The –700/–310 Fragment of the Apolipoprotein A-IV Gene Combined with the –890/–500 Apolipoprotein C-III Enhancer Is Sufficient to Direct a Pattern of Gene Expression Similar to That for the Endogenous Apolipoprotein A-IV Gene*

Spatial gene expression in the intestine is mediated by specific regulatory sequences. The three genes of the apoA-I/C-III/A-IV cluster are expressed in the intestine following cephalocaudal and crypt-to-villus axes. Previous studies have shown that the –780/–520 enhancer region of the apoC-III gene directs the expression of the apoA-I gene in both small intestinal villi and crypts, implying that other unidentified elements are necessary for a normal intestinal pattern of apoA-I gene expression. In this study, we have characterized transgenic mice expressing the chloramphenicol acetyltransferase gene under the control of different regions of the apoC-III and apoA-IV promoters. We found that the –890/+24 apoC-III promoter directed the expression of the reporter gene in crypts and villi and did not follow a cephalocaudal gradient of expression. In contrast, the −700/+10 apoA-IV promoter linked to the −500/−890 apoC-III enhancer directed the expression of the reporter gene in enterocytes with a pattern of expression similar to that of the endogenous apoA-IV gene. Furthermore, linkage of the −700/−310 apoA-IV distal promoter region to the −890/+24 apoC-III promoter was sufficient to restore the appropriate pattern of intestinal expression of the reporter gene. These findings demonstrate that the −700/−310 distal region of the apoA-IV promoter contains regulatory elements that, in combination with proximal promoter elements and the −500/−890 enhancer, are necessary and sufficient to restrict apoC-III and apoA-IV gene expression to villus enterocytes of the small intestine along the cephalocaudal axis.

The mammalian small intestine is lined with a constantly renewing epithelium that is compartmentalized into a proliferative undifferentiated zone located in intestinal crypts and a nonproliferative differentiated zone located in the villi. The epithelium is composed of four specialized cell types that arise from stem cells located just above the base of the crypts (1). Enterocytes, mucus-producing goblet cells, and enteroendocrine cells differentiate as they migrate to the top of the villi, whereas Paneth cells differentiate and migrate to the base of the crypts. Each lineage completes its differentiation program through an orderly migration (2,3). Enterocytes are the most abundant epithelial cells in the small intestine and express a variety of specific genes as they exit the crypt compartment. Despite the rapid renewal of the intestinal epithelium, numerous genes display a specific pattern of expression in enterocytes from the proximal to the distal intestine and from the crypt to the villus tip (4,5).
Several studies performed with transgenic mice expressing a human gene or a reporter gene have established that spatial gene expression in the intestine is supported by specific regulatory sequences (6 -11). Transcription of the intestinal fatty acid-binding protein (FABP-I) 1 gene is strictly confined to the intestinal epithelium. The FABP-I promoter (Ϫ103/ϩ28) is sufficient to direct transcription of the gene along the duodenumto-colon axis. However, upstream sequences are needed to confine FABP-I expression to differentiated enterocytes of the villus. In particular, a 20-bp element located between nucleotides -263 and -244 of the promoter prevents FABP-I expression in crypt cells (6,10). Liver and intestinal human apoB gene expression is governed by distinct regulatory regions. Intestinal expression requires a very distant element located between 33 and 70 kb 5Ј upstream from the apoB gene (11). Similarly, the Ϫ256/ϩ22 proximal promoter of the human apoA-I gene is sufficient to direct its hepatic transcription. However, the sequences responsible for the intestinal expression reside 9 kb downstream from the human apoA-I gene (12).
The apoA-I gene is located on chromosome 11 in a cluster that also contains the apoC-III and apoA-IV genes. The human apoA-I gene is expressed at similar levels in the intestine and liver, whereas the human apoC-III gene is expressed predominantly in the liver and to a lesser extent in the intestine (13). The apoA-IV gene is mainly expressed in the intestine in humans and non-human primates. ApoA-IV is also expressed in the liver in mice (13,14). The apoA-I and apoA-IV genes are transcribed in the same direction, whereas the apoC-III gene is transcribed in the opposite direction. The apoC-III/A-IV intergenic region therefore constitutes a common 6.6-kb 5Ј-flanking sequence for these two genes.
