The sequence reported in this article has been deposited in the DDBJ/EMBL/GenBank database (accession no. AF286219)
The murine homologue of the 25-hydroxyvitamin D [25(OH)D] 1α-hydroxylase gene [1α(OH)ase; Cyp27b1], which is mutated in humans with vitamin D-dependent rickets type I (VDDR-I; also known as pseudovitamin D-deficiency rickets [PDDR]) was cloned and characterized. Like the human, the mouse gene has nine exons, and the exon-intron organization is well conserved. By interspecific backcross analysis, the Cyp27b1 gene was mapped to 70.5 cM on mouse Chr 10. This is in a region syntenic with human Chr 12q13.1-q13.3 to which the human 1α(OH)ase gene was previously mapped. Kidney expression of the 1α(OH)ase was localized to cortical tubules and was higher in the adult mouse than in the fetus, consistent with the increased role of its product as a circulating hormone postnatally. Prenatally, the 1α(OH)ase gene, together with the vitamin D receptor (VDR) gene, was expressed in embryonic stem cells, and expression of 1α(OH)ase in bone and intestine was higher in the fetus than in the adult. These observations suggest that 1,25-dihydroxyvitamin D [1,25(OH)2D] plays a role in fetal development. In view of the fact that humans lacking 1α(OH)ase have apparently normal prenatal development, this may point to functional redundancy in the fetal vitamin D system, which now can be explored further in mouse models in which the 1α(OH)ase gene has been deleted.
VITAMIN D has a profound effect on mineral ion homeostasis as well as on the growth and differentiation of a number of tissues.(1–4) The most potent form of vitamin D is the metabolite 1,25-dihydroxyvitamin D [1,25(OH)2D],(5–7) which exerts its genomic action via a nuclear receptor.(8,9) The synthesis of 1,25(OH)2D from its precursor 25-hydroxyvitamin D [25(OH)D] is mediated by the mitochondrial enzyme 25(OH)D 1α-hydroxylase [1α(OH)ase]. Recently, mouse,(10) rat,(11,12) and human(13,14) complementary DNAs (cDNAs) and the human gene(11,14,15) encoding the enzyme have been cloned. Parathyroid hormone (PTH) and calcitonin exert a positive regulatory effect, whereas 1,25(OH)2D exerts a negative effect on transcription of the 1α(OH)ase gene.(16,17)
Although a number of cell lines and tissues can synthesize 1,25(OH)2D, the kidney is the principal site of the circulating hormone. Furthermore, although a variety of tissues are targets for 1,25(OH)2D action, bone and intestine are particularly important for normal growth and calcium homeostasis.
In the present study, we have cloned the mouse gene encoding the 1α(OH)ase and have examined its developmental expression in mouse kidney, bone, and intestine and correlated this with expression of the vitamin D receptor (VDR) and downstream effector molecules such as the calbindins. Our results show that the 1α(OH)ase, along with the VDR, is expressed in embryonic stem cells, suggesting a role of components of the vitamin D system in the very earliest phases of development of the embryo. Kidney expression of the 1α(OH)ase was higher in the adult than in the fetus, consistent with the increased role of its product as a circulating hormone in the postnatal environment. In contrast, expression of the 1α(OH)ase gene in bone and intestine was higher in the fetus than in the adult, suggesting that conditions to facilitate the role of 1,25(OH)2D as an autocrine/paracrine regulator of growth and differentiation may predominate prenatally.
MATERIALS AND METHODS
Isolation of bacteriophage recombinant clones encoding the mouse 25(OH)D3 1α(OH)ase gene
A full-length mouse 25(OH)D3 1α(OH)ase cDNA was obtained by reverse-transcription polymerase chain reaction (RT-PCR) of mouse kidney cDNA using primers (Table 1) based on the published sequence.(10) A 129svj mouse Lambda FIX II genomic library (Stratagene, Inc., La Jolla, CA, USA) was screened with the [3H]-labeled cDNA by standard procedures.(18) Independent positive clones were plaque purified; the genomic inserts were isolated by digestion with XhoI and cloned into the PCR 2.1 vector (Invitrogen, Inc., Carlsbad, CA, USA). The sequence of the inserts was determined by a semiautomated method with an ABI 373 sequencer (Sheldon Biotechnology Center, McGill University, Montreal, Quebec, Canada).
