The Journal of Nutrition Vol. 128 No. 11 November 1998, pp. 1869-1877

Vitamin A Up-Regulates Expression of Bone-Type Alkaline Phosphatase in Rat Small Intestinal Crypt Cell Line and Fetal Rat Small Intestine^1,2

Takeshi Nikawa, Kazuhito Rokutan³, Kayo Nanba, Kaori Tokuoka, Shigetada Teshima, Michael J. Engle^*, David H. Alpers^*, and Kyoichi Kishi

Department of Nutrition, School of Medicine, The University of Tokushima, Tokushima 770-8503, Japan and ^* Department of Internal Medicine, Division of Gastroenterology, Washington University School of Medicine, St. Louis, MO 63110

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Vitamin A is a potent inducer for liver/bone/kidney alkaline phosphatase (L/B/K ALP) in a variety of tissues. However, the evidence for induction of L/B/K ALP by vitamin A in small intestine is limited. In this study, we investigated the influence of vitamin A on L/B/K ALP expression in rat small intestinal crypt IEC-6 cells and fetal rat small intestine. Treatment of IEC-6 cells with all-trans retinoic acid (RA) increased the levels of activity, protein and mRNA of L/B/K ALP, whereas enterocyte-specific proteins, including intestinal ALP, sucrase-isomaltase and glucose transporter-2, were not induced. The reverse transcription-polymerase chain reaction technique revealed that this L/B/K ALP transcript had the bone-type but not the liver-type leader exon. IEC-6 cells constitutively expressed mRNAs of all subtypes of retinoic acid receptor (RAR) and retinoid X receptor (RXR) at varied concentrations. Among these receptor mRNAs, RAR mRNA quickly responded to RA treatment, and the level was doubled within 4 h. Gel mobility shift assay showed that RA induced an RXRE-binding activity in IEC-6 cells. The L/B/K ALP transcript, expressed in fetal rat small intestine, also contained the bone-type leader exon. Intragastric administration of 10 mg retinyl acetate to pregnant rats from gestational d 7 to 15 increased the levels of this transcript and enzyme in 15-d fetal rat small intestine. Our results suggest that vitamin A may be an important regulator for L/B/K ALP expression in fetal rat small intestine as well as in IEC-6 cells.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Alkaline phosphatases (ALP; EC 3.1.3.1)⁴ are a group of enzymes that hydrolyze phosphomonoester bonds in various organic compounds. There are at least four distinct forms of ALP, i.e., intestinal, placental, germ cell and liver/bone/kidney (L/B/K) ALP (Trowsdale et al. 1990). The L/B/K ALP is ubiquitously expressed; it is present at higher levels in the kidney, bone, liver and ovary, and at lower levels in many other tissues (Studer et al. 1991). The L/B/K isozyme is considered to play a role in the mineralization of the extracellular matrix in the bone (Swarup et al. 1981). However, it has recently been reported that mice lacking L/B/K ALP have no gross defect in skeletal formation, and die rather from seizure due to defective metabolism of vitamin B-6 (Waymire et al. 1995). Thus, the physiologic functions of L/B/K ALP are not completely understood. In fetal rats, this tissue nonspecific phosphatase is expressed in the first and second gestational phases in surface intestinal epithelial cells (Komoda et al. 1986). It was also shown that the small intestine of the mice lacking the L/B/K ALP gene contained large amounts of gas, never observed in normal animals, suggesting that L/B/K ALP may play an important role in the development of small intestine (Narisawa et al. 1997).

A number of investigators have reported that all-trans retinoic acid (RA) up-regulates L/B/K ALP in a variety of cells, such as neutrophils (Sato et al. 1989), osteoblasts (Zhou et al. 1991) and fibroblasts (Reese et al. 1992). However, the effect of RA on L/B/K ALP induction in small intestinal epithelial cells has not been studied in detail. RA is a potent morphogen and differentiation factor in a variety of tissues, including fetal small intestine. More recently, Plateroti et al. (1997) reported that RA stimulated the interaction between mesenchymal and epithelial cells during development of rat small intestine, and that treatment with RA enhanced the mesenchymal cell-mediated differentiation of intestinal epithelial cells.

In this study, we examined the effect of retinoids on L/B/K ALP expression in rat immature intestinal epithelial IEC-6 cells (Quaroni et al. 1979) and in fetal rat small intestine. Because rat L/B/K ALP has two alternative leader exons, E1 (bone)- and E2 (liver)-type, and the expression of bone- or liver-type ALP mRNA has been suggested to be regulated in a tissue-specific manner (Toh et al. 1989), we examined whether the tissue nonspecific ALP gene preferentially used one of the promoters.

	MATERIALS AND METHODS

Abstract Introduction Methods Results Discussion References

Reagents and media.  Fetal calf serum (FCS) and Dulbecco's modified Eagle's medium (DMEM) were obtained from Flow Laboratories, McLean, VA. RA and insulin from bovine pancreas were purchased from Sigma Chemical, St. Louis, MO. RA was dissolved in ethanol at 10 mmol/L (stock solution) and stored at 20°C in the dark. Human recombinant transforming growth factor-1 (TGF-) and epidermal growth factor (EGF) from mouse salivary gland (receptor grade) were purchased from King Brewery, Tokyo, Japan and Collaborative Research, Bedford, MA, respectively. An enhanced chemiluminescence Western blotting detection system, poly(dI-dC)·poly(dI-dC), [-³²P]ATP (>220 TBq/mmol) and T4 polynucleotide kinase were purchased from Amersham, Little Chalfont, UK. [-³²P]dCTP (>110 TBq/mmol) was obtained from ICN Pharmaceuticals, Costa Mesa, CA. Antisera and cDNAs for rat L/B/K ALP and rat intestinal ALP were prepared as described previously (Seetharam et al. 1986). Mouse retinoic acid receptor (RAR) ,  and  cDNAs (Zelent et al.1989) and mouse retinoid X receptor (RXR) ,  and  cDNAs (Durand et al. 1992) were kindly provided by Dr. P. Chambon, Centre National de la Recherche Scientifique, Strasbourg, France. Antiserum against rat sucrase-isomaltase and a cDNA for rat glucose transporter-2 (GLUT-2) were kind gifts from Dr. K. Miyamoto, University of Tokushima, Japan (Miyamoto et al. 1992). Rat sucrase-isomaltase cDNA was kindly provided by Dr. P. G. Traber, University of Pennsylvania (Traber 1990). IEC-6 cells, mouse F9 teratocarcinoma cells and a cDNA for human glyceraldehyde-3-phosphate dehydrogenase (GAPDH, ATCC 57091) were purchased from the American Type Culture Collection, Rockville, MD.

