Proc. Natl. Acad. Sci. USA Vol. 90, pp. 11356-11360, December 1993 Developmental Biology Identification of the promoter and a transcriptional enhancer of the gene encoding L-CAM, a calcium-dependent cell adhesion molecule (gene expression/regulatory sequences/morphogenesis/cadherins) BARBARA C. SORKIN, FREDERICK S. JONES, BRUCE A. CUNNINGHAM, AND GERALD M. EDELMAN The Department of Neurobiology, The Scripps Research Institute, 10666 North Torrey Pines Road, La Jolla, CA 92037 Contributed by Gerald M. Edelman, August 20, 1993

ABSTRACT L-CAM is a calcium-dependent cell adhesion expressed in the dermis, but rather in the adjacent epidermis molecule that is expressed in a characteristic place-dependent (6). pattern during development. Previous studies of ectopic ex- L-CAM mediates calcium-dependent cell-cell adhesion (7) pression of the chicken L-CAM gene under the control of via a homophilic mechanism; i.e., L-CAM on one cell binds heterologous promoters in transgenic mice suggested that directly to L-CAM on apposing cells (8). In all tissues where cis-acting sequences controlling the spatiotemporal expression the molecule is expressed, L-CAM is detected as a trans- patterns ofL-CAM were present within the gene itself. We have membrane protein of 124 kDa (7). Mammalian proteins now examined the L-CAM gene for sequences that control its uvomorulin (9), E-cadherin (10), cell-CAM 120/80 (11), and expression and have found an enhancer within the second Arcl (12) are similar to L-CAM in their biochemical and intron of the gene. A 2.5-kb Kpn I-EcoRI fragment from the functional properties and tissue distributions. The calcium- intron acted as an enhancer of a simian virus 40 minimal dependent CAMs also include B-cadherin (13), M-cadherin promoter driving a chloramphenicol acetyltransferase (CAT) (14), N-cadherin (15), P-cadherin (16), R-cadherin (17), and reporter gene and produced 14.0-fold induction of CAT activ- T-cadherin (18). ity in MDCK cells. To narrow down the region responsible for The chicken L-CAM gene is "10 kb in length and contains enhancer activity and to determine whether the enhancer could 16 exons (19). A search for the upstream sequences that function in a cell type-specific manner, a number of smaller regulate expression of the L-CAM gene revealed another restriction fragments from the intron were tested for activity in gene, designated K-CAM, encoding a calcium-dependent two chicken cell lines, the LMH hepatoma line, which produces CAM very similar to L-CAM only 700 bp upstream of the high levels of L-CAM, and the SL-29 fibroblast line, which L-CAM translation initiation codon (20). Sequence analyses produces little, if any, L-CAM. Four L-CAM enhancer plas- and comparisons have indicated that K-CAM and B-cadherin mids containing shorter segments derived from the intron (13) are the same molecule. showed enhanced CAT activity levels (between 9.4- and 16.5- Studies of the ectopic expression of the chicken L-CAM fold) in extracts from transfected LMH cells but not from SL-29 gene in transgenic mice directed by either the rat insulin cells. DNA sequence analysis of the L-CAM enhancer region promoter or the mouse neurofilament promoter suggested revealed putative binding sites for the transcription factors that the L-CAM gene contained cis-regulatory sequences SP1, E2A, and AP-2. In addition, LE-9, the smallest L-CAM within its introns (21). This conclusion was prompted by the enhancer segment (310 bp), contained a consensus binding site observation that, although its expression was directed to for the liver-enriched POU-homeodomain transcription factor, tissue sites normally expected for the heterologous promot- HNF-1. Tests of upstream sequences showed that a 630-bp ers, the chicken L-CAM gene was also expressed in some fragment, corresponding to nearly