Since the three genes are expressed at different levels in the liver and intestine, this gene cluster represents an interesting model to decipher the molecular mechanisms involved in the determination of tissue-specific expression. The intestinal expression of the cluster is entirely restricted to enterocytes as they emerge from the crypt (15,16) and decreases from the proximal to the distal small intestine (14). A preliminary report indicated that the apoC-III/A-IV intergenic region allows the intestinal expression of the apoA-IV gene in transgenic mice (17). More recently, Bisaha et al. (16) have demonstrated that the Ϫ890/Ϫ500 apoC-III enhancer is sufficient to direct the intestinal expression of the apoA-I gene. Based on these in vivo studies and previous in vitro promoter studies performed by us and others, we hypothesized that common regulatory sequences control the intestinal expression of the three genes of the cluster.
In this study, we generated transgenic mice expressing the CAT reporter gene under the control of specific regulatory sequences of the apoA-IV/C-III intergenic region. Analysis of these mouse lines showed that the Ϫ700/Ϫ310 apoA-IV promoter in combination with the Ϫ500/Ϫ890 apoC-III enhancer is sufficient for correct gene expression in the enterocytes along the proximal-to-distal and crypt-to-villus axes.

MATERIALS AND METHODS
Generation of Transgenic Mice-The different transgenes used in this study are shown in Fig. 1. The transgenes C3-CAT and eC3-A4-CAT were obtained by digestion of the pUCSH-CAT plasmid containing the Ϫ890/ϩ24 5Ј-flanking region of the human apoC-III gene or the -700/ϩ10 apoA-IV promoter region fused with the Ϫ500/Ϫ890 apoC-III enhancer region, respectively, with XbaI and BamHI (see Fig. 1A). These plasmids have previously been described (18,19).
The transgene dA4-C3-CAT was obtained from a plasmid in which the human Ϫ890/ϩ24 apoC-III promoter region fused upstream from the CAT reporter gene was linked in the opposite direction to the Ϫ700/ϩ10 apoA-IV promoter region fused with the lacZ reporter gene. This vector was constructed as follows. The Ϫ700/ϩ10 apoA-IV promoter region was amplified by polymerase chain reaction using nucleotide primers from Ϫ700 to Ϫ680 (coding strand) and from ϩ10 to Ϫ10 (noncoding strand) containing a SalI and a HindIII restriction site, respectively. The resulting apoA-IV fragment was cloned upstream from the lacZ gene fused to a nuclear localization sequence (20); the C3-CAT fragment was then inserted in the BamHI and XbaI sites, upstream from the apoA-IV promoter in the opposite direction. The dA4-C3-CAT transgene was excised from the plasmid by digestion with BamHI and SmaI at a site located at nucleotide -310 in the apoA-IV promoter. Transgenes were dissolved at a concentration of 4 ng/l and microinjected into fertilized eggs from C57BL/6J ϫ CBA/J females mated with males of the same strain using established procedures (21).
Characterization of Transgenic Mice-DNA was extracted from the tails of 10 -15-day-old pups and then analyzed by polymerase chain reaction using oligonucleotide primers corresponding to sequences ϩ163 to ϩ189 (coding strand) and ϩ696 to ϩ670 (noncoding strand) of the CAT gene. Founder mice were then further analyzed by Southern blotting, and the copy number was estimated by densitometric scanning of autoradiograms as described previously (21). Positive founder F 0 mice were outbred to generate lines of heterozygous mice.
CAT Assay-Individual tissue samples were homogenized and assayed for CAT activity as described previously (22,23). The concentration of soluble protein was determined by the Bio-Rad protein assay. The percentage of chloramphenicol converted to acetylated forms was determined either by densitometric scanning of autoradiograms or by scraping individual spots from the thin-layer chromatogram and counting in a scintillation counter. CAT activity is expressed as pmol of acetyl chloramphenicol generated per min/mg of protein after subtracting the background for each tissue from control mice, which do not express the CAT gene.
Preparation of Radiolabeled Probes-Specific 300-bp cDNAs encoding the mouse apoC-III and apoA-IV genes were obtained by reverse transcription-polymerase chain reaction amplification and subcloned in the pBluescript KS plasmid. Oligonucleotides 5Ј-AGCCCAAGCTT-ϩ578 ATGCAGCCCCGGACGCTCCTCA ϩ599 -3Ј (coding strand) and 5Ј-G-CTCTAGA ϩ2051 TCACGACTCATAGCTGGAGTTGG ϩ2028 -3Ј (noncoding strand) and 5Ј-AGCCCAAGCTT ϩ1 AATCTGCACAGGGACACAGGTA-CA ϩ24 -3Ј (coding strand) and 5Ј-GCTCTAGA ϩ300 TAGCACCCCAAGTT-TGTCCTGGA ϩ277 -3Ј (noncoding strand) were used for the amplification of apoC-III and apoA-IV sequences, respectively (24, 25). The two cDNAs were digested with XbaI and HindIII and ligated into the pBluescript KS vector that had previously been digested with XbaI/ HindIII. A 265-bp fragment of the CAT gene was obtained from pUCSH-CAT by digestion with HindIII and EcoRI and cloned into the pBluescript SK vector.