Analyses of cell growth and enzyme activities in IEC-6 cells.  IEC-6 cells were maintained in tissue culture flasks at 37°C with 5% CO₂/95% air in DMEM, supplemented with 5% FCS and 100 iu/L insulin. F9 cells were cultured and treated with RA, as described previously (Nikawa et al. 1995). For analysis of cell growth, IEC-6 cells were plated in DMEM containing 0.1% FCS at a density of 1 × 10⁷ cells/L for each well of a 24-well culture plate. After overnight culture, cell growth was stimulated by addition of 5% FCS or 0.1 mg/L EGF, and 1µmol/L RA or 80 pmol/L TGF- was simultaneously added (d 0). Cell proliferation was monitored by counting the number of cells.

After IEC-6 cells were treated with 10 nmol/L and 1 µmol/L RA, or with 80 pmol/L TGF- in DMEM containing 0.1% FCS, cells were washed three times with ice-cold PBS, scraped with a rubber policeman, and stored in PBS at 80°C until analyses. The stored cells were disrupted on ice by sonication for 10 s before measurement of enzyme activities. ALP activity in the whole-cell extracts was measured using p-nitrophenyl phosphate as a substrate according to the method of Hausamen et al. (1967). Sucrase-isomaltase activity was determined according to the method of Dahlqvist (1984). Protein concentration was measured according to the method of Lowry et al. (1951) with bovine serum albumin (BSA) as a standard.

Immunoblot analysis.  The whole-cell extracts from IEC-6 cells and tissue proteins from fetal rat small intestine were subjected to 8% SDS-PAGE (40 µg protein per lane) and transferred to a polyvinylidene difluoride membrane. The membrane was blocked with 5% BSA and then incubated for 1 h at 25°C in PBS with a 1:1000 dilution of rabbit antiserum against rat L/B/K ALP, rat intestinal ALP or rat sucrase-isomaltase. Bound antibodies were detected using an enhanced chemiluminescence system.

Northern blot analysis.  Total RNA was isolated from IEC-6 cells with an acid guanidinium thiocyanate-phenol-chloroform mixture (Isogen; Nippon Gene, Tokyo, Japan). Total RNA (20 µg) was separated in a 1% agarose gel, blotted and UV-crosslinked to a Hybond N⁺ nylon membrane (Amersham). Prehybridization of membranes was carried out for 4 h at 42°C in buffer containing 50% formamide, 6X SSC (1X SSC consists of 0.15 mol/L NaCl and 15 mmol/L sodium citrate), 1% SDS, 5X Denhardt's solution (1X Denhardt`s solution consists of 0.02% Ficoll, 0.02% BSA and 0.02% polyvinylpyrrolidone) and 50 mg/L heat-denatured, sheared salmon sperm DNA. Hybridization was performed overnight at 48°C in a mixture, containing 4 vol of prehybridization solution, 1 vol of 50% dextran sulfate, and 1-5 × 10⁹ cpm/L of the appropriate cDNA probe. The isolated cDNA fragments were labeled with [-³²P]dCTP by the random primer method (Amersham Megaprimer Kit). Membranes were washed once in 2X SSC/0.1% SDS, followed by two washes (each in 0.5X SSC/0.1% SDS and 0.1X SSC/0.1% SDS). Washing was carried out at 60°C for homologous probes and at 50-55°C for heterologous probes. Membranes were exposed to Kodak X-ray films for the appropriate times, and then films were developed. Autoradiograph signals were quantitated by densitometric analysis. Each mRNA level was standardized by that of GAPDH mRNA.

Reverse transcription and polymerase chain reaction.  Polymerase chain reacrtion (PCR) amplification of the L/B/K ALP transcript in total RNA was performed after reverse transcription (RT) reaction using an RNA-PCR kit (Takara, Tokyo, Japan). The antisense downstream oligonucleotide was obtained in exon 3 (5'-CAGTGTCAGCCGTTAATTGAC-3' complementary to nucleotides -79 to -59 of the rat L/B/K ALP cDNA), and sense upstream oligonucleotide was specific for exon 1 (5'-AGAGTAGGCTCCCGCCACG-3', nucleotides -167 to -149 of the rat L/B/K ALP gene) or exon 2 (5'-ATCGTGCTGACCTTGCCACA-3', nucleotides -304 to -285 of the rat gene). For -actin amplification, two oligonucleotides specific for the rat -actin gene (5'-GCGCTCGTCGTCGACAACGG-3', nucleotides 19 to 38, and 5'-GATGGCTACGTACATGGCTG-3', complementary to nucleotides 389 to 408 of the rat -actin cDNA) were used. After 40 cycles (0.5 min at 94°C followed by 0.5 min at 55°C and 1.5 min at 72°C in a DNA thermal cycler), the resulting samples were electrophoresed on a 1.5% agarose gel. The separated products were transferred onto nylon membranes and subjected to Southern blot analysis using ³²P-labeled oligonucleotides specific for exon 1 (5'-TGCAGGCAGGAGTGGGGA-3', complementary to nucleotides -107 to -90 of the rat L/B/K ALP gene), exon 2 (5'-TGAGGCTGAAATGGTCT-3', complementary to nucleotides -106 to -90 of the rat L/B/K ALP gene) and -actin (5'-GGCACCACACTTTCTACA-3', nucleotides 257 to 274 of the -actin cDNA). These oligonucleotides were radiolabeled with [-³²P]ATP by 5'-end labeling with T4 polynucleotide kinase.

