW 3 ' RACE 5 ' RACE 97product (359 bp) was detected by Southern analysis, with the core product asprobe. The product was purified by gel electrophoresis and MERMAID kit andsubjected to a second round of amplification using the same primers. A secondround of amplification was necessary to produce enough DNA for blunt endcloning. After the second-round amplification, the product was cloned into theSmal site of pUC19. Subsequently, the 5'RACE technique was used to amplify acDNA fragment spanning part of the core product and the 5' end of the RNA (Fig.13). cDNA was synthesized by a specific primer (B1) complementary to the bovinecalbindin-D9k mRNA sequence. After adding a homopolymer dA tail to the 3' endof the cDNA strand, the PCR was carried out with R2 primer, complementary tothe rat RNA and a nonspecific primer NotI-d(T) 18 in first-round amplification.After 30 cycles of amplification, the PCR product (294 bp) was visualized byethidium bromide staining, purified, and cloned. The sequences of the two RACEproducts were determined. The 5' and 3'RACE products overlapped with the corefragment and were combined representing the full-length cDNA sequence.Fig. 14 shows the nucleotide and derived amino acid sequence of theporcine calbindin-D9k. Total length of the cDNA excluding the poly(A) tail was 443nucleotides. The cDNA contained a start codon at position 58, a terminationsignal at 295, and a poly(A) tail preceded by the polyadenylation signal AATAAA.98Fig. 14. Nucleotide sequence of the porcine calbindin-D 9k cDNA. Thecombined nucleotide sequence of the 3'RACE and 5'RACE clones are shown withthe position numbers on each site. The coding region is shown in bold and thepolyadenylation signal is underlined. The derived amino acid sequence is shownunderneath in the single letter code. The sequence of porcine calbindin-D 9k cDNAhas been deposited in the EMBL/GenBank data base (accession no. L13068)ACACACTCCTATTTGATTCTTCCAGCTGCCGCACGACTGGACGACCAGGACACCAAAATG^1 ^ +^ +^ +^ +^ +^ + 60TGTGTGAGGATAAACTAAGAAGGTCGACGGCGTGCTGACCTGCTGGTCCTGTGG .ria-fACN -AGTGCCCAAAAGTCTCCTGCAGAACTGAAGAGCATTTTTGAAAAATATGCAGCCAAAGAA^61 ^ +^ +^ +^ +^ +^ + 120TCACGGGrri-rCAGAGGACGTCTTGACTTCTCGTAAAAACI -1-1-rrATACGTCGGTTTCTTS A Q K SPAELK S IF E K Y A AK EGGGGATCCAAACCAGCTGTCGAAGGAGGAGCTGAAGCAACTGATTCAGGCTGAATTCCCC^121 ^ +^ +^ +^ +^ +^ + 180CCCCTAGGTTTGGTCGACAGCTTCCTCCTCGACTTCGTTGACTAAGTCCGACTTAAGGGGG DPNQL SKEELK Q L I Q A E F PAGTTTACTGAAAGGTCCGAGAACCCTAGATGACCTCTTTCAGGAACTGGACAAGAATGGA^181 ^ +^ +^ +^ +^ +^ + 240TCAAATGACTTTCCAGGCTCTTGGGATCTACTGGAGAAAGTCCTTGACCTGTTCTTACCTS L^ GL K PR TLDDLF QELDK N G -GATGGAGAAGTTAGTTTTGAAGAATTCCAGGTGTTAGTGAAAAAGATATCCCAGTGAAGG^241 ^ +^ +^ +^ +^ +^ + 300CTACCTCTTCAATCAAAACTTCTTAAGGTCCACAATCACTI -1-1-ECTATAGGGTCAuri CC•D GE^ FEEEF QVLVKK IS Q*AAAAACACGCAACAGTCCCATTTAAGGACTGGAAGGGATGCGGCCCTGCTCTGTACAAAA301 ^ +^ +^ +^ +^ +^ + 360TTTTTGTGCGTTGTCAGGGTAAATTCCTGACCTTCCCTACGCCGGGACGAGACATG1 -1-1-1.CCACCCATGTCTCTGCTGAATTCTCGGTGATTCTAATGATCCAGCAGTGAGGTCTAATTA361 ^ +^ +^ +^ +^ +^+ 420GGTGGGTACAGAGACGACTTAAGAGCCACTAAGATTACTAGGTCGTCACTCCAGATTAATATAAAGAACTCCTTAGACACGTC421  ^ 443TATTTCTTGAGGAATCTGTGCAG99In Fig.11, the comparison of the cDNA sequences of porcine, bovine, and ratcalbindin-D9ks is illustrated. The sequence identity was calculated for the codingand the 5' and 3' non-coding regions. The porcine sequence revealed 81.3%identity with rat, and 90.0% with bovine cDNA within the coding region.Sequence identities within the 5' and 3' non-coding regions were substantiallylower. The 3' non-coding region of porcine cDNA was similar in length to thebovine and human cDNAs in that they all have an additional 40-50 nucleotideswhen compared to the rat sequence. The porcine sequence was published in 1992(Jeung et al., 1992b).1003.3 Expression of human calbindin-D9k mRNANorthern analysis of human RNA was performed using the 3'RACE cloneas probe. With 20 jig total RNA a weak signal was detectable in duodenum, butnot in placenta, uterus, kidney, pectoral muscle, or brain (data not shown). Fig.15 shows the autoradiograph of a Northern blot using 10 jag of humanpoly(A)+RNA from a 2.5-year-old child and a 32-year-old adult as well as 20 jigtotal RNA from rat duodenum. The human probe hybridized specifically to a RNAof about 600 nucleotides in human RNA. The level of calbindin-D9k wasapproximately 5-fold higher in the child duodenum than in the adult tissue.When an identical blot was probed with a rat calbindin-D9k cDNA, a strong signalwas detected only in the rat RNA (Fig. 15). The human probe also hybridizedintensely to porcine RNA (data not shown).After no expression of calbindin-D 9k was found by northern blot analysis inuterus and placenta using both total and poly adenylated, RNA reversetranscriptase/PCR and S1-nuclease analysis was performed. Several combinationsof primers, derived from the human mRNA sequence were used. A positive signalwas found only in duodenum (Fig. 16). In uterus and placenta (first and secondtrimester and term) no signal was detected. S1 nuclease analysis was performedwith duodenal (positive control) and term placental RNA. No protected fragmentwas found in term placenta using the 3'RACE clone as probe (data not shown).101Fig. 15. Northern blot analysis of human and rat intestinal RNA. 10 pg ofpoly(A)±RNA from a 2.5-year-old child (1, 4) and 32-year-old adult (2, 5), and 20jug of total RNA from rat intestine (3, 6) were analyzed. The blot was hybridizedwith the isolated human (A) or with a rat (B) cDNA probe.A  28S-18S-CaBP-5102Fig. 16. PCR detection of human and porcine calbindin-D 9k mRNAexpression.(A) RT/PCR was performed with H5 and H4 primers to analyze expression ofhuman calbindin-D 9k. Ten % of the PCR reaction was analyzed on 6% PAGE andstained with ethidium bromide. The gel shows the PCR products (324 bp) asindicated alongside with size markers. The expected products were obtained fromduodenum (2) but not uterus (1), 1st trimester (3), 2nd trimester (4), and termplacenta (5).(B) The amplified products of the correct size (129 bp) were obtained from porcineileum (1), kidney (2), uterus; mature (3) and immature (4), duodenum (6), jejunum(7). Liver (5) and lung (8) showed no expression.(A)^ (B)1 2 3 M 45^1 2 3 4 5 6 7 8 M-324 by -129 by1033.4 Expression of porcine calbindin-D9k mRNANorthern analysis of porcine RNA was performed using both 5' and 3' RACEclones as probes. When 20 jug total RNA was analyzed, a strong signal wasdetected in intestine (duodenum > jejunum > ileum) after overnight exposure (Fig.17); but not in lung, liver, spleen, pancreas, aorta, or brain. Very low expressionof calbindin-D9k mRNA was detected in the immature uterus (from a local porkfarm) and kidney after 7-day exposure (Fig. 17). Uterus from a mature pig (froma local slaughter house) revealed a substantial higher calbindin-D 9k mRNA levelthan the immature tissue (weanling piglet). The autoradiographs revealed asingle band of approximately 600 nucleotides. When porcine RNA was analyzedalong with rat, bovine, and ovine RNA, all species gave a signal of similar size(data not shown).Reverse transcriptase/PCR analysis was also performed with porcine RNA.Expression of calbindin-D9k was detected in duodenum, jejunum, ileum, kidney,uterus (mature and immature). Liver and lung tissue did not show anyexpression (Fig. 16). These results were in agreement with the northern blot data.104Fig. 17. Northern analysis of porcine RNA. Twenty lag of total RNA fromduodenum (1), jejunum (2), ileum (3), kidney (4), and uterus from mature pig (5)as well as uterus from immature pig (6) were analyzed. The blot was hybridizedwith the isolated porcine cDNA probe and exposed overnight as shown in panelA or for 7 days (panel B). Ethidium bromide-stained ribosomal bands are shownin the lower panel.(1) A B 1^2^3^4^5^628S-18S-CaBP-(2) A B1 2 3^4 5 628S18S —„Jr" Iwo1053.5 Identification of the transcription initiation site of the human andporcine calbindin-D9k.The transcription start site of the human calbindin-D 9k gene wasdetermined by primer extension analysis. For RNA-dependent extension, a 24-mer synthetic oligonucleotide with the anti-sense sequence was used as theprimer. When this product was analyzed next to a set of dideoxy chaintermination reactions performed using the M13 vector as molecular weight ladder,the site of transcription initiation could be located. The prominent reversetranscription product was mapped to 103 by from the primer on the human cDNAsequence (Fig. 18). The same extension product was observed when human totalRNA from adult duodenum or from the 2.5 year old child was used in thisexperiment. The 5'-end of the product corresponded to the marked C* residue inthe sequencing ladder. In addition no product was detected in a controlexperiment where total RNA from human spleen was substituted for duodenum.For the porcine calbindin-D 9k gene, the transcription initiation site was alsodetermined by primer extension analysis using the same primer as shown inFig.18 (The sequences in this region of human and porcine calbindin-D 9k mRNAare identical). The prominent single product corresponded to the C* residue asin the human primer extension experiment (Fig. 18). A negative control RNAfrom porcine skeletal muscle gave no extension product. These results confirmed106that the transcription of the human and porcine calbindin-D9k genes initiate at asingle distinct site which size corresponded to the predicted positions from the5'RACE clones.107Fig. 18. Determination of the transcription initiation site of the humanand porcine calbindin-D9k.(A) Schematic illustration of the primers used for primer extension. In the center,the mRNA and primer are indicated. On the side the obtained and on the rightthe predicted extension product is shown.(B) Autoradiograph of the human primer extension experiment. A sequencingreactions from M13 vector DNA was run in parallel with the primer extension.The arrow indicates the transcription initiation site. On the right, the M13sequence used as size marker is shown with the residue marked (*) correspondingto the size of the extension product. For the reactions, 20 pg of total RNA fromhuman duodenum (child (1) and adult (2)) and spleen (3) were extended with a 3213end labelled oligonucleotide corresponding to bases 103 to 123 of the humancalbindin-D9k cDNA.(C) Autoradiograph of porcine primer extension. Duodenum (1), skeletal muscle(2).(A)-r Cap siteproduct 103 nt 102 ntprimertMOM(C)AC G T123' 5'T AT AG CT AG CA TG CC* GG CG CA TT AA TA TC G5' 3'108Fig. 18(cont). Determination of the transcription initiation site of thehuman and porcine calbindin-D 9k.(B)^AC G T1233' 5'j* T AT AG CT AG C44^A TG CC* G*A. G CA TA TC GA TT AG C5' 3'1093.6 Genomic Southern blot analysis.Southern blot analysis was undertaken to detect the human calbindin-D 9kgene in genomic DNA and to determine whether the restriction sites are conservedwhen compared to the rat gene. Under high stringency conditions the DNAprepared from human placental tissue was digested with EcoRI, BamHI, HindIIIrestriction enzymes, size fractionated, and hybridized with a riboprobe synthesizedfrom the human 3'RACE clone. As shown in Fig. 19, the human calbindin-D 9kgene is located on a 13 kb HindIII and 6.5 kb EcoRI fragment. No band wasdetected in the BamHI digest.In order to locate the 5' part of the human calbindin-D 9k gene in a genomicSouthern, hybridization with a genomic DNA fragment containing parts of exonsI and II and intron A was performed. This fragment was obtained by genomicPCR amplification which is described in detail in chapter 3.8. Using this proberesulted in intense background hybridization obscuring any specific fragments onthe autoradiograph (data not shown). This could subsequently be explained by thepresence of an Alu type repetitive sequence in intron A, commonly found in thehuman genome.110Fig. 19. Southern blot analysis with the human calbindin-D 9k 3'RACEriboprobe. Human genomic DNA (10 pg) was digested with EcoRI, BamHI, andHindIII, respectively. DNA was size-fractionated by electrophoresis on 0.8%agarose, and transferred onto nylon membrane. The filter was probed under highstringency conditions with the labelled human calbindin-D9k 3'RACE clone.EcoRI BamHI HindIII12 kb7 kb6 kb1113.7 Chromosome assignmentBIOSMAP Somatic Cell Hybrid DNAs were used to locate the calbindin-D 9kgene to a unique human chromosome by a given set of amplification primers.BIOSMAP Somatic Cell Hybrid PCRable DNAs include DNA from a panel of 25human-hamster somatic cell hybrids and control DNAs from human and hamstercell lines. Assignment of a target sequence to a human chromosome wasaccomplished by comparing human specific amplified products in an ethidiumbromide stained gel with the listed chromosomal content of the hybrid cell panel.PCR involving specific primers for the human calbindin-D 9k gene produced thecorresponding DNA fragments with the control human cell line and three hybridcell lines (9, 16, and 20), but no product was obtained with a hamster cell line or22 other hybrid cell lines. Comparison with the provided panel of humanchromosomal components of each cell line identified chromosome X as carrier ofthe calbindin-D9k gene. Fig. 20 shows the results of agarose gel electrophoresisof the PCR products. The bands corresponding to a 525 by product are consistentwith the calculated size from the nucleotide sequence.112Fig. 20. Chromosome assignment of the human calbindin-D 9k gene. Fiftyng hamster/human somatic cell hybrid DNA was subjected to 30 cycles of PCRusing primers corresponding to positions 234-254 and 738-758 of the genomicsequence. The PCR products were electrophoresed on agarose and stained withethidium bromide. The left lane indicates molecular weight standards. Thepredicted size of the product is 524 bp. Lane 1, hamster DNA; lane 2, onerepresentative hybrid (SM212, #1) of 22 without product formation; lane 3, humanDNA; lane 4-6, hybrids SM803 (#9), SM909 (#16), SM968 (#20)(commonchromosome X). The human chromosome content of each cell line (#1-25) isshown in Table 2. MW 1 2 3 4 5 61.0 kb516/506bp —394bp-.61•111..—524113Table 2. Chromosome contents of a human x hamster hybrid andassignment of the human calbindin-D 9k gene. The gene content of eachhybrid was determined by PCR. The presence or absence of the human gene isindicated by +, -, respectively. For the human chromosomes, the symbols mean:+, chromosome present in at least 70%;(+), chromosome present in 40-60%;(-),chromosome present in 5-30%; d, multiple deletions in5q; D, deleted at 5g15.1-5g15.2; -, chromosome absent.Hybrids"*"--.........................Cal bindin-D9 kHuman chromosome1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 X Y(1) - d - - - -(2) - - .(3) - - - - - - - - - -(4) - - - + - (+) (-)(5) - - - - (+) - + + - + + -(6) - - + - - -(7) - - - D - -(8) - - - D - - - (+) + -(9) + - - (-) - - - - - - -(10) - - - - - - + -(11) + - - - -(12) (-) - (-) - - (+) -(13) - - - - - - - -(14) (+) . - - - - -(15) D - - (-) - -(16) - - D -(17) - - - - - . .(18) - - - -(19) - - - + - - - - - -(20) * - - - - + - - - - -(21) - - - + - -(22) (+) -(23) - + - - + . -(24) - - - (+) - (-) . . . .