Both sense and antisense mouse apoA-IV and apoC-III RNA probes (size: 300 bp) were generated using T3 and T7 RNA polymerases, respectively (Promega). Sense and antisense CAT riboprobes were synthesized with T7 and T3 polymerases, respectively. All probes were labeled using 35 S-UTP.
In Situ Hybridization-Adult mice were killed by cervical dislocation, and the entire small intestines were rapidly removed and divided into three parts representing the proximal, middle, and distal regions of the small intestine. The samples were fixed in 2% paraformaldehyde in phosphate-buffered saline, pH 7.2, and embedded in paraffin. Sections (4 m thick) were mounted on glass slides.
In situ hybridization was performed by a modification of the method of Sassoon and Rosenthal (26). Sections of small intestine were hybridized with 200,000 cpm of probe/slide at 42°C overnight. Tissues were washed for 30 min in 5ϫ SSC (1ϫ SSC ϭ 0.15 M NaCl and 0.015 M sodium citrate, pH 7.0) and 10 mM dithiothreitol at 50°C, two times for 20 min in 2ϫ SSC and 50% formamide at 60°C, and for 10 min in 1ϫ SSC at 37°C and then treated for 30 min with RNase A. Subsequent washes in 1ϫ and 0.1ϫ SSC at 37°C were followed by dehydration in graded ethanol for total desiccation. The washed slides were dipped into Kodak NTB2 emulsion, stored in the dark at Ϫ20°C for 8 -16 days, and then developed.
RNA Preparation and Analysis-Total RNA from intestinal segments was extracted (RNazol kit, Bioprobe Systems), separated on 1.2% denaturing formaldehyde-agarose gels, and transferred onto Hybond N ϩ membranes (Amersham Pharmacia Biotech). The membranes were hybridized with the [ 32 P]dCTP-labeled mouse apoC-III and apoA-IV cDNA probes in 50% formamide, 5ϫ SSC, 5ϫ Denhardt's solution, 0.04% pyrophosphate, 5 mM sodium P i , and 0.1 mg/ml herring sperm DNA at 42°C. Washings were performed in 1ϫ SSC and 0.1% SDS at 65°C. The quality and the amount of RNA samples were estimated using an 18 S RNA probe. Autoradiograms were scanned using an NIH Imager analysis program.
Histoenzymology-Intestinal CAT activity was measured in the small intestine using the technique of Donoghue et al. (27,28). Additional acetyl-CoA was added after 12 h of incubation. Slides were prepared from tissues of nontransgenic mice. Mice were analyzed and utilized as controls. Tissues were embedded in paraffin, and sections (4 m thick) were stained using the periodic acid-Schiff technique to identify goblet cells and the Grimelius silver method to identify enteroendocrine cells.

RESULTS
Generation of Transgenic Mice-We generated transgenic mice expressing the CAT reporter gene under the control of either the human -890/ϩ24 apoC-III promoter (C3-CAT) (Fig.  1B) or the human -700/ϩ10 apoA-IV promoter fused to the -500/-890 apoC-III enhancer in the opposite direction to the apoA-IV promoter in accordance with the organization of the apoA-I/C-III/A-IV gene cluster (eC3-A4-CAT) (Fig. 1B). The number of transgene copies incorporated into the genome of each transgenic founder was determined by Southern blot analysis in comparison with increasing amounts of the transgene diluted in nontransgenic DNA ( Fig. 1C and Table I).
