The maltose binding protein of Escherichia coli is secreted into the external periplasmic compartment of the cell by virtue of an amino-terminal signal sequence. Using DNA sequencing, we have determined the precise nature of mutations in the signal sequence which prevent the export of the maltose binding protein, causing it to accumulate in the cytoplasm in its precursor form. In most cases, the change of a single hydrophobic or uncharged amino acid to a charged amino acid within the signal sequence is sufficient to block the secretion process.

In previous studies, a cyanogen bromide peptide derived from amino-acid residues 3-92 of f3-galactosidase (EC 3.2.1.23; ,1-D-galactoside galactohydrolase) was shown to have a-donor activity in intracistronic a-complementation. We have now isolated the defective fl-galactosidase a-acceptor protein from the deletion mutant strain M15 of Escherichia coli and find that it lacks residues 11-41 of f3-galactosidase. This is demonstrated by the isolation and sequence determination of a cyanogen bromide peptide from the M15 protein, which is identical to the corresponding peptide from ,8-galactosidase except for the missing amino acids. We conclude that the a-donor peptide restores the region missing in the M15 protein.Intracistronic complementation is the association between differently altered mutant proteins or protein fragments from the same cistron to give biologically functional protein. For f3-galactosidase (EC 3.2.1.23; ,3-D-galactoside galactohydrolase), the enzyme specified by the Z gene of the lactose operon in Escherichia coli, a-complementation was described by Ullmann, Jacob, and Monod in 1968 (1). Map positions of mutations in complementing strains showed that mutant strains with an intact operator-proximal region would complement a mutant strain such as M15 with a deletion in the same area. This operator-proximal 1/4 or 1/5 of the genetic length of the Z gene, which they defined as the a region, specifies the amino-terminal portion of the f3-galactosidase polypeptide chain (2).Morrison and Zipser reported that a soluble peptide fraction with a-donor activity towards extracts of M15 could be obtained by autoclave treatment of extracts of a variety of wildtype and mutant strains. The active material had a molecular weight of about 7400 (3). Similarly, Lin et al. in this laboratory found that cyanogen bromide cleavage of ,3-galactosidase also yields a peptide with a-donor activity (4). The peptide has now been isolated and determined to be the fragment derived from residues 3-92 of ,B-galactosidase (5).As an approach towards understanding the molecular nature of the interaction between this peptide (CB2) and defective 3-galactosidase from strain M15, we now report the isolation of M15 protein and the identification of the missing portion of the sequence. M15 protein, although it lacks enzyme activity, has a strong substrate binding site (6) and, therefore, could be purified with the aid of affinity chromatography. From the pure protein a cyanogen bromide peptide (AMl5CB2) was obtained which was identical to CB2 except Abbreviations: dansyl, 5-dimethylamino-naphthalene-l-sulfonyl; CB and T refer to cyanogen bromide and tryptic peptides, respectively. * To whom requests for reprints should be sent. Bacterial Strains. E. coli K12 strain M15 (2320), chromosomally lac IZM15, carries a spontaneous deletion mutation located at the extreme operator-proximal end of the Z gene (10). It was obtained from J. Beckwith. Strain RV/F'M15, supplied by A. Ullmann, is chromosomally deleted of the lac operon and carries...