(25) - - D - _ -1153.8 Cloning of the human calbindin-D9k intron A regionTo clone a genomic fragment of the human gene for calbindin-D 9k, the PCRtechnique was employed. Genomic DNA was used as substrate for amplificationwith a pair of primers (H5 from exonl and H2 from exonII, assuming a similargenomic structure as in the rat gene). The DNA was size fractionated on agaroseand analyzed by Southern blot hybridization (Fig. 21). The blot was hybridizedwith an oligonucleotide which lies in between the primers H5 and H2 in the cDNAsequence. A single prominent band hybridized to the probe. Size-selected DNAfrom the gel was cloned into the pUC19 plasmid and subjected to sequenceanalysis. The sequence showed the predicted overlaps with the human cDNAsequence at both ends, indicating exon I and II separated by intron A. Thecomplete sequence of this clone is presented in Fig. 22. The primers used foramplification are underlined. Within this region of the human calbindin-D 9k genethere is a single 769 by intron.516/506 by1.0 kb1.0 kb516/506 by394 by116Fig. 21. Southern blot analysis of human and porcine genomic PCRproducts. Genomic DNA (100 and 200 ng) was boiled in PCR buffer andsubjected to 30 cycles of PCR as described in Experimental Procedures. FollowingPCR, 10% of each sample was analyzed on a 1.0% agarose gel, and then stainedwith ethidium bromide. PCR samples were loaded below as follows: M; molecularmarker; negative control without DNA (1); amplified products from 100 ng (2) and200 ng (3) genomic DNA. DNA was transferred to nylon membrane and thenhybridized with a 32P end labelled oligonucleotide located inside the PCR primerpair. (A) human PCR (B) porcine PCR(A)M 1 2 3^M 1 2 31.0 kb-516/506 by(B)1 2 3 M^1 2 3 M117Fig. 22. Nucleotide sequence of the human calbindin-D 9k intron A region.Exonic sequences are bold, PCR primers are underlined. The boxed sequencerepresents a segment with high identity to an estrogen response element (ERE)located at the corresponding region of the rat calbindin-D 9k gene.1 CTCCTCTTTG ATTCTTCTAG CTGTTTCACT ATTGGGCAAC GTTAGTG 51 TGATirTGGT ACTACCTAGA101 CCrTTTAGAA TGGTGTGAAA151 GCCATCTGGA GATATTCCCT201 TATTCTTCAA GACCACAGCC251 GTAGAATGTC ATTAATTACT301 CCTAACAGAT GTACTCTAAG351 CTTTTTTTTT TTTTTTTTTT401 CTGGAGTGCA GTGGCATCGC451 AGCGATTCTC CTGCCTCGGT501 CCACCACGCC TGGCTAATTT551 CCAGGCTTCC ATGGTCTCGA601 GGCCTCCCAA AGTGCTGCTG651 CTAAGATGCT TATTTTTTAA701 CCCCGACCAT TCTCrr12CC751 GCCCAATGGC ACTAAAATCC801 TCAGCliiii GGTAAGTTTAGCTCCTTCTG TTTGCATTAA TACTCTGCTTGATAATGGTC TGGAGAGAAG GTGAGiliATGTAACTTAGA AGCTAGAAGC TGTGGTCAAGTGCAGAGTGC AGATCAGGTA CATAAAAAGTAATAAAAATG TTTGGCTCCA GGGACCATCTAACAGACATT TAACATACTA TCTAAGATGTTTGCAGCGGA GGCTCGCTCT GTCACCCAGGACTCACTGCA ACCTCCGCCT CCCGGGTTCACTTCCAAGTA GATGGGACTA TAGCGACGCATTGTATAGAG ACAGGGTCTC ACCATGTTGGACTCCTGACT TCAAGTGATC CACCCTCCTTGGATTACAGG CATGAGCTAC CGTGCCTGGCACATGTTAGT AGTTACTGGT GCCATTCCGATTATTCTTCC CCATTTATGC CAAGGAAATGTGCAGACAAC CCTCAGCACA AAACTGTACTIIIICTTCCC AATTTATAAG ACACCAGAATAGGAACTGAA GAGGATTTTT GAAAAATATGGACCAGTTGT CAAAGGATGA ACTGAAGCTA851 GAGTACTAAA AAGTCTCCTG901 CAGCCAAAGA AGGTGATCCA951 TTGATTCAGG CTGAATTCCC CAG1183.9 Cloning of the porcine calbindin-D9k intron A region.Genomic DNA was prepared from porcine muscle using a previouslydescribed protocol for genomic Southern blot analysis. For amplification of theporcine calbindin-D9k, PCR (30 cycles) was carried out with P5 and H2 primers.A Southern blot of the PCR product was hybridized with an oligonucleotide probeinside the amplification primers. Fig. 21 shows the DNA fragment amplified fromporcine genomic DNA. This amplified porcine DNA was cloned into the Smal siteof pUC19 vector, and sequenced. The sequences at the 5' and 3' ends of this clonerevealed the expected segments from the porcine calbindin-D9k cDNA, representingexons I and II, separated by intron A (1,061 bp). The complete sequence is shownin Fig. 23. The genomic sequence was compared to the GenBank/EMBL database.A number of porcine genes scored with high sequence identity due to a shortinterspersed repeat. This repeat is commonly found in genomic and 3' non codingmRNA sequences in the pig. Highest scores were found in the genes for FSH-0subunit, the sarcoplasmic/ endoplasmic-reticulum calcium pump; the inhibin [3AmRNA 3' non-coding region also carries this repeat. Interestingly, when theintron A regions of the porcine and human calbindin-D 9k genes were compared twosegments of relative high sequence identity were detected. The first segmentcovers the 5' end of the intron and stretches over 356 by with 80.3% sequenceidentity. The other segment is located at the 3' part and scores 71.0 % over 152bp. The intron sequence was also searched for a motif resembling the ERE found119at +51 in the intron A region of the rat calbindin-D 9k gene (Krisinger et al., 1988).rat calbindin-D9k ERE +51:^GGTCA GGG TGATCporcine calbindin-D9k^GGTCA GTG GGATTThis sequence was subjected to gel retardation analysis along with several othersequences of interest as described in 3.12.120Fig. 23. Nucleotide sequence of the porcine calbindin-D 9k intron A region.Exonic sequences are bold, PCR primers are underlined. The boxed sequencerepresents a segment with high identity to an estrogen response element (ERE)located at the corresponding region of the rat calbindin-D 9k gene. A partialsequence of porcine calbindin-D 9k gene has been deposited in the EMBL/GenBankdata base (accession no. L13067)1 ATTCCTATTT GATTCTTCCA GCTGCCGCAC GACTGGACGA CCAIGGTAAGT51 GGGATTTGAG TGCTACCCGC ACCTCCTCCA GCTGCCTTTC AGAACGGCGG101 GAAAGAACAA AGTGGGTTTG TGCCCATCTA151 AGAGGCTAGA AGCTGTGGTC GAGTTTCTTT201 GCAGAGCAGT TACATGAAAA GTGAAGAATG251 GTTTGGCGCC AGAGACTCTC TTCTAACAGA301 TGAATGTACA ATCTAAGATG ATTCTTCTAT351 TTTGTCiiii TACAGCTGTA CCTGGGACAT401 GTCGAATCTG AGTTGCCAGC TGCCAGCTTA451 CCGAGCCGAG TCTGCCATCT ACACCCCAGC501 ACCCACTGAG CGAGGCCCAG GGATCGAACC551 CAGGTTATTA ACCTGCTTGA GCCACAAAAG601 TTTTTTTTTA AAAGGTTGTT AGCAGGAGTT651 ATCAACCCGA CCAGTATCAG CAAGAGGGGG701 TCAGTAGGTT AAGGAATCTG CATGGCCTTG751 GACTCAGCTC GGGATCCAGC GTGGATGTGG801 TCTGATTCAA CCCCTAGCCC GGGAACIIVC851 TTAAAAAGGA CTGGAAAAAA AAAAAAAAAA901 CTGGTGCATT GTCGTTCCCA CAGTTCCCCG951 TTTATCCTCA CCCTCCCCTT TTCTTTATTC1001 ACGGCCCCCC CCCCCCACGC CAACGCACTC1051 AGTATAAAAC CCCACTTCGG CiiiiiGGTA1101 TACAGGACAC CAAAATGAGT GCCCAAAAGT1151 ATTTTTGAAA AATATGCAGC CAAAGAAGGG1201 GGAGGACGTG AAGCAACTGA TTCAGTCTGA 121AAAACGTCAC CGGTAACTTCAAGACCACAG CCTACAGAGTTCCAAGTTAC TAGTAAAAGTTGTACTCGAA GAACAGACACATTAAATATA TGTATArrriATGAAGTTCC CAGGCTAGGGCACCACAGGC CACGTTGGATTCAAGGCCAC GCGACCCTTAGAATCCTCAT GGATACTAGTGAAACTCTAA GAIIiiiiiiCCCACTGTGG CTCGGTGGTAGCTTCATTCC CTGGCCTCACAGCTGTGGTT GTAGATGGCATGTAGCGCGG CGGCTACAGCCATATGCCAT GGATGCAGCCAAAAAAGGTG TGAGCAGTCACCCCGTCTCT CCTTCTCCTCTCTCTCACTG TGCCAGGGAAACCTCCCGCT GGCAATGATAAGTTCATGTC TTTCTCCATTCTCCTGCAGA ACTGAAGAGCGATCCAAACC AGCTGTCGAAATT CCCAG1223.10 Isolation of human calbindin-D 9k geneInitially leucocyte and placental genomic libraries were screened with thehuman cDNA probe using a random primed labelling procedure. These attemptsfailed to detect any positive clones. Since the human intron A clone representsa fragment of genomic DNA, it could have been used to screen a human genomicDNA library. However, although this probe was generated specifically from thehuman calbindin-D9k gene, many cross-hybridizing genomic DNA clones were alsodetected. As mentioned above, this was caused by the presence of a commonrepetitive sequence within intron A. (This was found after initial screeningexperiments.) As an alternative, the above mentioned libraries were screenedwith a riboprobe synthesized from the 3'RACE clone, which was able to detect asingle copy of the calbindin-D 9k gene on a genomic Southern blot. However, thisscreening yielded no positive clones.After the calbindin-D9k gene had been localized to the X-chromosome achromosome specific library was screened with the 3'RACE riboprobe. The librarycontained DNA that was partially digested with Sau3AI to give an average sizeof 10 - 22 kb. Twelve clones were positive in the first round of screening, 3recombinant clones were isolated after 2 successive cycles of screening at lowdensity. X-genomic DNA was prepared and digested with a number of restrictionenzymes. A Southern blot of the positive clones was hybridized with the human123intron A clone targeting the 5' end as well as the 3'RACE clone locating the 3' endof the gene. Each of the clones was shown to be identical by restriction analysisand Southern blot. This clone spanned a region of almost 12 kb of genomic DNA,with the human calbindin-D9k gene sequence lying approximately in the middle.The restriction map reveals 5 fragments for EcoRI digestion. Two EcoRIfragments hybridized with the 3'RACE cDNA (5.6 kb) and intron A (2.1 kb),respectively. This indicated that the 5.6 kb EcoRI fragment corresponds to the3'part of the human calbindin-D 9k gene and the 2.0 kb EcoRI fragment to the 5'part of the gene (Fig. 24). Each of the EcoRI fragments was subcloned into pUC19and sequenced by the Sanger method. The structure of the human calbindin-D9kgene and the flanking regions are schematically diagrammed in Fig. 25. The geneis contained within 4.5 kb after comparison to the cDNA sequence and was foundto consist of 3 exons flanked by 2 introns with the exon-intron boundary beinglocated at the linker between the two calcium binding domains. Exon I (49 bp)contains only 5' non-coding region of the mRNA, while the coding sequence of themRNA is distributed between exons II (145) and III (264). This overall structureis identical to the rat calbindin-D 9k gene. The data are presented in Fig. 25 asa schematic map of the genomic clone including the sequencing strategy andEcoRI restriction sites.- 5.6 kb- 2.4 kb- 2.0 kb- 1.4 kb- 1.0 kb124Fig. 24. Southern blot analysis of the human calbindin-D 9k clone. DNA ofthe genomic clone was digested by either EcoRI or EcoRI/HindlII enzymes,electrophoresed on a 0.8% agarose gel, transferred onto nylon membrane andhybridized to either the human intron A or the 3'RACE clone. (1) Ethidiumbromide stained gel. (2) Autoradiograph after hybridization.(1) M HinEco HinEco(2) M intron A^3'RACE- 5.6 kb- 2.4 kb20 2515100^5Ei5'^I^Fig. 25. Organization and sequencing strategy of the human calbindin-D 9k gene. The EcoRI sites areindicated as El. Exons are indicated as open boxes numbered I-III. The arrows indicate the priming sites for thesequencing reactions.30^35^40^45 Kbkm,v mingEi ^t 13'•■•••••.Eii III Insert -403-i■-■-+:1-0-041.-► 040.-•••11-°^-.41•Dfb51,0665(24.6605t.^ ,..teataeao-awnrsarre =MN2^3^4^5^6^7^8^9^10^11 Kb1263.11 Analysis of the human calbindin-D9k gene sequenceA computer software system (Genetics Computer Group, Inc.) was used toanalyze the genomic sequences. The human calbindin-D 9k gene sequence wascompared to the entire data base of GenBank and EMBL (European MolecularBiology Laboratory) using the FASTA program. Sequence identities between thehuman and rat/porcine calbindin-D9k genes were searched by the BESTFITprogram. The promoter regions of the human and rat genes were mapped withthe "tfsites" database, which contains 2,036 recognition sites for transcriptionfactors. This program detects only perfect matches to the filed patterns. In orderto find sequences closely related to a pattern, a BESTFIT was performed with: allknown steroid hormone responsive elements and the various forms of the vitaminD response elements.The putative transcription factor binding sites of the promoter region areshown in Fig. 26. A schematic alignment of the human, rat and pig genomicsequences with the identical regions as well as prominent repeats are illustratedin Fig. 27. The entire DNA sequence of the human calbindin-D 9k gene is shownin Fig. 28.When the calbindin-D9k gene was compared to known sequences of the databases it was noted that there are four regions of high identity. These regions were127Fig. 26. Organization of the human calbindin-D 9k promoter region.Negative numbering starts at the transcription initiation site. Sequences oftranscription factor binding sites are shown and the binding factor named. Avitamin D responsive elements like (DRE-like) repeat is marked at -1,213.-1324-1204pit-iDRE-like :AGGTGATATAAGGCA:--WAP-1084 cGAATGTCTACTTATTCA4 TTTAA.--964EBP-844 L'TCTGAAGC^WAP-724 'TTAAA:^-604WAP-484 TAAA^AP-2 ^CCCAAAGTGC,CAT WAP-364 ALGAAA^ TTTAA4G^-244EBPCTTGe^ciEBP CAT CAT-124 ---irCTTAAGC1^ AA'ikj GAR---TATAERE-like elementf.CTATAAA4^engrailed-4+1 EXON I^(50-nt) 6qTTAGTINTRON AGTGATTAluk—r"-(===1 Owl*80.3%)^(11.099I II Alu^ III^Alupighuman(71.9%)IratC65.5%) (71.89'I^IFig. 27. Schematic comparison of the genomic sequences of porcine, human and rat calbindin-D 9k gene.The upper line provides a size scale. Repetitive sequences are marked as open boxes. Homologies for the intronicand flanking regions are indicated in % within the dotted lines. The exons were used to align the three genes andgaps introduced for optimal alignment.0^625^1250^1875^2500^3125^3750^4375^5000^5625^6324129Fig. 28. Nucleotide sequence of the human calbindin-D 9k gene. The TATAsequence is marked as an open box. Exonic regions are underlined. The aminoacid sequence is indicated in the single letter code. The broken underlinesequence denotes a consensus found near 3' termination sites. The humancalbindin-D9k gene sequence has been deposited in the EMBL/GenBank data base(accession no. L13042)130-1324 GAATTCTATT AGATTTCCCA TTTGTTCTCC AATGTAAATA AGGCCCTAGA-1274 ATTTTAGAGA TGAGAAACAT CAGGGAGAAA TACAGTCCCA ATAGTCCAGT-1224 CCTCAGTTTG TAGGTGATAT AAGGCAGAGT CAGGGAAAAT TCTATAAGTG-1174 TGGAAGTGTA CAAGTGTTTC AAAGCTGTAA AGTGCTCTTA AAATCACCTG-1124 ACCCCTTCTG TGAGTTACAT CAGCATTAGA TCATGTCCTT GGAATGTCTA-1074 CTTATTCATG GAACTCTTCA GTAGAATCAA TTTCCTTGGC AATTTTAAAT-1024 CATGGCAGTT ACAAAAAGTT TGAAGTATTT CAACTGAACA AGAGTAAGTC-974 GAGTGTTCTC ACAAGTTTTC GAATGATCAC GATATATATG GTAAGGAAAA-924 GGATAATGAG AAATAAAAAG CAGACACAGG ACCAACAAAT AATGAAAGAG-874 CTGAAAAAAT GACCATGTTA CTCAATATTG ATTTGCCAAA TAAAACCACA-824 AATTACTGTG TATCTTCTTT CCACTTCTTA AACCTGCTTC TGAAGCAATA-774 CTTCCATTCA TTCATTCATG TATTCAAATA TTTAAGCTGC TAATATGTAA-724 AGACATCAGA AGGACTTCAA TGGCCATACC GTAATTTTTT AAAACTATGT-674 AATATTGTCC TTCTATGACA GC/1'111111C TTTTTTACAG ACAGGATCTC-624 ACTCTGTCAC CCAGGCTGGA GTGCAGTGGC ACAATCATGG CTCACTGTTG-574 CCTCAACCTC CCAGGGTAAG GAATCCTCCC ACCTCAGCCT CTACAGTAGC-524 TGGGACTACA GGGATGTACC ATCACACCTG GCTGATTTTT TAAArrrriC-474 AGTAGAAATG TCTTGCTGTG TTGACCAGGC TGGTATCAAA CTCCTCACCT-424 CAAGCAATCC TCCTCCTGGC CTCCCAAAGT GCTGGATTAC AGGCATGTTA-374 CCGCACAGGG CAGCATTTAT iiiliCTGAT ATGAAGAAAG TTAiiiiiii-324 AAAGCCTAAT ATGCAACTAA ATTGAAAAGA TCCTCTATAT CTAAGCTTTA-274 TAAATGTGAA CTGTGAGCAA ATATAATTCT ACATTGGAGG GAAAATATAA-224 TCTTCAGGGG GGAAAAAAAT CCAAGCTTAC GACTCTGAAC CATGAAGCAA-174 CTGTCGAGAT-124 CAGTCTTAAG-74 CAACAAAAGG-24 iaTGTCAT27 GCTGTTTCAC77 AGCTCCTTCT127 AGATAATGGT177 TGTAACTTAG227 CTGCAGAGTG277 TAATAAAAAT327 GAACAGACAT377 TTTGCAGCGG427 CACTCACTGC477 TCTTCCAAGT527 TTTGTATAGA577 AACTCCTGAC627 GGGATTACAG677 AACATGTTAG727 CTTATTCTTC777 CTGCAGACAA827 ATTTTCTTCC8779279771027GAGGAACTGAE E LAGACCAGTTGD Q LCCAGTTTACTPSLLGATGGTGGGAGGACCTTTACCCTGAGTTTCAAATTTTCATAATTCTGCTCTATTGGGCAAGTTTGCATTACTGGAGAGAAAAGCTAGAAGCAGATCAGGTGTTTGGCTCCTTAACATACTAGGCTCGCTCAACCTCCGCCAGATGGGACTGACAGGGTCTTTCAAGTGATGCATGAGCTATAGTTACTGGCCCATTTATGCCCTCAGCACCAATTTATAAAGAGGAI-riaKRIFTCAAAGGATGS K DCAAAGTAAGTKGGGAGGGAGGCTTGAAGGGACATTTCAGAATGGATCTTGCAAAAACCATTAATCAGGGTGAACGCGGCAACCAGGTTAGTATACTCTGCTGGTGAGTTTACTGTGGTCAAACATAAAAAGAGGGACCATCATCTAAGATGTGTCACCCAGTCCCGGGTTCATAGCGACGCCACCATGTTGCCACCCTCCTCCGTGCCTGGTGCCATTCCGCCAAGGAAATAAAACTGTACGACACCAGAATGAAAAATATE K YAACTGAAGCTELKLGGCCATCCGCGAGGGGAGAGGGACACATCAGGCAAGCTCTCCTTCCTTATAATAATTACCGCGTGCCCGTACTCCTCTTTGTGATTTTGGTCCriaaAGATGCCATCTGGGTATTCTTCATGTAGAATGTTCCTAACAGATCIIIIIIVTGCTGGAGTGCAAGCGATTCTACCACCACGCGCCAGGCTTCTGGCCTCCCACCTAAGATGCACCCCGACCAGGCCCAATGGTTCAGCliirTGAGTACTAAMSTKGCAGCCAAAGA A KATTGATTCAGL I QAGAGCCCACTGACTCCGGTAGGTGGGiiiiCAGAAAACAAGGGGTTCATGCTTAAAATGGCAAGAjaGATTCTTCTATACTACCTAGATGGTGTGAAAGATATTCCCAGACCACAGCCATTAATTACTGTACTCTAATTTTTTTTTTAGTGGCATCGCCTGCCTCGGCTGGCTAATTCATGGTCTCGAAGTGCTGCTTTATTTTTTATTCTCTTTTCCACTAAAATCTGGTAAGTTTAAAGTCTCCT K S PAAGGTGATCCEGDPGCTGAATTCCA E FTAATGGGACTGAGCCTTATACTTCCCCTTAGGGCCAAGCC1311077 GGGACCGTCG1127 ATTAAGGCCA1177 TTCTATCCCC TCCTCCAGGC CTCCTGCAGC CTCAGCACAG GACTCCGCAA1227 GCCTAAGGCC ATAATAAATG GAAGCAATGC CAAGTTATGC CTGGAAGACC1277 CACAAGTCAC CAAGCAAACC CAATAAGGCC TTCAGCATGG ATTCAGTACT1327 TCTGTGCCAG GCTTTAGGAA AACAAAGAGG ACTGCAAAAT AGTGTTAGCT1377 GGCCTTGGAC AGGCTTCATT TGTGTATGTA AAATCCCTCA ACAAGACCAA1427 CACTTACTAG CAGTCTGTTA GGAAAACCCA CTGACAAATT GCCTTGTCCA1477 GCACTCATTT ATATCTAAAA GCATTTACAT TTGTGTTCAT CACTTAAATT1527 TGTGTTAATG AGGCAACCGT GCATTTCTAA ACAACTTTCA AGAAAGGACT1577 AAATTTCTGG CCATAAAGTT TCACATCAAA ATCACCCCAG AGTTGTCAAA1627 ATCAAAATGC ATTCAAGGCA AACAGAGTAC AGAGAAGAGC TGATTCGTCC1677 TCACTCCAAT ATACAGTACT GTCTTTCCTT TTGCTAAAGA AAAAGCTGAG1727 TCAGACACTT GCCAAGATGG TAAGGCAACT TACTCCAGGA GGGCCATGGC1777 GGACCCATAC AGGGACCACT GAAACGTGTT CTTGCCATGG GGAGAGAGAT1827 TGGGCTCAAC TCCAGAAAGG ATAAGTGGGG GTTTATGACC ATAGAGCAGG1877 ATGGGGGGGG CCAGTGATGA AAATTACCAA GGAAATTATC CAGATAAGGA1927 GCTTCTGGCT AAATTGCCTT GATAGGATTC TAGCTGAAGG CAGGCCAGGT1977 GATCAAATAT TAAGGGTAGT CAGATACAAG GACAGAGAGG GAAGCCCAAG2017 GTCAGGGCCT AGTCGAGGCA GAGGGAACTT CAAAGGAGAG AGTCATTGTC2077 AIIIIGCAAT TAAAAACCAT CTTCACAACA GAAGTGACTA TGGAATTGTT2127 TGTATGACTC TGTCACTATT TAGCTCTAAC TTCAGGAAAG AACATTTCTT2177 CAAATArrri AACCAGCTCT CCAGArrriA ACCAGAAATG TCCCTTCTGC2227 CAATATGACA ATTGGAAAGA CTTCTGTTTT TAATACTAAA AAAGAAAAAA2277 ATCTGAATTC AATGACCCCA GAATCTGATA TATTCTCATT TTCTCATAGT2327 TGAAATATAG AAGAACTTGT AAAAACACTT AAAGTCACTA GCAGAAAAAA2377 GGAMIlliA ATTAAACAGC AAAACAAGTT AGTGAAAAAA TTTAAAATTT2427 TAAAAGACCC TTTArrrriC TAAAGGAAGA AAAATAATCC ATCTAACAGG2477 AAAGCCTAGG CCAGGTACTC AAACAAGTTG TTTCAAAAGG TAACACGGGT1322527 AACiirrECT ATA'rrriATG TCTTTCCTAC AAAAAGTACG TAAGACATAA2577 TCTGGAACGA GGTAGGAAGA CTAATGGTAA GTTGTGAAAG TCTACTTAAC2627 ATGTACCTGA TTATATTCCC ATAAGGTTTC CATTTACTAA GG'rrr12CCCA2677 TAATATGTTT ACATATATCA TCACCATGCA TGGTTTAGAA TCTAGTAAAG2727 GCTTCAGGAA AACAGTGTTT CATATTCATA AATCAAAGGC TGAATAAGTA2777 ACTTACAGTT TGGTCTTGTA ATTAAAGCAA AGTGCTAAGT GTTTTTGTAG2827 GAATCTTAGG GGTTAAAATA TATCTTCCAT AGAGATGTGC CTGCACAAAA2877 TTGTGAATCT ACTTAGCACT ACTGAACTGT ATGCTTAAAA ATGATTAAGA2927 GCGTAAATTC AATGTTATGT GrrrrrLACC ACAATTTATA TATACATATA2977 TATATATAAT ATAGTCTAAT TTGAAACTGA CAAACTGAAC TGTTATTTCA3027 ATCGGTTTTC TAAGTAGTAT TTTGCACAGA TAGACTCCTA CCTGAAAGTC3077 CAAGTAGTAA TGGCTTTAAG ATTTAGAGTT TTATGGTAGA GACAAAAATT3127 AACTCCACTT AAGACACTGC AGGAAGTCCA GTAGTCAAAA AAGCCTGCAC3177 ACGATCTCAC ACGACATCTC AAGCAGTGGT TCTCCAAGTG TGGTCCCAGA3227 GCAGCAGCAT TGGCACCACC TGGGAAGCTG TTGGAAAATA TAGATTCTCC3277 AGCTTGAACA ACACAGCAAG ACCTTGTCTC TATTAAAACT TTTAAAAAAT3327 TAGCCAGGTG TGGTGCACAC GCCTTTAGTC CCAGCTACTC GGGAGGCTGA3377 GGCAGGAGGA TCATTTGAGC CCAGGAGGTG GAAGCTGCAG TGGGCTGTGA3427 TCATGCCCAC TGCACTCCAG CCTGGGTGAC AGAGCAAGAC CTTGTGTACA3477 GATTCTCGGG CCCCACCCCA GGCCTAGTGA ATCAGAAACT CTGAGTGGGG3527 CTGGCAGTCT GTATTTACTA AGGTCAGCAA GCCTATCGTG TGATTCTGAT3577 GCACTGAAAG CATGGGAACC ATTGAGCCAA AGTCAIIIIG ATAGCTCAAC3627 GTTTGGAAAA ACTGGGCGGT GTTAGTTAAG GGGTTGAGCT CTGGAGTCAA3677 CGTGTGACCC CAACCCCTCT GTTAACCAGC TGTGACACTC AGTAAGTGAC3727 TGCCTGTCTC TGAGCCTGTT TCCCCACCTG TAAAGTAGAT CTCCCTCATA3777 GGGTGGTCGG AAGGATTACA GAAGGAGCCC TCCCTGCCTG CTGCCTCATG3827 TAATATGTCA GCCCTAGCCC TGGAGCAGGA GACAAACCCA GCTTCCAGAA1333877 AACCTTGGGG CAGGGGTCCT CAACTGTTTG TGTCGTGGGC AATCCAGTGA3927 AGCCCACAGA TCCCTTCTCA CAAGAATGTT TTTAGAGGCA TGGAGGAAAA3977 AGACCCAAGA TCGCATAAGA AAACAACTGT ATTAAAATAC GCATTGTATT4027 AAAATAACAG ATGTGTGATA TGGAGGAAAA AAAGGTAGCC AGTTCTATGA4077 CACGGAGCAA TACAACAGCT AAAGCTGGTT TCTCTCTGTT AGCTATTAGA4127 TGCATGTCAC TGGAAACCCT GAGCTGCCTT CTTGGCACAC AGACCCCAAG4177 GCCTCTAATT TCTGAGAATG TGTAGCAGTA ACTGATTCAT TTTTTCTCCC4227 CTCCCTTCTC CCATCCIIii TTATGTAAAA CAGGGTCCAA ACACCCTAGAG P N T L D4277 TGATCTCTTT