Gel mobility shift assay.  Preparation of whole-cell extracts and gel mobility shift assay were performed, as described previously (Hirakawa et al. 1996). After IEC-6 cells were untreated or treated with RA or TGF- for the indicated days, cells were washed three times with cold PBS. The cells were harvested in PBS with a rubber policeman and collected in microcentrifuge tubes. After centrifugation at 3000 × g for 5 s, pelleted cells were rapidly frozen in liquid nitrogen. The frozen pellets were suspended in lysis buffer, consisting of 20 mmol/L HEPES (pH 7.9), 25% glycerol, 0.42 mol/L NaCl, 1.5 mmol/L MgCl₂, 0.2 mmol/L EDTA, 0.5 mmol/L phenylmethylsulphonyl fluoride, 0.5 mmol/L leupeptin, 5 mmol/L trans-epoxysuccinyl-l-leucylamido-(4-guanidino)butane, 0.3 mmol/L aprotinin, 1 mmol/L pepstatin and 0.5 mmol/L dithiothreitol. Whole-cell extracts were obtained by ultracentrifugation at 100,000 × g for 5 min at 4°C. The supernatants were dialyzed against binding reaction buffer, consisting of 10 mmol/L Tris-HCl (pH 7.8), 50 mmol/L NaCl, 1 mmol/L EDTA, 8% glycerol, 1 mmol/L dithiothreitol, 5 mmol/L MgCl₂ and 0.5 g/L BSA, and stored at 80°C. A synthetic double-stranded oligonucleotide (5`-AGCTAGGTCAGAGCCCAGCT-3'), encoding a RXRE motif similar to the putative RXRE located in the promoter of the L/B/K ALP exon 2 gene (corresponding to nucleotides -479 to -492 upstream of exon 2) (Toh et al. 1989), was used to detect RXRE-binding activities. The RXRE oligonucleotide was radiolabeled with [-<sup>32</sup>P]ATP by 5`-end labeling with T4 polynucleotide kinase. The radiolabeled oligonucleotide was annealed to the complementary strand. The synthetic double-stranded oligonucleotide, encoding nucleotides of the AP-1 binding site (5'-TTGGGGTGACATCATGGGCTA-3'), was used as a non-self-competitor. Whole-cell extracts (10 µg protein) were mixed with 0.1 ng of ³²P-labeled oligonucleotide and 1 mg of poly(dI-dC)·poly(dI-dC) in the binding reaction buffer. Binding reaction was performed for 30 min at 25°C. The RXRE-containing complexes were separated in 5% nondenaturing polyacrylamide gels, and electrophoresis was carried out with 0.5X Tris-borate-EDTA (1X Tris-borate-EDTA consisted of 45 mmol/L Tris-borate, pH 8.3, and 12 mmol/L EDTA) at 180 V for 2 h at 4°C.

Treatment of pregnant rats with retinyl acetate.  Adult male and female Sprague-Dawley rats weighing 250-350 g were purchased from Japan SLC (Shizuoka, Japan). Rats were housed in a room maintained at 23°C on a 12-h light:dark cycle and were allowed free access to standard nonpurified MF diets⁵ (Oriental Yeast, Osaka, Japan) and water for 2 wk. Male and female rats were mated, and vaginal smear was monitored every day. Day 0 was defined by a sperm-positive vaginal smear or vaginal plug. Retinyl acetate (10 mg) dissolved in 100 µL of corn oil was delivered intragastrically to pregnant rats by a blunt stainless feeding tube (Isis, Osaka, Japan) every 2 d after d 7. It has been reported that RA occasionally caused necrosis of placenta, whereas retinyl acetate passes safely through the placenta (Takahashi et al. 1975). Therefore, we treated with retinyl acetate instead of RA. Pregnant rats were killed by cervical dislocation 4 h after the vitamin A administration on d 15. Fetal body weights were measured, and small intestines were removed from fetuses and maternal rats. Total RNA was extracted from fetal small intestine as described above. Small intestinal mucosa of fetal or maternal rats was scraped with a slide glass and collected in microcentrifuge tubes. The mucosa was homogenized in ice-cold PBS with a teflon homogenizer. After centrifugation of the homogenates at 12,000 × g for 15 min at 4°C, supernatants were used to determine the activities of ALP and sucrase-isomaltase, and the levels of these enzymes by Western blot analysis.

These experiments were performed according to the international guiding principles for biomedical research involving animals and were allowed by The Committee of The University of Tokushima for Experimental Animals.

Statistical analysis.  All data are expressed as means ± SD for 3-8 individual samples per group and analyzed by one-way ANOVA using SPSS (release 6.1, SPSS Japan, Tokyo, Japan). Differences between means were tested by Scheffé's test. A P-value <0.01 was considered to be significant.

	RESULTS

Abstract Introduction Methods Results Discussion References

Effect of RA on cell growth.  We first tested whether RA affected proliferation of IEC-6 cells. The number of cells remained constant over 5 d when they were cultured in 0.1% FCS-containing DMEM (Fig. 1). Cell growth was stimulated by 5% FCS and was significantly suppressed by the addition of 1 mmol/L RA or 80 pmol/L TGF- (Fig. 1A). Cell growth stimulated by EGF was also suppressed by incubation with RA or TGF- (Fig. 1B). Thus, RA had an antiproliferative action on IEC-6 cells similar to that of TGF-. Simultaneous addition of RA and TGF- did not additionally suppress the proliferation of cells (data not shown).

View larger version (20K): [in this window] [in a new window]   Fig 1. Effect of retinoic acid (RA) or transforming growth factor-1 (TGF-) on IEC-6 cell growth. IEC-6 cells (1 × 104 cells) were plated in each well of 24-well culture plates and cultured in Dulbecco's modified Eagle's medium (DMEM) containing 0.1% fetal calf serum (FCS). After overnight culturing, the cell growth was stimulated by the addition of (A) 5% FCS or (B) 0.1 mg/L epidermal growth factor (EGF). At the same time, cells were untreated or treated with 1 mmol/L RA or 80 pmol/L TGF- (d 0). The number of cells was counted on d 1, 3 and 5. The numbers of cells cultured in 0.1% FCS-containing DMEM are shown by open circles. Each culture medium was changed to a fresh one at the times indicated by arrows. Values are means ± SD, n = 8. *Significantly <5% FCS- or EGF-treated cells, P < 0.01.

Induction of L/B/K ALP by RA in IEC-6 cells.  Low levels of ALP activity were detected in IEC-6 cells. The activity was not changed by increasing the passage of cells and not affected by changing the concentration of FCS: 0.06 ± 0.01 iu/mg protein (n = 8) in 0.1% FCS-containing DMEM and 0.07 ± 0.02 iu/mg protein (n = 8) in 5% FCS medium. As shown in Figure 2, RA increased ALP activity on d 5-7 in a dose-dependent manner, whereas TGF- did not. There was no synergistic effect between RA and TGF- on the induction of ALP activity in IEC-6 cells (data not shown).

View larger version (16K): [in this window] [in a new window]   Fig 2. The effect of retinoic acid (RA) or transforming growth factor-1 (TGF-) on alkaline phosphatase (ALP) activity in IEC-6 cells. After treatment with RA at 10 nmol/L or at 1 µmol/L, or 80 pmol/L TGF- for the indicated days, ALP activity was measured as described. The activity was also measured in untreated control cells. Each medium was changed to a fresh one at the times indicated by arrows. Values are mean ± SD, n = 8. *Significantly greater than control cells, P < 0.01.