the entire intergenic region tissues where the heterologous promoter was not expected to between L-CAM and its neighboring CAM gene, K-CAM, be active, but where L-CAM is normally expressed (kidney, could function as a promoter. In combination with the L-CAM liver, intestine, and lung). These results were reproduced in enhancer, this fragment directed cell type-specific expression of at least two independently derived pedigrees for each L-CAM the CAT reporter gene in LMH cells at a level comparable to construct. This suggested that regulatory elements within the that observed with enhancer constructs using the simian virus introns of the L-CAM gene, acting in conjunction with the 40 minimal promoter. These combined observations derme a ectopic promoter, were creating combinatorial patterns of promoter and an enhancer for the chicken L-CAM gene. They expression independent of the chromosomal integration site raise the possibility that these cis-acting regulatory sequences of the transgene. may be instrumental in directing specific place-dependent In the present study, we describe the identification and expression of the L-CAM gene in the chicken. characterization ofan enhancer within the second and largest intron of the L-CAM gene.* In addition, we show that the intergenic region between the K-CAM and L-CAM genes The liver cell adhesion molecule (L-CAM) is expressed early functions as a promoter. When combined with the enhancer, in chicken embryonic development and persists in nonneural this region was found to regulate expression of a reporter epithelia in the adult (1). During development, changes in gene in a cell type-specific manner. L-CAM expression are spatiotemporally correlated with in- ductive events (2). Consistent with this observation, trans- fection and antibody perturbation studies (3-5) indicate that MATERIALS AND METHODS L-CAM plays a critical role in morphogenesis, particularly in The chicken fibroblast cell line SL-29 and the Madin-Darby the formation of epithelia. For example, antibodies to canine cell lines were obtained from the L-CAM disrupt feather development by altering pattern kidney (MDCK) formation in the dermis, despite the fact that L-CAM is not Abbreviations: SV40, simian virus 40; CAT, chloramphenicol ace- tyltransferase; CAM, cell adhesion molecule; L-CAM, liver CAM; The publication costs of this article were defrayed in part by page charge N-CAM, neural CAM. payment. This article must therefore be hereby marked "advertisement" *The sequence reported in this paper has been deposited in the in accordance with 18 U.S.C. §1734 solely to indicate this fact. GenBank data base (accession no. U02633). 11356 Downloaded by guest on September 30, 2021 Developmental Biology: Sorkin et aL Proc. Natl. Acad. Sci. USA 90 (1993) 11357 American Type Culture Collection and were cultured in K-CAM L-CAM Dulbecco's modified Eagle's medium containing 10%o (vol/ 16 2 vol) fetal bovine serum. The chicken hepatocellular carcino- ma-derived cell line LMH (22) was grown in dishes coated with a gelatin substrate in Waymouth's MB752/1 medium containing 10%o fetal bovine serum. 1 - 3 Western blots were performed as described (23). Confluent 60-mm dishes of cells were washed twice with phosphate- buffered saline containing 0.5 mM CaCl2 and 0.5 mM MgC92, and each dish was extracted with 0.5 ml of 2x Laemmli sample buffer (24). Samples were resolved by electrophoresis on a 7.5% polyacrylamide gel and transferred to nitrocellu- lose (23). L-CAM was detected on blot transfers using rabbit anti-L-CAM IgG followed by incubation with 1251-labeled protein A. 12 3 4 Enhancer constructs were prepared (25) from segments of the second intron of the L-CAM gene. Ten different restric- tion fragments (LE-1-LE-10) were inserted into theXba I site downstream ofthe chloramphenicol