Table Table 1.. Primers Used in This Study
Interspecific backcross mice and gene mapping
C3H/HeJ-gld and Mus spretus (Spain) mice and [(C3H/HeJ-gld × M. spretus)F1 × C3H/HeJ-gld] interspecific backcross mice were bred and maintained as previously described.(19,20)M. spretus was chosen as the second parent in this cross because of the relative ease of detection of informative restriction fragment length variants (RFLV) in comparison with crosses using conventional laboratory strains.
DNA isolated from mouse organs by standard techniques was digested with restriction endonucleases and 10-μg samples were electrophoresed in 0.9% agarose gels. DNA was transferred to Nytran membranes (Schleicher & Schull, Inc., Keene, NH, USA), hybridized with the [32P]-labeled 1618-base pair (bp) 1α(OH)ase cDNA (Table 1) at 65°C, and washed under stringent conditions, all as described previously.(21) Gene linkage was determined by segregation analysis.(22) Gene order was determined by analyzing all haplotypes and minimizing crossover frequency between all genes that were determined to be within a linkage group. This method resulted in determination of the most likely gene order.(23)
Embryonic stem cell culture
The ES-R1 cell line(24) that has been adapted to grow in the absence of feeder cells was used. Cells were grown in gelatinized flasks in Dulbecco's modified Eagle's medium (Gibco-BRL, Rockville, MD, USA) supplemented with 1000 U/ml leukemia inhibitory factor, 15% heat-inactivated fetal calf serum, 0.2 mM L-glutamine, 0.1 mM nonessential amino acids, and antibiotics (50 U/ml penicillin and 50 μg/ml streptomycin). The medium was changed every day and the cells were passaged every 3 days.
RT-PCR of 1α(OH)ase, VDR, calbindin D9k, and calbindin D28k messenger RNA
Total RNA was extracted from fetal and adult mouse kidneys, intestine, and bone and ES cells, using Trizol (Gibco-BRL, Rockville, MD, USA). Five-microgram aliquots of DNAse-treated RNA were reverse-transcribed and the cDNA was PCR-amplified using sequence-specific primers (Table 1). Thirty-two cycles [denaturation 94°C, 1 minute; annealing, 1α(OH)ase, 59°C, VDR, calbindin D9k, and calbindin D28k, 54°C, all for 1 minute; extension, 72°C, 1 minute] were performed with a programmable thermocycler (GeneAmp PCR System 9600; PE Applied Biosystems, Foster City, CA, USA). Aliquots of the PCR reactions were electrophoresed through ethidium bromide-stained 1% agarose gels. PCR products were sequenced to confirm their identity.
In situ hybridization
All animal experiments were carried out in compliance with and approved by the Institutional Animal Care and Use Committee. Kidneys and tibias were taken from 3-month-old CD-1 mice (Charles River, St. Constant, Montréal, Quebec, Canada) and 18-day-old fetal mice (delivered from timed pregnant mothers by caesarian section) and immersed in liquid nitrogen. Frozen sections of the tissues were collected on poly-L-lysine-coated glass slides, which were subsequently treated with 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.4) for 15–20 minutes and then with proteinase K (10 μg/ml) in 10 mM Tris-HCl (pH 8.0) at 37°C for 3-5 minutes. Sections were acetylated by incubation for 10 minutes with 0.25% acetic anhydride in 0.1 M triethanolamine (pH 8.0).
To generate probes for in situ hybridization, a 309-bp fragment of the mouse 1α(OH)ase cDNA was PCR-amplified using the primer pair described in Table 1. The PCR product initially cloned into the PCR 2.1 vector was released by KpnI and XhoI and ligated into the pBluescript II SK(+) vector. Digoxigenin-labeled sense and antisense complementary RNA probes were generated from the linearized plasmid using T7 and T3 RNA polymerases, respectively.
Hybridization buffer (50% formamide, 10 mM Tris-HCl [pH 7.6], 100 μg/ml yeast transfer RNA [tRNA], 1× Denhardt's solution, 10% dextran sulfate, 600 mM NaCl, 0.25% sodium dodecyl sulfate [SDS], and 1 mM EDTA) was preheated at 90°C for 10 minutes. The concentration of cRNA probe was adjusted to 0.1–1.0 μg/ml and hybridization of sections was performed overnight at 56°C in a humidified chamber. Sections were washed with 50% formamide in 2× SSC at 55°C for 30 minutes and treated with TNE [10 mM Tris-HCl (pH 8.0), 0.5 M NaCl, and 1 mM EDTA] solution at 37°C. Sections were preincubated with 2% blocking reagent and incubated with alkaline phosphatase-conjugated sheep antidigoxigenin antibody at a dilution of 1:1000 for immunodetection of the probe. Visualization was performed using nitroblue tetrazolium salt and 5-bromo-4-chloro-3-indolylphosphate.
