Search

We have cloned and characterized a mouse cDNA coding for LFB3, a DNA binding protein containing an extra‐large homeodomain. The first 315 amino acids of LFB3 are highly homologous to the DNA binding domain of LFB1, a regulatory protein involved in the expression of several liver‐specific genes. LFB3 is a transcriptional activator which binds to DNA as a dimer and forms heterodimers with LFB1 both in vitro and in vivo. However, LFB3 expression seems not to be directly correlated with the liver‐specific phenotype, since it is detected in dedifferentiated hepatoma cell lines which express neither LFB1 nor several liver‐specific genes. LFB3 expression starts before that of LFB1 during mouse and rat development, and is strongly increased upon retinoic acid induced differentiation of F9 embryonic carcinoma cells. LFB3 and LFB1 are expressed in the epithelial component of many organs of endodermal and mesodermal origin, suggesting that they may play a more general role associated with the differentiation of specialized epithelia.

We present a novel general method for localized mutagenesis. The DNA segment to be mutagenized is inserted in the beta‐galactosidase gene of a M13‐lac vector, generally causing loss of beta‐galactosidase function by generation of frameshifts or nonsense codons. Mutations in the inserted DNA which restore beta‐galactosidase function are readily detected and analyzed. The application of this method to the promoter of an eukaryotic (Caenorhabditis elegans) tRNAPro gene has allowed the isolation of several mutants altered in transcription.

A segment of the human alpha 1‐antitrypsin (alpha 1AT) 5′‐flanking region comprising nucleotides −137 to −37 from the start of transcription is sufficient to drive liver‐specific transcription from the homologous alpha 1AT promoter and from the heterologous SV40 promoter. In this paper we characterize two proteins, LF‐A1 and LF‐B1, whose ability to bind wild‐type and mutant alpha 1AT promoter segments correlates with the ability of these segments to activate transcription in vivo. DNase I protection and methylation interference analysis reveals that LF‐A1 recognizes sequences present in the regulatory region of the human alpha 1‐antitrypsin, apolipoprotein A1 and haptoglobin‐related genes. These sequences share a common 5′ TGG/A A/C CC 3′ motif. LF‐B1 binds to the palindrome 5′ TGGTTAAT/ATTCACCA 3′ which is present in the human alpha 1‐antitrypsin gene between positions −78 and −62 from the start of transcription. LF‐B1 also recognizes a related sequence present in the human albumin gene between −66 and −50. These results suggest that LF‐A1 and LF‐B1 are common positive trans‐acting factors which are required for the expression of several genes in the hepatocyte.

Human genomic clones of the haptoglobin Hp1F and the ‘haptoglobin related’ gene (Hpr) have been isolated. The two genes are adjacent, spanning a region of approximately 21 kb. A comparison of their coding sequences shows that Hpr differs from Hp1F at 28 codons. Northern blot and primer elongation analyses with human liver RNA show that the haptoglobin gene Hp1F appears to be transcribed some 1000‐fold less in fetal than in adult liver. In adult liver the amount of Hpr mRNA is at the lower limit of detection, therefore the extent of its expression is at most less than 1000‐fold that of the Hp1F gene. No Hpr mRNA can be detected in fetal liver.

During the acute phase (AP) reaction the expression of a series of liver‐specific genes coding for secretory proteins is either stimulated or suppressed by different cytokines released by activated monocytes. Transforming growth factor beta (TGF‐beta) is a cytokine that, first identified for its ability to regulate cellular growth, has been gradually recognized to modulate several other functions. We have investigated the effect of TGF‐beta on the expression of acute phase genes in liver cells. We found that TGF‐beta selectively induces a specific decreases in the amount of mRNAs of genes negatively regulated during AP reaction, like albumin and apolipoprotein A‐I (ApoA‐I). The inhibitory effect of TGF‐beta on the expression of negative AP genes is primarily post‐transcriptional and it is very likely to be mediated via an enhancement of the turnover of both albumin and ApoA‐I mRNAs.