acetyltransferase (CAT) gene in the pCAT-Promoter vector (Promega). For promoter constructs, a 630-bp Mbo I-Kpn I fragment, corresponding to almost the entire intergenic region between the K-CAM and L-CAM genes, was cloned into the Xba I site upstream ofthe CAT gene in the promoterless CAT vector, pCAT-Basic (Promega). To assay the enhancer in the presence of the L-CAM promoter, the Sst I fragment from the L-CAM FIG. 1. A DNA fragmnent from the second intron of the L-CAM second intron (LE-6) was cloned into the BamHI site down- gene confers activation of SV40 promoter-mediated CAT gene stream of the CAT gene. expression in MDCK cells. (Upper) Diagram of the 16th exon (filled Cells cultured in 60-mm dishes were transfected in dupli- box) ofthe K-CAM gene and the first three exons (open boxes) ofthe cate with 25 .g ofCAT construct, 30 pl ofLipofectin reagent L-CAM gene. The 2.5-kb Kpn I-EcoRI fragment of the L-CAM (BRL) in serum-free medium (Optimem; GIBCO/BRL), and second intron was inserted into the Xba I site in the pCAT-Promoter 1 ug of the ,-galactosidase expression vector pCMVb vector. The vector contains an SV40) promoter (hatched) followed by (Promega). After a minimum of5 h, the medium was replaced the CAT gene. (Lower) One hundred-millimeter dis'hes of MDCK cells were transfected with pCAT-Promoter (lane 1), pCAT- with that containing 10%o serum, and cells were allowed to Promoter containing the 2.5-kb intron fragment in the foreword grow an additional 18-60 h. ,B-Galactosidase and CAT assays orientation (lane 2), pCAT-Promoter containing the 2.5-kb intron were performed as described (25). To normalize for differ- fragment in the reverse orientation (lane 3), and pSV2CAT, a ences in transfection efficiency, equal amounts of 3-galacto- construct in which the CAT gene is driven by both the SV40promoter sidase activity were used for each CAT assay. The 14C- and SV40 enhancer (lane 4). Cells were harvested after 60 h and labeled chloramphenicol and acetylated products were re- assayed for CAT activity. Acetylated forms ofchloramphenicol were solved by thin-layer chromatography and detected by resolved by thin-layer chromatography and detected by autoradiog- exposure to film or by using a Phosphorlmager (Molecular raphy. Dynamics). All results shown are the average ofat least three To test whether the L-CAM enhancer could function in a separate experiments and are given as the ratio of CAT cell type-specific manner in cell lines derived from the activity observed for the L-CAM enhancer construct to chicken, cell lines that showed both high and low levels of the basal activity of the pCAT-Basic or pCAT-Promoter L-CAM expression were selected by using immunoblot anal- vector. with antibodies. We found that the chicken DNA sequencing of the L-CAM second intron was deter- ysis L-CAM mined on fragments cloned in pBluescript vectors (Strata- LMH hepatoma cell line (22) expressed high levels ofL-CAM gene) using the Sequenase system (United States Biochem- protein, whereas the chicken SL-29 fibroblast cell line did not (Fig. 2A). These two cell lines were then used to assay the ical) and aligned using the Staden software (26). 2.5-kb Kpn I-EcoRi fragment that showed enhancer activity in MDCK cells and a number of smaller fragments (LE-1- RESULTS LE-10) derived from it for enhancer activity. The 2.5-kb Kpn To determine whether regions of the L-CAM gene had I-EcoRI fragment stimulated CAT activity 12-fold over the enhancer activity, we inserted various fragments of the gene levels produced by the vector control in LMH cells but into the pCAT-Promoter vector, in which expression of a showed only a modest increase of 1.8-fold in SL-29 cells (Fig. CAT reporter gene is driven by a minimal simian virus 40 2B; LE-1). Three other constructs containing smaller frag- (SV40) promoter. Enhancer