Identification and restriction mapping of genomic clones
At the initial screening of 500,000 plaques, using the full-length 1α(OH)ase cDNA as probe, 10 strongly positive phage plaques were identified. After two rounds of plaque purification, phage inserts were assessed by restriction enzyme mapping and shown to comprise two types of clone designated λg1α(OH)ase m1 and λg1α(OH)ase m2 (Fig. 1), which together encoded the entire mouse 1α(OH)ase gene and flanking sequences. The inserts of both clones were sequenced.
DNA sequence analysis and structural organization of the mouse 1α(OH)ase gene
The nucleotide sequence of the mouse 1α(OH)ase gene is shown in Fig. 2. It contains nine exons (and is therefore similar to the human gene which also has a large exon 9) and covers 4.5 kilobases (kb) of mouse genomic DNA. The coding regions of the mouse and human genes are over 80% identical, and the sizes of the exons and placement of the introns are well conserved (Fig. 3). In both species the hormone binding domain is encoded by exons 6 and 7, and the heme binding domain is encoded by exon 8.
Mouse chromosomal localization and fine mapping
To determine the chromosomal location of the 1α(OH)ase gene, we analyzed a panel of DNA samples from an interspecific cross that has been characterized for over 800 genetic markers throughout the genome. The genetic markers included in this map span between 50 and 80 centimorgans on each mouse autosome and the X chromosome (Chr). Initially, DNA from the two parental mice (C3H/HeJ-gld and [C3H/HeJ-gld × M. spretus]F1) were digested with various restriction endonuclease and hybridized with the 1α(OH)ase probe to determine RFLVs to allow haplotype analyses. An informative EcoRI RFLV was detected: C3HeJ-gld, 11.0 kb and 4.5 kb; M. spretus, 6.8 kb.
Comparison of the haplotype distribution of the 1α(OH)ase RFLV indicated that this gene cosegregated in 111 of 114 meiotic events with the microsatellite marker D10Mit71 locus on mouse Chr 10 (Fig. 4). The haplotype distribution among other genes localized to mouse Chr 10 is shown in Fig. 4. The best gene order(23) ± SD(22) indicated was D10Mit71-2.6 ± 1.5 centiMorgan-lfng-3.5 ± 1.8 centiMorgan-Cyp27b1. [The official gene symbol for the 1α(OH)ase gene is Cyp27b1.]
Developmental expression of the 1a(OH)ase gene and related members of the vitamin D system
We initially assessed whether the 1α(OH)ase gene could be detected in embryonic stem cells. As determined by RT-PCR, the messenger RNA (mRNA) for the enzyme synthesizing 1,25(OH)2D was readily expressed in these cells (Fig. 5). The results described in this and subsequent sections are representative of experiments that were repeated three times on different preparations. In view of the fact that genomic actions of 1,25(OH)2D are mediated by the VDR, we also examined whether VDR mRNA was present. RT-PCR showed the expression of this key component of the vitamin D system. Consequently, the vitamin D system might function at the earliest stages of embryonic development.
We next examined the relative expression of the mouse 1α(OH)ase in fetal versus adult kidney, bone, and intestine by RT-PCR (Fig. 6). The intensity of expression of the glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene, which was used as a control, was comparable in all tissues. In kidney, a 2.5-fold higher level of expression of the 1α(OH)ase was found in the adult as compared with the fetus (Fig. 6). In contrast, in bone and intestine, the 1α(OH)ase expression levels were considerably higher in the fetus as compared with the adult. Thus, in adult bone expression levels were 10-fold greater than in fetal bone, and although the 1α(OH)ase mRNA was readily detected in fetal intestine, it was virtually absent in adult intestine.