CAAGAACTGG ACAAGAATGG AGATGGAGAA GTTAGrrriGDLF QELD K N G DGE VSFE4327 AAGAATTCCA AGTATTAGTA AAAAAGATAT CCAGTGAAAG GAGAAAACAA E F Q VLV KKIS Q*4377 AATAGAACCC TGAGCACTGG AGGAAGAGCG CCTGTGCTGT GGTCTTATCC 4427 TATGTGGAAT CCCCCAAAGT CTCTGGTTTA ATTCTTTGCA ATTATAATAA 4477 CCTGGCTGTG AGGTTCAGTT ATTATTAATA AAGAAATTAT TAGACATACC4527 TTACTTTGTT AAAGTACTGA CCTCATAACA TAATACGTGA CTTGAAAGTA4577 ACCTTAGTCT GATTCAACTA ATTCTATAGA TTTAAAAAAG AGAGAGAGAG4627 AGAGAGAGAG AGGGAGGCCT GAGGCAGGAA GAACACTTGA GCCCCCGAGG4677 TCAAGGTTGC AGTGAGACAT GATCACGCCA CTGCACTCCA GCCTGGACCA4727 CAGAGGGAGA CCCTGTCTAA AAATAAAACA AAATAAAAAT AAAAATAAAA4777 AAATAAAATT GTGAGGGCCA AGATAGCTCT GTGATCCTGC CGAAAAGAAT4827 CCAGAAATGC CAAACCAACT CCCTTCCTCC ACIIIICACC AGAATAACCC4877 TGTTGCTTGC CTCTCCTTCC TTACCTCCTC AGTTGCTCAG GTCATCTCAA4927 TTTCACTCTG ATCTCACACT ATCCACTAGC GAGCAAGAAA CACTTCACAA4977 CAGACAGTTG ACGTCATGCA TCAC134135dispersed throughout the gene. All four segments were identified as Alu repeatsequences. Two of these repeats were full sized Alu sequences (350 bp), one beingin the 5' flaking region (at -354 to -652) and the other part of intron A (+356 to+757). The other two repeats are truncated (150) and found in intron B (+3,277to +3,476) and the 3' flanking regions (+4,640 to 4,806). The closest related genescompared to the 5' Alu repeats were the ENO2 (gene for neuron specific y-enolase)and the prothrombin gene. Closely related repeats compared to the 3' repeatswere identified in the gene for the 14 kD I3-galactoside-binding lectin andlymphocyte receptor. Between the human and rat genes four separated regionshowed some degree of sequence identity. The 5' flanking region revealed 71.9%sequence identity over a length of 1,324 by with several gaps in both genesoptimizing the alignment. The 3' end of intron A gave a 66.7% sequence identityover a length of 175 bp. At the end of intron B a region of 209 by was detectedwith 65.5% sequence identity. Another stretch of 72 by in the 3' flanking regionscored 71.8% sequence identity. The comparison of the human and porcine intronA regions indicated two segments of relative high sequence identity. The 5' partof this sequence gave a score of 80.3% identity over 356 bp. The 3' end of theintron gave 71.0% sequence identity over 152 bp. The above mentioned Alu repeatof the human intron A region was not part of these homologous segments.Computer analysis of the porcine intron A identified a short interspersed repeatfound in many porcine genes.136A number of putative transcription factor binding sites was found in the 5'flanking region of the gene. At position -28 from the transcription initiation sitea perfect canonical TATA box marks the gene's promoter. The context of thissequence which is preceded by AC also makes this element a binding site for theengrailed factor (Ohkuma et al., 1990). Three CCAAT boxes were localized inclose vicinity of this site (-71, -105, -340). A binding site for CArG (Miwa et al.,1987), a muscle specific transcription factor was detected at -85. Also, threebinding sites for EBP20-protein (Costa et al., 1988), which can vary in sequencewere mapped at -121, -150, and -784. A total of four binding sites for WAP (wheyacidic protein) (Lubon and Hennighausen, 1987) were found at -328, -486, -686,and -1,031. An AP-2 (enhancer-binding protein AP-2) (Mitchell et al., 1987)recognition site is encoded at -402. Position -1,084 constitutes a pit-1 (pituitaryspecific, POU factor) binding element (Elsholtz et al., 1990).A search for steroid hormone type response elements using the ERE, GRE,RARE, and THE (estrogen-, glucocorticoid-, retinoic acid-, thyroid hormone-response element) consensus sequences identified no candidate for such a bindingsite. A comparison to each of the ERE-half sites (GGTCA and TGACC) scored 5and 7 matches throughout the entire genomic sequence. When the rat calbindin-D9k ERE was used in a BESTFIT analysis, two homologous segments were locatedat +50 and +3,607.137rat calbindin-D9k ERE +51: GGTCA GGG TGATChuman calbindin-D9k +50: GGTTA GTG TGATThuman calbindin-D 9k +3,607: AGTCA TTT TGATAThe sequence at +50 corresponds to the same location as the rat calbindin-D 9kERE and was used for further studies described in the next chapter.The published vitamin D response elements (VDRE) were also comparedwith the human gene sequence. An element resembling the VDRE of the humanosteocalcin gene (Ozono et al., 1990) with 9 out of 12 matching nucleotides wasnoted at -1,213.human osteocalcin VDRE -491: GGGTGA ACG GGGGCAhuman calbindin-D9k -1,213: AGGTGA TAT AAGGCASix perfect matches for the 6 nucleotide half sites of the GRE consensus weredetected. These sites were dispersed over the whole genomic sequence.All intron-exon splice junctions could be identified on the basis of the cDNAsequence and the GT-AG rule established by Mount (1982). The nucleotidessurrounding the 5'- splice junction read AG/GTAAG-TTATA which is in agreementwith the 5'splice consensus. At the 3'-end of the intron the splice junction138consensus is composed of two sequence elements: a pyrimidine tract (mainly T)followed by ACAG/G-T at the junction. Both of these sequence elements are inagreement with those previously published.1393.12 Estrogen receptor binding assay to an ERE-like sequence in thehuman and porcine calbindin-D9k genesComputer analysis of the human and porcine genomic sequences revealeda motif in both genes similar to the ERE found in the rat calbindin-D 9k gene(Darwish et al., 1991). This sequence is also situated at the same location in thegenes. The rat sequence resembles an ERE in the vitellogenin gene and has beenshown to bind the estrogen receptor and mediate estrogenic regulation of this genein the rat uterus (Darwish et al., 1991). These sequences are shown in Fig. 6.The human sequence differed in two relevant positions from the rat sequence.Therefore, two mutants of the human sequence were synthesized restoring thenucleotide found in the rat element (m ERE dr, right arm human specific; m EREdl, left arm human specific). A palindromic, double point mutant of thevitellogenin ERE (m ERE) served as negative control.Binding of the estrogen receptor to these sequences was analyzed using agel retardation assay. Cytosol extract from immature rat uterus was used as asource for estrogen receptors. Heat treatment of the estrogen receptor in vitrocauses a so heat transformation, and the transformed receptor binds DNA moretightly. Therefore, part of the uterine cytosol was used after heat treatment. Theresults of these gel retardation experiments are shown in Fig. 29. When thecytosol extract was included in the incubation mixture, a DNA-protein complex140Fig. 29. Gel retardation assay using the porcine element, vitellogeninERE, rat calbindin-D9k ERE, a mutant ERE, human calbindin-D9k ERE, mERE dl and m ERE dr. Approximately 3 fmol of each labelled 45-nucleotidefragment was incubated with 1 jig of poly(dI.dC) and without cytosol (-), with 10jig of cytosol (+) and with 10 pg of heat-treated cytosol (+H). The arrow indicatesthe retardation complex.Pig^Vitellogenin^Rat^m ERE^Human^in ERE dl^m ERE dr -^+ +H^-^+^+H + +H^+^+11 +^+H^-^+^+11^-^+^+11fre142was detected with the vitellogenin ERE, rat calbindin-D9k ERE, m ERE dl and mERE dr; but no such complex was detected with the m ERE, human and pigsequences. The amount of formed complex was significantly increased when heattreated cytosol was used instead of non heat treated cytosol (Kumar andChambon, 1988). Therefore, the human calbindin-D9k sequence with twonucleotides changed from the rat ERE failed to bind the estrogen receptor.This preliminary experiment also shows that exchange of either sequenceat left or right arm of the motif to the rat ERE restores receptor binding, althoughthe affinity is noticeable lower. The porcine motif also failed to bind receptor inthis assay. When antiserum raised against the rat estrogen receptor was includedin the incubation the retardation complex was shifted to higher molecular weightindicating the specificity of protein DNA binding (Fig. 30).143Fig. 30. Gel retardation in the presence of estrogen receptor antiserum.Lanes 4-9 are positive controls using the vitellogenin ERE without cytosol (4),with cytosol (5) and decreasing amounts of antiserum (lane 6, 1:5; lane 7, 1:25;lane 8, 1:125; lane 9, 1:625). All other elements were run in sets of three, fromleft to right: without cytosol, with cytosol and with cytosol and antiserum (1:5dilution). The arrow marks the retardation complex. The * indicates theDNA/receptor/antibody complex.Rat^m ERE^Human^m ERE dl^m ERE dr1^2^3^4^5^6^7^8^9^10^11^12^13^14^15^16^17 18^19^20^21^22^23 24Pig VitellogeninfreeDNA145N. DISCUSSION4.1 Cloning of the human calbindin-D9k cDNAIn previous experiments I was unable to use a rat cDNA probe to detectcalbindin-D9k in human RNA, whilst detection was possible in other species suchas porcine, bovine, and ovine mRNA (data not shown). Therefore, conventionalscreening methods with a suitable cDNA library were not feasible for cloning thehuman calbindin-D 9k. Instead, PCR methodology was used to amplify fragmentsof the human calbindin-D 9k mRNA (Ohara et al., 1989). Primers were designedafter the N- and C- terminal loop regions of the calcium binding domains, whichare highly conserved between species. Successful amplification of the targetsequence was dependent on sufficient identity of the primers used for PCR.Initially primer combinations based on the rat (Darwish et al., 1987) sequencewere tried without success. Using additional primers derived from a bovine(Kumar et al., 1989) sequence in combination with modified PCR conditions (lowerannealing temperature, slower change in temperature profiles) finally allowed theformation of a human specific product. As determined later in the project, theprimers used for core product amplification had 3 mismatches each within the 20nucleotides. The fact that these mismatches from the actual human targetoccurred in the center rather than at the 3' end probably explains why they146annealed sufficiently. It was very important to prove that the product was in facthuman and not of any other origin such as rat or porcine via contamination of thePCR process. This was accomplished by restriction digest, northern and Southernanalysis.With the human core product at hand one could perform specific northernanalysis of human RNA. Expression was found as expected in duodenal RNA. Inuterus and placenta no message for calbindin-D 9k was detectable. Therefore, itbecame unfeasible to screen available cDNA libraries. There is no human cDNAlibrary from duodenum available to date. Construction of a cDNA library ishampered by the availability of tissue, especially fresh enough to isolate non-degraded, high quality RNA. Despite the substantial degradation of the duodenalRNA available to us, PCR amplification was accomplished. Instead ofconstructing a library it was chosen to try amplification by the anchored PCRtechnology (also called RACE) to generate a full length cDNA clone. Using thesequence information obtained by the core PCR cloning it was possible tosynthesize human specific primers for the anchored PCR approach. Porcine RNAwas used as a cross species control to establish core and anchored PCR proceduresusing high quality RNA of an unknown sequence. In fact all three steps, coreamplification as well 5' and 3' RACE procedures proved to work rather efficientlywith the porcine material. Poor human RNA quality was probably the reason forlow efficient amplification during core and RACE procedures. Despite these147problems, it was finally possible to produce sufficient amounts of humanamplification products to allow their molecular cloning. The product of 30-cyclePCR amplifications were cloned into a pUC19 vector, multiple isolates of the sameallele obtained, and their sequences compared. In 3 separate clones, nomismatched bases were identified in the human and porcine calbindin-D 9ksequences.Identity between the rat and human cDNA sequences is lower than betweenrat (Darwish et at, 1987) and bovine (Kumar et at, 1987) or porcine calbindin-D9k .The low sequence identity combined with the difficulty to obtain high quality, non-degraded RNA from human intestine may explain the lack of detection of humancalbindin-D9k using a rat cDNA probe.Interestingly, the human calbindin-D 9k has an extra 54 nucleotides insertedin the 3'non-coding region, when aligned to the rat RNA. All of the 3'non-codingregion in the rat and human RNAs are encoded in a single exon of the respectivegenes. Therefore, different RNA splicing is ruled out. It is noteworthy that themRNAs for bovine and porcine calbindin-D 9k have a similar but not identicalinserted region compared to the rat sequence.The human calbindin-D9k shows highest sequence identity to the porcineand bovine proteins (88.6%), followed by rat (78.5%) and murine (75.9%) calbindin-148D9ks. This relatively high sequence identity of the coding region was expected,especially in the region encoding the 17 C-terminal amino acids. This part of thesecond calcium-binding domain is identical to the porcine and bovine proteins,while the two rodent proteins differ in four positions in this region. There arefour unique amino acids found in human at position 3 (threonine --> alanine), 12(arginine —> serine or glycine), 30 (aspartic acid -+ glutamic acid), and 48(asparagine —> arginine or serine). The two loop regions of the calcium bindingdomains reveal the highest amino acid conservation, while the helix and linkerregions are more variable. Each of these domains consists of a loop of 12-14amino acids which contribute 6-8 oxygen atoms to the coordination of the calciumion. The loop is flanked by two a-helices which, in a helix-loop-helix arrangement,is known as EF-hand (Kretsinger and Nockold, 1973). Putative EF-hands couldbe predicted from the derived amino acids from the human cDNA sequence. Thehuman calbindin-D9k has two potential calcium-ion binding sites, one of which liesbetween residues 10 and 39 and the second between residues 50 and 77. The C-terminal loop (residues 58-69) corresponds to an almost perfect EF-hand sequence,with five residues containing carboxylic acid derivatives in their side chains(aspartic acid 58, asparagine 60, aspartic acid 62, glutamic acid 64, serine 66,glutamic acid 69) in the exact positions of the loop region that are thought to beimportant in calcium-ion binding by the bovine calbindin-D 9k (Szebenyi et al.,1981).149The N-terminal EF-hand loop (alanine 18, glutamic acid 21, aspartic acid23, glycine 26) of the human calbindin-D 9k is less likely to bind calcium ions thanthe C-terminal structure. In the human calbindin-D 9k, this potential binding sitecontains glycine instead of the aspartic acid found in other species such as rat(Darwish et al., 1987), bovine (Kumar et al., 1989), and porcine calbindin-D 9k. Thecalcium loop of an EF-hand contains amino acids with side chain oxygens (Ser,Thr, Asp, Asn, Glu, or Gln) which serve to coordinate the bound calcium ion(Kretsinger et al., 1982). The calcium binds to the EF-hand with an approximateKd=1-10 jaM, although the exact calcium-binding affinity of each binding site mayvary significantly from protein to protein. Binding of calcium to the EF-handinduces a conformational change in the proteins and their potential activity(Baudier and Gerard, 1983). Other calcium-binding proteins such as calmodulin(Babu et al., 1985) and troponin C (Satyshur et al., 1988) contain four EF-handswhile the parvalbumin (Swain et al., 1989) contains three EF-hands per monomermolecule. The S-100 protein family is the only EF-hand-containing protein familywith two EF-hands. The C-terminal EF-hand in the S-100-related proteinscontains 12 amino acids and is similar in sequence to that found in the calmodulinprotein family. However, the N-terminal EF-hand in each of the S-100-relatedproteins contains 14 amino acids instead of 12. This variation produces asignificant decrease in the calcium affinity of the N-terminal EF-hand whencompared to the C-terminal EF-hand (Kretsinger 1982). The calcium bindingaffinity of the C-terminal EF-hand is approximately K d=20-50 illVI while the150calcium binding affinity of the N-terminal EF-hand is approximately Kd=200-500AM. Interestingly, the N-terminal EF-hand is conserved in the calbindin-D 9k ofall species known so far. This observation argues for an important and conservedfunction in spite of its low calcium binding affinity.Cloning of the human calbindin-D9k cDNA as described in this thesis hasbeen published in July 1992 (Jeung et al. 1992a). In June 1992 another Group inthe UK used the same approach and cloned the identical sequence (Howard et al.,1992).4.2 Cloning of the porcine calbindin -D9k cDNAThe anchored PCR technique was also used to clone the full-length cDNAencoding the calbindin-D9k from porcine duodenum. The cDNA structure wassimilar to that of the bovine (Kumar et al., 1989). The porcine cDNA clone wascomposed of 57 nucleotides 5' non-coding region, 237 nucleotides coding region for79 amino acids, and 149 nucleotides 3' non-coding region. The previouslypublished cDNA clone for the bovine calbindin-D 9k, which was isolated byconventional screening techniques, has only 16 nucleotides 5' non-coding region.It is likely that this bovine clone does not represent the entire 5' noncoding region.Generally the 5' RACE method produces full-length cDNA products. Interestingly151it was found that the porcine cDNA, like the bovine and human cDNAs, has anadditional 40-50 nucleotides inserted