constructs were tested initially in ments of this region of the L-CAM second intron conferred MDCK cells. This cell line has been shown to express the enhanced CAT gene expression and activity on the SV40 mammalian E-cadherin/uvomorulin (27), a molecule closely minimal promoter in LMH cells but not in SL-29 cells. These related to L-CAM. When a 2.5-kb Kpn I-EcoRI fragment constructs, designated LE-6, LE-7, and LE-9 contained from the second intron of the gene was inserted into the successively smaller fragments of the Kpn I-EcoRI fragment pCAT-Promoter vector (Fig. 1), it showed significantly ele- (1.35 kb, 930 bp, and 310 bp, respectively); all showed vated CAT activity relative to that observed for the vector significant elevation of CAT activity over the vector control alone after transfection of these plasmids into MDCK cells (13.0-, 16.5-, and 9.4-fold, respectively; Fig. 2B). Compari- (compare lanes 2 and 3 to lane 1 in Fig. 1 Lower). The activity sons of the activities from the different enhancer constructs ofthe intron fragment was slightly higher when assayed in its suggested that segments of the 2.5-kb Kpn I-EcoRI fr-agment normal genomic orientation than when assayed in the oppo- (LE-1) contained DNA sequences that may inhibit expres- site orientation (compare lane 2 to lane 3 in Fig. 1 Lower). sion of the reporter gene. For example, the 930-bp Sst I-Xho Downloaded by guest on September 30, 2021 11358 Developmental Biology: Sorkin et al. Proc. Natl. Acad. Sci. USA 90 (1993)

CAT Activity:

A B V, , I'-r, t vMI', Xl I -,c fTt, E(rR Ix-fold over _- |1. vector i i i i I control SL-29 LMH 8L-29 LMH 205- LE-1 I- I- 1.8 12.0

LE-2 I -- I 0. 8 1. 0 116- LE-3 I .I 1.2 1. 2 1.5 3. 8 94- -p LE-4 I I 2 . 6 1. 6 LE-5 I I

_waft LE-6 i I 2 .4 13.0 68- LE-7 1. 5 16. 5 LE-8 2.5 1. 0

LE-9 1. 4 9 .4 LE-lO 0. 8 2.4

FIG. 2. (A) Expression of L-CAM in chicken cell lines. Extracts from confluent 60-mm dishes of LMH and SL-29 cells were subjected to immunoblot analysis with polyclonal anti-L-CAM antibody. Migration of standard proteins ofdifferent molecular sizes (kDa) is indicated at left. Mature L-CAM migrates at 124 kDa; faster migrating bands are probably proteolytic fiagments ofL-CAM (7). (B) Enhancer activity ofrestriction fragments ofthe L-CAM second intron assayed in two different chicken cell lines, SL-29 and LMH. A partial restriction map ofthe 2.5-kb Kpn I-EcoRI fragment ofthe L-CAM second intron is shown. L-CAM enhancer plasmids LE-1-LE-10 were constructed by insertion ofthe indicated DNA fragment into the Xba I site downstream of the CAT gene in pCAT-Promoter. Enhancer constructs were transfected into either LMH hepatoma cells or SL-29 fibroblasts. Extracts were assayed for /3-galactosidase activity, normalized, and assayed for CAT activity. Levels of CAT activity were quantitated after exposure of the thin-layer chromatograms to a PhosphorImager screen. Activity was taken as the ratio of the sum above background ofthe pixel values for the acetylated forms ofchloramphenicol to the same value for pCAT-Promoter (without insert) from the same experiment. Equal areas were used for all comparisons. All transfections were done in duplicate, and all values shown are the average of at least three different experiments.