To determine which isomer of ALP contributed to the increase of ALP activity in IEC-6 cells, Western blot analysis was performed with antisera against rat L/B/K and intestinal ALP. Untreated IEC-6 cells contained low levels of a 75-kDa protein corresponding to rat L/B/K ALP (Fig. 3, lanes 1-4). RA at 1 µmol/L induced the 75-kDa L/B/K ALP protein on d 3-7 (lanes 8-10). In contrast, TGF- did not increase the amount of L/B/K ALP protein in IEC-6 cells (lanes 5-7) and it did not affect ALP activity. Intestinal ALP protein was not detected in IEC-6 cells before or after treatment with RA or TGF- (Fig. 4A). These results suggested that RA induced selectively L/B/K ALP in IEC-6 cells, but not intestinal ALP.

View larger version (43K): [in this window] [in a new window]   Fig 3. Western blot analysis of liver/bone/kidney alkaline phosphatase (L/B/K ALP) protein in IEC-6 cells. IEC-6 cells were untreated (lanes 1-4), or treated with 80 pmol/L transforming growth factor-1 (TGF-) (lanes 5-8) or 1 µmol/L retinoic acid (RA) (lanes 8-10) for 0, 3, 5 or 7 d. Each medium was changed to a fresh one every 3 d. Whole-cell proteins extracted from the cells were subjected to 8% SDS-PAGE (40 µg of protein per lane) and transferred to a membrane. Western blot analysis with antibody against L/B/K ALP was performed as described. Similar results were obtained in three separate experiments. M.W.STD., molecular weight standards.

View larger version (27K): [in this window] [in a new window]   Fig 4. Western and Northern blot analyses of enterocyte-specific proteins in IEC-6 cells. IEC-6 cells were cultured with 1 µmol/L retinoic acid (RA) and 80 pmol/L transforming growth factor-1 (TGF-) for the indicated days. Each medium was changed to a fresh one every 3 d. The extracted cell proteins were subjected to 8% SDS-PAGE (40 µg of protein per lane) and transferred to a membrane. Tissue proteins (10 µg protein) prepared from adult rat kidney or small intesinal mucosa were used as positive controls. Western blot analysis with antibody against (A) intestinal ALP or (B) sucrase-isomaltase was performed as described. Total RNA (20 µg), extracted from IEC-6 cells, rat kidney or rat liver, was subjected to 1% agarose/0.6 mol/L formaldehyde gel electrophoresis and transferred to a nylon filter. (C) Northern hybridization with a rat glucose transporter-2 (GLUT-2) cDNA was performed as described. Similar results were obtained in three separate experiments. M.W.STD., molecular weight standards.

We also examined the effect of RA and TGF- on other enterocyte-related proteins, sucrase-isomaltase and GLUT-2. We could not detect the activity, protein or mRNA of sucrase-isomaltase even after treatment with RA (Fig. 4B and data not shown). No TGF--induced up-regulation of this enzyme was found by Western blot analysis (Fig. 4B). In addition, GLUT-2 mRNA was not observed by Northern blotting before or after addition of RA and TGF- (Fig. 4C).

Detection of bone-type ALP transcript in IEC-6 cells.  The concentrations of L/B/K ALP mRNA in IEC-6 cells were measured by Northern blot analysis. Untreated IEC-6 cells contained low levels of L/B/K ALP mRNA (Fig. 5A, lanes 1-4). After treatment with TGF-, the transcript level was transiently elevated (lanes 5-8), although TGF- did not increase ALP activity and protein (Figs. 2, 3 and 4). Treatment of the cells with RA increased the concentration of L/B/K ALP mRNA progressively from d 1 to 5 (lanes 9-12). The level of intestinal ALP mRNA was also measured by Northern blot analysis, but the mRNA was not detected even in the presence of RA or TGF- (data not shown).

View larger version (41K): [in this window] [in a new window]   Fig 5. Liver/bone/kidney alkaline phosphatase (L/B/K ALP) transcripts in IEC-6 cells. (A) The level of L/B/K ALP mRNA was measured by Northern blot analysis. IEC-6 cells were cultured without any supplement (lanes 1-4), or with 80 pmol/L transforming growth factor-1 (TGF-) (lanes 5-8) or 1 µmol/L retinoic acid (RA) (lanes 9-12) for the indicated days. Each medium was changed to a fresh one every 3 d. Total RNA (20 µg) was subjected to 1% agarose/0.6 mol/L formaldehyde gel electrophoresis and transferred to a nylon filter. Northern hybridization with a rat L/B/K ALP cDNA was performed as described. Similar results were obtained in three separate experiments. (B) The bone- and liver-type leader sequences in L/B/K ALP transcripts were detected by reverse transcription-polymerase chain reaction (RT-PCR). Total RNA (1 µg) was prepared from IEC-6 cells untreated or treated with 1 µmol/L RA for 3 d. RT-PCR was performed using oligonucleotide couples specific for L/B/K ALP exon 1, 2 and -actin. The products of the amplification were run on a 1.5% agarose gel, blotted on a nylon membrane and hybridized to 32P-labeled synthetic oligonucleotides specific for ALP exons 1, 2 and -actin as described. These oligonucleotides used for the hybridization were distinct and internal to the couple of primers for the amplification step. The expected sizes of the amplified products are 109, 246 and 390 bp for L/B/K ALP exon 1, 2 and -actin, respectively. Similar results were obtained in three separate experiments.

The rat L/B/K ALP gene contains two leader exons (exon 1 and exon 2), giving rise to two alternatively spliced mRNAs in the 5'-untranslated region. Rat exon 1 and exon 2 were transcribed in bone-type and liver-type transcripts, respectively (Toh et al. 1989). Because the cDNA probe used for the Northern blot analysis did not distinguish the two transcripts, RT-PCR was designed to selectively amplify the exon 1- or exon 2-containing transcript. As shown in Figure 5B (lanes 3 and 4), the exon 2-containing transcript was not detected in untreated and RA-treated IEC-6 cells. In contrast, the exon 1-containing transcript was amplified in untreated cells (lane 1). Even after treatment with RA, only the bone-type transcript was amplified (lane 2), but not the liver-type transcript (lane 4).

