Vol. 185, No. 2, 1992 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

June 15, 1992 Pages 663-669

MOLECULAR CLONING AND CHROMOSOMAL ASSIGNMENT OF HUMAN CALBINDIN-D9K

ALISON HOWARD’, STEPHEN LEGON’, NIGEL K SPURR3, AND JULIAN RF WALTERS’*

‘Gastroenterology Unit, Departments of Medicine and *Chemical Pathology, Royal Postgraduate Medical School, Hammersmith Hospital, London W12 ONN, UK

3Human Genetic Resources Laboratory, Imperial Cancer Research Fund, Clare Hall Laboratories, South Mimms, Herts EN6 3LD, UK

Received April 22, 1992

Human calbindin-D9k, the vitamin D-dependent calcium binding protein, has been cloned and sequenced following initial amplification of intestinal cDNA sequences by the polymerase chain reaction using mixed oligonucleotide primers. The derived amino acid sequence of 79 residues has a calculated molecular weight of 90 15 and is 89% homologous with the bovine and porcine sequences. Probing DNA from human-rodent somatic cell hybrids mapped human calbindin-D9k to chromosome Xp. A single abundant mRNA transcript was detectable in proximal small intestine but not in kidney, uterus or placenta. 0 1992 Academic Press, 1°C.

Calbindin-D9k, the vitamin D-dependent Ca*‘-binding protein, (CaBP9k) was first described in rat

intestine [I] and its concentration in that tissue has subsequently been shown to correlate well with

transcellular, vitamin D-dependent Ca” absorption (for review see [2-41). CaBP9k is thought to

increase Ca*+ absorption by buffering Ca*’ in the cytoplasm but also increases ATP-dependent Ca*+

transport in duodenal basolateral membrane vesicles [S], and can bind to the regulatory calmodulin

binding domain of the plasma membrane Ca*‘-pump [6]. In keeping with this role in calcium

absorption, concentrations of CaBP9k are highest in duodenal villus enterocytes; it has also been

described in rodents in placenta, yolk sac, uterus, kidney, teeth, and bone matrix [3,4].

Amino acid sequences for CaBP from pig [7], cow [8], rat [9] and mouse [lo] were directly

determined and indicated two high-affinity Ca*‘-binding sites (EF-hands). Nucleotide sequences have

been published for the rat cDNA and gene [11,12] and also for the bovine cDNA [13]. Recent work

has demonstrated DNase I hypersensitive sites [ 141 and a la,2Sdihydroxycholecalciferol

*To whom correspondence should be addressed at Gastroenterology Unit, Royal Postgraduate Medical School, Du Cane Rd., London W12 ONN, United Kingdom.

Abbreviations:CaBPgK, calbindin-D9k; PCR, polymerase chain reaction; bp, base pairs.

0006-291X/92 $4.00

663 Copyright 8 I992 by Academic Press, Inc.

All rights of reproduction in any form reserved.

Vol. 185, No. 2, 1992 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

(1,25(0H),D)-response element in the promoter region of the rat gene [ 151. Expression of the protein

is regulated in a tissue-specific manner at both transcriptional and post-transcriptional levels by

1,25(OH),D, glucocorticoids, estrogens and possibly calcium [ 16-201.

Despite this extensive study in other species, investigations of intestinal Ca”-binding proteins in

humans have been limited. A protein was described in human duodenal mucosa with similar Ca”-

binding properties to those in rat and pig but with a molecular weight of 12,000 - 13,000 [21]. A

10 kDa protein was purified from human small intestinal cytosol; antibodies to this protein were not

immunoreactivity with pig or rat duodenal cytosolic proteins [22]. The distribution of this protein and

its correlation with 1,25(OH),D levels [23] suggested it might be similar to CaBP9k in other species.

One other study using a rat CaBP9k cDNA probe demonstrated cross-hybridization to various human

fetal tissues [24] but neither the protein structure, nor the cDNA for CaBP9k have previously been

demonstrated in humans. Using the polymerase chain reaction (PCR), we have amplified and cloned

cDNA sequences from human intestinal mucosa encoding CaBP9k, and now describe the sequence,

chromosomal localization and expression in human tissues.

EXPERIMENTAL

Proximal small intestinal mucosa was obtained from patients undergoing intestinal anastomosis after pancreatic surgery. Additional colon and kidney samples were from uninvolved areas in patients having resection for cancer; uterine and placental tissue were from an elective full-term Qesarian section. RNA was prepared by acid guanidinium-isothiocyanate-phenol-chloroform extraction with LiCl precipitation; poly A+ RNA was obtained by affinity chromatography on an oligo-dT cellulose column; and cDNA for PCR amplification was synthesised by AMV reverse transcriptase using random hexanucleotide primers using standard methods as previously described [25].