I fagment, which had the highest activity observed for any from the K-CAM-L-CAM intergenic region was inserted of the constructs tested (Fig. 2B; LE-7), showed little if any upstream of the CAT gene in pCAT-Basic in its normal enhancer activity when 0.5 kb of 5' intronic sequence was genomic orientation and assayed in the presence of the added to it (Fig. 2B, compare LE-4 to LE-7). L-CAM enhancer, CAT activity in LMH cells was 10.4-fold To examine the L-CAM enhancer for potential transcrip- greater than that observed using pCAT-Basic (Fig. 4). When tion factor binding sites, we determined the DNA sequence similarly assayed in SL-29 cells, activity remained low (Fig. of the Kpn I-EcoRI fragment of the second intron (Fig. 3). 4). The same pCAT-Basic/L-CAM enhancer construct con- The sequence indicates that the segment showing the greatest taining the 630-bp promoter fragment in the reverse orienta- enhancer activity (LE-7) contains an assortment of putative tion showed no detectable CAT activity in either LMH or binding sites for transcription factors known to regulate gene SL-29 cells (Fig. 4). These data support the interpretation that expression in other systems. These sites include three se- the small segment of DNA between the K-CAM and L-CAM quences that resemble the consensus AP-2 binding sequence genes can function as a promoter and, in conjunction with the (28), four E2A recognition sites (29), and two consensus L-CAM enhancer, can yield cell type-specific expression of binding sites for Spl (30). Moreover, the 310-bp LE-9 en- the gene. hancer, which showed high activity in LMH hepatoma cells, contained a consensus binding site for the liver-eniched POU-homeodomain transcription factor, HNF-1 (31). DISCUSSION We have previously shown that upstream of the L-CAM Our previous studies (21) of transgenic mice expressing gene, there is a homologous gene, K-CAM (20, 32), which genomic L-CAM constructs driven by the insulin or neuro- encodes B-cadherin (13). To test whether the small segment filament promoters showed that L-CAM was present not only of DNA between the chicken K-CAM and L-CAM genes in the tissues expected for these promoters but also in a functioned as a promoter, we inserted a 630-bp Mbo I-Kpn number of sites where the L-CAM gene is normally ex- I genomic fragment from this intergenic segment into the pressed. These results suggested that some of the DNA promoterless CAT vector, pCAT-Basic. This fragment was regulatory elements responsible for site-specific expression then tested for its ability to direct expression of the CAT of L-CAM were nested inside of the gene, probably within reporter gene in the presence or absence of the L-CAM introns. We have now identified and characterized a cell enhancer. As shown in Fig. 4, CAT activity in LMH cells type-specific enhancer within the second intron of the transfected with a construct containing the 630-bp K-CAM- chicken L-CAM gene as well as the L-CAM promoter in the L-CAM intergenic region alone was 1.4-fold greater than that intergenic region between the K-CAM and L-CAM genes. observed in cells transfected with pCAT-Basic. Little pro- The enhancer, a 2.5-kb Kpn I-EcoRI segment of the moter activity was noted in SL-29 cells. No promoter activity second intron in the L-CAM gene, conferred activation of was observed in either cell line when the 630-bp fragment was CAT reporter gene activity driven by the SV40 minimal inserted into vector in the reverse orientation (data not promoter in MDCK cells, which express the closely related shown). mammalian CAM, E-cadherin/uvomorulin (27). To study the As would be expected for an enhancer without a promoter, activity of the chicken L-CAM enhancer in cell lines derived the 1.35-kb LE-6 L-CAM enhancer had no detectable activity from the same species, we analyzed the expression of the in either cell line when inserted into the promoterless pCAT- protein in chicken cell lines. Liver was the source from which Basic vector (Fig. 4). In contrast, when the 630-bp fragment L-CAM was originally purified (7), and this tissue expresses Downloaded by guest on September 30, 2021 D-Ovevelopmental Biology: Sorkin et al. Proc. Natl. Acad. Sci. USA 90 (1993) 11359