RA-induced changes of RAR and RXR mRNA levels in IEC-6 cells.  To characterize the effects of RA on IEC-6 cells, the expressions of RAR,  or  and RXR,  or  in IEC-6 cells were measured by Northern blot analyses, and these levels were compared with those in F9 cells (Fig. 6). The mRNAs of RAR, RAR and RXR were expressed at higher levels in IEC-6 cells, similar to observations in F9 cells (Fig. 6A). F9 cells expressed abundant RXR and RXR mRNAs, whereas IEC-6 cells contained these mRNAs at lower concentrations. The cDNA probe for RAR hybridized two RNA bands with molecular masses of 3.6 and 2.2 kbp in IEC-6 cells, and the 2.2-kbp transcript was not detected in F9 cells. In the absence of RA, the level of each receptor mRNA remained constant during the experimental period. The time-dependent change of each mRNA level after treatment with RA was quantified by densitometric analysis (Fig. 6B). RA did not change the levels of RAR or RXR,  and  mRNAs. In contrast, RA significantly increased the concentrations of RAR and . The RAR mRNA level was doubled within 4 h after addition of RA and continued to increase until 72 h. The response of RAR was less remarkable; a doubling was first observed at 48 h. We also examined the effect of TGF- on RAR and RXR mRNA concentrations, but it did not affect these levels (data not shown).

View larger version (37K): [in this window] [in a new window]   Fig 6. Time-dependent changes in retinoic acid receptor (RAR) and retinoid X receptor (RXR) mRNA concentrations after treatment with RA. (A) IEC-6 cells were treated with 1 µmol/L RA for 0, 4, 8, 16, 24, 48 or 72 h, and then total RNAs were extracted. Total RNA was also extracted from F9 cells treated with 1 µmol/L RA for 1 d. Samples of 20 µg RNA were subjected to 1% agarose/0.6 mol/L formaldehyde gel electrophoresis and transferred to a nylon filter. The filters were hybridized with the 32P-labeled cDNA probes for the mouse RAR and RXR as described. Similar results were obtained in three separate experiments. (B) Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA level was also measured by Northern blot analysis using a cDNA for human GAPDH and was used to quantify the mRNA level of -, - or -subtype of RAR and RXR by densitometric analysis. Values are means ± SD, n = 3. *Significantly greater than cells before treatment, (time 0; P <0.01).

RXRE-binding activity in IEC-6 cells.  It has not been documented whether the promoter of rat L/B/K ALP exon 1 gene has the RAR response element (RARE) or RXR response element (RXRE) motif, whereas a putative RXRE sequence is present in the promoter region of exon 2. However, gel mobility shift assay showed that RA increased an RXRE-binding activity in IEC-6 cells. Untreated IEC-6 cells did not have any specific RXRE-binding activity (Fig. 7, lanes 1 and 4). Treatment with RA for 1 or 3 d produced a specific protein-DNA complex (lanes 2, 5 and 6). This binding activity was inhibited by addition of excess unlabeled RXRE oligonucleotide (lanes 3, 7 and 8), but not by a nucleotide encoding the AP-1 binding site (non-self) (lanes 9 and 10). TGF- did not promote any specific RXRE-binding activity on d 1 (data not shown) and 3 (lane 11).

View larger version (80K): [in this window] [in a new window]   Fig 7. Detection of a retinoid X receptor response element (RXRE)-binding activity after treatment with retinoic acid (RA). IEC-6 cells were cultured without any supplement (lanes 1 and 4), or with 1 µmol/L RA (lanes 2, 3, 5-10) or 80 pmol/L transforming growth factor-1 (TGF-) (lane 11) in 0.1% fetal calf serum (FCS)-containing Dulbecco's modified Eagle's medium (DMEM). Whole-cell proteins were extracted on the indicated days, and gel mobility shift assay was performed as described. Lanes 7 and 8 contained a 10- and 100-fold molar excess of the unlabeled RXRE oligonucleotide, respectively. Lanes 9 and 10 contained a 10- and 100-fold molar excess of the unlabeled non-self oligonucleotide encoding the AP-1 binding site, respectively. An RXRE-protein complex is indicated by "RXRE." A nonspecific DNA-protein complex is denoted by "NS." Similar results were obtained in three separate experiments.

Effect of retinyl acetate on L/B/K ALP expression in fetal rat intestine.  After pregnant rats were treated with 10 mg retinyl acetate every 2 d from d 7 of gestation, fetal rats were removed from the dams on d 15. About 15 fetuses were present per pregnant rat, and the number of fetuses was not different between retinyl acetate- and vehicle-treated rats. However, the weights of fetuses of retinyl acetate-treated rats (3.48 ± 0.25 g) were significantly less (P < 0.01) than those of control rats (5.42 ± 0.38 g). Furthermore, several fetuses (2 or 3 per pregnant rat) in retinyl acetate-treated rats showed a malformation such as shortening of the extremities or enlargement of the head.

Tissue proteins and total RNA were prepared from the small intestine of fetuses with no macroscopic malformation; Western blot analyses of L/B/K and intestinal ALP isozymes and Northern blot analysis for L/B/K ALP were performed. As shown in Figure 8A, supplementation of retinyl acetate to dams increased L/B/K ALP protein in 15-d fetal small intestine, compared with that in the fetal intestine of vehicle-treated dams. In contrast, intestinal ALP protein was not detected in the tissues of vehicle- or retinyl acetate-treated dams (Fig. 8B). Intestinal tissues from retinyl acetate-treated fetal rats contained more abundant L/B/K ALP mRNA than those from vehicle-treated fetuses (Fig. 8C). We also tested the effects of retinyl acetate on enzyme activities in small intestine of maternal rats. Administration of retinyl acetate did not affect the activities of ALP and sucrase-isomaltase in small intestines of maternal animals: ALP activity was 0.99 ± 0.25 mIU/mg protein (n = 4) in retinyl acetate-treated rats and 0.81 ± 0.51 mIU/mg protein (n = 4) in vehicle-treated rats; sucrase-isomaltase activity was 240 ± 34 mIU/mg protein (n) in retinyl acetate-treated rats and 290 ± 56 mIU/mg protein (n) in vehicle-treated rats.