We have constructed a plasmid, pLM8, containing the coding sequence of the mature human apoA‐1 fused to the coding sequence of the IgG‐binding domains of protein A (PA) from Staphylococcus aureus. The hybrid gene is transcribed in Escherichia coli under the control of a heat‐sensitive repressor, leading to the synthesis of large amounts of hybrid protein (apoA‐1–PA). The hybrid protein was purified by denaturation with urea and alkali, renaturation and affinity chromatography on an IgG Sepharose column. ApoA‐1–PA is soluble and has an Mr of 316 kd, as determined by gel filtration. This is five times the monomer size of 62 kd, predicted from the sequence and found by SDS‐PAGE analysis. Cell surface binding activity of the hybrid protein was tested using two different cell types (J774 macrophages and Fao hepatocytes) and compared to human high density lipoprotein (HDL). High‐affinity binding was found for both ligands in both cell lines (Kd = 3.4 X 10(‐8)M in Fao cells, 4.9 X 10(‐8) M in J774 cells for apoA‐1–PA and 3.0 X 10(‐8) M in Fao cells, 2.8 X 10(‐8) M in J774 cells for HDL), with approximately 2 X 10(5) high‐affinity binding sites per cell. ApoA‐1–PA and HDL effectively competed with each other for binding to the cell surface. Additionally, they both bound to a 110‐kd polypeptide on a ligand blot, identifying an HDL receptor. The binding parameters of HDL were very similar to those of apoA‐1–PA.(ABSTRACT TRUNCATED AT 250 WORDS)

A human cDNA library was constructed using M13 derivative vectors. The simple and rapid procedures for sequencing single‐stranded DNA by the dideoxy chain termination method allowed a screening of individual clones directly by DNA sequence analysis. Some of these clones were identified as coding for: serum albumin, alpha1‐antitrypsin, retinol‐binding protein, prothrombin, haptoglobin, and metallothionein. Furthermore, a clone coding for aldolase B was tentatively identified on the basis of high sequence homology with rabbit muscle aldolase.

NF1‐like proteins play a role in transcription of liver‐specific genes. A DNA‐binding protein, recognizing half of the canonical NF1 binding site (TGGCA) present on the human albumin and retinol‐binding protein genes, has been purified from rat liver. Several peptides deriving from a tryptic digest of the purified protein were sequenced and the sequence was used to synthesize specific oligonucleotides. Two overlapping cDNA clones were obtained from a rat‐liver cDNA library; their sequence reveals an open reading frame coding for 505 amino acids, including all the peptides sequenced from the purified protein. The DNA‐binding domain, most likely located within the first 250 amino acids, is highly homologous to the sequence of CTF/NF1 purified from HeLa cells. Northern analysis reveals several mRNA species present in different combinations in various rat tissues.

Interleukin‐6 (IL‐6) triggers the formation of a high affinity receptor complex with the ligand binding subunit IL‐6Ralpha and the signal transducing chain gp130. Since the intracytoplasmic region of the IL‐6Ralpha does not contribute to signaling, soluble forms of the extracytoplasmic domain (sIL‐6Ralpha), potentiate IL‐6 bioactivity and induce a cytokine‐responsive status in cells expressing gp130 only. This observation, together with the detection of high levels of circulating soluble human IL‐6Ralpha (shIL‐6Ralpha) in sera, suggests that the hIL‐6‐shIL‐6Ralpha complex is an alternative form of the cytokine. Here we describe the generation of human IL‐6 (hIL‐6) variants with strongly enhanced shIL‐6Ralpha binding activity and bioactivity. Homology modeling and site‐directed mutagenesis of hIL‐6 suggested that the binding interface for hIL‐6Ralpha is constituted by the C‐terminal portion of the D‐helix and residues contained in the AB loop. Four libraries of hIL‐6 mutants were generated by each time fully randomizing four different amino acids in the predicted AB loop. These libraries were displayed monovalently on filamentous phage surface and sorted separately for binding to immobilized shIL‐6Ralpha. Mutants were selected which, when expressed as soluble proteins, showed a 10‐ to 40‐fold improvement in shIL‐6Ralpha binding; a further increase (up to 70‐fold) was achieved by combining variants isolated from different libraries. Interestingly, high affinity hIL‐6 variants show strongly enhanced bioactivity on cells expressing gp13O in the presence of shIL‐6Ralpha at concentrations similar to those normally found in human sera.

