We have constructed a series of plasmid vectors (pBAD vectors) containing the P BAD promoter of the araBAD (arabinose) operon and the gene encoding the positive and negative regulator of this promoter, araC. Using the phoA gene and phoA fusions to monitor expression in these vectors, we show that the ratio of induction/ repression can be 1,200-fold, compared with 50-fold for P TAC -based vectors. phoA expression can be modulated over a wide range of inducer (arabinose) concentrations and reduced to extremely low levels by the presence of glucose, which represses expression. Also, the kinetics of induction and repression are very rapid and significantly affected by the ara allele in the host strain. Thus, the use of this system which can be efficiently and rapidly turned on and off allows the study of important aspects of bacterial physiology in a very simple manner and without changes of temperature. We have exploited the tight regulation of the P BAD promoter to study the phenotypes of null mutations of essential genes and explored the use of pBAD vectors as an expression system.In bacterial physiology studies, it is often useful to express a cloned gene from an inducible promoter and assess the effect of the expression or depletion of the gene product in mutants lacking the chromosomal gene. In these situations, it is highly desirable to use a system that can be efficiently shut off. This is particularly the case when the phenotype caused by the absence of a protein can be obscured by leakiness from a partially repressed promoter or when even low levels of a protein are detrimental to the cell. Also, it would be desirable to modulate the expression system to achieve synthesis levels similar to those of the wild-type gene. However, the available repertoire of Escherichia coli expression systems usually produce high levels of the corresponding cloned gene product (4,13,18,45,48) and in many cases still produce substantial levels of synthesis in uninduced or repressed conditions (4,13,15,16,48,49). These systems include controllable expression vectors based on the strong inducible promoters P LAC (48), P TAC (13), P TRC (4), P L and P R (18), and P T7 (45). Some are better repressed than others, but induction of expression requires changes of temperature (18, 45) and produces very high levels of protein, resulting in conditions detrimental to cell growth and viability (17, 45).We have been studying the function of several essential genes that encode membrane proteins involved in cell division of E. coli (7a, 22). To analyze their role, we have sought to deplete cells of the proteins and then examine the phenotype of cells so depleted under conditions that would minimize alterations in cell physiology. For these purposes, we wished to use a plasmid that would satisfy the following two conditions: (i) the synthesis of the proteins should be shut off rapidly and efficiently without changes of temperature, and (ii) expression before depletion should not produce very high levels of protein, which itself may give a phe...

Over the last 10 years, our knowledge ofextracellular proteolysis has progressed dramatically. Different enzymatic cascades cooperate to achieve extracellular matrix (ECM)I degradation, and a number of participant proteins have been characterized and cloned. Physiological inhibitors have been identified for most of these enzymes. Also, the concept of focused proteolysis, through binding of enzymes and inhibitors to specific regions ofthe extracellular milieu, has received broad experimental support. Finally, the biosynthesis of many of the relevant proteases and inhibitors has been shown to be under the control of hormones and growth factors. Plasminogen activators (PAs) and their inhibitors (PAIs) are thought to be key participants in the balance ofproteolytic and antiproteolytic activities that regulates matrix turnover. This article summarizes the evidence that supports this contention, discusses the role of PAspecific cell surface binding sites, and also draws attention to a number of instances in which the presence of PAs cannot be reconciled with an exclusive function in ECM degradation.The plasminogen activator/plasmin system ENZYMES PAs are serine proteases of tryptic specificity. Two enzymes, differing mostly in the domain organization and function of their noncatalytic regions, have been identified in mammals: urokinase-type PA (uPA) and tissue-type PA (tPA) (1). They are the products of distinct genes, and are secreted as singlechain (sc) proteins; whereas sc tPA is active, sc uPA is essentially inactive (pro-uPA) (2). Cleavage of pro-uPA by plasmin, kallikrein, Factor XIIa or cathepsin B (3) yields the disulfidelinked two-chain active enzyme. The two PAs have distinct targeting determinants in their noncatalytic regions: the "growth factor domain" of uPA directs the binding of the enzyme (and that ofpro-uPA) to a plasma membrane receptor (4,5), whereas other structural domains in tPA (the "finger" region and the "kringles") allow its binding to fibrin and other It is impossible to select a small set of references that would provide appropriate coverage of the entire field discussed. We have thus included references to reviews and to recent papers that can be used to trace back earlier work. We apologize to all our colleagues whose contributions are not adequately referenced, as well as to our readers.Receivedfor components of the ECM (6). These and perhaps additional interactions, for instance with heparinlike molecules, could have a dramatic effect on the focusing of PA-controlled proteolysis (7). The different extracellular addressing ofthe two PAs suggests that they play different biological roles.Plasminogen is the prefered substrate for PAs, but other molecules may also be cleaved by one or the other PA; for instance, avian uPA exerts a plasmin-independent effect on the morphology of chick fibroblasts (8). Plasminogen is present in plasma and extracellular fluids at a 1-2 AM concentration, in the range of the Km of the activation reaction. It can associate with fibrin and other proteins via lysi...

The secretion of plasminogen activators has been implicated in the controlled extracellular proteolysis that accompanies cell migration and tissue remodeling. We found that the human plasminogen activator urokinase (Uk) (Mr 55,000 form) binds rapidly, specifically, and with high affinity to fresh human blood monocytes and to cells of the monocyte line U937. Upon binding, Mr 55,000 Uk was observed to confer high plasminogen activator activity to the cells. Binding of the enzyme did not require a functional catalytic site (located on the B chain of the protein) but did require the noncatalytic A chain of Mr 55,000 Uk, since P/r 33,000 Uk did not bind. These results demonstrate the presence of a membrane receptor for Uk on monocytes and show a hitherto unknown function for the A chain of Uk: binding of secreted enzyme to its receptor results in Uk acting as a membrane protease. This localizes plasminogen activation near the cell surface, an optimal site to facilitate cell migration.

Protein disulfide bond formation in Escherichia coli requires the periplasmic protein DsbA. We describe here mutations in the gene for a second protein, DsbB, which is also necessary for disulfide bond formation. Evidence suggests that DsbB may act by reoxidizing DsbA, thereby regenerating its ability to donate its disulde bond to target proteins. We propose that DsbB, an integral membrane protein, may be involved in transducing redox potential across the cytoplasmic membrane.

We show that the two-component signal transduction system of Escherichia coli, CpxA-CpxR, controls the expression of genes encoding cell envelope proteins involved in protein folding and degradation. These findings are based on three lines of evidence. First, activation of the Cpx pathway induces 5-to 10-fold the synthesis of DsbA, required for disulfide bond formation, and DegP, a major periplasmic protease. Second, using electrophoretic mobility shift and DNase I protection assays, we have shown that phosphorylated CpxR binds to elements upstream of the transcription start sites of dsbA, degP, and ppiA (rotA), the latter coding for a peptidyl-prolyl cis/trans isomerase. Third, we have demonstrated increased in vivo transcription of all three genes, dsbA, degP, and ppiA, when the Cpx pathway is activated. We have identified a putative CpxR consensus binding site that is found upstream of a number of other E. coli genes. These findings suggest a potentially extensive Cpx regulon including genes transcribed by ¢r 7° and (r r, which encode factors involved in protein folding as well as other cellular functions.