GGTACCAAGGACAAAGGTGTTCCTGTTGGGATAGGCTTGGTGCGTGGTAAAGTGGCATAGAAATACAGTTGTGGGGGTGGGCTGGGTGTGTGGCACCGAG 236 Kpnl GGTGGCCQTGGGGAACAAAGCCGTGGCTAAGTGCCCGTGTGCCCTGCGTGGTGTCAGGGCCGGCTGCGCTCTGCCCAGTTCAGTGCTGGTGGTCGGGGGC 336 ACCCAGAGCTGTGCCTACGTGTCCCQGCTCCCCGAGCAGCCQGGTGGTGCAGGTGCCGGGGCGAGGGCGTGCGGGGGCAGTGGGTGCCTCTCCCQGCAGC 436 ACCGAGGCGGGCGGCGCTGACTCAGGCGGAGCGGGCTTGTTCGGGCCGGCAGGCTGGGTGTGGGGCATGGAGCCCGGCCCTGAGCCTGCACCGAGCCCQC 536 GTCCCQGCACCGAGCCCQQTCGGCACCAAGCTCACAGCCCQ8ACTGAGTGCCCATCCTGGTGCAGAGCCGTGCGTGGGTGCAGGCGGTGGTTCTTGGGA 636 AGGGTGGCTGAAGTCCAGTCTGAGTTTCGCCCACGCTGGGGAGTATCGGGGCTGGGAGCTGAACCCTGCACTGTACCTAACGCATTCCATGTTTGTG 736 AGCTTAATGGTTAATCATTTACCAAAGCTGTGCAGCGCCAGCAGGCTGTGCQCAd'GGGCACTGCCTGACACCAAGGTGACGGGCAGCTGCG;TGCCAAGCG 836 HNF-1 E2A GCACTTCCCCAGCACACGCGTTGGCTCCTAGCTGGAAATCCCCGGCTGCCGAGCACCACCGCTTCCCGTCTGTGGGGA"CAGAGGTGGTGTGTGGG 936 _ OAAP2 _ ____ AGGAGGGTGTCTGGGTCTGCCCACGCGTGCCQGCGCCGTGCCGCAATGCCTGGCCQQAGAGCATCCTGCTGTGCCCTfCCGTGCCGTGCAGCCCTTCCMAAI 1036.LW.vv GCCCCQTTTGTGGCACAGCACGGGCAGGCTGATGAGCTGCCCTGTAATTATCCGATGAGCTGAGCGCCGTGTTTGCAGTCAGAAGGTCTGGATTGGCCCCI 1136 ACGCAGAGCTGGAGTGGTGAACATCGTGGGTGCTCCAGGGGCCAGGTTGGGTGCAGCTCTGCGGCTCCTCGACGGCGCTGCAGCTGCCCGGCTGATGA 1236 E2A n orrsrrr t rssrrrsjkt-t- Ee,fx I I~A.J.IA4VtA1Ivrrr eerfrrslIIUAI -FtlI,IUIJ'...IAMtMUII*tAI..A~AJt..AMtI.AXXA Ave ik x srve,rxAf t,rvrtse rrsrrXAIA.,s...JLrvLI.JozbA4zr ilAlerrlitSAU.LIA.--bI1.3. 1212JJ GCTGCCCCACACCAAGGGGAGGACAGACAGCCCTGCCCCGCATCCCTGGGGCTGTGCTCAGCCTCTTTTACTGCGCAAAGGGTCCCCAGGTAGGGGATG 1436 .... E2A... GGATGGCAGATCCCAGGAAGGAGCTGTCAAGGAGCCCTAGCTGTGCGGGTACGGCACCTTATTTGGGGAACACAGGTTTGCAGCQGGTGTGAQCTCC 1536 ___ CCAGTTGCGTGGCTCCCGGATCAGCCCCAAAATGGGAGAAACCTCGAG______- E______xGTCTGAAGTCACGGTGAAACTCGGTGACGTCACTCCCTGGCAGGACGTGTG______1636 ~~Xhol- GCTCACGGCGTTCTTGCAGCACTTTCACCCCGGCTGCTCATGGGCTGACGTGTGCCTCCCCCCGCCGCCQGCTTCCGCCCCGCCACAGCCGCGCTTGCAC 1736 AACCTGTGTGGGGACAGCGGGCAGCCTTGGGATGGTGGCACTTGGCTGTGCCTGTGACTCACCCTGCGCCCQGCCCQGTGAGACACTTCCCTGTGTCAC 1836 AGCAGAATGGGGGTAGCTTAGTGTCCCCCCAGGCCCAGCCCQGCGTGGAGGGCTCACCTCCCCGCCCCAGCTTTGAGCCGGCAATTCACGGATTTTGCCC 1936 CGTGGCCAAACGCGTTTGAGGGTCCQGGAAATCAGGAGCGCTCGGTTGCTCTGCAGCGCAGGAGCCCCTGAGCTCGACCCCCGCCCCQCGCTTTGTCGAG 2036 GCCCCACTGGCTCTGCGCCTCGGGGCAGAACGGGGCTGGGAGGGGCTGGCTTTCAAGTGTGGAGCTGGAAAAACAG,GTTGTGCGG5GGGXCGCCCCAGAG 2136 AGGGGGGAGAGCGGGACGCTGAGCCGTGGTGCGGGGCCGG,GTGCTGAGCACGGTGTGCTCGCTCCACCGCGTCGGTGCGGGGCCGGGTGCTGAGCACGGT 2236 GTGCTCGCTXCACCGCGTCGGTGCGGGGCCGGGTGCTGACACGGTGTGCTCGCTCGCCGCGTCGGTGCGG#GGCCGGGTGCTGACG,ACGGTGTGCTCGCTC 2336 GCCGCGTCGGTGCGGGGCCGGGTGCTGAGCACGGTGCGTTCCGCCGGCCQGGTGCCAGGCGCCGGGTGCCCGQAGGTATGGGGCACCTCCCXACTGCGAG 2436 CCCACGGTGCGGGAGCGCCCGGCATGCCCCAGCACGCCCTGXXACXXGGCACCACCGGCTGTCGXGCCCGGCCGGGGGTCCCGCCTGGGTTGGCACCGC 2536 CCCGCCGGGAGGCGGCACXGCGGTGAGCGAGCAGCCCCCAGCCCGCTCTGCCQTCGCCTCCCCCTTGCTGTTGCCTCTGAATTC 2620

FIG. 3. Nucleotide sequence of the 2.5-kb Kpn I-EcoRI fragent of the L-CAM second intron. Numbering corresponds to position in the intron, taking the first nucleotide ofthe intron as 1. Restriction endonuclease cleavage sites shown in Fig. 2 are indicated with a dashed underline. Consensus recognition sequences fortranscription factors found in the region ofthe intron showing the strongest transcriptional activation (LE-7, boxed; see Fig. 2) are indicated by solid underlines. high levels of the protein, whereas neither fibroblasts nor Drosophila connectin gene (36) have been suggested as fibroblastic cell lines normally express L-CAM or its putative targets for different homeodomain-containing transcription mammalian homologues. In accord with this distribution of factors during embryogenesis (37, 38). It will be ofinterest to L-CAM, a hepatoma cell (LMH) expressed high levels of determine whether L-CAM and other genes encoding calci- L-CAM, whereas a fibroblastic cell line (SL-29) did not. um-dependent CAMs are regulated differentially from Consistent with these patterns of L-CAM protein synthesis, N-CAM by homeobox gene products during embryogenesis. the enhancer was