Transcription of the human haptoglobin (Hp) gene is induced by interleukin‐6 (IL‐6) in the human hepatoma cell line Hep3B. Cis‐acting elements responsible for this response are localized within the first 186 bp of the 5′‐flanking region. Site‐specific mutants of the Hp promoter fused to the chloramphenicol acetyl transferase (CAT) gene were analysed by transient transfection into uninduced and IL‐6‐treated Hep3B cells. We identified three regions, A, B and C, defined by mutation, which are important for the IL‐6 response. Band shift experiments using nuclear extracts from untreated or IL‐6‐treated cells revealed the presence of IL‐6‐inducible DNA binding activities when DNA fragments containing the A or the C sequences were used. Competition experiments showed that both sequences bind to the same nuclear factors. Polymers of oligonucleotides containing either the A or the C regions confer IL‐6 responsiveness to a truncated SV40 promoter. The B region forms several complexes with specific DNA‐binding proteins different from those which bind to the A and C region. The B region complexes are identical in nuclear extracts from IL‐6‐treated and untreated cells. While important for IL‐6 induction in the context of the haptoglobin promoter, the B site does not confer IL‐6 inducibility to the SV40 promoter. Our results indicate that the IL‐6 response of the haptoglobin promoter is dependent on the presence of multiple, partly redundant, cis‐acting elements.

We have cloned a segment of cDNA from human liver coding for an apoferritin subunit, probably an H chain. Sequence comparison with the available protein sequence shows that our clone corresponds to a ferritin subunit present as a minor species in human spleen and placenta, but as major species in HeLa cells. Northern blot analysis shows the existence of only one band of similar size in human liver, HeLa cells, Daudi lymphoma and Hep3B hepatoma cell lines. In contrast, Southern blot analysis provides evidence for a multigene family.

The transcription factor LFB1/HNF1 from rat liver nuclei is a 628 amino acid protein that functions as a dimer binding to the inverted palindrome GTTAATN‐ATTAAC consensus site. We have crystallized a 99 residue protein containing the homeodomain portion of LFB1, and solved its structure using X‐ray diffraction data to 2.8 A resolution. The topology and orientation of the helices is essentially the same as that found in the engrailed, MAT alpha 2 and Antennapedia homeodomains, even though the LFB1 homeodomain contains 21 more residues. The 21 residue insertion is found in an extension of helix 2 and consequent lengthening of the connecting loop between helix 2 and helix 3. Comparison with the engrailed homeodomain‐DNA complex indicates that the mode of interaction with DNA is similar in both proteins, with a number of conserved contacts in the major groove. The extra 21 residues of the LFB1 homeodomain are not involved in DNA binding. Binding of the LFB1 dimer to a B‐DNA palindromic consensus sequence requires either a conformational change of the DNA (presumably bending), or a rearrangement of the subunits relative to the DNA.