within the 3' non-coding region. Theinsertion sites in all three species are similar but not identical. As discussed withregard to the human cDNA, it is considered unlikely that this inserted region inthe 3' end of the mRNA is caused by different splicing. Based on the results of thepresent studies it rather appears that the rodent (rat and possibly mouse) genehas this particular region deleted. It is not known what event might have causedthe loss of that region in the rat gene.The porcine calbindin-D9k has been previously purified and sequenced frompig duodenum (Hormann et al., 1979). The deduced amino acid sequence of theclone here differed in one position from the reported sequence. Residue number25 had been determined as aspartic acid using amino acid sequencing techniques.The cDNA-derived amino acid at this position is asparagine. A similardiscrepancy was found when the rat cDNA was cloned and compared to the ratamino acid sequence. It is suggested that the asparagine residue is the authenticamino acid, and a loss of the amide group during protein purification andsequencing resulted in an aspartic acid residue at this position.The porcine calbindin-D9k has two potential calcium binding sites, one ofwhich is located between residues 10 and 39 and the second between residues 50and 77. The N-terminal EF-hand loop (alanine 18, glutamic acid 21, aspartic acid15223, glycine 26) of porcine calbindin-D 9k is identical with other known species. TheC-terminal loop (residues 58-69) corresponds to an almost perfect EF-handsequence (Kretsinger, 1982), with five resides containing carboxylic acidderivatives in their side chains (aspartic acid 58, asparagine 60, 62, glutamic acid64, serine 66, glutamic acid 69) in the exact positions of the loop region. Thenumber 62 residue is a unique amino acid for the porcine calbindin-D9k comparedto other species. In any case, although this amino acid belongs to the EF-handsequence where calcium is coordinated, this small change is probably notstructurally important, as both asparagine and aspartic acid have oxygen-containing side-chains and are both calcium ligands.4.3 Calbindin-D9k gene expression in human tissuesThis is the first time that expression of human calbindin-D 9k was studiedusing a specific cDNA probe. In a previous report by Thomasset's group (Brun etal., 1987) a purified rat cDNA probe was used and a signal was described in fetaltissues such as duodenum, lung, kidney and thymus. In the present experimentthese results were not reproducible and as shown in Fig. 15 there appears to beno cross hybridization of rat and human calbindin-D 9k under standard stringencyconditions as they were used in these experiments. Based on the degree ofidentity between the human and rat sequences cross hybridization should occur.153In fact, with large amounts of rat duodenal RNA (20 pg) the human probe doesgive a weak signal on northern blots. One has to note that in the rat duodenumexpression of calbindin-D 9k is very high. The protein can amount to 1-3 % of totalcytosolic proteins (Shimura and Wasserman 1984). On the other hand using anorthern blot of human RNA does not produce a signal with the rat probe. Thisis probably caused by the much lower expression level in combination with themismatches of the probe.Other tissues in which calbindin-D 9k is highly expressed in the rat areplacenta and uterus (Krisinger et al., 1992a; Krisinger et al., 1992b). Northernblots of total and poly (A+) RNA from these human tissues failed to demonstratecalbindin-D9k expression. These results were confirmed by both S1-nucleaseanalysis and RT/PCR. These techniques are extremely sensitive and it istherefore concluded that unlike other mammalian species studied so far there isa lack of expression in these reproductive tissues in human. Subsequently,Howard et al. (1992) also reported a lack of expression in placenta and uterususing their cDNA probe and PCR techniques. This tissue specific difference ingene expression is rather unexpected. Given the high degree with which thisprotein is conserved throughout other species and the high level of expression inboth placenta and uterus it may play an important functional role. Currently,there is no definite explanation for the lack of placental and uterine expression inhumans. With regard to uterine expression, the mechanism responsible for this154tissue specific expression is known. The calbindin-D 9k gene is regulated by anestrogen response element mediating the steroid control in uterus (Darwish et al.1991). A possible disruption of this regulatory element within the human genecould be the cause of the tissue specific lack of gene expression. This aspect isdiscussed in more detail in the chapter describing the analysis of the humancalbindin-D9k gene (4.7). The regulation of placental calbindin-D 9k remains largelyunknown. A control by vitamin D metabolites as in the duodenum has been ruledout (Garel et al., 1980). Estrogen control of placental calbindin-D 9k expression hasnot been studied yet. The pattern of expression during the course of gestation,where an extreme increase of expression occurs between days 16-20 of gestationmay favour a positive effect of estradiol on calbindin-D 9k expression (Krisinger etal., 1992b). This is also in agreement with the expression of the estradiol receptorin placenta. Although the estrogen control of placental calbindin-D 9k is speculativeit could provide an explanation for the observed lack of expression in human,along with the silence of the gene in uterus.The information on human calbindin-D 9k at present is very limited. Staun'sgroup (Staun, 1986) in Denmark is the only laboratory that has purified calbindin-D9k from human intestine and generated polyclonal antibodies. These antibodieshave not been used to detect the protein in other tissues than duodenum andkidney. In a study with biopsy materials from human small intestine, Staunreported a decline of calbindin-D9k levels with age (15 - 126 months) (Staun et al.,1551991), which does not continue during adulthood (aged 20 - 89 years) (Staun et al.,1988). In the present study it was also found that the expression levels in asample from a 2.5 year old child were approximately five fold higher than in a 32year adult. Although there are very limited and non quantitative data, they arein line with the expression profile in the rat (Armbrecht et al., 1989). In thisanimal calbindin-D 9k expression has been studied extensively. There is an agedependent decrease of expression which is concomitant with a decline in vitaminD receptor numbers (Eberling et al., 1992). There is also a decrease of calciumtransport actively in the intestine with age. The cause of these age dependentchanges are not clear at present.4.4 Calbindin-D9k gene expression in porcine tissuesThe steady-state level of calbindin-D9k mRNA was investigated in varioustissues using northern analysis. Total RNA from duodenum, jejunum, ileum,kidney, uterus, liver, lung, heart, pancreas, spleen, aorta, ovary, fimbriae, cervix,and brain was analyzed. Porcine calbindin-D 9k was highly expressed in intestine(duodenum > jejunum > ileum), as has been shown in the rat (Leonard et al.,1984). Calbindin-D9k mRNA was also detected at much lower levels in the kidneyand uterus. Detection of calbindin-D 9k in the kidney is consistent with finding bybiochemical techniques showing the expression of both calbindin-D 28k (the major156renal calbindin) and calbindin-D9k. In contrast to this finding Ambrecht et al.(1989) could not detect calbindin-D 9k expression in the rat kidney with a cDNAprobe.Recent studies have shown that renal calbindin-D9k was present in the 18-day-old rat fetus suggesting a role for this protein in the regulation of calciummetabolism in the prenatal period. Autoradiographic studies have previouslyindicated that 1,25(OH) 2D3 receptors are localized in the neonatal and adult ratkidney in the regions of the distal nephrone where calbindin-D 9k is also localized(Stumpf et al., 1980). These findings suggest a role for this protein in vitamin Dregulated calcium reabsorption processes in the kidney during development (Liand Christakos, 1991).Expression of calbindin-D9k in the uterus was similar to renal expressionand much lower than intestinal levels. Two tissue samples were analyzed. Onespecimen was retrieved from a local slaughter house and the other from aweanling piglet at a local pork farm. The calbindin-D9k mRNA level in the matureslaughter house animal was substantially higher than in the weanling. Uterinecalbindin-D9k expression has mainly been studied in the rat (Krisinger et al.,1992a,b). A tight regulation by estradiol and its receptor is found during theestrous cycle (Krisinger et al., 1992a). The steady state mRNA levels within thefour day cycle varies tremendously from non detectable during diestrus, over low157levels at estrus and ten fold higher levels in proestrus. In fact proestrus mRNAlevels in uterus are as high as those in duodenum A similar dramatic regulationis described during pregnancy and lactation (Krisinger et al., 1992b). It is alsonoteworthy that in ovariectomized (Darwish et al., 1991) as well as immature rats(3-week-old) (L'Horset et al., 1990) there is no calbindin-D 9k expression. It is notsurprising to find very low calbindin-D9k mRNA levels in the mature uteruscollected at the slaughter house. It can be assumed that a similar estrogeniccontrol of the calbindin-D 9k gene exists in the pig uterus. Therefore, chances offinding low levels at a random time during the cycle are relatively high. Thereis no information available with regard to the estrous cycle of the pig from theslaughter house. It is somewhat surprising to find calbindin-D 9k expressed is theweanling uterus. It will have to be examined in the future with detailedexperiments whether in the pig there is a different hormonal mechanismsregulating calbindin-D9k expression in this tissue.4.5 Cloning of the human calbindin-D9k gene and its analysisIn order to understand the mechanisms which direct the specific expressionof calbindin-D9k in the various tissue specific manners discussed earlier, it isnecessary to identify and analyze the corresponding gene. Ultimately the genomicsequence of this locus will determine which of the general and tissue specifictranscription factors control the activity of its expression.158Initial screening attempts of a genomic library using the human cDNAprobe were not successful. Therefore, PCR amplification of intronic sequences wascarried out. Primers were designed assuming a similar genomic structure as inthe rat gene. Amplification of intron B with specific human primer was notaccomplished. This may be due to the unexpected large size of this intron in thehuman gene (1.8 kb in rat vs. 3.3 kb in human). Primers annealing in exons I andII were able to amplify the intron A region of the human gene. After the producthad been cloned and sequenced it was possible to synthesize additional intronspecific primers for additional PCR experiments which were used for thechromosomal localization.Chromosome localization: The identification of the chromosomal location of thecalbindin-D9k gene can be useful with regard to information about the evolutionof this calcium binding protein and it may also become of interest if a geneticdisorder involves the particular chromosome and sublocation.For chromosomal location analysis DNA from hamster-human somatic cellhybrids was used for PCR amplification with primers located in the human intronA. DNA from 3 of the cell lines contained a PCR fragment which showed theexpected size predicted from the genomic DNA sequence. Comparison with theprovided panel of human chromosomal components of each cell line identifiedchromosome X as carrier of the calbindin-D 9k gene. Subsequently, Howard et al.(1992) also localized the gene on the X-chromosome by Southern blot analysis.159The existing sequence data on calcium binding proteins demonstrate thatthe relationship between calbindin-D 9k and other members is quite distant. Evenbetween the human calbindin-D9k and -28k there is no sequence identity. In viewof this evolutionary distance among these various calcium binding proteins, it isnot surprising that their genes are widely dispersed among chromosomes.Chromosomal locations of various human calcium binding proteins are indicatedin Table 3.. For example calbindin-D28k (Varghese et al., 1988) is located onchromosome 8, parvalbumin (Berchtold et al., 1987) on 22, calretinin (Parmentieret al., 1989) on 16, S-10013 protein (Allore et al., 1988) on 21, and calcyclin (Ferraiet al., 1987) on chromosome 1. Sequence analysis carried out by Kretsinger et al.(1980) has shown that the S-100 protein is the closest relative of calbindin-D 9k .Preliminary evidence indicated that S-100 related sequences are on many differenthuman chromosomes. Perhaps, separation of the calcium binding protein geneshas favoured the evolution of different developmental and/or regulatory functions.The evolution of mammals appears to have brought about a new alternative forcalbindin-D28k found in all vertebrates in the form of calbindin-D9k. The twoproteins probably share the same overall function. This may be reflected by thefact that mammals express two different calbindins while other species such asayes, reptilia and amphibia possess only the 28k form of calbindin. It appearsthat the latter species use calbindin-D28k instead of the 9 Kd protein for intestinalcalcium transport. Similarly, the egg shell gland (uterus) expresses the calbindin-D28k protein.Table 3. Chromosome localization of calcium binding proteins.Protein Human chromosomeParvalbumin 22Calbindin D-9k XCalbindin D-28k 8Calretinin 16S-10(43 21Calcyclin 1Myosin alkali light chains (MLC-1 and MLC-3) 2Myosin light chain (MLC-1 emb/A) 17MRP-8 and MRP-16 1Sorcin 7Sorcin-related gene 4Calmodulin pseudogene 17160161Even the hormonal regulation of calbindin-D 28k in the avian intestine (vitamin D)and eggshell gland (estrogens) (Corradino et al., 1968; Bar and Hurwitz, 1973;Fullmer et al., 1976) is analogous to calbindin-D 9k regulation in the mammalianintestine and uterus (Krisinger et al., 1992a,b).Some of the genetic disorders assigned to chromosome X include hereditaryhypophosphatemia and hypophosphatemia II (Glorieux and Scriver 1972; Scriver,1974). These diseases affect phosphate-, vitamin D-, and calcium-metabolism.This finding led to speculation on a possible causal relationship between X-linkedhypophosphatemic rickets (HYP), the most common form of familialhypophosphatemic rachitic disease and calbindin-D 9k. Machler et al. (1986), usingrestriction length polymorphism, have shown that the human HYP gene locus ismapped to the Xp22.1-p22.2 region. Recently, the HYP gene has been sublocalizedbetween the flanking markers DXS257 (telomeric) and DXS41 (centromeric) (Readet al., 1986). Linkage study of calbindin-D 9k, using somatic cell hybrids, to definethe genetic map of the X-chromosome suggests a location on the short arm(Howard et al., 1992). Unfortunately, studies for HYP have been constrained bythe limited number of patients samples for linkage analysis, the lack of sufficientpolymorphic probes in the region, and the relatively large distances between theflanking markers and the HYP gene locus. As a consequence, furtherinvestigations concerning the relevance of calbindin-D 9k gene to define the geneticmap of the region or as a tightly flanking marker, may ultimately provide162information concerning the role of calbindin-D9k in this disease. Upon request, oneof the subclones for the human calbindin-D 9k gene was provided to Dr. Oudet, atINSERM in Paris, France who is pursuing the linkage study of the HYP locus.Structure of the human calbindin-D 9ksBLe: Before one can be certain of having acomplete genomic fragment including the 5' regulatory region, the site oftranscription initiation has to be defined. Since the cDNA was isolated by the5'RACE technique it was expected to be full length. Primer extension analysisconfirmed that the human calbindin-D 9k gene contains a single transcriptioninitiation site at the predicted location. The first transcribed nucleotide is a Cresidue. It has been observed that most, but not all, eukaryotic genetranscriptions are initiated by an A residue surrounded by pyrimidines. ThemRNA cap site of this gene is a C residue, which is flanked by one pyrimidine onboth sides, and different from the rat gene. Twenty eight nucleotides upstreamfrom this location a perfect TATA box consensus (TATAAA) is detected whichprovides further conformation of this site as be the genuine transcription initiationsite.To identify the human calbindin-D 9k gene directly from genomic DNASouthern blot analysis was performed. Genomic DNA cleaved with 3 restrictionendonucleases and probed with the 3' RACE clone indicated that the humangenome contains a single calbindin-D9k gene; which had also been shown for the163rat. Because this analysis was performed with the 3'RACE clone only the 3'endof the gene was targeted (including the entire exon III). The 6.5 kb EcoRI and 13kb HindIII fragments are in agreement with the sequence of the genomic clone.No BamHI site was found in or near the calbindin-D 9k gene which may explain thelack of a signal in this particular digest of the genomic Southern blot.A genomic clone was isolated that contained the human calbindin-D 9k gene.Comparison of the genomic sequences with that of the human calbindin-D 9k cDNAindicated that the human gene contains two introns. The localization of