View larger version (33K): [in this window] [in a new window]   Fig 8. Liver/bone/kidney alkaline phosphatase (L/B/K ALP) in fetal rat intestine. Pregnant rats were administered 10 mg of retinyl acetate intragastrically every 2 d from gestational d 7 to 15, as described. Proteins prepared from fetal rat small intesine were subjected to 8% SDS-PAGE (40 µg of protein per lane) and transferred to a membrane. Western blot analysis with antibody against L/B/K (A) or intestinal ALP (B) were performed as described. Mucosal proteins (10 µg protein), prepared from adult rat small intesine, were used as a positive control. Total RNA (20 µg) was extracted from the fetal rat intestine, subjected to 1% agarose/0.6 mol/L formaldehyde gel electrophoresis and transferred to a nylon filter. (C) Northern hybridization with a rat L/B/K ALP cDNA was performed as described. (D) For determination of the leader exon of L/B/K ALP transcript in fetal rat intestine, reverse transcription-polymerase chain reaction (RT-PCR) amplification with oligonucleotide couples specific for L/B/K ALP exon 1, 2 or -actin was performed using the total RNA (1 µg). The products of the amplification were run on a 1.5% agarose gel, blotted on a nylon membrane and hybridized to 32P-labeled synthetic oligonucleotides specific for ALP exon 1, 2 and -actin, as described. The amplified products of bone-type ALP (109 bp) and -actin (390 bp) are indicated by arrows. The expected size of the exon 2-containing ALP transcript is 246 bp. Similar results were obtained in three separate experiments. M.W.STD., molecular weight standards.

The RT-PCR technique demonstrated that the L/B/K ALP transcript, expressed in intestine of untreated fetuses, had the exon 1 (bone-type) leader sequence (Fig. 8D, lanes 1 and 2), whereas the exon 2 (liver-type)-containing transcript was not detected (lanes 5 and 6). Even after treatment with retinyl acetate, only the bone-type ALP transcripts were amplified (lane 3, 4, 7 and 8).

	DISCUSSION

Abstract Introduction Methods Results Discussion References

In several intestinal epithelial cell lines with a mature enterocyte phenotype, RA was shown to induce distinct proteins, i.e., cellular retinal binding protein II in human Caco-2 adenocarcinoma cells (Levin and Davis 1997) and carcinoembryonic antigen in human HT-29 adenocarcinoma cells (Friedman et al. 1987). The IEC-6 cell line, derived from neonatal rat small intestine and resembling immature crypt cells by immunologic and morphologic criteria (Quaroni et al. 1979), has been used to study the mechanism of differentiation of small intestinal epithelial cells in vitro. We first examined the effect of RA on the induction of intestinal ALP in IEC-6 cells. The intestinal isoform of ALP is an enterocyte brush border enzyme. The intestinal crypt cells do not express intestinal ALP, but the expression of intestinal ALP is induced and increased during their differentiation and migration toward the tip of the villus. We found that RA did not induce intestinal ALP isozyme, but it up-regulated the expression of L/B/K ALP in IEC-6 cells. Several differentiation factors, including extracellular matrix (Carroll et al. 1988) and vitamin D (Jeng et al. 1994), were previously examined to determine whether they could induce intestinal ALP in IEC-6 cells. However, none of these factors induced intestinal ALP, although they stimulated the expression of the L/B/K isozyme. L/B/K ALP is expressed in surface epithelial cells and crypt cells during the first and second gestational phases (Komoda et al. 1986). These observations led us to consider that IEC-6 cells may possess features characteristic of fetal intestinal crypt cells, and the RA-mediated regulation of L/B/K ALP expression may be operating in the fetal small intestine.

We also confirmed that RA did not induce other enterocyte-specific proteins such as sucrase-isomaltase and GLUT-2. Recently, Suh and Traber (1996) showed that when IEC-6 cells were transfected with a mouse intestine-specific homeodomain Cdx-2 protein, they could express sucrase-isomaltase and further differentiate into absorptive cells, goblet-like cells and enteroendocrine cells. Thus, IEC-6 cells may be multipotent cells; however, they may require distinct differentiation factors, including Cdx-2, to mature into different cell types characteristic of the small intestine. RA alone may not be able to induce the critical factors required for the expression of enterocyte-specific proteins.

RA was able to increase activity, protein and mRNA of L/B/K ALP. In contrast, TGF- caused only a transient expression of the L/B/K ALP mRNA without the induction of L/B/K ALP protein. The similar effects of RA and TGF- on L/B/K ALP induction were also reported in rat preosteoblastic UMR 201 cells (Zhou et al. 1994). It was suggested that RA could stabilize the L/B/K ALP mRNA in addition to its stimulatory action on the transcription, whereas TGF- did not have such protective action. TGF- was reported to increase sucrase activity in the cells (Kurokowa et al. 1987). However, TGF- did not induce any enterocyte-specific proteins examined in IEC-6 cells.

Rat L/B/K ALP has been reported to have two leader exons, exon 1 (the bone-type) or exon 2 (the liver-type), and its transcription may be under two different types of regulation in a tissue-specific manner. For instance, bone-type mRNA is dominant in human osteosarcoma cells (Weiss et al. 1986) and neutrophils (Sato et al. 1994), whereas liver-type mRNA is preferentially expressed in rat liver and kidney (Zernik et al. 1991). Northern blot and RT-PCR analyses demonstrated that IEC-6 cells constitutively expressed low levels of the ALP transcript containing leader exon 1, and that RA up-regulated the expression of this bone-type ALP transcript. RA exerts its biological actions through binding to the nuclear transcription factors, RAR and RXR (Durand et al. 1992, Zelent et al. 1989). They bind to a specific DNA motif, termed RARE or RXRE. Apparent RARE sequence has not been documented in the 5'-flanking regions of rat L/B/K ALP exon 1 and 2 genes. In response to RA, IEC-6 cells produced an RXRE-binding activity in IEC-6 cells on d 1-3. However, an RXRE is located in the upstream of liver-type, but not bone-type exon. Studer et al. (1991) also reported that RA induced L/B/K ALP protein by increasing the amount of the mRNA containing exon 1A (corresponding to rat exon 1) in mouse F9 teratocarcinoma cells, although an RARE or RXRE has not been reported to be present in the 5'-flanking region of this exon. These results indicated that RA-dependent additional factor(s) may be required for the induction of the bone-type ALP. Thus, the underlying mechanisms for the RA-mediated expression of the rat bone-type ALP gene remain to be elucidated.

To further characterize the RA-mediated response in IEC-6 cells, the expression of RAR and RXR in the cells was measured. IEC-6 cells constitutively expressed all three subtype mRNAs of both receptor types. RA increased the levels of RAR and RAR mRNAs, whereas the expressions of the other receptor mRNAs were not affected. In response to RA treatment, the RAR mRNA expression was significantly increased within 4 h, whereas a significant increase in RAR mRNA was first observed 48 h after treatment with RA. The time-dependent change in the mRNA level of RAR was also correlated with the expression of L/B/K ALP mRNA. This pattern of RAR expression in IEC-6 cells is in agreement with the report (Plateroti et al. 1993) that RAR transcript was expressed in the isolated intestine of 17-d rat embryos, and its level increased in response to RA.