CAT activity was determined in tissue extracts from adult transgenic mice of each line (Table I). In all cases, CAT activity was mainly detected in the liver and small intestine, although ectopic expression of the CAT gene was observed in the heart of C3-CAT transgenic mice. Despite differences in the level of expression of the transgene, tissular distribution of CAT activity was similar in the different lines produced with the same transgene. This distribution differed in different transgenes. In C3-CAT transgenic mice, the CAT activity in any part of the intestine did not exceed one-third of the liver activity measured. In contrast, in eC3-A4-CAT transgenic mice, the level of CAT activity in the proximal and middle regions of the small intestine was similar to that in the liver. From each construct, two mouse lines exhibiting the highest CAT activity were fur-ther studied: lines V and W of C3-CAT transgenic mice and lines Zf and Ze of eC3-A4-CAT transgenic mice.
Spatial Pattern of C3-CAT and eC3-A4-CAT Expression in the Small Intestine-The spatial pattern of expression of the transgenes in the intestine along the crypt-to-villus and cephalocaudal axes was further determined and compared with that of endogenous mouse apoC-III and apoA-IV genes (Fig. 2). CAT activity did not vary along the small intestine when CAT gene expression was controlled by the -890/ϩ24 apoC-III promoter ( Fig. 2A). In contrast, CAT activity displayed a decreasing cephalocaudal gradient in eC3-A4-CAT mice ( Fig. 2A). This expression pattern mimicked that of endogenous mouse apoA-IV and apoC-III mRNAs (Fig. 2B).
Transgenic expression along the crypt-to-villus axis in the intestine was analyzed by in situ hybridization (Fig. 3). To demonstrate the specificity of the signal, an antisense and a sense CAT riboprobe were first hybridized to sections of nontransgenic and transgenic jejunum, respectively, as negative controls (Fig. 3, a and b). Using dark-field microscopy, a few scattered grains representing the background signal were seen with both probes. The CAT reporter gene driven by the human Ϫ890/ϩ24 apoC-III promoter was expressed markedly in the crypt cells and lightly in the villus epithelial cells of the transgenic mice (Fig. 3d). The endogenous mouse apoC-III gene was expressed similarly in the crypt and villus epithelial cells (Fig.  3c). In contrast to C3-CAT transgenic mice, eC3-A4-CAT transgenic mice exhibited a pattern of CAT mRNA expression in the crypt-to-villus unit that was strikingly similar to that of the endogenous mouse apoA-IV mRNA (Fig. 3, e and f). These results suggest that the Ϫ700/ϩ10 apoA-IV promoter contains a regulatory region that, in combination with the Ϫ500/Ϫ890 apoC-III enhancer, restricts the expression of the reporter gene to the villus, thus reproducing in the transgenic mice the cryptto-villus gradient of expression that is observed for the endogenous apoA-IV gene.
The ApoA-IV Distal Promoter Confers a Cephalocaudal and Crypt-to-Villus Expression Gradient to C3-CAT Transgenes-To determine which region of the apoA-IV promoter was responsible for the expression pattern along both the cephalocaudal and crypt-to-villus axes, we generated additional mouse lines expressing the reporter CAT transgene under the control of the -890/ϩ24 apoC-III promoter fused to the -310/ Ϫ700 apoA-IV distal promoter region (Fig. 1B). Three founders were identified by polymerase chain reaction and were analyzed by Southern blotting (Fig. 1C).
As in C3-CAT transgenic mice, CAT activity was much higher in the liver than in the intestine of dA4-C3-CAT mice. However, in the latter, we observed a decreasing gradient of CAT activity from the proximal to the distal region of the intestine (Fig. 4). This pattern of expression differed from that of C3-CAT transgenic mice (Fig. 2). Thus, the addition of the -310/-700 apoA-IV distal promoter region to the -890/ϩ24 apoC-III promoter restored the cephalocaudal pattern of expression observed for the endogenous apoC-III gene.
The expression of the transgene in the crypt-to-villus unit was visualized by histochemical staining of the nuclei of CATexpressing cells in the small intestine. No staining was observed in control mice (Fig. 5, a and aЈ) or in the lamina propria of transgenic mice (Fig. 5). CAT histochemistry revealed a pattern of expression similar to that observed previously by in allowed the determination of the copy number in each line. Tail DNA from a normal mouse was used for a negative control (control). Three transgenic mouse lines were analyzed for the C3-CAT transgene (V, W, and Y), the eC3-A4-CAT transgene (Zf, Zg, and Ze), and the dA4-C3-CAT transgene (A, B, and C).

FIG. 1. Maps of the different transgenes used to generate transgenic mice and Southern blot analysis to determine transgene integration in the different transgenic mice.