The three‐dimensional backbone fold of a polypeptide fragment from the rat LFB1/HNF1 transcription factor was determined by nuclear magnetic resonance (NMR) spectroscopy in solution. This fragment contains an amino acid sequence that is approximately 22% homologous to the well known homeodomains, but which contains 81 amino acid residues as compared with 60 residues in ‘typical’ homeodomains. For the present studies we used a recombinant 99 amino acid polypeptide containing this sequence in positions 10–90, which was uniformly labelled with 15N and also doubly labelled with 15N and 13C. The NMR structure of this polypeptide contains three alpha‐helices comprising the residues 18–29, 36–50 and 71–84, a loop formed by residues 30–35, and a long stretch of non‐regular secondary structure linking the second and third helices. The relative location and orientation of the helices is very similar to that in the Antennapedia (Antp) homeodomain structure, despite the fact that helix II is elongated by about one turn. This confirms a recently advanced hypothesis based on sequence comparisons that this polypeptide segment of LFB1/HNF1 might represent a homeodomain‐like structural element. The helix‐turn‐helix motif, which has been shown to comprise the DNA recognition helix in the Antp homeodomain, can readily be recognized in the LFB1/HNF1 homeodomain, in spite of an extensive modification of the primary structure. The two residues of the tight turn in the Antp homeodomain are replaced by a 23 residue linker region between the two helices in LFB1/HNF1, which bulges out from the rest of the molecule and thus enables the formation of a non‐classical helix‐‐turn‐‐helix motif.

It has been found that 334 bases of the 5′ flanking region of the human retinol‐binding protein (RBP) gene contain sufficient information to direct accurate and specific transcription in human hepatoma but not in HeLa cells. Dissection of this region reveals the existence of at least three distinct controlling elements: an enhancer, which can activate promoters in a variety of cell lines and is therefore non‐tissue specific; a negative cis‐acting element, which apparently binds a repressor molecule not present or non‐functional in hepatoma cell lines; a promoter element. Transcription is confined to hepatoma cells only when both the negative element and the promoter element are present.

The region from −137 to −2 of the human alpha 1‐antitrypsin (alpha 1AT) promoter directs liver‐specific in vitro transcription. Two cis‐acting elements, A and B, have been identified within this segment by site‐directed mutagenesis. Competition with synthetic oligonucleotides corresponding either to the A or to the B sequence inhibits transcription from the wild‐type promoter in vitro. Cis‐linked A and B elements mediate liver‐specific transcription from a truncated HSV‐TK promoter in vitro. Five different proteins, LF‐A1, LF‐A2, LF‐B1, LF‐B2 and LF‐C, bind to the alpha 1AT promoter in liver extracts. LF‐A1 and LF‐B1 are positive transcriptional factors which bind to the A and B elements respectively. Their absence in spleen provides an explanation for the liver specificity of transcription. A protein similar to LF‐B2 is present in spleen. Binding of LF‐B1 and LF‐B2 to the alpha 1AT promoter is mutually exclusive, suggesting that LF‐B2 might be a repressor.

The 5′ flanking region of the human alpha 1‐antitrypsin (alpha 1‐AT) gene contains cis‐acting signals for liver‐specific expression and, when fused to a reporter gene, is able to drive the expression of this gene specifically in liver cells. Here we report the results of a functional dissection of the alpha 1‐AT regulatory region. The expression of the bacterial chloramphenicol‐transacetylase (CAT) gene, fused to a set of alpha 1‐AT 5′ flanking regions shortened by progressive deletions or mutated by base pair substitutions, has been compared by transfection in HepG2 (hepatocyte) and HeLa (non‐hepatocyte) human cell lines. A minimal tissue‐specific element has been identified between the nucleotides −137 and −37 (from the transcriptional start site). This DNA segment activates the heterologous SV40 promoter in hepatoma cell lines but not in HeLa cells. This element contains at least two regions referred to as the A (‐125/‐100) and B (‐84/‐70) domains, both essential for transcription. There are at least two other regulatory domains located upstream of the ‘minimal element’; the most active of these is located between positions −261 and −210 from the cap site. These upstream elements activate the heterologous SV40 early promoter both in hepatoma cell lines and in HeLa cells. Upon fractionation of rat liver nuclear extracts two proteins have been identified, alpha 1TF‐A and alpha 1TF‐B, which bind specifically to the A and B domains respectively. Transcriptionally inactive A and B domain mutants are not able to bind these proteins.