We found that the protein and transcript of L/B/K ALP, expressed in 15-d fetal rat small intestine, were further accumulated when pregnant rats were intragastrically administered 10 mg of retinyl acetate per rat from gestational d 7 to 15. Furthermore, this L/B/K ALP transcript contained the bone-type leader exon. In contrast, Western blot analysis with an antibody against intestinal ALP revealed that intestinal ALP was not expressed in 15-d fetal rat small intestine and was not induced even after administration of retinyl acetate. Subsequently, administration of retinyl acetate did not affect ALP activity in maternal rat small intestine. Considering these observations, vitamin A may be an important regulator for expression of L/B/K ALP in fetal rat small intestine as well as in IEC-6 cells, but not for intestinal-type ALP expression.

For studying the vitamin A transport system in rat placenta and the effect of vitamin A supplementation on fetal development, high dosages of retinyl acetate (>20 mg per rat) were usually administered to pregnant rats in previous reports. Lorente and Miller (1977) reported that treatment of pregnant rats with 41.3 mg retinyl acetate per day during the second gestational phase increased the content of fetal liver>10-fold. Takeyama et al. (1996) reported that treatment of pregnant rats with an excess of RA (10 mg) was necessary to up-regulate the expression of RAR mRNAs in embryo. On the basis of these findings, we treated pregnant rats with 10 mg retinyl acetate every 2 d to induce the effects of vitamin A on fetuses and to avoid the toxic/teratogenic effects as much as possible. However, this dosage of retinyl acetate is likely to still be excessive because it caused limb malformations and head enlargement in the fetuses of 10-20% of dams. We examined the effects of the retinyl acetate supplementation on ALP induction in the small intestine using fetuses with no macroscopic malformations.

Based on the pattern of the expression of ALP isozymes, IEC-6 cells seemed to have features characteristic of fetal small intestinal cells because rat fetal intestine expressed the bone-type isozyme and its expression was also enhanced by vitamin A supplementation. RAR mRNA was expressed only in fetal rat small intestine, but not in adult rat small intestine, also supporting our hypothesis. Therefore, IEC-6 cells may be a useful model system to further study the role of vitamin A-mediated up-regulation of L/B/K ALP in normally developing fetal small intestine. In the fetal rat small intestine, L/B/K ALP is expressed primarily in the surface cells lining the primitive gut during the first gestational phase; the cells lining the newly formed crypts during the second phase also express L/B/K ALP (Komoda et al. 1986). Intestinal ALP first appears during the third gestational phase. Once this isoform appears in enterocytes on villi, the L/B/K ALP expression rapidly and reciprocally decreases. Recently, Narisawa et al. (1997) reported that mice lacking the L/B/K ALP gene showed abnormal movement of the small intestine, suggesting that L/B/K ALP may be involved in the development of the small intestine. Our results suggest that deficiency or excess supplementation of vitamin A during the early gestational phases might modify the pattern of ALP isozyme expressions in fetal rat small intestine and might affect its development.

	FOOTNOTES

Manuscript received 29 April 1998. Initial reviews completed 1 June 1998. Revision accepted 11 July 1998.

	ACKNOWLEDGMENTS

We thank P. Chambon, Laboratoire de Genetique Moleculaire des Eucaryotes-U.184, Centre National de la Recherche Scientifique, Strasbourg, France, who kindly provided plasmids containing RAR and RXR cDNAs. We are grateful to P. G. Traber, Gastroenterology Division, University of Pennsylvania, Philadelphia, PA, for a kind gift of the plasmid pRSl-1. We also thank K. Miyamoto, Division of Clinical Nutrition, The University of Tokushima, Japan, for providing us with antiserum against rat sucrase-isomaltase and a cDNA for rat GLUT-2.