theintrons in the human gene corresponds to the exact position of those in the ratgene (Krisinger et al., 1988). As in the rodent, the two calcium binding sites areencoded by two separate exons. However, the sequence and size of the two humanintrons are quite different. Overall the human introns are 2-3 times larger thanthe rat counterparts. This observation is to be expected because generally,intervening sequences are not well conserved among different species.Genomic-cDNA sequence discrepancies: One difference was noted between the nucleotide sequence of the gene andcDNA. The 5' end of the cDNA as determined from the 5'RACE clone had threeadditional A residues that were not found in the genomic sequence. Howard etal. (1991) also published a cDNA sequence from a RACE clone that matched the164genomic sequence described here. This difference is not due to a sequencing errorand unlikely to represent a polymorphism but is more likely explained by aformation of a PCR/cloning artefact.Computer analysis of the 5' flanking region: Computer analysis of the 5' flanking region of the calbindin-D 9k gene wasperformed in order identify typical eukaryotic upstream promoter elements. Anumber of such regulatory sequences were identified. The human calbindin-D 9kgene has a TATA box type promoter accompanied by CCAAT boxes as has beendescribed for a large number of genes including the rat calbindin-D 9k .Compilations and analyses of the eukaryotic polymerase II promoter sequenceswere recently published by Bucher and Trifonov (1986) and Nussinov et al. (1986).Both noted that the most striking signal is the short DNA sequence motif locatedabout 30 by in front of the transcription start site. This TATAAA sequence (socalled TATA box) functions as the binding site for the TATA factor, which in turnpositions the RNA polymerase II onto the initiation site. The CCAAT boxsequence motif is found in most eukaryotic promoters between -50 and -129 andis required for efficient transcription (Benoist et al., 1980). The human calbindin-D9k promoter exhibits two of these motifs within the defined boundary. Anothermotif overlaps the TATA box and represents a site for the "engrailed" protein.This factor acts as a competitor with the TATA binding protein and has been165characterized in drosophila (Ohkuma et al., 1990). Another general transcriptionfactor binding site termed AP-2 (Mitchell et al., 1987) was located around -400 andis likely to be involved in transcription of this promoter (TATA box). Thecomputer search will only identify perfect matches with the consensus sequence.Therefore, these sites are mentioned in the discussion although a putativefunction is not obvious. The remaining sites are all tissue specific motif such asthe CArG, a muscle specific transcription factor; the EBP20-protein (Miwa et al.,1987), a liver specific protein (Costa et al., 1988); the WAP (whey acidic protein)(Lubon and Hennighausen, 1987), a protein specific for the mammary gland; a pit-I (Elsholtz et al., 1990), which is a pituitary specific factor. The possiblesignificance of these binding sites for the regulation of the calbindin-D 9k genewould be speculative and needs further studies. With regard to the WAP bindingsite it may be added that at least in the rat mammary gland calbindin-D 9k is notexpressed. This tissue could be a reasonable site for expression of this calciumbinding protein in light of the large amounts of calcium being transferred intomilk during lactation.Steroid regulatory elements: It has been demonstrated that the rat calbindin-D 9kgene is under direct transcriptional control of 17-13 estradiol (uterus) (Darwish etal., 1991) and 1,25(OH)2D3 (intestine) (Darwish et al., 1992) mediated byrespective cis-acting sequences. There is also some information suggesting anegative effect on intestinal expression by glucocorticoid (Huang et al., 1989) and166uterine expression by progesterone (Bruns et al., 1988b). Therefore, regulatorysites for the vitamin D-, 17-flestradiol-, and progesterone receptors were searchedfor.Molecular analysis has identified the cis-acting sequences required for theinduction of vitamin D regulated genes. Only four such elements have beencharacterized today (Noda et al., 1990; Demay et al., 1990; Ozono et al., 1990;Darwish et al., 1992). Their sequence identity is not as high as found with othersteroid hormone response elements. Therefore, all four sequences were used ina computer search to identify related motifs. Recently Darwish et al. (1992), haveused binding and transfection experiments to define a 15 by element within ratcalbindin-D9k upstream region that is necessary for the regulation of geneexpression in response to vitamin D. It appears that the rat calbindin-D9k elementis quite distinct from the VDRE in the osteocalcin genes. The VDREs of thehuman (Ozono et al., 1990) and rat (Demay et al., 1990) osteocalcin genes havebeen characterized. They represent direct repeats spaced by three nucleotides.The human and rat osteocalcin VDREs differ in two positions but still are bothfunctional. The mouse osteopontin gene carries a VDRE (Noda et al., 1990) thatis yet again different from the three aforementioned ones. Given the variabilityof the VDREs, a motif found at - 1,213 in the human calbindin-D9k gene may bea good candidate. This position would be further upstream than the VDRE in therat calbindin-D9k. This could easily be caused by the insertion of the Alu repeat167sequence located in the 5' flanking region of the human gene. In this thesis it hasbeen proven that the calbindin-D9k gene is expressed in the human duodenum.With all likelihood this expression is mediated by the vitamin D receptor inanalogy to other mammals Thus it is anticipated that the human gene doesindeed harbour a VDRE which will have to be identified by transfection analysisof the flanking regions.When the gene was searched for sequence identities to the ERE consensusno perfect match was identified, although, the BESTFIT program located twomotifs with high sequence identity at +50 and +3,607. The sequence requirementsfor an ERE to be active have been studied in great detail. Based on theinformation available it is unlikely that these two sequences maintain bindingaffinity for the estradiol receptor. In addition to the questionable nucleotidesequence, the position of the second motif at the end of intron B makes anysignificance of this element unlikely. It was decided that the first sequencedeserved more attentions for two reasons: First the location +50 is very close tothe promoter region and in similar context as an active ERE in the rat gene;second the sequence of the human intron A has very little identity to the ratintron A (Krisinger et al., 1988), with the motif at +50 being the closest matchoverall. Thus this sequence was analyzed in more detail as described in the nextchapter (section 4.7).168The consensus sequences for the GRE and PRE (progesterone responseelement) are identical and represent a 15 nucleotide motif with two specific setsof 6 base pairs separated by three variable positions. There was no perfect matchor even closely sequence similar region to the full GRE/PRE within the humancalbindin-D9k gene. A search using the "tfsites" as data base detected 6 perfectmatching half sites to the GRE, 3 for each site of the palindrome. A search for theERE half sites was carried out as well. The ERE half sites consist of only 5nucleotides. Therefore it was not surprising to find a total of 12 matches (5 and7, respectively) throughout the 6,324 nucleotides of genomic sequence. So far,although cluster of half sites act as ERE in a far upstream position of ovalbumingene (Kato et al., 1992), no gene has be reported which is controlled by one halfsite of any response element in entire gene or by half sites spaced by largesegments of DNA. Thus the relevance of this matching half sites is questionable.With respect to the biological effects of glucocorticoids and progesterone onthe calbindin-D9k gene, the limited data suggest a negative effect on expression inintestine (Huang et al., 1989) and uterus (Bruns et al., 1988b). A decrease ofcalbindin-D9k mRNA levels in the rat intestine has reported by Huang et al., 1989.There is no evidence that would indicate that this effect is caused by a directinfluence on the calbindin-D 9k gene. Similarly, the negative control ofprogesterone on uterine calbindin-D 9k has been reported and it is not clearwhether this steroid directly effects the calbindin-D 9k gene or a secondary response169is involved. The information on negative steroid hormone response elements isscarce. There is a total of 11 negative response elements published for theglucocorticoid receptor (reviewed by Akerblom and Mellon, 1991). The derivedconsensus sequence from this group of genes was compared with both the humanand rat calbindin-D9k genes. No substantial homology was found.Chromatin structure: S1 nuclease or DNase I sensitivity is often describedwithin the regulatory regions of genes (Nikol and Felsenfeld, 1983). Because oftheir unusual DNA structure (for a review see Wells, 1988), it has been suggestedthat they could play an important role in the regulation of gene transcription.The rat calbindin-D 9k gene has been studied by the DNase I-hypersensitivitytechnique (Perret et al., 1991). DNase I-hypersensitive sites (HS) were found inthe chromatin of both tissues, those expressing calbindin-D 9k and those which donot. The nuclei were treated with DNase I, digested with restriction enzymes andanalyzed by Southern analysis. The authors have shown that duodenalchromatin, where the gene is strongly expressed, contained one major HS (HS4)and three minor HSs (HS2, HS3 and HS6) near the promoter region. The patternof HSs in duodenal chromatin was unchanged after stimulation with 1,25(OH) 2D3 .The liver chromatin contained one major HS1 identical to duodenal HS1 and oneadditional HSO which could explain the absence of calbindin-D 9k expression. Thehypersensitivity of HS5 including the transcription start site and TATA box, iscorrelated with gene activity, since it is only present in chromatin where the gene170is actively expressed (duodenum and uterus). Analysis of HS4 showed that it wasmuch less sensitive in uterine chromatin than in duodenal chromatin, whichindicates that it marks the binding site for important regulatory factors implicatedin the duodenum-specific expression of calbindin-D9k. Fine mapping of any ofthese regions has not been carried out yet and specific motifs potentially involvedin vitamin D and/or estrogen regulation of the gene are not known at present.DNase I-hypersensitivity studies of the human calbindin-D 9k gene will be able toidentify protected patterns associated with tissue specific expression. Inparticular with respect to the findings of mRNA expression in human uterus, itwill be of interest whether the promoter region reveals a different protectionpattern in duodenal compared to uterine nuclei.Alu repeats: Two typical and two truncated Alu repetitive elements (reviewed byKariya et al., 1987) were found in the sequenced region of the calbindin-D 9k gene.The two full size Alu repeats surround the promoter region at a distance of about300 bp. The two truncated forms are found in reverse orientation in intron B andabout 100 by down stream of exon III. The significance of this repetitive sequenceis not known. Alu repeats are very abundant; there are about 300,000 to 500,000copies per haploid human genome, amounting to 3 to 6% of the total mass of theDNA. Sequences equivalent to the Alu have been described in the genomes ofother primates, as well as in rodents. The human Alu sequences areapproximately 300 by long and consist of two directly repeating monomer units.171They are commonly displayed with the first monomer, of approximately 120 bp,on the right hand side followed by an A-rich stretch (central A-rich region), andthe second monomer, of approximately 150 bp, that is larger than the firstmonomer due to a 31 by insertion. Finally, an A-rich stretch is found at theextreme right. The Alu repeats of the human calbindin-D9k gene gave highhomology scores when the gene sequence was compared to the GenBank/EMBLdatabases, using the FASTA program of GCG. Ullu and Tschudi (1984) and Chenet al. (1985) have suggested that Alu sequences may represent defective 7SL RNAmolecules that have been reverse-transcribed and inserted into the genome.Various speculations have been made, e.g., that they are the start sites in thenormal cellular DNA replication (Jelinek et al., 1980; Ariga, 1984), modulators ofchromatin structure (Duncan et al., 1981), hot spots in recombination(Jagadeeswaran et al., 1982; Rogers, 1985), inhibitors of gene conversion (Hess etal., 1983) or mRNA stabilizer in the cytoplasm (Calabretta et al., 1981; Robertsonand Dickson, 1984). However, no convincing experimental evidence supportingany one of these speculations has yet been reported. Some Alu members havebeen reported to be transcribed by RNA polymerase III from a polymerase IIIpromoter that is located inside the Alu sequence. Two of the Alu sequences ofhuman calbindin-D9k are found in introns and will be transcribed by RNApolymerase II collinearly with protein-coding region, and later removed duringmRNA maturation as described by Sharp (1983) and Allan et al. (1983).172Splice junction: Splicing of precursor mRNA of vertebrate genes involves twocleavage-ligation reactions, resulting in the ligation of two exons. The processinvolves formation of a branched circular RNA, with the cleaved 3' terminus of theintron ligated to the adenosine nucleotide within the splice branch site located inthe same intron near the splice acceptor site. The sequences of splice donor,branch, and acceptor sites are well conserved (reviewed by Padgett et al., 1986).The consensus sequence for the 5' splice site is 5'-AG-GTAAGT-3' where the GTis invariant. In the human calbindin-D 9k gene, the sequences for the 5' splice siteare 5'-AG/GTTAG-3' in the intron A and 5'-AG/GTAAG in intron B, and thereforein agreement with this model. The consensus sequence for the 3'splice site is(T/C)„N(C/T)AG-G/A where only the AG is strictly conserved in all functionalintrons. At each position of the polypyrimidine tract purines can occur with a lowfrequency. This polypyrimidine tract can be observed at the 3' end of the twointrons in human calbindin-D 9k gene preceding the CAG sequence of the 3' splicesite.3'end sequences: The highly conserved hexanucleotide sequence AATAAA isrequired for polyadenylation of mRNAs transcribed by RNA polymerase II(McLauchlan et al., 1985), a point mutation within this sequence abolishes theability to form correctly terminated mRNA (Montell et al., 1983). a-thalassaemiais a genetic defect due to an altered AATAAA signal (Higgs et al., 1983).However, it is clear that sequences apart from the AATAAA signal are required173for the poly adenylation process since this sequence is present at sites other thanat mRNA 3' termini. The human calbindin-D 9k DNA gene carries a conservedsequence located downstream from the poly A site at 37 nucleotides from theAATAAA signal which is similar to a sequence noted by McLauchlan et al. 1985.This sequence (YGTGTTYY) (Y: C or T) is present downstream from many genes.The preferred location for this sequence is 24-30 nucleotides downstream from theAATAAA signal. Recent studies suggest that this conserved element is likely toplay an important role in the formation of the 3' termini of mRNAs transcribedfrom a wide variety of mammalian genes (Manley JL., 1988).4.6 Analysis of the porcine intron A sequenceThe genomic sequence determined from the porcine calbindin-D9k intronrevealed the presence of a typical short interspersed repeat found in many otherpig genes. Besides this, there were two regions at the most 5' and 3' part of theintron which were quite homologous to the human intron A. Search for a generaltranscription factor binding site did not indicate any motifs with possibleimplications for activity of the nearby promoter. The intron A region of the ratgene has been shown to contain an ERE (Darwish et al., 1991). The identicallocation, at the boundary of exon I and intron A in the pig gene features asequence with 7 out of 10 matching nucleotides to this ERE. Therefore this174element was probed for its binding affinity for the estrogen receptor as describedin 4.6..4.7 Estrogen receptor binding assay with an ERE-like sequence in thehuman and porcine calbindin-D9k genes.Most of the studies on calbindin-D 9k were carried out in the rat model. Therat gene is subject to tissue specific expression as has been described in numerousin vitro and in vivo experiments. The regulation of calbindin-D9k levels in uterusand fallopian tube was reported by Delorme et al. (1983) and Bruns et al. (1985)to be under the control of estrogen. When 1713-estradiol was administratered toovariectomized female rats, a significant increase in the levels of calbindin-D 9k wasdetected in these tissues (Delorme et al., 1983; Bruns et al., 1988b). 