In order to investigate the mechanism of expression of the human alpha 1‐antitrypsin (alpha 1‐AT) gene in macrophages, we have characterized the alpha 1‐AT transcriptional units in these cells and discovered that there is a macrophage‐specific promoter located approximately 2000 bp upstream of the hepatocyte‐specific promoter. Transcription from the two alpha 1‐AT promoters is mutually exclusive: the macrophage promoter is silent in hepatocytes and the hepatocyte promoter is silent in macrophages. In addition, in macrophages two distinct mRNAs are generated transcript by alternative splicing. These results suggest that alpha 1‐AT gene transcription responds to two different cell‐specific regulatory mechanisms.

The promoter proximal sequences of a group of liver‐specific genes including that of albumin interact with the same hepato‐specific factor named HNF1, LFB1, APF or HP1, a distant member of the homeoprotein family. A distinct protein, termed variant HNF1 (vHNF1), of lower mol. wt but displaying identical sequence specificity is found both in dedifferentiated variants and in an extinguished somatic hybrid that fail to express most or all of the tissue‐specific traits, including albumin. We show here that HNF1 transcripts are present only in differentiated hepatoma cells. No transcripts are detected in dedifferentiated variants or in the extinguished cell hybrid, strongly suggesting that the vHNF1 protein is encoded by a distinct gene. Finally, HNF1 transcripts reappear in revertants to the hepatic phenotype. Run‐on transcription analysis in isolated nuclei demonstrates that the expression of HNF1 in these cell lines is regulated primarily at the transcriptional level. Contrary to HNF1, the mRNAs coding for two other nuclear factors involved in albumin transcription, C/EBP and NF1, do not follow the distribution of albumin transcripts in these cell lines. These results indicate that extinction in somatic hybrids or loss of expression upon dedifferentiation of liver‐specific genes possessing an HNF1 recognition site is caused, at least in part, by a block in HNF1 gene expression.

LFB1/HNF1 regulates the hepatocyte‐specific transcription of several genes, binding as a dimer to cis‐acting elements that match the inverted palindrome GTTAATNATTAAC. The DNA binding domain of LFB1/HNF1 is characterized by a unique tripartite structure that includes an unusually long homeodomain (domain C), a region related to the POU‐specific A‐box (domain B) and a short N‐terminal dimerization domain (domain A). We report that a recombinant peptide corresponding to the isolated homeodomain of LFB1/HNF1 binds as a monomer to a half‐palindrome binding site, but shows diminished sequence specificity. Domain B, in addition to the homeodomain, is required and sufficient for proper recognition of LFB1/HNF1‐responsive sites. A protein consisting of only these latter two domains is a monomer in solution, but forms dimers upon DNA binding. The protein‐protein contacts established within the bound dimer restrain the orientation of the two homeodomains with respect to one another, thus contributing in a critical fashion to the recognition of the dyad symmetry‐related LFB1/HNF1 sites. The DNA‐independent dimerization domain (domain A) is required to increase the affinity of DNA binding, but does not influence the dimer geometry.

NF1 is a DNA‐binding protein involved in initiation of adenovirus DNA replication as well as in modulating the rate of transcription initiation of genes containing the sequence TGGCA. We show here that recombinant NF1 expressed via vaccinia virus is transported into the nucleus and binds to its cognate sequences with the same specificity as NF1 purified from HeLa cells. Furthermore, the recombinant NF1 forms oligomers in solution and binds as a dimer to palindromic as well as half‐site sequences. NF1 expressed via vaccinia virus stimulates the initiation of adenovirus replication in vitro. The N‐terminal 240 amino acids of the protein are sufficient for full DNA‐binding activity as well as stimulation of adenovirus replication. By analysis of several NF1 mutants translated in vitro, we also define the minimal DNA‐binding domain and localize the region responsible for DNA binding on the N‐terminal and for oligomerization on the C‐terminal side of this domain.