Amplification of the cDNA by PCR was performed using Taq DNA polymerase (Amersham), with buffer and nucleotide conditions recommended by the supplier and consisted of 30 cycles of denaturation at 95°C for 30 set, annealing at 50°C for 3 min and extension at 70°C for 30 sec. Three separate sets of primers were used. Initially, mixed oligonucleotide primers were used based on regions of conserved amino acids in the rat, cow, pig and mouse CaBP9k sequences. These were:

5’-(X)GC(A,C,G,T)AA(A,G)GA(A,G)GG(A,C,G,~GA(C,~CC-3’ =A 5’-(X)TC(A,G)(A,T)A(A,G)CT(A,C,G,T)AC(C,T)TC(A,C,G,T)CC-3’ = B

where (X) = added linkers at the 5’ ends. Both mixed primers had a degeneracy of 256-fold and were used at a concentration of 27 t&ml.

To obtain the 3’ end of the sequence, two rounds of anchor PCR were performed based on the methods described for rapid amplification of cDNA ends (RACE) [26]. In the first round (30 cycles), the mixed primer A was used as the upstream primer and oligo-dT with a linker as the downstream primer, and in the second (also 30 cycles), a specific upstream primer (C in Figure 1) and the linker were used. To obtain the 5’ end of the sequence, cDNA was A-tailed using terminal deoxynucleotide transferase. A similar strategy to that described above for the 3’-end was followed using the mixed primer B and oligo-dT with linker in the first round and the specific primer D (in Fig. 1) and the linker in the second round. PCR products were extracted, identified on agarose gels, cloned in Ml3 and sequenced using protocols described before [25].

Chromosomal localization was determined using DNA from human-rodent hybrid cell-lines. DNA was restricted with EcoRl, electrophoresed, blotted onto Hybond N membranes and hybridized with a 149 bp probe cloned from the initial PCR amplification and labelled using the 3’ oligonucleotide as a primer [25].

664

Vol. 185, No. 2, 1992 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

mRNA from various tissues was fractionated on formaldehyde/agarose gels, blotted and hybridized with the 149 bp probe described above. Alternatively, synthetic 33-mer oligonucleotides were labelled by terminal deoxynucleotide transferase. Hybridization was overnight at 60°C in 250 mM Na phosphate buffer, pH 7.2, 2.5 mM EDTA, 25 uM aurin tricarboxylic acid, 5% w/v SDS, 10% dextran sulphate and 0.5% dried milk powder. Filters were washed at 60°C for 60 min in 0.25 M Na phosphate buffer, pH 7.2, 1mM EDTA, and 2% SDS, followed by 0.2 x SSC with 0.2% SDS at 60% for 60 min.

RESULTS AND DISCUSSION

Amplification and sequencirrg

The nucleotide and deduced amino acid sequence of human calbindin-D9k is shown in Figure 1. This

was derived from sequencing multiple independent clones obtained by three different PCR

amplifications of cDNA prepared from human intestinal poly A+ RNA. The initial strategy using the

pair of mixed primers derived from the conserved sequences in other species gave a product of 149 bp.

Sequencing confirmed this region to be homologous with CaBP9k sequences of the other species. The

remainder of the 5’ and 3’ coding and non-coding sequences were each obtained with two rounds of

PCR amplification as described in Methods using the specific primers indicated for the second round

amplifications.

In addition to obtaining clones extending over the full 5’ or 3’ regions indicated, other shorter clones

were found which ended at one of the several runs of A(n) in the 3’, or T(n) in the 5’ region.

10 20 30 40 50 60 70 l * * * * * * * * *

80 90 l * * * * * * *

AAACTCCTCT TTGATTCTTC TAGCTGTTTC ACTATTGGGC AACCAGACAC CAGA ATG AGT ACT AAA AAG TCT CCT GAG GM CTG AAG AGG

TTTGAGGAGA AACTAAGAAG ATCGACAAAG TGATAACCCG TTGGTCTGTG GTCT TAC TCA TGA T T T TTC AGA GGA CTC CTT GAC TTC TCC Met Ser Thr Lys Lys Ser Pro Glu Glu Leu Lys Arg,