1,25(OH)2D3,the major inducing factor in intestine did not affect calbindin-D9k expression inuterus. Conversely, estrogen has no effect on the expression of calbindin-D 9k inthe intestine. Moreover, Darwish et al. (1991) have demonstrated that animperfect palindromic sequence at position +51 from the transcription sitefunctions as an ERE. This element is capable of binding to the estrogen receptorand induces transcriptional activity in response to estrogen.175As described in 4.5., computer analysis of the human calbindin-D9k generevealed the presence of a sequence similar the rat ERE at position +50 from thetranscription initiation site. This element resides at the exon I-intron A borderand differs by only 2 essential nucleotides from the rat calbindin-D 9k ERE. Botharms of this motif have a C changed into a T residue. This particular sequencewas determined both from the PCR amplified intron A clone as well as thegenomic clone from the X-chromosomal library. The porcine intron bears a relatedsequence at the same location which has one further change of a T to A residue.Experiments were carried out to determine whether these sequences arestill able to bind the ER. The rat calbindin -D9k and Xenopus vitellogenin EREswere used as positive controls in gel retardation assays. A DNA/ER complex wasdetected with the EREs, but no such binding activity was observed with thehuman or porcine sequence. Next, it was examined whether one of the detectedsequence differences would suffice to eliminate ER binding. The mutantsgenerated with each of the two deviation were tested concomitantly with the othersequences. Both mutants showed binding activity. The retarded complexesmigrated comparable to those formed with the positive controls. In theseexperiments crude rat uterine cytosol was used as a source of ER. Previousreports have shown that the same complexes are obtained using purified ER orcrude cellular extract. It has also been demonstrated that the formation of thesecomplexes is enhanced after heat treatment of the receptor or cytosol (Kumar and176Chambon, 1988). This is thought to be caused by heat transformation of the ERincreasing its affinity for DNA. The two mutants used in the present experimentsalso showed enhanced binding of the transformed cytosol. The involvement of theER in all detected complexes was confirmed by the use of a rat ER antiserumcausing an additional retardation of these complexes. Future experiments withthe two constructs m ERE dr and dl have to verify whether they are in factfunctional as EREs. The weak receptor binding detected with these sequencesmay not allow enhancer activity on the promoter. This point can be addressed infuture transfection experiments with the two mutants.In the present experiment human and porcine sequences were analyzedusing a rat cytosol preparation as ER source. Therefore, it may be a concern thatthe rat ER cannot bind to a DNA element from a different species. Previousstudies have shown that the EREs from species as different as Xenopus, chickenand human are identical. This is due to the high conservation of the receptor'sDNA binding domain. The amino acid sequences of the human (Green et al.,1986) and rat (Koike et al., 1987) ERs were compared and found to bear nodifference. The pig ER sequence is unknown at present, however it is unlikelythat the DNA binding domain is any different from other species.The specific sequence requirements of a functional ERE have been studiedin different laboratories using point mutated elements (Kumar and Chambon,1771988; Ponglikitmongkol et al., 1990). None of the sequences, whether naturallyoccurring or artificial mutants matched the regions under consideration in thepresent experiments. Klein-Hitpass et al. (1988) suggested that a purine/purineor a pyrimidine/pyrimidine exchange in the conserved 13-base pair inverted repeatof the ERE would be permissible, but a purine/pyrimidine exchange was found todecrease the inducibility of a reporter gene. Both differences between the rat EREand the human motif are such pyrimidine/pyrimidine exchanges. A singlenucleotide change in the left arm of the rat ERE from an C to T reduced ERbinding substantially. These results suggest that since no natural ERE has a Tat this position, recognition of the DNA double helix by ER may be hindered bythe presence of a T residue at this position. Nevertheless, binding still occurs andit remains to be seen whether this correlates with enhancer activity. A singlenucleotide change in the right arm of the rat ERE from a C to T residue had onlya similar effect on ER binding. This switch occurs at the far 3' position of theERE.Recently, Savouret et al. (1991) have analyzed steroid responsive elementsof the rabbit progesterone receptor gene. Three binding sites for the ER weredetected. Two of those were in the 5' flanking region and one at +698 covering thetranslation start site. Transfection assays revealed that the downstream sequenceis in fact the major ERE controlling the expression of this gene. The twoupstream sequences could not render a construct responsive to estrogens in spite178of a strong binding affinity of one such element. The upstream element (GGTCANNN CGATT) which showed moderate binding affinity is similar to the right armmutant used in the present experiments (GGTCA NNN TGATT). The 3' terminalT residue in both motifs deviates from the rat ERE (GGTCA NNN TGATC). Themotif found in the progesterone receptor gene has one additional deviationcompared to the right arm mutant. Therefore, the right arm mutant would havetolerated one additional deviation from the rat ERE in terms of ER binding.The aim of these binding experiments was to determine whether the humanand porcine intron A's harbour a functional ERE as described for the ratcalbindin-D9k gene. Based on the mRNA expression analysis of human andporcine tissues, one might have predicted to find a functional ERE in the porcineand not in the human gene. It turned out however, that neither of the genescarried an ERE at the same location as the rat gene does. The data indicatinguterine expression of the porcine calbindin-D9k raise the question whether thereis an ERE in the 5'flanking region of this gene. Most EREs are actually foundbetween -300 and -2,000 upstream of regulated genes, although, in case of the ratcalbindin-D9k no ERE has been found in the upstream region up to -2,200. Itappears that the rat gene is regulated only through the intronic ERE. To addressthis issue in the porcine model a further analysis of the flanking region isnecessary. With regard to the human calbindin-D 9k gene it was expected to seeno functional ERE in the intron A location. Therefore, this finding is in179agreement with the lack of uterine expression in human. Also, there was no ERE-like sequence detected within the 1,324 by 5' flanking region as well as in theentire gene. There is still the possibility of an ERE located further upstream ofthe human gene. Although, in light of the mRNA data, one may conclude that thehuman calbindin-D9k gene has lost its responsiveness to estrogen throughdisruption of the regulatory site within the gene. In the case of the porcine genehowever, the present data suggest further studies probing the steroid hormoneregulation of the calbindin-D 9k gene.180V. SUMMARY AND CONCLUSIONSThe aim of this thesis was to characterize human calbindin-D 9k in terms ofits structure and regulation at the genomic level. Over the past 20 yearsnumerous studies have gathered information on this protein in variousmammalian species. The bulk of data was generated using a rat model. Theprimary structure at the protein and cDNA level, as well as the regulation at thegenomic level have been clarified to a great extent. At present, the precisefunction of calbindin-D 9k remains speculative. Calbindin-D 9k facilitates calciumabsorption in the intestine and is apparently is controlled by 1,25(OH) 2D3 .Calcium transfer across the placenta may also be related to calbindin-D 9kexpression in this tissue controlled by as yet unknown factors. The uterineexpression of calbindin-D9k is clearly dependent on 170-estradiol. A function of theprotein in endometrium and myometrium is purely speculative. The currenthypothesis is that this highly expressed calcium binding protein affectsmyometrial calcium concentrations and thereby influences uterine activity. Thefact that there is maximal expression of calbindin-D 9k at term may indicate thatthis particular function is most important at parturition.In human physiology many aspects of gene expression and function arequite similar to those in the rat. With regard to calbindin-D9k as a factor involvedin the vitamin D controlled calcium homeostasis, the two species appear to be181closely related. For the human placenta and uterus there was no informationrelating to calbindin-D 9k available at all. In the pig model intestinal and placentalcalbindin-D9k have been studied and appear to play a similar role as in the ratmodel. The experiments described in this thesis aimed at characterizing theanalogy between rat and human calbindin-D 9ks with regard to their structure androle in calcium metabolism. Analysis of placental and uterine calbindin-D 9kaddressed the question of whether expression and function are also similar to therat model. In particular, the potential role of uterine calbindin-D 9k duringpregnancy could have important implications for human gestation. This becomesevident in the high frequency of dysfunction of uterine contractility in human (andother primate) pregnancy. This pathophysiological condition is unique to human(and primate) pregnancy and often leads to idiopathic preterm labour. Theprematurely born infant faces severe health problems, some them lasting alifetime.Experiments of this thesis determined the complete primary structure ofthe calbindin-D9k cDNA in pig and human. The cDNAs were used to describe thetissue specific expression of the gene in both species. The transcription initiationsite for both genes was identified. A partial porcine and complete human genomicclone was isolated. The entire structure of the human gene was determined andfirst experiments were carried out to probe the regulation of expression.In summary, results of this thesis confirmed that both the human and pig182carry a homologous gene to the rat calbindin-D 9k. The structural homologies asdetermined at the cDNA and genomic level were as expected. The regulationhowever was not identical in both species. While in the pig expression ofcalbindin-D9k was found in uterus (present study), in humans calbindin-D 9k isabsent in this reproductive tissue. This represents an important aspectdemonstrating that the analogies to the rat model cannot be applied to humanphysiology in all tissues. The data provided here represent preliminary evidencethat the lack of calbindin-D9k expression in the human uterus (and placenta) isdue to a change in the steroid regulatory element of the gene. Furtherexperiments are needed to clarify steroid regulation of the human and porcinecalbindin-D9k genes to address this difference of the pig and human regulationdata.Furthermore, the human calbindin-D 9k gene was localized to the X-chromosome. This will allow future association of this gene with any geneticdisorder linked to a neighbouring region on this chromosome. Hypophosphatemicrickets is a genetic disorder mapped to Xp.22.1.-2. The genomic clone isolated inthe presented study has already been used by a collaborator in a localization blot.The results localized the human calbindin-D9k gene to the Xp22.2 to telomereregion. Therefore, an association of calbindin-D 9k with this disease cannot beruled out.183The most significant finding of these experiments is the lack of calbindin-D9k expression in human reproductive tissues (uterus and placenta). Theimplications for human physiology are to be determined. Obviously theseimplications are difficult to asses at present because the functions of placental anduterine calbindin-D9k are unknown. New experimental models are needed totackle the problem of functional analysis of this protein. These models mayinclude in vitro myometrial and placental tissues and transgenic animals. It isof interest to test whether other primates also lack calbindin-D 9k expression inreproductive tissues. There should be no technical problem to use the humancDNA probe isolated in this project to answer this question. The rhesus monkeyis a primate model used to study preterm labour and would provide theopportunity to include analysis of uterine calbindin-D9k under those experimentalconditions.184VI. REFERENCESAkerblom I and Mellon P 1991 Repression of gene expression by steroid andthyroid hormones, in Nuclear hormone receptrs, Parker, M., Ed., Academic press,New York, 175Allan M, Lanyon WG, Paul J 1983 Multiple origins of transcription in the 4.5 kbupstream of the e-globin gene. Cell 35:187Allore R, O'Hanlon D, Price R, Neilson K, Willard HF, Cox DR, Marks A, DunnRJ 1988 Gene encoding the subunit of 5100 protein is on chromosome 21:implications for Down Syndrome. Science 239:1311Alpers DH, Lee SW, Avioli LV 1972 Identification of two calcium binding proteinsin human small intestine. Gastroenterology 62:559Ariga H 1984 Replication of cloned DNA containing the Alu family sequenceduring cell extract-promoting simian virus 40 DNA synthesis. Mol Cell Biol 4:1476Armbrecht HJ, Boltz M, Strong R, Richardson A, Bruns MEH 1989 Christakos SExpression of calbindin-D decreases with age in intestine and kidney.Endocrinology 125(6):2950Arnold BM, Kovacs K, Murray TM 1976 Cellular localization of intestinal calciumbinding protein in pig duodenum. Digestion 14:77Arnold BM, Kuttner M, Wilis DM, Hitchman AJ, Harrison I, Murray TM 1975Radioimmunoassay studies of intestinal calcium binding protein in pig:distribution of intestinal calcium binding protein in pig tissues. Can J Pharmacol53:1135Babu YS, Sacks JS, Greenhough TJ, Bugg CE, Means AR, Cook WJ 1985 Threedimensional structure of calmodulin. Nature 315:37Baimbridge KG, Miller JJ, Parkes CO 1982 Calcium binding protein distributionin rat brain. Brain Res 239:1519Balmain N, Brehier A, Cuisinier-Gleizes P, Mathieu H 1986a Evidence for thepresence of Calbindin-D 28k (CaBP2Sk) in the tibial growth cartilages of rats. CellTissue Res 245:331Balmain N, Tisserand-Jochem E, Thomasset M, Cuisinier-Gleizes P, Mathieu H1986b Vitamin D-dependent calcium binding protein (CaBP 9k) in rat growthcartilage. Histochemistry 84:161185Bar A, Hurwitz S 1973 Uterine calcium binding protein in the laying fowl. CompBiochem Physiol 45A:579Bar A, Rosenberg J, Hurwitz S 1984 The lack of relationship between vitamin D3metabolites and calcium binding protein in the eggshell gland of laying birds.Comp Biochem Physiol 78:75Baudier J, Gerard D 1983 Ions binding to S-100 proteins: structural changesinduced by calcium and zinc on S-100a and S-1001:1 proteins. Biochem 22:3360Belanger LF 1957 The mineralization of rat enamel in the light of 'Caautoradiography and microincineration. J Dent Res 36:595Benoist C, O'Hare K, Breathnach R, Chambon P 1980 The ovalbumin gene-sequence of putative control regions. Nucleic Acids Res 8:127Berchtold MW, Epstein P, Means AR 1987 Structural organization and chromosalassignment of the parvalbumin gene. J Biol Chem 262:8696Berdal A, Balmain N, Brehier A, Thomasset M, Hotton D, Cuisinier-Gleizes P,Mathieu H, Developmental pattern and vitamin D-dependency of the calbindinsCaBP 9kDa and CaBP 28kDa in the ameloblasts of rodent teeth: animmunocytochemical study. In Abstracts of the 7th Workshop on Vitamin D,Rancho Mirage, CA, 1988, p 120Bortell R, Owen TA, Bidwell JP, Gavazzo P, Breen E, Wijnen JV, DeLuca HF,Stein JL, Lian JB, Stein GS 1992 Vitamin D-responsive protein-DNA interactionsat multiple promoter regulatory elements that contribute to the level of ratosteocalcin gene expression. Proc Natl Acad Sci USA 89:6119Brun P, Dupret JM, Perret C, Thomasset M, Mathieu H 1987 Vitamin D-dependent calcium-binding proteins (CaBPs) in human fetuses: comparativedistribution of 9K CaBP mRNA and 28K CaBP during development. Pediatr Res21(4):362Bruns ME, Burnett SH, Damjanov A, Damjanov I, Bruns DE 1988a The presenceof immunoreactive 9 kilodalton calbindin in mouse embryonal carcinoma cells. JBone Min Res 3[Suppl]: RS151 (Abstract 330)Bruns MEH, Fausto A, Avioli L 1978 Placental calcium binding protein in rats.J Biol Chem 253:3186186Bruns ME, Kleeman E, Bruns DE 1986 Vitamin D-dependent calcium bindingprotein in mouse yold sac: biochemical and immunochemical properties andresponses to 1,25 dihydroxycholecalciferol. J Biol Chem 261:7485Brims ME, Kleeman E, Mills SE, Bruns DE, Herr JC 1985 Immunocytochemicallocalization of vitamin D-dependent calcium binding protein in mouse placentaand yolk sac. Anat Rec 213:514Brims ME, Overpeck JC, Smith GC, Hirsch GN, Mills SE, Brims DE 1988bVitamin D-dependent calcium binding protein in rat uterus: differential effects ofestrogen, tamoxifen, progesterone and pregnancy on accumulation and cellularlocalization. Endocrinology 122:2371Bruns MEH, Vollmer S, Wallshein V, Bruns DE 1981 Vitamin D-dependentcalcium binding protein: immunochemical studies and synthesis by placentaltissue in vitro. J Biol Chem 256:4649Bruns MEH, Wallshein V, Bruns DE 1982 Regulation of calcium binding proteinin mouse placenta and intestine. Am J Physiol 243:E47Bucher P and Trifonov EN 1986 Compilation and analysis of eukaryotic Pol IIpromoter sequences. Nucleic acids Res 14:10009Calabretta B, Robberson DL, Maiwel AL, Saunders GF 1981 mRNA in humancells contains sequences complementary to the Alu family of repeated DNA. ProcNatl Acad Sci USA 78:6003Chapman M, Pond W, Taylor AN, Korkk L, Rivers J 1972 Calcium binding factorand rickets in guinea pig. J Anim Sci 35:283Chen PJ, Cywinski A, Taylor JM 1985 Reverse transcription of 7SL RNA by anavian retrovirus. J Viol 54:278Christakos S, Brunette MG, Norman AW 1981 Localization of immunoreactivevitamin D-dependent calcium binding protein in chick nephron. Endocrinology109:322Christakos S, Friedlander