To complement these studies of the enzyme synthesizing 1,25(OH)2D, we examined the mRNA expression of mediators of 1,25(OH)2D action, namely, the VDR and the calbindins D9k and D28k.(25) VDR levels were higher in the kidney than in bone but remained relatively constant in these two tissues whether in fetal or adult stages of life (Fig. 6). In contrast, VDR expression was approximately 3-fold lower in fetal than in adult intestine. Expression of the vitamin D target gene, calbindin D9k was detected in fetal and adult kidney and intestine with lesser expression in bone (Fig. 6). There was little difference in the expression level between fetal and adult kidney or between fetal and adult intestine. In bone, calbindin D9k expression was low, although adult bone showed about 2-fold more expression as compared with fetal tissue. Calbindin D28k was detected in the fetal kidney and in the fetal intestine but was virtually absent in the adult bone and intestine. (Fig. 6) Levels were highest in fetal and adult kidney, increasing slightly in the adult kidney. Therefore, the development pattern of calbindin D28k resembled that of 1α(OH)ase more than that of calbindin D9k in these tissues.
Localization of 1α(OH)ase gene expression in kidney and bone
We next examined the cellular localization of the 1α(OH)ase mRNA by in situ hybridization. In murine fetal kidney, expression of the 1α(OH)ase gene was observed in kidney, expression of the 1α(OH)ase gene was observed in the cortical renal tubules (Fig. 7). No expression was detected in the renal glomeruli or in the medulla.
In the fetal skeleton, 1α(OH)ase gene expression was observed in both the growth plate cartilage and the bone tissue (Fig. 8). In epiphyseal growth plate, a positive hybridization signal was observed in chondrocytes in the proliferating zone ( Figs. 8 and 9). Expression of the 1α(OH)ase gene also was seen in osteoblasts lining mixed spicules in the metaphysis ( Figs. 8 and 9) and the osteoblasts in the periosteum. No 1α(OH)ase mRNA was observed in osteoclasts.
We have isolated and characterized the mouse homologue of the 1α(OH)ase gene. The amino acid sequence predicted by the open reading frame corresponds exactly to that encoded by the previously reported mouse 1α(OH)ase cDNA(10) and begins with an ATG (methionine) codon in a perfect Kozak consensus sequence.(26) The similarity in exon-intron structure, comparable sizes of coding exons, and >80% identity of the coding nucleotide sequence shared between the mouse and human genes indicates that the 1α(OH)ase has been highly conserved throughout mammalian evolution.
By interspecific backcross analysis, the 1α(OH)ase gene (official gene symbol Cyp27b1) was mapped to position 70.5 cM on mouse chromosome 10. This is within a region (position 66.0 cM-74.0 cM) syntenic with a part of human chromosome 12 within which the human 1α(OH)ase gene has been mapped (12q13.1-q13.3(11) and to which the pseudovitamin D-deficiency rickets (PDDR) or vitamin D-dependent rickets type I (VDDR-I) locus had previously been linked.
The 1α(OH)ase is a type 1 P450 enzyme(27) located in mitochondria and receives electrons from the reduced form of nicotinamide adenine dinucleotide phosphate (NADPH) via flavoprotein ferredoxin reductase and ferredoxin. The substrate specificity lies in the P450 moiety. All mammalian mitochondrial P450 enzymes identified thus far are involved in critical steps of steroid or sterol biosynthesis including synthesis of the vitamin D sterols, 25(OH)D, 24,25(OH)2D, and 1,25(OH)2D.(4) When the amino acid sequence of the mouse 1α(OH)ase is compared with those of the other related type 1 P450 enzymes, there is near perfect alignment of the sites interrupted by introns. Consequently, cloning and characterization of the mouse 1α(OH)ase gene sequence provides further evidence of the strongly conserved exon/intron organization of the type 1 P450 genes and provides additional support for the hypothesis that they all originated from a series of gene duplication events.(28)
Two of the major functions of 1,25(OH)2D involve mineral ion (notably calcium) homeostasis, and, directly or indirectly, regulation of growth and development. The importance of 1,25(OH)2D for mineral homeostasis may be most critical postnatally in mammals and 1,25(OH)2D acts as a circulating hormone in carrying out this function. In this context, the renal 1α(OH)ase appears crucial because, irrespective of other sites of 1α(OH)ase expression, loss of activity of the renal enzyme results in a sharp drop in circulating 1,25(OH)2D and a major negative impact on calcium homeostasis including a decline in intestinal calcium absorption. It is notable in this regard that our present studies revealed an increase in expression of the renal 1α(OH)ase and a decline in the intestinal enzyme in the shift from a fetal to adult environment. This is consistent with the observation that renal enzyme activity is needed for calcium absorption in postnatal mammals and that local 1,25(OH)2D synthesis in the intestine plays a minimal role in this metabolic function. On the other hand, levels of expression of the VDR increased in postnatal intestine reflecting the increasing importance of this organ as a site of 1,25(OH)2D action to maintain calcium homeostasis.