- - - - - - - - - - A - - - - - - - - - - ,

-----------c----------, 100 110 120 130 140 150 * 160

* 170

* * * l * * * * l 1 1 * e *

A T T T T T GM AAA 'PAT GCA GCC AAA GAA GGT GAT CCA GAC CAG T T G TCA AAG GAT GAA CTG AAG CTA T T G A T T CAG GCT GAA TAA AM CTT T T T ATA CGT CGG T T T CTT CCA CTA GGT CTG GTC MC AGT TTC CTA CTT GAC TTC GAT AAC TM GTC CGA CTT Ile Phe Glu Lys Tyr Ala Ala Lys Glu Gly Asp Pro Asp Gln Leu Ser Lys Asp Glu Lea, Lys Le,, Leu Ile Gln Ala Glu,

<---------&---- (------------D--------

180 190 200 210 220 230 240 * *

250 * * * l l * * * l l * * * *

TTC CCC AGT T T A CTC AAA GGT CCA MC ACC CTA GAT GAT CTC T T T CAA GAA CTG GAC AAG AAT GGA GAT GGA GAA G T T AGT MG GGG TCA AAT GAG T T T CCA GGT T T G TGG GAT CTA CTA GAG MA G T T CTT GAC CTG TTC T T A CCT CTA CCT CTT CM TCA Phe Pro Ser Leu Leu Lys Gly Pro Asn Thr Leu Asp Asp Leu Phe Gln Glu Leu Asp Lys Asn Gly Asp Gly Glu Val Ser,

( - -B--

260 270 280 290 300 310 320 *

330 340 * t * * * * l * * * * * * * * t l

T T T GAA GM TTC CM GTA T T A GTA AM AAG ATA TCC CAG TGA AGGAGA AAACAAAATA GAACCCTGAG CACTGGAGGA AGAGCGCTGT AAA CTT CTT AAG GTT CAT MT CAT T T T TTC T A T AGG GTC ACT TCCTCT T T T G T T T T A T CTTGGGACTC GTGACCTCCT TCTCGCGACA Phe Glu Glu Phe Gin Val Leu Val Lys Lys Ile Ser Gln End

350 360 370 380 390 400 410 420 430 440 * * * * * * * * * * * * * * * * * l * *

GCTGTGGTCT TATCCTATGT GGAATCCCCC AAAGTCTCTG GTTTAATTCT TTGCAATTAT AATAACCTGG CTGTGAGGTT CAGTTATTAT TAATAAAGAA CGACACCAGA ATAGGATACA CCTTAGGGGG TTTCAGAGAC CAAATTAAGA AACGTTAATA TTATTGGACC GACACTCCAA GTCAATAATA A T T A T T T C T T

450 * *

ATTACTAGAC ATAC - poly A TAATGATCTG TATG

&uLx.J. Nucleotide and amino acid sequence of human calbindiwD9k. The position of the mixed primers (A and B) and the specific primers (C and D) used in the PCR are indicated.

665

Vol. 185, No. 2, 1992 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

Calcium-binding site

N-terminal site

Helix Loop Helix Site of 1__-_____1______-_______l________l intron in d I 4 8 rat

Human MSTKKSPEE LKRIFEKY AAKEGDPDQLSKDE LKLLIQAE FPSLLK' Porcine [ 7 3 .AQ...A. .. S .......... ..N....E. ..Q ........... Bovine [13] ..A ...... ..G .......... ..N....E. .. ..L.T. ...... Rat 111,123 ..A ...... M.S..Q ....... ..N....E. .... ..S. ...... Mouse [lo] AE...A. M.S..Q ............ ..E. .... ..S. ......

Human GPNT LDDLFQEL DKNGDG EVSFEE FQVLVKKI SQ Porcine ..R. . . . ..I.. . . ..N. . . . . . . . . . . . . . . Bovine . .s. ..E..E.. . . . . . . .I.... . .._.... :: Rat ASS. ..N..K.. . . . . . . . ..Y.. .E.FF..L . . Mouse (Q)ASS. ..N..K.. . . . . . . . ..Y.. .EAFF..L . .

l!igu~Z. Comparison of the amino acid sequences of human calbindin-D9k and other species. Amino acids are shown where they differ from the human sequence. The sequences are arranged to show the two Ca*‘-binding domains with their helix-loop-helix structure and also the site of the intron in the rat and the alternative amino acid sequence in the mouse.

Additionally, four alternative sequences were obtained upstream from the 5’ end of the sequence

indicated in Fig. 1. Oligonucleotide probes synthesized to these alternative sequences failed to give

any signal on Northern blots of intestinal mRNA, unlike the strong signal obtained with a similar 33-

mer from the common 5’ untranslated sequence. Hence these alternative sequences were concluded

to be artefacts of the PCR (or reverse transciptase reaction) and so the transcript is shown numbered

from the first nucleotide of the common sequence.