EJ, Frandsen BR, Norman AW 1979 Development andapplication of a radioimmunoassay for chick intestinal calcium binding proteinand tissue distribution. Endocrinology 104:1495Christakos S, Gabrielides C, Rhoten WB 1989 Vitamin D-dependent calciumbinding proteins: chemistry, distribution, functional considerations, and molecularbiology. Endocr Rev 10:3187Christakos S, Malkowitz L, Sori A, Sperduto A, Feldman SC 1987 Calciumbinding protein in squid brain: biochemical similarity to the 28,000 Mr vitaminD-dependent calcium binding protein. J Neurochem 49:1427Christakos S, Rhoten WB, Feldman SC 1987 Rat calbindin-D28k: purification,quantitation, immunocytochemical localization and comparative aspects. MethodsEnzymol 139:534Clark SA, Dame MC, Kim YS, Stumpf WE, DeLuca HF 1985 1,25-Dihydroxyvitamin D3 in teeth of rats and humans: receptors and nuclearlocalization. Anat Rec 212:250Corradino RA, Wasserman RH, Pubolos MH, Chang SI 1968 Vitamin D3 inductionof a calcium binding protein in the uterus of the laying hen. Arch BiochemBiophys 125:378Costa RH, Grayson DR, Xanthopoulos KG, Darnell JR, JE 1988 A liver-specificDNA-binding protein recognizes multiple nucleotide sites in regulatory regions oftransthyretin, al-antitrypsin, albumin, and simian virus 40 genes. Proc Natl AcadSci USA 85:3840Coty WA 1980 A specific high affinity binding protein for 1,25-dihydroxyvitaminD in chick oviduct shell gland. Biochem Biophys Res Commun 93:285Coty WA, Scordato JC, McConkey CL 1982 Estrogen regulates 1,25-dihydroxyvitamin D receptor levels in the chick oviduct shell gland. In: NormanAW, Schaefer K, Herrath DV, Grigoleit HG (eds) Vitamin D Chemical Biochemicaland Clinical Endocrinology of Calcium Metabolism. DeGruyter, Berlin, p 121Darwish H, DeLuca H 1992 Identification of a 1,25-dihydroxyvitamin D3-responseelement in the 5'-fanking region of the rat calbindin-9k gene. Proc Natl Acad SciUSA 89:603Darwish H, Krisinger J, Furlow JD, Smith C, Murdoch FE, DeLuca HF 1991 Anestrogen-response element mediates the transcriptional regulation of calbindin-D 9kgene in rat uterus. J Biol Chem 266:551Darwish HM, Krisinger J, Strom M, DeLuca HF 1987 Molecular cloning of thecDNA and chromosomal gene for vitamin D-dependent calcium-binding protein ofrat intestine. Proc Natl Acad Sci USA 84:6108Davis WL, Jones RG 1981 Calcium lysosome in rachitic and vitamin D3 repletechick duodenal absorptive cells. Tissue Cell 13:381188Dechesne CJ, Escudero P, Lamande N, Thomasset M, Sans A 1987Immunohistochemical identification of neuron-specific enolase and calbindin investibular receptors of human fetuses. Acta Otolaryngol [Suppl] (Stockh) 436:69Delorme AC, Cassier P, Geny B, Mathieu H 1983 Immunocytochemical localizatonof vitamin D-dependent calcium binding protein in the yolk sac of the rat.Placenta 4:263Delorme AC, Dana JL, Acker MG, Ripoche MA, Mathieu H 1983 In rat uterus173-estradiol stimulates a calcium binding protein similar to duodenal vitamin D-dependent calcium binding protein. Endocrinology 113:1340Delorme AC, Danan JL, Mathieu H 1983 Biochemical evidence for the presenceof two vitamin D-dependent calcium binding proteins in mouse kidney. J BiolChem 258:1878DeLuca HF,Schnoes HK 1984 Vitamin D: metabolism and mechanism of action.Annu Rep Med Chem 19:179Demay MB, Gerardi JM, DeLuca HF, Kronenberg HM 1990 DNA sequences in therat osteocalcin gene that bind the 1,25-dihydroxyvitamin D3 receptor and conferresponsiveness to 1,25 dihydroxyvitamin D3. Proc Natl Acad Sci USA 87:369Desplan C, Thomasset M, Moukhtar MS 1983 Synthesis, molecular cloning andrestriction analysis of DNA complementary to vitamin D-dependent calciumbinding protein mRNA from rat duodenum J Biol Chem 258:2762DiFiglia M, Christakos S, Aronin N, Ultrastructural localization of calbindin-D 28kin rat and monkey basal ganglia including subcellular distribution with colloidalgold labeling. J Comp Neurol, in pressDrescher D, DeLuca HF 1971 Vitamin D stimulated calcium binding protein fromrat intestinal mucosa: purification and some properties. Biochemistry 101:2302Duncan CH, Jagadeeswaran P, Wang RRC, Weissman SM 1981 Structuralanalysis of templates and RNA polymerase III transcripts of Alu family sequencesinterspersed among the human (3-like globin genes. Gene 13:185Dupret J, Brun P, Perret C, Lomri N, Thomasset M, Cuisinier-Gleizes P 1987Transcriptional and posttranscriptional regulation ofvitamin D-dependent calciumbinding protein in rat duodenum by 1,25 dihydroxycholecalciferol. J Biol Chem262:16553189Eberling PR, Sandgren ME, DiMangno EP, Lane AW, DeLuca HF, Riggs L 1992Evidence of an age-related decrease in intestinal responsiveness to vitamin D:relationship between serum 1,25-dihydroxyvitamin D3 and intestinal vitamin Dreceptor concentrations in normal women. Endocrinology 75(1):176Elms TN, Taylor AN 1987 Calbindin-D 28k localization in rat molars duringodentogenesis. J Dent Res 66:1431Elsholtz H, Albert VR, Treacy MN, Rosenfeld MG 1990 A two-base change in aPOU factor-binding site switches pituitary-specific to lymphoid-specific geneexpression. Genes Dev 4:43Emtage JS, Lawson DEM, Kodicek E 1974 The response of the small intestine tovitamin D. Isolation and properties of chick intestinal polyribosomes. Biochem J140:239Evans T, Schon E, Gora-Maslak G, Patterson J, Efstratiadis A 1984 S1-hypersensitive sites in eukaryotic promoter regions. Nucleic Acids Res 12:8043Feher JJ 1983 Facilitated calcium diffusion by intestinal calcium binding protein.Am J Physiol 244:C303Feldman SC, Christakos S 1983 Vitamin D-dependent calcium binding protein inrat brain: biochemical and immunocytochemical characterization. Endocrinology112:290Ferrai SB, Calabretta B, de Riel JK, Battini R, Ghezzo F, Lauret E, Griffin C,Emanuel BS, Gurrieri F, Baserga R 1987 Structural and functional analysis of agrowth-regulated gene, the human calcyclin. J Biol Chem 262:8325Fontaine 0, Matsumoto T, Goodman DBP, Rasmussen H 1981 Liponomic controlof Ca' transport: relationship to mechanisms of action of 1,25 dihydroxyvitaminD3. Proc Natl Acad Sci USA 78:1751Freedman RA, MacLaughlin JD, Weiser MM 1981 Properties of Ca' uptake andrelease by Golgi membrane vesicles from rat intestine. Arch Biochem Biophys206:233Fullmer CS, Wasserman RH 1975 Isolation and partial characterization ofintestinal calcium binding protein from cow, pig, horse, guinea pig and chick.Biochim Biophys Acta 393:134Fullmer CS, Brindak ME, Bar A, Wasserman RH 1976 The purification of calciumbinding protein from the uterus of the laying hen. Proc Soc Exp Biol Med 152:237190Fullmer CS, Wasserman RH 1981 The amino acid sequence of bovine intestinalcalcium binding protein. J Biol Chem 256:5669Fullmer CS, Wasserman RH 1985 Vitamin D induced intestinal calcium bindingprotein from chick: tentative amino acid sequence. In: Norman AW, Schaefer K,Grigolet HG, Herrath DR (eds) Vitamin D: Chemical, Biochemical and ClinicalUpdate. deGruyter, New York, p 323Fullmer CS, Wasserman RH 1987 Chicken intestinal 28-kilodalton calbindin-D:complete amino acid sequence and structural considerations. Proc Natl Acad SciUSA 84:4772Garabedian M, Tanaka Y, Holick MF 1974 Response of intestinal calciumtransport and bone calcium mobilization to 1,25-dihydroxyvitamin D3 inthyroparathyroidectomized rats. Endocrinology 94:1022Garel JM 1983 Parathyroid hormone, calcitonin and mineral metabolism inmammalian fetus and neonate. In: Holick M, Anast C, Gray T (eds) PerinatalCalcium and Phosphorus Metabolism. Elsevier, Amsterdam, p 71Garel JM, Delorme AC, Marche P, Nguyen TM, Garabedian M 1980 Vitamin D3metabolite injections to thyroparathyroidectomized pregnant rats: Effects oncalcium-binding proteins of maternal duodenum and of fetoplacental unit.Endocrinology 109:284Glisin V, Crkvenjakov R, Byus C 1974 Ribonucleic acid isolated by cesium chloridecentrifugation. Biochem 13:2633Glorieux FH and Scriver CR 1972 Loss of parathyroid hormone-sensitivecomponent of phosphate transport in X-linked hypophosphatemia. Science 173:997Gona A, Pendurthi TK, Al-Rabiai S, Gona 0, Christakos S 1986Immunocytochemical localization and immunological characterization of vitaminD-dependent calcium binding protein in bull-frog cerebellum Brain Behav Evol29:176Green S, Walter P, Kumar V, Krust A, Bornert JM, Argos P, Chambon P 1986Human oestrogen receptor cDNA: sequence, expression and homology to v-erb-A.Nature 320:134Hall AK, Bishop JE, Norman AW 1987 Inhibitory and stimulatory effects ofdexamethasone and 1,25 dihydroxyvitamin D3 on chick intestinal calbindin-D28kand its mRNA. Mol Cell Endocrinology 51:25191Halloran BP, DeLuca HF 1981 Effect of vitamin D deficiency on skeletaldevelopment during early growth in the rat. Arch Biochem Biophys 209:7Herzberg 0, James MNG 1985 Structure of the calcium regulatory muscle proteintroponin C at 2.3 A resolution. Nature 313:653Hess A, Hess A (eds), The history of rickets. in "Rickets, including osteomalaciaand tetany" Lea & Febiger Philadelphia, 1929, 22Hess JF, Fox M, Schmid C, Shen CKJ 1983 Molecular evolution of the humanadult a-globin-like gene region:insertion and deletion of Alu family repeats andnon-Alu DNA sequences. Proc Natl Acad Sci USA 80:5970Hitchman AJW, Kerr MK, Harrison JE 1973 The purification of pig vitamin Dinduced intestinal calcium binding protein. Arch Biochem Biophys 155:221Hofman T, Kawakami M, Hitchman AJ, Harrison JE, Dorrington KJ 1979 Theamino acid sequence of porcine intestinal calcium binding protein. Can J Biochem57:737Hopp TP, Woods KR 1981 Prediction of protein antigenic determinants fromamino acid sequences. Proc Natl Acad Sci USA 78:3824Howard A, Legon S, Spurr NK, Waters JRF 1992 Molecular cloning andchromosomal assignment of human calbindin-D9k. Biochem Biophys Res Commun185(2):663Huang YC, Lee S, Stolz R, Gabrielides C, Pansini-Porta A, Bruns ME, Bruns D,Mifflin T, Pike JW, Christakos S 1989 Effect of hormones and development on theexpression of the rat 1,25 dihydroxyvitamin D3 receptor gene and comparison tocalbindin gene expression. J Biol Chem 264:17454Hunziker W 1986 The 28 kDa vitamin D-dependent calcium binding protein as asix domain structure. Proc Natl acad Sci USA 83:7578Hunziker W, Schrickel S 1988 Rat brain calbindin-D 28k: six domain structure andextensive amino acid homology with chicken calbindin-D 28k. Mol Endocrinol 2:465Hunziker W, Siebert P, King M, Stucki P, Dugaiczyk A, Norman AW 1983Molecular cloning of a vitamin D-dependent calcium binding protein mRNAsequence from chick intestine. Proc Natl Acad Sci USA 80:4228192Jagadeeswaran P, Than D, Forget BG, Weissman SM 1982 A gene deletion endingat the midpoint of a repetive DNA sequence in one form of hereditary persistenceof fetal haemoglobin. Nature 296:469Jande SS, Maler L, Lawson DEM 1981a Immunohistochemical mapping of vitaminD-dependent calcium binding protein in brain. Nature 294:765Jande SS, Tolnai S, Lawson DEM 198 lb Immunohistochemical localization ofvitamin D-dependent calcium binding protein in duodenum, kidney, uterus andcerebellum of chicken. Histochemistry 71:99Jelinek WR, Toomey TP, Leinwand L, Duncan CH, Biro PA, Choudary PV,Weissman SM, Rubin CM, Houck CM, Deininger PL, Schmid CW 1980Ubiquitous, interspersed repeated sequences in mammalian genomies.Proc NatlAcad Sci USA 77:1398Jeung EB, Krisinger J, Dann JL, Leung PCK 1992a Molecular cloning of the full-length cDNA encoding the human calbindin-D9k. FEBS lett 307(2):224Jeung EB, Krisinger J, Dann JL, Leung PCK 1992b Cloning of the porcinecalbindin-D9k complementary deoxyribonucleic acid by anchored polymerase chainreaction technique. Biol Reprod 47:502Kallfelz FA, Wasserman RH 1972 Correlation between 'Ca absorption andintestinal calcium binding activity in the golden hamster. Proc Soc Exp Biol Med125:54Kallfelz FA, Taylor AN, Wasserman RH 1967 Vitamin D-induced calcium bindingfactor in rat intestinal mucosa. Proc Soc Exp Biol Med 125:54Kariya Y, Kato K, Hayashizaki Y, Himeno S, Tarui S, Matsubara K 1987 Revisionof consensus sequence of human Alu repeats - a review. Gene 53:1Kato S, Tora L, Ymauchi J, Masuhige S, Bellard M, Chambon P 1992 A farupstream estrogen response element of the ovalbumin gene contains several half-palindromic 5'-TGACC-3' motifs acting synergistically. Cell 68:731Kessler MA, Lamm L, Jarnagin K, DeLuca HF 1986 1,25 Dihydroxy-vitamin D3stimulated mRNAs in rat small intestine. Arch Biochem Biophys 251:403Kim YS, Clark SA, Stumpf WE, DeLuca HF 1985 Nuclear uptake of 1,25dihydroxyvitamin D3 in developing rodent teeth. An autoradiographic study. AnatRec 212:301193Kim YS, Stumpf WE, Clark SA, Sar M, DeLuca HF 1983 Target cells for 1,25-dihydroxyvitamin D3 in developing rat incisor teeth. J Dent Res 62:58Kimura M 1983 The neutral theory of evolution. Cambridge Univ Press,Cambridge, pp20.King M, Norman AW 1986 Analysis of the mRNA coding for the chick vitamin Dinduced calbindin and its regulation by 1,25 dihydroxyvitamin D3. Arch BiochemBiophys 258:612Klein-Hitpass L, Kaling M, Ryffel GU 1988 Synergism of closely adjacentestrogen-responsive elements increases their regulatory potential. J Mol Biol201:537Koike S, Sakai M, Muramatsu M 1987 Molecular cloning and characterization ofrat estrogen receptor cDNA. Nucleic Acids Res 15: 2499Kowarski S, Cowen LA, Takahashi MT, Schachter D 1987 Tissue distribution andvitamin D dependence of IMCAL in the rat. Am J Physiol 253:6411Kowarski S, Schachter D 1975 Vitamin D-dependent particulate calcium bindingactivity and intestinal calcium transport. Am J Physiol 229:1198Krisinger J, Darwish H, Maeda N, DeLuca HF 1988 Structure and nucleotidesequence of the rat intestinal vitamin D-dependent calcium binding protein gene.Proc Natl Acad Sci USA 85:8988Krisinger J, Dann JL, Currie WD, Jeung EB, Leung PCK 1992a Calbindin-D9kmRNA is tightly regulated during the estrous cycle in the rat uterus. Mol CellEndocrinol 86:119Krisinger J, Dann JL, Jeung EB, Leung PCK 1992b Calbindin-D 9k gene expressionduring pregnancy and lactation in the rat. Mol Cell Encrinol 88:119Kretsinger RH, Mann JE, Simmonds JG. Model of facilitated diffusion of calciumby the intestinal calcium binding protein. In: Vitamin D: Chemical, Biochemicaland clinical Endocrinology of Calcium Metabolism, edited by A.W. Norman, K.Schaefer, D.V. Herrath, and H.-G. Grigoleit. New York: de Gruyter, 1982, p. 233-248.Kretsinger RH, Nockolds CE 1973 Carp muscle calcium binding protein: structuredetermination and general description. J Biol Chem 248:3313194Kubinoff MJ, Nellans NH 1985 Active calcium sequestration by intestinalmicrosomes. J Biol Chem 260:7824Kumar V, and Chambon P 1988 The estrogen receptor binds tightly to itsresponsive element as a ligand-induced homodimer. Cell 55:145Kumar R, Wieben R, Beecher SJ 1989 The molecular cloning of the cDNA forbovine vitamin D-dependent calcium-binding protein: structure of the full-lengthprotein and evidence for homologies with other calcium-binding proteins of thetroponin-C superfamily of proteins. Mol Endocrinol 3:427Lee YS, Taylor AN, Reimers TJ, Edelstein S, Fullmer CS, Wasserman RH 1987Calbindin in peripheral nerve cells is vitamin D dependent. Proc Natl Acad SciUSA 84:7344Legrand C, Brehier A, Clavel MC, Thomasset M, Rabie A 1988 Cholecalcin (28kDa CaBP) in rat cochlea: development in normal and hypothyroid animals: animmunocytochemical study. Brain Res 466:121Leonard WJ, Strauss SH, Go MF, Alpers DH, Gordon a 1984 Synthesis andcompartmentalization of rat intestinal vitamin-D-dependent calcium-bindingprotein. Eur J Biochem 139:561L'Horset A, Perret C, Brehier A, Thomasset M 1990 1713-estradiol stimulates thecalbindin-D9k (CaBP9k) gene expression at the transcriptional andposttranscriptional levels in the rat uterus. Endocrinology 127:2891Li H and Christakos S 1991 Differential regulation by 1,25 dihydroxyvitamin D3of calbindin-D9k and calbindin-D 28k gene expression in mouse kidney.Endocrinology 128:2844Lippiello L, Wasserman RH 1975 Fluorescent antibody localization of the vitaminD-dependent calcium binding protein in the oviduct of the laying hen. J HistochemCytochem 23:111Lubon HK and Hennighausen L 1987 Nuclear proteins from lactating mammaryglands bind to the promoter of a milk protein gene. Nucleic Acids Res 15(5):2103MacManus JP, Watson DC, Yaguchi M 1986 The purification and complete aminoacid sequence of the 9000 Mr Ca" binding protein from rat placenta: identity withthe vitamin D-intestinal Ca' binding protein. Biochem J 235:585195Machler M, Frey D, Gai A 1986 X-linked dominant hypophosphatemia is closelylinked to DNA markers DXS41 and DXS43 at Xp22. Hum Genet 73:271Manley JL 1988 Polyadenylation of mRNA precusors. Biochim Biophysic Acta950:1Marche P, DeLorme A, Cuisinier-Gleizes P 1978 Intestinal and placental calciumbinding protein in vitamin D-deprived or supplemented rats. Life Sci 23:2555Markose ER, Stein JL, Stein GS, Lian JB 1990 Vitamin D-mediated modificationin protein-DNA interactions at two promoter elements of the osteocalcin gene. ProNatl Acad Sci USA 87:1701Mathieu CL, Burnett SA, Mills SE, Overpeck JG, Bruns DE, Bruns ME 1988Gestational changes in calbindin-D 9k in rat uterus, yolk sac and placenta:implications for maternal-fetal calcium transport and uterine muscle fucntion.Proc Natl Acad Sci USA 86:3433Matsumoto T, Fontaine 0, Rasmussen H 1981 Effect of 1,25 dihydroxyvitamin D3phospholipid metabolism in chick duodenal mucosal cell. J Biol Chem 255:10834McLauchlan J, Gaffney D, Whitton JL, Clements JB 1985 The consensus sequenceYGTGTTYY located downstream from the AATAAA signal is required for efficientformation of mRNA 3'termini. Nucleic Acids Res 13(4):1347Minghetti PP, Cancela L, Fujisawa