By in situ hybridization, we localized 1α(OH)ase gene expression to renal cortical tubules. This localization is consistent with previous studies that have mapped functional 1α(OH)ase activity to the proximal tubule.(29,30) Additionally, the previous localization of the type 1 PTH/PTH-related protein (PTHrP) receptor to similar sites in renal cortical tubules(31) emphasizes the important regulation by PTH of the renal 1α(OH)ase in postnatal animals. VDR gene knockout mice have very high, 50-fold normal, levels of renal 1α(OH)ase transcripts.(10) As shown in the present study, the abundant expression of the VDR in kidney could reflect the important negative feedback exerted by 1,25(OH)2D on the renal 1α(OH)ase, although the mechanistic role played by the VDR in this process remains unclear. The proximal promoter of the 1α(OH)ase mediates repression by 1,25(OH)2D,(16,17) but it does not contain a consensus vitamin D response element (VDRE). At present, it is unclear whether a nonconsensus type VDRE might be involved or whether 1,25(OH)2D suppresses the 1α(OH)ase gene by an indirect mechanism.
Our in situ hybridization studies showed that the predominant skeletal sites of expression of the 1α(OH)ase were the proliferating growth plate chondrocytes and both periosteal and metaphyseal osteoblasts. Previous in vitro studies also have shown production of 1,25(OH)2D in chondrocytic(32) and osteoblastic cells.(33) Furthermore, in vitro studies have shown effects of 1,25(OH)2D3 on phenotypic indices of both chondrocytic and osteoblastic cells, such as alkaline phosphatase.(32) Consequently, our in situ studies have identified components of the vitamin D autocrine/paracrine system in skeletal long bones, which may be important for normal skeletal growth and maturation.
Inactivating mutations in the 1α(OH)ase and VDR give rise to the rare inborn errors of metabolism, VDDR-I/PDDR(34) and VDDR-II (also known as human vitamin D-resistant rickets [HVDRR]),(35) respectively, which are inherited in an autosomal recessive fashion. They share common clinical and biochemical features. Affected individuals are apparently normal at birth but develop severe rickets some months thereafter. Mice lacking the VDR gene(36,37) and therefore mimicking the VDDR-II syndrome also are apparently normal at birth but develop rickets after weaning. The skeletal defects can in large part be ameliorated by feeding the mice a special diet rich in calcium and phosphorus.(38,39) Earlier studies in humans with VDDR-II/HVDRR also had shown that the rachitic phenotype was reversible with calcium treatment.(40) Thus, the present findings of very early developmental expression of the 1α(OH)ase and VDR in embryonic stem cells and fetal tissues including the developing skeleton are paradoxical. On the one hand, this would suggest an important role for the vitamin D system in early growth and development. On the other hand, the lack of obvious growth or developmental deficits in the fetus in the human conditions described previously and the mouse VDR gene ablated model argue against this. Potentially, therefore, the vitamin D system is redundant with other gene products able to compensate for the lack of either the 1α(OH)ase or the VDR. For example, there may be a second VDR or some of the important growth regulatory (and other) actions of vitamin D metabolites may be mediated by nongenomic mechanisms.(41,42) Additionally, other cytochrome P450 enzymes besides the 1α(OH)ase may be able to catalyze the 1α-hydroxylation of 25(OH)D in the fetus. Thus, an alternative pathway has been reported for the mitochondrial cytochrome P450 CYP27, which normally carries out the hepatic 25-hydroxylation of vitamin D.(43,44)
With the cloning of the mouse 1α(OH)ase gene these issues can be explored further by developing a 1α(OH)ase gene knockout mouse model, either whole animal or tissue specific and then generating a double knockout of both the 1α(OH)ase and the VDR genes. Further studies will now be required to determine the precise mechanisms of action of the vitamin D autocrine/paracrine system components in bone, kidney, and gut, and their interaction with other mediators of growth and development.
This work was supported by grants MT-5775 and MT-9315 from the Medical Research Council of Canada (D.G. and G.N.H., respectively), grant 007311 from the National Cancer Institute of Canada (D.G.), and from the Kidney Foundation of Canada (G.N.H.).