The sequence comprises 454 bp. An open reading frame of 237 bp, coding for 79 ammo acids was

obtained. 54 bp of 5’ untranslated sequence and a 3’ untranslated region consisting of 163 bp ending

in a poly A tail were found. A polyadenylation signal sequence (AATAAA) was found 23 bp from

the 3’ end. Although an alternatively spliced transcript has been indicated by the amino acid sequences

in the mouse [lo], no evidence of this was found in man in sequences from more than twelve different

clones. In the open reading frame which is the same size as in the rat and cow, there is 80%

nucleotide homology with the rat sequence f11,12J and 87% with that of the cow 1131. The 5’ and

3’ untranslated regions have much lower homology. They are of similar but not identical length to

those in other species; in the rat, the 5’ untranslated region is 62 bp and the 3’ region is 107 bp

whereas in the cow the 3’ region is 159 bp.

The amino acid sequence is 89% homologous with that of the cow and pig (Fig. 2) but less with those

of rat (78%) or mouse (77%). Amino acid substitutions are generally conservative and differ in only

four positions from those found in one of the other species. The sequences in the putative Ca*+-

binding domains are particularly conserved though substitutions are common in the linker region. The

calculated molecular weight of the human protein is 9015. It should be noted that in other species,

Vol. 185, No. 2, 1992 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

direct amino acid sequencing has shown that the N-terminal of the protein may be modified by

acetylation after cleavage of the methionine [7] and serine [9,10] which could result in a molecular

weight of 8803 in man. As has been pointed out previously, limited amino acid homology

(approximately 30%) exists with other members of the EF-hand Ca’+-binding protein family,

particularly those with two Ca*+-binding sites such as S-100 proteins alpha and beta, calcyclin,

annexin, calpactin I, p9Ka, 42A and cystic fibrosis antigen [27].

Chromosomal assigwnerlt and localization

Southern blots of EcoRl-digested DNA from somatic cell hybrids and controls were probed with a

149 bp CaBP9k sequence and gave a characteristic pattern of bands in human genomic DNA which

was not seen in rodent DNA. Screening of a panel of somatic cell hybrids indicated that the CaBP9k

gene was on the human X chromosome. This was confirmed by a positive signal with DNA from the

cell hybrid Horlgx, which contains the human X chromosome alone [28]. DNA was negative from

cell-lines which contained only the long arm (IW15 [29], NIDl [30], and B2 [31]), indicating that the

gene for CaBP9k is located on the short arm. The further localization of CaBP9k on Xp requires

additional study, but preliminary data indicates that it is not at the locus for X-linked

hypophosphatemic rickets at Xp22.1 - Xp22.3 [32].

Expressiou of human CaBP9k

In RNA from human proximal intestinal mucosa, a single abundant transcript was demonstrated by

Northern blotting. Similar findings were obtained when the blots were probed with sequences obtained

kb

1.77 _

0.53 _

w. Northern blot of RNA from various human tissues probed with human CaBP9k. The gel was loaded with poly A+ RNA as follows: proximal small intestine, lane 1 - lOFg, lane 2 - 10 ng; colon, lane 3; placenta, lane 4; uterine myometrium, lane 5; endometrium, lane 6; and kidney, lane 7, all IOpg. The position of two markers from an RNA ladder (BRL) is shown. The autoradiogram is overexposed to show the difference between small intestine and other tissues.

Vol. 185, No. 2, 1992 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

from the PCR amplifications or with a synthetic 33-mer oligonucleotide constructed to bases 13 to

45 in the 5’ region. CaBP9k could not be detected in blots of RNA prepared from human colon,

placenta, uterine endometrium or myometrium, kidney (Fig. 3). PCR of placental cDNA using the

conditions described for intestine also did not amplify any CaBP9k sequences. Although one study

using a rat CaBP9k probe has found cross-hybridization in some other tissues in the human fetus and

placenta [24], it appears that CaBP9k is less widely distributed in human adults than in rodents. Its

lack of expression in the reproductive system may imply absence in the human genomic sequence of

the estrogen-responsive element recently described in rats [19]. The genomic structure and regulation

of expression of CaBP9k will be the subject of further study in humans and may differ from that in

rodents.

ACKNOWLEDGMENTS

We thank Drs. Clare Selden, and Neil Turner for gifts of tissue or RNA. This work was supported

by the Trustees of the Hammersmith and Queeu Charlotte’s Special Health Authority and by an MRC

Project grant to JRFW.

1. 2. 3. 4. 5. 6. 7.

8. 9. 10.

11. 12.

13. 14. 1.5. 16. 17.

18.

19.