A, schematic representation of the apoA-I/C-III/A-IV gene cluster. The direction of transcription for each gene is shown by an arrow. The length of this cluster of genes is ϳ15 kb. The proximal promoters of apoA-I and apoC-III suffice for in vivo hepatic expression (250 and 500 bp in length, respectively). The -500/-890 apoC-III enhancer is required for intestinal apoA-I expression. The proximal promoter of apoA-IV (700 bp) is practically inactive in HepG2 and Caco-2 cells (19). B, maps of CAT constructs driven by segments of the human apoC-III and/or apoA-IV promoter that were used to generate transgenic mice (21). The C3-CAT transgene retains the entire human -890/ϩ24 apoC-III promoter upstream from the CAT reporter gene. The eC3-A4-CAT corresponds to the -700/ϩ10 apoA-IV promoter linked to the -500/-890 apoC-III enhancer. The two promoter segments are in their "normal" opposite direction. The dA4-C3-CAT contains the -890/ϩ24 apoC-III promoter sequence fused to the Ϫ700/-310 apoA-IV distal gene promoter region. The two promoter segments are in their normal opposite direction. pb, base pair. C, Southern blot analysis of the different constructs in transgenic mice. EcoRI-digested mouse tail DNA (10 g) was loaded onto a 0.8% agarose gel and then transferred to a Hybond N ϩ membrane and hybridized with a CAT probe (see "Materials and Methods"). Comparison with known copy numbers of transgenic DNA (20 to 0.5) situ hybridization. Crypt and villus nuclei were stained in C3-CAT transgenic samples (Fig. 5, b and bЈ), whereas staining was restricted to the villus in eC3-A4-CAT transgenic mice (c and cЈ) as well as in dA4-C3-CAT transgenic samples (d and dЈ).
Thus, the addition of the Ϫ310/Ϫ700 apoA-IV distal promoter region restricted the expression of the dA4-C3-CAT transgene to the villus in a pattern similar to that of endogenous apoA-IV, but not of endogenous apoC-III.  After double staining the villus epithelial cells from dA4-C3-CAT intestine, neither enteroendocrine cells, visualized by Grimelius staining (Fig. 6, a and b), nor the goblet cells, visualized by periodic acid-Schiff staining (Fig. 6c), coincided with the specific staining of nuclei of CAT-expressing cells. These results suggest that the Ϫ310/Ϫ700 apoA-IV distal promoter region is sufficient to restrict gene expression to villus enterocytes along the cephalocaudal axis. DISCUSSION A preliminary report has shown that the entire intergenic region between the apoC-III and apoA-IV genes directs a pattern of expression of the transgene similar to that of endogenous apoA-IV (17). This expression was abolished with shorter constructs lacking the apoC-III promoter region. Bisaha et al. (16) showed that the apoC-III enhancer, located at nucleotide -520 upstream from the transcription initiation site of the apoC-III gene, is sufficient to direct the intestinal expression of the apoA-I gene, the third gene of the apoA-I/C-III/A-IV cluster, but not to restrict its expression to the villus. This prompted us to decipher the regulatory regions responsible for the accurate expression of apoA-IV by combining the apoC-III enhancer and the apoA-IV promoter. In our present study, we found that a combination of the Ϫ890/Ϫ500 apoC-III enhancer and the Ϫ700/Ϫ310 apoA-IV distal promoter allowed the intestinal expression of the apoA-IV gene in transgenic mice, specifically in the villus enterocytes, with a cephalocaudal gradient. Taken together, these results demonstrate that the appropriate lineage-specific crypt-to-villus and cephalocaudal patterns of human apoA-IV expression in transgenic mouse intestine require both the apoC-III enhancer and the apoA-IV distal promoter.
Our findings indicate that CAT activity in the intestine and liver in mice expressing the eC3-A4-CAT transgene reproduces the correct pattern of expression of endogenous mouse apoA-IV rather than that of human apoA-IV, which is predominantly expressed in the intestine. Lauer et al. (17) reported no expression of the human apoA-IV gene in the liver of transgenic mice expressing human apoA-IV genomic sequences containing 5Јflanking regions 0.3-7.7 kb long and a 3Ј-flanking region 1.5 kb long. These results, taken conjunction with our own, suggest that a hepatic silencer may reside downstream from nucleotide ϩ24 of the human apoA-IV gene. It is possible that either a silencer in the human apoA-IV promoter or differences in the nuclear activities between humans and rodents may account for this difference in tissue-specific expression of the apoA-IV gene. Similarly, the weak ectopic expression of the C3-CAT and eC3-A4-CAT transgenes in tissues other than those of the liver and intestine may reflect the lack of tissue-specific silencers in the Ϫ890/ϩ24 apoC-III and Ϫ700/ϩ10 apoA-IV promoter regions.