Eukaryotic tDNA promoters are composed of two essential regions contained within the coding sequence (Box A and Box B). Due to the highly conserved structure of prokaryotic and eukaryotic tRNA, most prokaryotic tRNA genes are expected to be active templates in eukaryotic transcriptional systems. In this paper we show that Escherichia coli tDNATyr is not transcribed in the nucleus of Xenopus laevis oocytes. By in vitro construction of hybrid molecules between inactive prokaryotic tDNATyr from E. coli, and active eukaryotic tDNAPro from Caenorhabditis elegans, we show that tDNATyr can be made into an active gene if its first third, including the Box A region, is replaced by that of the eukaryotic tDNA . These results suggest that an improper Box A sequence is responsible for the inactivity of the E. coli tRNATyr gene, and argue against the role of secondary and tertiary DNA conformations in RNA polymerase III transcription.

A strategy to identify disease‐specific epitopes from phage‐displayed random peptide libraries using human sera is described. Peptides on phage (phagotopes) that react with antibodies present in patient sera are purified from > 10(7) different sequences by affinity selection and immunological screening of plaques. Disease‐specific phagotopes can be identified out of this pool through an ‘antigen independent’ procedure which avails itself only of patient and normal human sera. Using this strategy, we have selected antigenic mimics (mimotopes) of two different epitopes from the human hepatitis B virus envelope protein (HBsAg). We could show that a humoral response to these mimotopes is widespread in the immunized population, suggesting that the strategy identifies phagotopes that have a potential role as diagnostic reagents. Immunization of mice with the selected phagotopes elicited a strong specific response against the HBsAg. These results open new inroads into disease‐related epitope discovery and provide the potential for vaccine development without a requirement for the use of, or even information about, the aetiological agent or its antigens.

Haptoglobin is a plasma protein scarcely present in fetal but abundant in adult serum, where it is present at a concentration of approximately 150 mg/100 ml. In this paper we show by run‐on experiments that the haptoglobin (Hp) gene is actively transcribed in adult but not in fetal liver nuclei. Studies with established cell lines indicate that the Hp gene is expressed in the hepatoma cells HepG2 but not in the hepatoma cell line Hep3B nor in HeLa cells. Plasmids carrying various segments of the 5′ flanking region of the Hp gene fused to the chloramphenicol acetyl transferase (CAT) gene direct CAT transcription when introduced into HepG2 but are inactive in Hep3B and in HeLa cells, thus behaving like the resident chromosomal Hp gene. Deletion analysis defines a region, upstream to the transcription initiation site, essential for cell‐specific expression. The Hp gene is induced in Hep3B cells by treatment with supernatant from LPS‐stimulated monocytes (SMS), in a manner mimicking the acute phase reaction. We characterize the DNA segment necessary and sufficient for cell‐specific expression of the Hp‐CAT constructions in HepG2 and show that the same segment is also sufficient for acute phase induction in Hep3B.

alpha 1‐acid glycoprotein (alpha AGP) is a well‐characterized human plasma protein. Its structural properties have been studied for many years but little is known about its function. Amino acid sequence analysis of purified human alpha AGP from plasma pooled from several individuals showed considerable heterogeneity. We have cloned the genomic DNA segment encoding alpha AGP and we show that it contains three adjacent alpha AGP coding regions, AGP‐A, B and B‘, identical in exon–intron organization but with slightly different coding potential. These results account for the heterogeneity observed by protein sequencing. Southern blot analysis indicates that the cloned cluster contains all the alpha AGP coding sequences present in the human genome. The larger majority of alpha AGP mRNA in human liver is transcribed from AGP‐A, whose promoter and cap site have been determined while the level of AGP‐B and B’ mRNA in human liver is very low. Using Hep3B hepatoma cells as a model system for the in vitro study of the acute phase reaction, we show that only AGP‐A is strongly induced by treatment with culture medium of LPS stimulated monocytes.