Kahfelz, F.A., Taylor, A.N. & Wasserman, R.H. (1967) Proc. Sot. Exp. Biol. Med. 127, 54-58 Bronner, F. (1988) Calcif. Tissue Int. 43, 133-137 Gross, M. & Kumar, R. (1990) Am. J. Physiol. 259, F195-F209 Christakos, S., Gill, R., Lee, S. & Li, H. (1992) J. Nurr. 122, 678-682 Walters, J.R.F. (1989) Am. J. Physiol. 256, G124-Cl28 James, P., Vorherr, T., Thulin, E., Forsen, S. L Carafoli, E. (1991) FEBS lett. 278, 15.5-1.59 Hofmauu, T., Kawakami, M., Hitchman, A.J.W., Harrison, I.E. & Domington, K.J. (1979) Can. J. Biochem. 57, 737-748 Fullmer, C.S. & Wasserman, R.H. (1981) J. Biol. Chem. 256, 5669-5674 MacManus, J.P., Watson, D.C. St Yaguchi, M. (1986) Biochem. J. 235, 585-95 Hunt, D.F., Yates, J.R., Shabanowitz, J., Bruns, M.E. & Bruns, D.E. (1989) J. Biol. Chem. 264, 6580-6586 Petret, C., Lomri, N., Gouhier, N., Auffray, C. & Thomasset, M. (1988) Eur. J. Biochem. 172, 43-51 Krisinger, J., Darwish, H., Maeda, N. & DeLuca, H.F. (1988) Proc. Natl. Acad. Sci. USA 85, 8988-8892 Kumar, R., Wieben, E. & Beecher, S.J. (lY89) Mol. Endocrinol. 3, 427-32 Petret, C., L’Horset, F. & Thomasset, M. (1991) Gene 108, 227-235 Darwish, H.M. & DeLuca, H.F. (1992) Proc. NatI. Acad. Sci. USA 89, 603-607 Perret, C., Desplan, C. & Thomasset, M. (1985) Eur. J. Biochem. 150, 211-217 Bruns, M.E., Overpeck, J.G., Smith, G.C., Hirsch, G.N., Mills, S.E. & Bruns, D.E. (1988) Endocrinology 122, 2371-2378 Huang, Y., Lee, S., Stolz, R., Gabrielides, C., Pansini-Porta, A., Bruns, M.E., et al. (1989) J. Biol. Chem. 264, 174.54-17461 Darwish, H., Krisinger, J., Furlow, J.D., Smith, C., Murdoch, F.E. & DeLuca, H.F. (19Yl) J. Biol. Chem. 266, 551-558

20. Brehier, A. & Thomasset, M. (1990) Endocrinology 127, 580-587 21. Hitchman, A.J.W. & Harrison, J.E. (1972) Can. J. Biochem. 50, 758-765 22. Staun, M., Sjostram, H. & Nor&, 0. (1986) Eur. J. Clin. Invest. 16, 468-472 23. Staun, M., Boesby, S., Daugaard, H. & Jarnum, S. (1988) Calcif. Tissue Int. 42, 20.5-209 24. Brun, P., Dupret, J.M., Perret, C., Thomasset M. & Mathieu, H. (1987) Pediatr. Res. 21, 362-67 2.5. Howard, A., Legon, S. & Walters, J.R.F. (1992) Biochem. Biophys. Res. Commun. 183, 499-505

REFERENCES

668

Vol. 185. No. 2, 1992 BIOCHEMICAL AND BIOPHYSlCAL RESEARCH COMMUNICATIONS

26. 27. 28.

29. 30.

31.

32.

Frohman, M.A., Dush, M.K. & Martin, G.R. (1988) Proc. Natl. Acad. Sci. USA 85, 8988-9002 Kligman, D. & Hilt, D.C. (1988) Trends in Biochem. Sci. 13, 437-443 Heisterkamp, N., Groffen, J., Stephenson, JR., Spun, N.K., Goodfellow, P.N., Solomon, E., ef al. (1982) Nature 299, 747-749 Nabholz, M., Miggiano, V. & Bodmer, W.F. (1969) Nature 223, 358-363 Spurr, N.K., Feder, J., Bodmer, W.F., Goodfellow, P.N., Solomon, E. & Cavalli-Sforza, L.L. (1986) Ann. Hum. Genet. 50, 145-152 Larizza, L., Rampoldi, E., Monura, A., Doneda, L., Miggiano, V. & Barlati, S. (1983) Cell Biol. Int. Rep., 7,325-332 Thakker, R.V., Read, A.P., Davies, K.E., Whyte, M.P., Weksberg, R., Glorieux, F. et al. (1987) J. Med. Genet. 24, 7.56-760

669