In vitro transfection assays in the Caco-2 cell line showed that the transcription of the apoA-IV gene is controlled by hepatocyte nuclear factor 4 (HNF4), which binds to its proximal promoter region and requires the presence of other transcription factors that recognize elements of the apoC-III enhancer (19,29). Similarly, the transcription of apoA-I and apoC-III genes is driven through a synergy between HNF4 and others factors that bind the apoC-III enhancer (16, 30 -34). HNF4 binds the hormone response elements located in the three proximal promoters and in the apoC-III enhancer. Bisaha et al. (16) have shown that a region of the apoC-III enhancer containing the HNF4-binding site is insufficient to drive in vivo the intestinal expression of the apoA-I gene. Nevertheless, this factor could actively participate in the determination of intestinal apoA-I/C-III/A-IV gene expression. An HNF4-binding site has also been described in the FABP-I proximal enhancer, which is essential for intestinal expression (35). Furthermore, the involvement of HNF4 in the onset of intestinal functions has been demonstrated by the interruption of intestine development under extinction of the HNF4 homolog in Drosophila (36).
The eC3-A4-CAT transgene, which retains the Ϫ890/Ϫ500 apoC-III enhancer and Ϫ700/ϩ10 apoA-IV promoter regions, was able to direct a pattern of CAT activity among intestinal segments that resembled both the pattern of endogenous mouse apoA-IV and that observed in rats and chickens (14,37). Furthermore, the expression of the reporter gene was restricted to villus cells in a manner similar to the expression pattern of the endogenous mouse apoA-IV gene and also to that observed in rats (38). These results suggest that the apoA-IV promoter contains an element prohibiting the transcription of the reporter gene in crypt epithelial cells and in the distal part of the small intestine.
As already discussed, the apoC-III promoter region was not sufficient to restrict intestinal expression along the crypt-tovillus and cephalocaudal gradients. The addition of the Ϫ700/ Ϫ310 apoA-IV distal promoter to the apoC-III enhancer allowed the expression of the reporter gene to mimic the cephalocaudal gradient displayed by endogenous mouse apoC-III. Furthermore, the Ϫ700/Ϫ310 apoA-IV promoter region retained the elements sufficient to restrict expression to villusassociated enterocytes. These findings indicate that the Ϫ310/ Ϫ700 apoA-IV promoter region confers two suppressor functions, one prohibiting gene expression in the distal small intestine and the other prohibiting gene expression in crypt epithelial cells.
The spatial patterns of gene expression in the intestine involve both positive and negative elements. This has also been reported for the rat FABP-I gene, the rat liver fatty acidbinding protein gene, and the sucrase-isomaltase gene ( Histochemical staining for CAT activity with no counterstaining was performed as described under "Materials and Methods." Sections were photographed with phase-contrast (a-d) and with bright-field (aЈ-dЈ). Nontransgenic proximal intestine was incubated in the complete staining mixture (a and aЈ). No staining was observed in either crypt (asterisks) or villus (arrows) epithelial cells. In transgenic sections, only nuclei were stained (visible as a black deposit). This staining appears to be specific since it was observed only in the transgenic mice (compare a and aЈ with the other panels). In the proximal section of the intestine from C3-CAT transgenic mice (b and bЈ), CAT staining was observed in both crypt and villus epithelial cells. The patterns of CAT staining along the crypt-to-villus axis of adult mice expressing the eC3-A4-CAT (c and cЈ) and dA4-C3-CAT (d and dЈ) constructs are indistinguishable and are restricted to villus epithelial cells.
FABP-I promoter that modulates the intestinal cellular and spatial expression of the FABP-I gene. This element acts as a suppressor of gene expression in the distal small intestine/ colon, as a suppressor of gene activation in the crypt, and as a suppressor of gene expression in the Paneth cell lineage (10). No sequence in the Ϫ700/Ϫ310 apoA-IV distal promoter signif-icantly matched this 20-bp element. Thus, it is reasonable to hypothesize that in both apoA-IV and FABP-I, a distinct element controls a similar appropriate pattern of gene expression. Whether these two different elements bind similar or different repressors of gene expression in crypts and the distal part of the intestine remains to be established.