The amino acid sequence of (3-galactosidase was determined. The protein contains 1021 amino acid residues in a single polypeptide chain. The subunit molecular weight calculated from the sequence is 116,248. The sequence determination, carried out mainly by conventional methods, was aided by complementation tests, by the use -of termination mutant strains, and by a new immunochemical method. The five residue sequence Thr-Pro-His-Pro-Ala appears twice within the polypeptide chain, but no other striking homologous features are evident.f3-Galactosidase (f3-D-galactoside galactohydrolase, EC 3.2.1.23) is specified by the first structural gene (lac Z) of the lac operon in Escherichia coli. Physical and chemical studies have shown that the protein is a tetramer of four identical, unusually long, polypeptide chains. Estimates of the size of the monomer have varied from about 1000 to 1200 amino acid residues; the value of 1170 has been assumed in the past (1, 2).Although the determination of t~he primary structure of ,Bgalactosidase was a major undertaking, it seemed warranted for a number of reasons. Sequence ianformation is important in order to correlate some of the extensive genetic data available on the Z gene with the protein, to investigate enzyme structure-function relationships, and to study the origin of this single protein and other proteins of the lac operon by examination for homology.The amino acid sequence of #-galactosidase has now been completed and is presented here.RESULTS AND DISCUSSION The amino acid sequence proposed for ,B-galactosidase is shown in Fig. 1. From the composition (Table 1) molecular weights of 116,248 for the monomer and 464,992 for the tetramer were calculated.The sequence was derived by studies of peptides obtained by cleavage of the protein with trypsin, chymotrypsin, and cyanogen bromide (CNBr). Structure determination was initiated by isolation of tryptic peptides (3,4) including the aminoand carboxyl-terminal fragments (5). Additional large peptides were obtained from a tryptic digest of ,B-galactosidase blocked at lysine residues with citraconic anhydride. Details of peptide isolation and sequence determination will be published elsewhere.Of the 24 unique peptides produced by cyanogen bromide treatment, 8 ranging in size from 2 to 15 residues were purified by standard techniques of paper electrophoresis and paper chromatography. The 16 larger peptides, containing 23 to 119 residues, were chromatographed at pH 5.0 on a O-carboxymethylcellulose column in 0.02 M ammonium acetate buffer containing 8 M urea and were eluted with a salt gradient (6). The elution position of these peptides can be seen in Fig. 2 in low yield which were not cleaved at certain methionine residues, or peptides derived by cleavage of the three aspartyl-prolyl bonds in fl-galactosidase. All peptides were purified further by gel filtration and, in some cases, by additional ion-exchange chromatography procedures (6). Criteria of purity included dansyl amino-terminal analysis, electrophoresis on 7.5% po...

Escherichia coli strains have been constructed in which lacZ, the gene for the cytoplasmic enzyme beta-galactosidase, is fused to lamB, the gene for an outer membrane protein. One such strain produces a beta-galactosidase which remains cytoplasmic even though it possesses the complete signal sequence of the lamB protein precursor at the amino-terminal end.

The lac operon of Escherichia coli spans approximately 5300 base pairs and includes the lacZ, lacY, and WA genes in addition to the operator, promoter, and transcription termination regions. We report here the sequence of the IacA gene and the region distal to it, confirming the sequence of thiogalactoside transacetylase and completing the sequence of the lac operon. The lacA gene is characterized by use of rare codons, suggesting an origin from a plasmid, transposon, or virus gene. UUG is the translation initiation codon. A preliminary examination of 3' end of the lac messenger in the region distal to 'the lacA gene indicates several endpoints. A predominant one is located at the 3' end of a G+C-rich hairpin structure, which may be involved in termination of transcription or in post-transcriptional processing.An 'open reading frame of 702 base pairs is present on the complementary strand downstream from kacA.Since its description more than 25 years ago, the lactose operon has been a model system of great usefulness in biology. Indeed, study of one or another aspect of the lactose operon has touched on many of the most significant questions of biology. For example, the fundamental question of the mechanisms involved in expression ofgenes was first studied in this system (1). The discovery of the lac repressor and its binding to an operator site on DNA (2, 3) was one of the first problems concerning protein-DNA interactions to be examined. Studies of P-galactosidase in relation to mutants in lacZ have been important in defining many aspects of gene-protein relationships (4). Studies with fragments of f3-galactosidase also have served as a model system for investigating protein-protein interactions (5). The lactose permease, the product of the second structural gene, lacy, was the first membrane transport protein studied extensively (6). Many fundamental concepts ofthe transport of molecules into the cell were derived from these studies. Thiogalactoside transacetylase, which is the gene product of lacA, the third structural gene, has been something of a mystery. The most reasonable interpretation of its function is that it is involved in detoxification (7). As indicated by the experiments of Andrews and Lin (7), cells containing the transacetylase are able to overcome inhibition of growth by thiogalactosides under certain conditions. It is not surprising, therefore, that structures of the components ofthe lactose operon have also been investigated intensively. The amino acid sequence of the lac repressor was determined many years ago (8), and the amino acid sequence ofI-galactosidase was reported in 1977 (9). Both of these proteins were examined by classical methods ofprotein chemistry. When methods for determining DNA sequences became available, the DNA sequence of the control elements of the lactose operon was one of the first to be studied (10).More recently, the DNA sequence of lacZ was determined (11), with the amino acid sequence of f-galactosidase as confirmation (12). The DNA sequence of lac'Y also...