G, Theofan G, Norman AW 1988 Molecularstructure of the chicken vitamin D-induced calbindin-D 28k gene reveals elevenexons, six Ce-binding domains and numerous promoter regulatory elements. MolEndocrinol 2:355Mitchell PJ, Wang C, Tjan R 1987 Positive and negative regulation oftranscription in vitro: enhancer-binding protein AP-2 is inhibited by SV40 Tantigen. Cell 50:847Miwa T, Boxer LM, Kedes L 1987 CArG boxes in the human cardiac a-actin geneare core binding sites for positive trans-acting regulatory factors. Proc Natl AcadSci USA 84:6702Moews PC, Kretsinger RH 1975 Refinement of the structure of carp musclecalcium binding protein parvalbumin by model building and difference Fourieranalysis. J Mol Biol 91:201Morrisey RL, Zolock DT, Bucci TJ, Bikle DD 1978 Immunoperoxidase localizationof vitamin D-dependent calcium binding protein. Histochem Cytochem 26:268196Moue PD, Holt M, Brehier A, Lomri A, thomasset M 1988 Immunocytochemicaldemonstration of 9k and 28k vitamin D dependent calcium binding protein inhuman and rat oteoblastic cells in culture. J Bone Mineral Res 3[Supp1]:5152(Abstract 336)Mount SM 1982 A catalogue of splice junction sequence. Nucleic Acids Res10(2):459Narbaitz R, Stumpf WE, Sar M, DeLuca HF 1982 The distal nephron in the chickembryo as a target tissue for 1, alpha 25 dihydroxycholecalciferol. Acta Anat(Basel) 112:208Navickis RJ, Katzenellenbogen BS, Nalbandov AV 1979 Effects of the sex steroidhormones and vitamin D3 on calcium binding protein in the chick shell gland. BiolReprod 21:1153Nemere I, Leathers V, Norman AW 1986 1,25-dihydroxyvitamin D3 mediatedintestinal calcium tranport. J Biol Chem 161:16106Nikol JM and Felsenfeld G 1983 DNA conformation at the 5'end of the chickenadult 13-globin gene. Cell 35:467Noda M, Vogel RL, Crag AM, Prahl J, DeLuca HF, Denhardt DT 1990Identification of a DNA sequence responsible for binding of the 1,25-dihydroxyvitamin D3 receptor and 1,25-dihydroxyvitamin D3 enhancement ofmouse secreted phosphoprotein 1 (SPP-1 or osteopontin) gene expression. ProcNatl Acad Sci USA 87:9995Nussinov R, Owenes J, Maizel JV 1986 Sequence signals in eukaryotic upstreamregions. Biochim Biophys Acta 866:109Oberholtzer JC, Baettger C, Summers MC, Matschinsky FM 1988 The 28 kDacalbindin is a major calcium binding protein in the basilar papilla of the chick.Proc Natl Acad Sci USA 85:3387Ohara 0, Dorit RL, Gilbert W 1989 One-sided polymerase chain reaction: theamplification of cDNA. Proc Natl Acad Sci USA 86:5673Ohkuma Y, Horikoshi M, Roeder RG, Desplan C 1990 Engrailed, a homeodomainprotein, can repress in vitro transcription by competition with the TATA box-binding protein transcription factor IID. Proc Natl Acad Sci USA 87:2289Ozono K, Liao J, Kerner SA, Scott RA, Pike JW 1990 The vitamin D-responsiveelement in the human osteocalcin gene. J Biol Chem 265:21881197Padgett RA, Grabowski PJ, Konarska MM, Seller S, Sharp PA 1986 Splicing ofmessenger RNA precursors. Annu Rev Biochem 55:1119Pan LC, Price PA 1984 The effect of transcriptional inhibitors on the bone gamma-carboxyglutamic acid protein response to 1,25(OH)2D3 in osteosarcoma cells. J BiolChem 259:5844Parmentier M, DeVijlder JJM, Muir E, Szirer C, Islam MQ, Geurts van Kessel A,Lawson DEM, Vassart G 1989 The human calbindin 27-kDa gene: Structuralorganization of the 5' and 3' regions, chromosome assignment, and restrictionfragment length polymorphism. Genomics 4:309Parmentier M, Ghysens M, Rypen F, Lawson DEM, Pasteels JL, Pochet R 1987Calbindin in vertebrate classes• immunohistochemical localization and westernblot analysis. Gen Comp Endocrinol 65:399Parmentier M, Lawson DEM, Vassart G 1987 Human 27 kDa calbindincomplementary DNA sequence. Eur J Biochem 170:207Parmentier M, Szipirer J, Levan G, Vassart G 1989 The human genes forcalbindin 27 and 29 kDa proteins are located on chromosomes 8 and 16,respectively. Cytogenet Cell Genet 52:85Pasteels B, Parmentier M, Lawson DEM, Verstappen A, Pochet R 1987 Calciumbinding protein immunoreactivity in pigeon retina. Invest Ophthalmol Vis Sci28:658Pasteels JL, Pochet R, Surardt L, Hubeau C, Chirnoaga M, Parmentier M, LawsonDEM 1986 Ultrastructural localization of brain vitamin D-dependent calciumbinding protein. Brain Res 384:294Perret C, Desplan C, Brehier A, Thomasset M 1985 Characterization of rat 9-kDacholecalcin (CaBP) messenger RNA using a complementary DNA. Absence ofhomology with 28-kDa cholecalcin mRNA. Eur J Biochem 148:61Perret C, Desplan C, Thomasset M 1985 Cholecalcin (a 9k Da cholecalciferolinduced calcium binding protein) mRNA distribution and induction by calcitriolin the digestive tract. Eur J Biochem 150:211Perret C, L'Horset F, Thomasset M 1991 DNase I-hypersensitive sites areassociated, in a tissue-specific manner, with expression of the calbindin-D9k-encoding gene. Gene 108:227198Perret C, Lomri N, Gouhier N, Auffray C, Thomasset M 1988 The rat vitamin D-dependent calcium binding protein (9-kDa CaBP) gene: complete nucleotidesequence and structural organization. Eur J Biochem 172:43Perret C, Lomri N, Thomasset M 1990 Evolution of the "EF" family of calcium-binding proteins. Adv Exp Med Biol 269:17Pike J, Gooze LJ, Haussler MR 1980 Biochemical evidence for 1,25,(OH) 2D3receptor macromolecule in parathyroid, pancreatic, pituitary and placental tissue.Life Sci 26:407Ponglikitmongkol M, White JH, Chambon P 1990 Synergistic activation oftrnscription by the human estrogen receptor bound to tandem reponsive elements.EMBO J 9:2221Price PA, Baukol SA 1981 1,25-Dihydroxyvitamin D3 increases serum levels of thevitamin K dependent bone protein. Biochem Biophys Res Commun 99:928Puschett JB, Fernandez PC, Boyle ET, Gray RW, Omdahl JL, DeLuca HF 1972The acute renal tubular effects of 1,25 dihydroxyvitamin D3. Proc Soc Exp BiolMed 141:379Puschett JB, Morenz I, Kurnick WS 1972 Evidence for a direct action ofcholecalciferol and 25-hydroxycholecalciferol on the renal transport of phosphate,sodium and calcium J Clin Invest 51:373Rabie A, Thomasset M, Legrand C 1983 Immunocytochemical detection of acalcium binding protein in the cochlear and vestibular hair cells of the rat. CellTissue Res 232:691Rabie A, Thomasset M, Parkes CO, Clavel MC 1985 Immunocytochemicaldetection of 28,000 MW calcium binding protein in horizontal cells of the ratretina. Cell Tissue Res 240:493Raisz LG, Trammel CL, Holick MF 1972 1,25 dihydroxycholecalciferol: a potentstimulator of bone resorption in tissue culture. Science 175:768Read AP, Thakker RV, Davies KE 1986 Mapping of human X-linkedhypophosphatemic rickets by multilocus linkage analysis. Hum Genet 73:267Reynolds JJ, Holick MF, DeLuca HF 1974 The effects of vitamin D analogoes onbone resorption. Calif Tissue Res 15:333199Rhoten WB, Bruns ME, Christakos S 1985 Presence and localization of twovitamin D-dependent calcium binding proteins in kidneys of higher vertebrates.Endocrinology 117:674Rhoten WB, Christakos S 1981 Immunocytochemical localization of vitamin D-dependent calcium binding protein in mammalian nephron. Endocrinology 109:981Rhoten WB, Gona 0, Christakos S 1986 Calcium binding protein (28,000 mrcalbindin-D28k) in kidneys of the bullfrogRana catesbeiana during metamorphosis.Anat Rec 186:127Rhoten WB, Lubit B, Christakos S 1984 Avian and mammalian vitamin D-dependent calcium-binding protein in reptilian nephron. Gen Comp Endocrinol55:96Riggle PC, Bruns ME, Walters MR, Immunoblot analysis of 'Ca binding activitiesat 9KDa in rat lung and prostate. In Abstracts of the 7th Workshop on VitaminD, Rancho Mirage, CA, 1988, p 114Robertson HD and Dickson E 1984 Structure and distribution of Alu familysequences or their analogs within heterogeneous nuclear RNA of HeLa KB and Lcells. Mol Cell Biol 4:310Rogers J 1985 Oncogene chromosome breakpoints and Alu sequence. Nature317:559Roman A, Brisson P, Pasteels B, Demol S, Pochet R, Collen JP 1988 Pineal-retinalmolecular relationships; immunocytochemical evidence of calbindin-D 27ka inpineal transducers. Brain Res 442:33Roth J, Baeten D, Norman AW, Garcia-Segura LM 1981 Specific neurons in chickcentral nervous system stain with an antibody against intestinal vitamin D-dependent calcium binding protein. Brain Res 222:452Roth J, Brown D, Norman AW, Orci L 1982 Localization of the vitamin D-dependent calcium binding protein in mammalian kidney Am J Physiol 243:F243Roth J, Thorens B, Hunziker W, Norman AW, Orci L 1981 Vitamin D-dependentcalcium binding protein: immunocytochemical localization in chick kidney. Science214:197Sampon HW, Matthews DL, Martin JH, Kunin AS 1970 An electron microscopiclocalization of calcium in the small intestine of normal, rachitic and vitamin D-treated rats. Calcif Tissue Res 5:305200San A, Etchecopar B, Brehier A, Thomasset M 1986 Immunocytochemicaldetection of vitamin D-dependent calcium binding protein (CaBP 28k) in vestibularsensory hair cells and vestibular ganglion neurons of the cat. Brain Res 364:190Sanger F, Nicklen S, Coulson AR 1977 DNA sequencing with chain-terminatinginhibitors. Proc Natl Acad Sci USA 74:5463Satyshur KA, Rao ST, Pyzalska D, Drendel W, Greaser M, Sundaralingam M 1988Refined structure of chicken skeletal muscle troponin C in the two-calcium stateat 2 °A ressolution. J Biol Chem 263:1628Savouret JF, Baily A, Misrahi M, Rauch C, Redeuih G, Chauchereau A, MilgromE 1991 Characterization of the hormone responsive element involved in theregulation of the progesterone receptor gene 10(7):1875Schneeberger PR, Norman AW, Heizmann CW 1985 Parvalbumin and vitamin D-dependent calcium binding protein (Mr 28,000): comparison of their localizationin normal and ruchitic rats. Neurosci Lett 59:97Sharp PA 1983 Conversion of RNA to DNA in mammals:Alu-like elements andpseudogenes. Nature 301:471Shimura F, Wasserman RH 1984 Membrane associated vitamin D-induced calciumbinding protein (CaBP): quantification by a radioiramunoassay and evidence fora specific CaBP in purified intestinal brush borders. Endocrinology 115:1964Schreiner DS, Jande SS, Lawson DEM 1985 Target cells of vitamin D in theretina. Acta Anat (Basel) 121:153Schreiner DS, Jande SS, Parkes CO, Lawson DEM, Thomasset M 1983Immunocytochemical demonstration of two vitamin D-dependent calcium bindingproteins in mammalian kidney. Acta Anat (Basel) 117:1Scriver CR 1974 Rickets and the pathogenesis of impaired tubular transport ofphosphate and other solutes. Am J Med 57:43Sebrell WH, Harris RS (eds) in "The vitamins", Academic, NY, 1954, volII:131Sonnenberg J, Luine VL, Krey L, Christakos S 1986 1,25 Dihydroxyvitamin D3treatment results in increased choline acetyltransferase activity in specific brainnuclei. Endocrinology 118:1433201Sonnenberg J, Pansini AR, Christakos S 1984 Vitamin D-dependent rat renalcalcium binding protein: development of a radioimmunoassay tissue distributionand immunological identification. Endocrinology 115:640Staun M 1986 Enzyme-linked immunoadsorbent assay (ELISA) for calcium-binding protein of human small intestine. A quantitative method for measuringcalcium-binding protein in small intestinal biopsies. Clin Chem Acta 159:239Staun M, Boesby S, Daugaard H, Jarnum S 1988 Calcium-binding protein inhuman duodenal biopsies. Calcif Tissue Int 42:205Staun M, Paerregaard A, Krasilnikoff PA 1991 Measurement of calbindin-D 9k insmall intestinal biopsy specimens of children. J Ped Gastr Nutr 12:328Stumpf WE, Sar M, Narbaitz R, Ried FA, DeLuca HF, Tanka Y 1980 Cellular andsubcellular localization in 1,25,(OH)2D3 in rat kidney: comparison with localizationof parathyroid hormone and estradiol. Proc Natl Acad Sci USA 77:1149Stumpf WE, Sar M, Reid FA, Tanaka Y, DeLuca HF 1979 Target cells for 1,25-dihydroxyvitamin D3 in intestinal tract, stomach, kidney, skin, pituitary andparathyroid. Science 209:1188Stern PH, Mavreas T, Trummel CL 1976 Bone-resorbing of analogues of 25-hydroxycholecalciferol and 1,25-dihydroxycholecalciferol: effects of side chainmodification and stereoisomerization on responses of fetal rat bones in vitro. MolPharmacol 12:879Suda S, Takahashi N, Shinki T, Horiuchi N, Yamaguchi A, Yoshiki S, EnomatoS, Suda T 1985 1,25 Dihydroxyvitamin D3 receptors and their action in embryonicchick. Calcif Tissue Int 37:82Sundarlinggam M, Bergstrom P, Strasburg G, Rao ST, Rarychowdhury P, GreaserM, Wang BC 1985 Molecular structure of troponin C from chicken skeletal muscleat 3-angstrom resolution. Science 227:945Swain A, Anima S, Kretsinger RH 1989 Restrained least square refinement ofnative (calcium) and cadmium-substituted carp parvalbumin using X-raycrystallographic data to 1.6 °A resolution. J Biol Chem 261:8761Szebenyi DME, Abendorf SK, Moffat K 1981 Structure of vitamin D-dependentcalcium-binding protein from bovine intestine. Nature 294:327Szebenyi DME, Moffat K 1986 The refined structure of vitamin D-dependentcalcium binding protein from bovine intestine. J Biol Chem 261:8761202Szebenyi DME, Obendorf SK, Moffat K 1981 Structure of vitamin D-dependentcalcium binding protein from bovine intestine. Nature 294:327Takagi T, Najiri M, Konishi K, Maruyama K, Nonomura 1986 Amino acidsequence of vitamin D-dependent calcium binding protein from bovine cerebellum.FEBS Lett 201:41Taylor AN 1981 Immunocytochemical localization of vitamin D-induced calciumbinding protein. J Histochem Cytochem 29:65Taylor AN 1984 Tooth formation in the 28,000 dalton vitamin D-dependentcalcium binding protein: an immunochemical study. J Histochem Cytochem 32:159Taylor AN, Gleason WA, Lankford GL 1984a Immunocytochemical localization ofrat intestinal vitamin D dependent calcium binding protein. J HistochemCytochem 32:153Taylor AN, Gleason WH, Lankford GL 1984b Rat intestinal vitamin D-dependentcalcium binding protein immunocytochemical localization in incisor ameloblasts.J Dent Res 63:94Taylor AN, Inpanbutr N 1988 Vitamin D mediated intestinal calcium transport:ultrastructural distribution of calbindin-D28k following calcitriol. In: Bronner F,Peterlik M (eds) Cellular Calcium and Phosphate Transport in Health andDisease. Alan R Liss Inc, New York, p 109Taylor AN, Inpanbutr N, Deftos LJ 1987 Localization of calbindin-D 28k incalcitonin containing cells of chick ultimobranchial glands. Anat Rec 219:86Taylor AN, McIntosh JE, Bordeau JE 1982 Immunocytochemical localization ofvitamin D-dependent calcium binding protein in renal tubules of rabbit, rat andchick. Kidney Int 21:765Taylor AN, Wasserman RH, Jowsey J 1968 A vitamin D dependent calciumbinding protein in canine intestinal mucosa. Fed Proc 27:675Theofan G, Nguyen AP, Norman AW 1986 Regulation of calbindin-D 28k geneexpression by 1,25 dihydroxyvitamin D3 is correlated to receptor occupancy. J BiolChem 261:16943Thomasset M, Parkes CO, Cuisinier-Gleizes P 1982 Rat calcium binding protein:distribution, development and vitamin D dependence. Am J Physiol 243:E483203Thorens B, Roth J, Norman AW, Perrelet A, Orci L 1982 Immunocytochemicallocalization of vitamin D dependent calcium binding protein in chick duodenum.J Cell Biol 94:201Ullu E and Tsudi C 1984 Alu sequences are processed 7SL RNA genes Nature312:171Varghese S, Christakos S, 1,25-dihydroxyvitamin D3 induces transcription of therat calbindin-D28k gene. Program of the 70th Annual Meeting of The EndocrineSociety, New Orleans, LA, 1988, p 138 (Abstract 470)Varghese S, Lee S, Huang YC, Christakos S 1988 Analysis of rat vitamin D-dependent calbindin-D 28k gene expression. J Biol Chem 263:9776Verstappen A, Parmentier M, Chirnoaga M, Lawson DEM, Pasteels JL, Pochet R1986 Vitamin D-dependent calcium binding protein immunoreactivity in humanretina. Opthalmic res 18:209Walters MR 1981 An estrogen stimulated 1,25 dihydroxyvitamin D3 receptor inrat uterus. Biochem Biophys Res Commun 103:721Wassef M, Zanetta JP, Brehier A, Satelo C 1985 Transient biochemicalcompartmentalization of Purkinje cells during early cerebellar development. DevBiol 111:129Wasserman RH, Corradino RA, Taylor AN 1968 Vitamin D-dependent calciumbinding protein: purification and some properties. J Biol Chem 243:3978Wasserman RH, Fullmer CS 1983 Calcium transport proteins, calcium absorptionand vitamin D. Annu Rev Physiol 45:375Wasserman RH, Fullmer CS 1989 On the molecular mechanism of intestinalcalcium transport. Advances in Experimental Medicine and Biology 249:45Wasserman RH, Fullmer CS, Taylor AN 1977 The vitamin D-dependent calciumbinding proteins In: Lawson DEM (ed) Vitamin D. Academic Press, London, p 133Wasserman RH, Taylor AN 1966 Vitamin D3 induced calcium binding protein inchick intestinal mucosa. Science 152:791Wasserman RH, Taylor AN 1971 Evidence for a vitamin D3 induced calciumbinding protein in new world primates. Proc Soc Exp Biol Med 136:25Wells RD 1988 Unusual DNA structure. J Biol Chem 263:1095204Wilson PW, Harding M, Lawson DEM 1985 Putative amino and sequence of chickcalcium binding protein deduced from a complementary DNA sequence. NucleicAcids Res 13:8867Wilson PW, Rogers J, Harding M, Pohl V, Pattyn G, Lawson DEM 1988 Structureof chick chromosomal genes for calbindin and calretinin. J Mol Biol 200:615Wood TL, Kobayashi Y, Franz G, Varghese S, Christakos S, Tobin AJ 1988Molecular cloning of mammalian 28,000 Mr vitamin D-dependent calcium bindingprotein (calbindin-D28k) in rodent kidney and brain. DNA 7:585Yamamoto NY, Kawanobe Y, Takahashi H, Shimazawa E, Kimura S, Ogata E1984 Vitamin D deficiency and renal calcium transport in the rat J Clin Invest74:507Zhou XY, Dempster DW, Marion SL, Pike JW, Haussler MR, Clemens TL 1986Bone vitamin D-dependent calcium binding protein is localized in chondrocytes ofgrowth plate cartilage. Calcif Tissue Int 38:244


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items