active in LMH cells but not in SL-29 cells. Immediately upstream of the L-CAM gene, the K-CAM A number of smaller intron fragments (LE-6, LE-7, and gene (20) encodes a calcium-dependent CAM, B-cadherin LE-9; see Fig. 2B) were sufficient to direct SV40 promoter- (13), which is similar to L-CAM. In the present studies we mediated CAT reporter gene activity in LMH hepatoma cells showed that agenomic fragment that starts 30 bp downstream but not in SL-29 cells. The smallest of these, LE-9, a 310-bp ofthe polyadenylylation site in the K-CAM gene and ends in Sst I-Mlu I DNA fragment, contained a number of potential the 5' untranslated region ofthe L-CAM gene can function as binding sites for transcription factors, including AP-2, E2A, a promoter, particularly when combined with the L-CAM and Spl. This DNA fragment also contains a consensus enhancer. The L-CAM promoter/enhancer combination recognition site for the liver-enriched POU-homeodomain could direct expression of a reporter gene in a cell type- protein, HNF-1 (31). specific manner and at levels comparable to those observed This last observation prompts the suggestion that HNF-1 with SV40 promoter/L-CAM enhancer constructs. More- and other homeodomain transcription factors may be in- over, promoter activity from this intergenic region was volved in place-dependent patterns of expression of the orientation-sensitive, consistent with the classical definition L-CAM gene during embryonic development. Recently, we of a promoter. have provided evidence that other adhesion molecules such The analysis of the L-CAM promoter and enhancer se- as neural CAM (N-CAM) and the extracellular matrix protein quences described here should provide a firm basis for cytotactin/tenascin are targets for regulation by different further experiments in transgenic animals. These should homeodomain-containing transcription factors (33, 34). allow a better understanding of the factors that control the N-CAM is paradigmatic of the immunoglobulin-related su- place-dependent expression ofthis calcium-dependent adhe- perfamily of Ca2+-independent CAMs. L-CAM is one of the sion molecule during embryogenesis. earliest studied members of the Ca2+-dependent CAMs or cadherins. Members ofboth families can appear on the same The authors are grateful to Ms. Maria Samson and Ms. Eunice cell during embryogenesis, but as different tissues emerge, Huang for excellent technical assistance. This work was supported the place-dependent expression of these molecules diverges in part by U.S. Public Health Service Grants AG09326 and HD09635 (35). Other adhesion molecules such as that encoded by the to G.M.E. and HD16550 to B.A.C. Downloaded by guest on September 30, 2021 11360 Developmental Biology: Sorkin et al. Proc. Natl. Acad. Sci. USA 90 (1993) K-CAM L-CAM B. A. & Gallin, W. J. (1987) Proc. Natl. Acad. Sci. USA 84, 16 2 8502-8506. 9. Hyafil, F., Morello, D., Babinet, C. & Jacob, F. (1980) Cell 21, 927-934. 10. Yoshida-Noro, C., Suzuki, M. & Takeichi, M. (1984) Dev. Biol. 1H3 101, 19-27. 1 3 11. Damsky, C. H., Richa, J., Solter, D., Knudsen, K. & Buck, P e C. A. (1983) Cell 34, 455-466. CAT Activity: x-fold over 12. Imhof, B. A., Volimers, H. P., Goodman, S. L. & Birchmeier, vector control W. (1983) Cell 35, 667-675. 13. Napolitano, E. W., Venstrom, K., Wheeler, E. F. & Rei- SL-29 LHII chardt, L. F. (1991) J. Cell Biol. 113, 893-905. 14. Donalies, M., Cramer, M., Ringwald, M. & Starzinski-Powitz, A. (1991) Proc. Natl. Acad. Sci. USA 88, 8024-8028. 0.8 0.9 15. Hatta, K., Nose, A., Nagafuchi, M. & Takeichi, M. (1988) J. Cell Biol. 106, 873-881. 0.4 1.4 16. Nose, A. & Takeichi, M. (1986) J. Cell Biol. 103, 2649-2658. 17. Inuzuka, H., Miyatani, S. & Takeichi, M. (1991) Neuron 2, 69-79. _ 1.6 10.4 18. Ranscht, B. & Bronner-Fraser, M. (1991) Development 111, 15-22. 