	LITERATURE CITED

Abstract Introduction Methods Results Discussion References

• Bieri J. G. Potassium requirement of the growing rat. J. Nutr. 1977; 107:1394-1398
• Carroll K., Wong T. T., Drabik D. L., Chang E. B. Differentiation of rat small intestinal epithelial cells by extracellular matrix. Am. J. Physiol. 1988; 254:G355-G360[Abstract/Free Full Text]
• Dahlqvist A. Assay of intestinal disaccharidases. Scand. J. Clin. Lab. Investig. 1984; 44:169-172[Medline]
• Durand B., Saunders M., Leroy P., Leid M., Chambon P. All-trans and 9-cis retinoic acid induction of CRABPII transcription is mediated by RAR-RXR heterodimers bound to DR1 and DR2 repeated motifs. Cell 1992; 71:73-85[Medline]
• Friedman J., Seger M., Levinsky H., Allalouf D. Modulation of carcinoembryonic antigen release by HT-29 colon carcinoma line in the presence of different agents. Experimentia 1987; 43:1121-1122[Medline]
• Hausamen T. U., Helger R., Rick W., Gross W. Optimal conditions for the determination of serum alkaline phosphatase by a new kinetic method. Clin. Chim. Acta 1967; 15:241-245
• Hirakawa T., Rokutan K., Nikawa T., Kishi K. Geranylgeranylacetone induces heat shock proteins in cultured guinea pig gastric mucosal cells and rat gastric mucosa. Gastroenterology 1996; 111:345-357[Medline]
• Jeng Y.-J., Watson C. S., Thomas M. L. Identification of vitamin D-stimulated alkaline phosphatase in IEC-6 cells, a rat small intestine crypt cell line. Exp. Cell Res. 1994; 212:338-343[Medline]
• Komoda T., Koyama I., Nagata A., Sakagishi Y., DeSchryver-Kecskemeti K., Alpers D. H. Ontogenic and phylogenic studies of intestinal, hepatic, and placental alkaline phosphatase. Gastroenterology 1986; 91:277-286[Medline]
• Kurokowa M., Lynch K., Podolsky D. K. Effects of growth factors on an intestinal epithelial cell line: transforming growth factor  inhibits proliferation and stimulates differentiation. Biochem. Biophys. Res. Commun. 1987; 142:775-782[Medline]
• Levin M. S., Davis A. E. Retinoic acid increases cellular retinol binding protein II mRNA and retinol uptake in the human intestinal Caco-2 cell line. J. Nutr. 1997; 127:13-17[Abstract/Free Full Text]
• Lorente C. A., Miller S. A. Fetal and maternal vitamin A levels in tissues of hypervitaminotic A rats and rabbits. J. Nutr. 1977; 107:1816-1821
• Lowry O. H., Rosebrough N. J., Farr A. L., Randall R. J. Protein measurement with the folin phenol reagent. J. Biol. Chem. 1951; 193:265-275[Free Full Text]
• Miyamoto K., Takagi T., Fujii T., Matsubara T., Hase K., Taketani Y., Oka T., Minami H., Nakabou Y. Role of liver-type glucose transporter (GLUT2) in transport across the basolateral membrane in rat jejunum. FEBS Lett. 1992; 314:466-470[Medline]
• Narisawa S., Frohlander N., Millan J. L. Inactivation of two mouse alkaline phosphatase genes and establishment of a model of infantile hypophosphatasia. Dev. Dyn. 1997; 208:432-446[Medline]
• Nikawa T., Schulz W. A., van den Brink C. E., Hanusch M., van der Saag P., Stahl W., Sies H. Efficacy of all-trans--carotene, canthaxanthin, and all-trans-, 9-cis-, and 4-oxoretinoic acids in including differentiation of an F9 embryonal carcinoma RAR-lacZ reporter cell line. Arch. Biochem. Biophys. 1995; 316:665-672[Medline]
• Plateroti M., Freund J. N., Leberquier C., Kedinger M. Mesenchyme-mediated effects of retinoic acid during rat intestinal development. J. Cell Sci. 1997; 110:1227-1238[Abstract]
• Plateroti M., Sambuy Y., Nobili F., Bises G., Perozzi G. Expression of epithelial markers and retinoid-binding proteins in retinol- or retinoic acid-treated intestinal cells in vitro. Exp. Cell Res. 1993; 208:137-147[Medline]
• Quaroni A., Wands J., Trelstad R. L., Isselbacher K. J. Epithelioid cell cultures from rat small intestine: characterization by morphologic and immunologic criteria. J. Cell Biol. 1979; 80:248-265[Abstract/Free Full Text]
• Reese D. H., Larsen R. A., Hornicek F. J. Control of alkaline phosphatase activity in C3H10T1/2 cells: role of retinoic acid and cell density. J. Cell. Physiol. 1992; 151:239-48[Medline]
• Sato N., Takahashi Y., Asano S. Preferential usage of the bone-type leader sequence for the transcripts of liver/bone/kidney-type alkaline phosphatase gene in neutrophilic granulocytes. Blood 1994; 83:1093-1101[Abstract/Free Full Text]
• Sato N., Takatani O., Koeffler H. P., Sato H, Asano S., Takaku F. Modulation by retinoids and interferons of alkaline phosphatase activity in granulocytes induced by granulocyte colony-stimulating factor. Exp. Hematol. 1989; 17:258-262[Medline]
• Seetharam S., Sussman N. L., Komoda T., Alpers D. H. The mechanism of elevated alkaline phosphatase activity after bile duct ligation in the rat. Hepatology 1986; 6:374-380[Medline]
• Studer M., Terao M., Giann M., Garattini E. Characterization of a second promoter for the mouse liver/bone/kidney-type alkaline phosphatase gene: cell and tissue specific expression. Biochem. Biophys. Res. Commun. 1991; 179:1352-1360[Medline]
• Suh E. R., Traber P. G. An intestine-specific homeobox gene regulates proliferation and differentiation. Mol. Cell. Biol. 1996; 16:619-625[Abstract]
• Swarup G., Cohen S., Garbers D. L. Selective dephosphorylation of proteins containing phosphotyrosine by alkaline phosphatases. J. Biol. Chem. 1981; 156:8197-8201
• Takahashi Y. I., Smith J. E., Winick M., Goodman D. S. Vitamin A deficiency and fetal growth and development in the rat. J. Nutr. 1975; 105:1299-1310
• Takeyama K., Kojima R., Ohashi T., Sato T., Mano H., Masushige S., Kato S. Retinoic acid differentially up-regulates the gene expression of retinoic acid receptor a and g isoforms in embryo and adult rats. Biochem. Biophys. Res. Commun. 1996; 222:395-400[Medline]
• Toh Y., Yamamoto M., Endo H., Misumi Y., Ikehara Y. Isolation and characterization of a rat liver alkaline phosphatase gene: a single gene with two promoters. Eur. J. Biochem. 1989; 182:231-237[Medline]
• Traber P. G. Regulation of sucrase-isomaltase gene expression along the crypt-villus axis of rat small intestine. Biochem. Biophys. Res. Commun. 1990; 173:765-773[Medline]
• Trowsdale J., Martin D., Bicknell D., Campbell I. Alkaline phosphatases. Biochem. Soc. Trans. 1990; 18:178-180[Medline]
• Waymire K. G., Mahuren J. D., Jaje J. M., Guilarte T. R., Coburn S. P., MacGregor G. R. Mice lacking tissue non-specific alkaline phosphatase die from seizures due to defective metabolism of vitamin B-6. Nat. Genet. 1995; 11:45-51[Medline]
• Weiss M. J., Henthorn P. S., Lafferty M. A., Slaughter C., Raducha M., Harris H. Isolation and characterization of a cDNA encoding a human liver/bone/kidney-type alkaline phosphatase. Proc. Natl. Acad. Sci. U.S.A. 1986; 83:182-189
• Zelent A., Krust A., Petkovich M., Kastner P., Chambon P. Cloning of murine  and  retinoic acid receptors and a novel receptor  predominantly expressed in skin. Nature (Lond.) 1989; 339:714-717[Medline]
• Zernik J., Kream B., Twarog K. Tissue-specific and dexamethasone-inducible expression of alkaline phosphatase from alternative promoters of the rat bone/liver/kidney/placenta gene. Biochem. Biophys. Res. Commun. 1991; 176:1149-1155[Medline]
• Zhou H., Hammonds R. G. Jr., Findlay D. M., Fuller P. J., Martin T. J., Ng K. W. Retinoic acid modulation of mRNA levels in malignant, nontransformed, and immortalized osteoblasts. J. Bone Miner. Res. 1991; 6:767-777[Medline]
• Zhou H., Manji S. S., Findlay D. M., Martin J., Heath J. K., Ng K. W. Novel action of retinoic acid. J. Biol. Chem. 1994; 269:22433-22439[Abstract/Free Full Text]

0022-3166/98 $3.00 ©1998 American Society for Nutritional Sciences