A major challenge in basic and applied biological research is the engineering of small proteins with pre‐determined structures and novel functions. In a limited number of cases, this has been achieved by de novo design. An alternative combinatorial approach is based on the construction of large libraries of random peptides and on methods for the selection of the desired molecules. Here we describe a successful combination of both the rational design and the combinatorial approaches for developing proteins with useful biological functions, in this case the construction of a specific inhibitor of the cytokine human interleukin‐6. In previous work, the ‘minibody’, a 61 residue polypeptide consisting of a beta‐pleated framework and two hypervariable regions, was designed, synthesized and expressed on f1 phage surface. We report the construction of a repertoire of 50 million minibodies displayed on phage in which the hypervariable regions have been randomized. One polypeptide which binds tightly and specifically to human interleukin‐6 was isolated from this collection of minibody mutants. This particular minibody is an effective inhibitor of the cytokine's biological activity. The approach described here could in principle be applied to other molecular targets.

A genomic clone of the most abundant U1 RNA genes from Xenopus laevis was isolated from erythrocyte DNA and sequenced. Two different U1 RNA genes, U1A and U1B, are encoded in an HindIII 1.5‐kb fragment and both are expressed after microinjection in Xenopus oocytes. Deletions and site‐directed mutagenesis of the clones revealed two promoter elements in the U1B gene; one, located 250‐220 nucleotides upstream from the 5′ terminus of mature U1 RNA, functions as an activator, yielding a 10‐fold promotion of transcription; the other, located 60‐50 nucleotides upstream of the cap site, functions as an essential element for promotion of transcription. The U1A gene contained only the latter element in the cloned fragment. Homologous sequences can be identified in several U RNA genes of X. laevis.

Previous work in transfected cell lines and in nuclear extracts has led to the identification of two cis‐acting elements important for transcription of the human alpha‐1‐antitrypsin (A1AT) gene, which bind to two liver specific trans‐acting factors, LF‐A1 and LF‐B1. Mutations EM3 and PM1, which abolish the binding of LF‐A1 and LF‐B1 respectively, drastically reduce transcription activity of the A1AT gene in vitro and in cell culture. The same mutants have now been introduced in a larger DNA context and their effect has been tested in transgenic mice. A stretch of DNA was constructed which carries two transcriptional units: 18 kb of the human retinol binding protein (RBP) gene, driving the expression of the bacterial chloramphenicol acetyl transferase, linked to 17.5 kb containing the entire A1AT coding sequence with additional 5′ and 3′ flanking sequences. Transcription from the RBP promoter was shown to predominate in liver, and could be used as an internal marker of ‘active copy number’. Mutations in the A1AT gene promoter were introduced by homologous recombination in bacterial cells. The results show that base pair substitutions in the binding site for LF‐A1 and LF‐B1 drastically reduce transcription in non‐hepatic adult tissues, yolk sac, and fetal liver, whereas only LF‐B1 binding site mutations have a marked, albeit variable, effect in adult liver.

Human plasma retinol binding protein (RBP) is coded by a single gene and is specifically synthesized in the liver. We have characterized a lambda clone, from a human DNA library, carrying the gene coding for plasma RBP. Southern blot analysis and DNA sequencing show that the gene is composed of six exons and five introns. Primer elongation and S1 mapping experiments allowed the definition of the initiation of transcription and the identification of the putative promoter. The 5′‐flanking region of the RBP gene was fused upstream to the coding sequence of the bacterial enzyme chloramphenicol acetyl transferase (CAT): the chimeric gene was introduced, by calcium phosphate precipitation, into the human hepatoma cell line Hep G2 and into HeLa cells. Efficient expression of CAT was obtained only in Hep G2. Primer elongation analysis of the RNA extracted from transfected Hep G2 showed that initiation of transcription of the transfected chimeric gene occurs at a position identical to that of the natural gene. Transcriptional analysis of Bal31 deletions from the 3′ end of the RBP 5′‐flanking DNA allowed the identification of the RBP gene promoter.