19. Sorkin, B. C., Hemperly, J. J., Edelman, G. M. & Cunning- 0.9 0.8 ham, B. A. (1988) Proc. Natl. Acad. Sci. USA 85, 7617-7621. 20. Sorkin, B. C., Gallin, W. J., Edelman, G. M. & Cunningham, FIG. 4. The intergenic region between the K-CAM and L-CAM B. A. (1991) Proc. Natl. Acad. Sci. USA 88, 11545-11549. genes has promoter activity. The region indicated by an arrow 21. Begemann, M., Tan, S. S., Cunningham, B. A. & Edelman, designated with "p" corresponds to the 630-bp Mbo I-Kpn I DNA G. M. (1990) Proc. Natl. Acad. Sci. USA 87, 9042-9046. fragment from the intergenic region used to assay for promoter 22. Kawaguchi, T., Nomura, K., Hirayama, Y. & Kitagawa, T. activity. The region indicated by an arrow designated "e" corre- (1987) Cancer Res. 47, 4460-4464. sponds to the 1.35-kb Sst I (LE-6) fragment from the L-CAM second 23. Towbin, H., Staehelin, T. & Gordon, J. (1979) Proc. Natl. intron (see Fig. 2) and was used to assay for enhancement of Acad. Sci. USA 76, 4350-4354. promoter activity. Four CAT reporter plasmids were constructed in 24. Laemmli, U. K. (1970) Nature (London) 227, 680-685. the promoterless vector pCAT-Basic: a construct containing the 25. Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982) Molecular L-CAM enhancer without the promoter segment, a construct with Cloning: A Laboratory Manual (Cold Spring Harbor Lab. the promoter inserted in the orientation normally found in the Press, Plainview, NY). L-CAM gene, a construct with the promoter in the same orientation 26. Staden, R. (1982) Nucleic Acids Res. 10, 2952-2961. together with the L-CAM enhancer, and a construct with the 27. Behrens, J., Birchmeier, W., Goodman, S. L. & Imhof, B. A. promoter inserted in the opposite orientation together with the (1985) J. Cell Biol. 101, 1307-1315. L-CAM enhancer. Transfections of cells, CAT assays, and quanti- 28. Mitchell, P. J., Wang, C. & Tjian, R. (1987) Cell 50, 847-861. tations were performed as described in the legend for Fig. 2. Activity 29. Murre, C., McCaw, P. S. & Baltimore, D. (1989) Cell 56, was taken as the ratio of the sum above background of the pixel 777-783. values for the acetylated forms ofchloramphenicol to the same value 30. Briggs, M. R., Kadonaga, J. T., Bell, S. P. & Tjian, R. (1986) for pCAT-Basic. Science 234, 47-52. 31. Courtois, G., Morgan, J. G., Campbell, L., Fourel, G. & 1. Thiery, J.-P., Delouvee, A., Gallin, W. J., Cunningham, B. A. Crabtree, G. R. (1987) Nature (London) 238, 688-692. & Edelman, G. M. (1984) Dev. Biol. 102, 61-78. 32. Gaily, J. A. & Edelman, G. M. (1992) Proc. Natl. Acad. Sci. 2. Crossin, K. L., Chuong, C.-M. & Edelman, G. M. (1985) Proc. USA 89, 3276-3279. Natl. Acad. Sci. USA 82, 6942-6946. 33. Jones, F. S., Prediger, E. A., Bittner, D. A., DeRobertis, 3. Gumbiner, B. & K. J. Cell Biol. 457-468. E. M. & Edelman, G. M. (1992) Proc. Natl. Acad. Sci. USA 89, Simons, (1986) 102, 2086-2090. 4. Mege, R.-M., Matsuzaki, F,, Gailin, W. J., Goldberg, J. I., 34. Jones, F. S., Chalepakis, G., Gruss, P. & Edelman, G. M. Cunningham, B. A. & Edelman, G. M. (1988) Proc. Natl. (1992) Proc. Natl. Acad. Sci. USA 89, 2091-2095. Acad. Sci. USA 85, 7274-7278. 35. Edelman, G. M. & Crossin, K. L. (1991) Annu. Rev. Biochem. 5. McNeill, H., Ozawa, M., Kemler, R. & Nelson, W. J. (1990) 60, 155-190. Cell 62, 309-316. 36. Nose, A., Mahajan, V. B. & Goodman, C. S. (1992) Cell 70, 6. Gallin, W. J., Chuong, C.-M., Finkel, L. H. & Edelman, G. M. 553-567. (1986) Proc. Natl. Acad. Sci. USA 83, 8235-8239. 37. Gould, A. P. & White, R. A. H. (1992) Development 116, 7. Galin, W. J., Edelman, G. M. & Cunningham, B. A. (1983) 1163-1174. Proc. Natl. Acad. Sci. USA 80, 1038-1042. 38. Edelman, G. M. & Jones, F. S. (1993) J. Biol. Chem., 268, 8. Edelman, G. M., Murray, B. A., Mege, R.-M., Cunningham, 20683-20686. Downloaded by guest on September 30, 2021