Partial amino acid sequence of apolipoprotein(a) shows that it is homologous to plasminogen.

Apolipoprotein(a) [apo(a)] is a glycoprotein with Mr approximately equal to 280,000 that is disulfide linked to apolipoprotein B in lipoprotein(a) particles. Elevated plasma levels of lipoprotein(a) are correlated with atherosclerosis. Partial amino acid sequence of apo(a) shows that it has striking homology to plasminogen. Plasminogen is a plasma serine protease zymogen that consists of five homologous and tandemly repeated domains called kringles and a trypsin-like protease domain. The amino-terminal sequence obtained for apo(a) is homologous to the beginning of kringle 4 but not the amino terminus of plasminogen. Apo(a) was subjected to limited proteolysis by trypsin or V8 protease, and fragments generated were isolated and sequenced. Sequences obtained from several of these fragments are highly (77-100%) homologous to plasminogen residues 391-421, which reside within kringle 4. Analysis of these internal apo(a) sequences revealed that apo(a) may contain at least two kringle 4-like domains. A sequence obtained from another tryptic fragment also shows homology to the end of kringle 4 and the beginning of kringle 5. Sequence data obtained from two tryptic fragments show homology with the protease domain of plasminogen. One of these sequences is homologous to the sequences surrounding the activation site of plasminogen. Plasminogen is activated by the cleavage of a specific arginine residue by urokinase and tissue plasminogen activator; however, the corresponding site in apo(a) is a serine that would not be cleaved by tissue plasminogen activator or urokinase. Using a plasmin-specific assay, no proteolytic activity could be demonstrated for lipoprotein(a) particles. These results suggest that apo(a) contains kringle-like domains and an inactive protease domain.     

Delayed loss of cholesterol from a localized lipoprotein depot in apolipoprotein A-I-deficient mice

The anti-atherogenic role of high density lipoprotein is well known even though the mechanism has not been established. In this study, we have used a novel model system to test whether removal of lipoprotein cholesterol from a localized depot will be affected by apolipoprotein A-I (apo A-I) deficiency. We compared the egress of cholesterol injected in the form of cationized low density lipoprotein into the rectus femoris muscle of apo A-I K-O and control mice. When the injected lipoprotein had been labeled with [³H]cholesterol, the t½ of labeled cholesterol loss from the muscle was about 4 days in controls and more than 7 days in apo A-I K-O mice. The loss of cholesterol mass had an initial slow (about 4 days) and a later more rapid component; after day 4, the disappearance curves for apo A-I K-O and controls began to diverge, and by day 7, the loss of injected cholesterol was significantly slower in apo A-I K-O than in controls. The injected lipoprotein cholesterol is about 70% in esterified form and undergoes hydrolysis, which by day 4 was similar in control and apo A-I K-O mice. The efflux potential of serum from control and apo A-I K-O mice was studied using media containing 2% native or delipidated serum. A significantly lower efflux of [³H]cholesterol from macrophages was found with native and delipidated serum from apo A-I K-O mice. In conclusion, these findings show that lack of apo A-I results in a delay in cholesterol loss from a localized depot in vivo and from macrophages in culture. These results provide support for the thesis that anti-atherogenicity of high density lipoprotein is related in part to its role in cholesterol removal.     

Lipoprotein(a) and atherosclerosis: New perspectives on the mechanism of action of an enigmatic lipoprotein

Although elevated plasma concentrations of lipoprotein(a) (Lp(a)) have been identified as a risk factor for coronary heart disease, the pathophysiologic and physiologic roles of Lp(a) continue to elude basic researchers and clinicians alike. Lp(a) is a challenging lipoprotein to study because it has a complex structure consisting of a low-density lipoprotein-like moiety to which is covalently attached the unique glycoprotein apolipoprotein(a) (apo(a)). Apo(a) contains multiply repeated kringle domains that are similar to a sequence found in the fibrinolytic proenzyme plasminogen; differing numbers of kringle sequences in apo(a) give rise to Lp(a) isoform size heterogeneity. In addition to elevated plasma concentrations of Lp(a), apo(a) isoform size has been identified as a risk factor for coronary heart disease, although studies addressing this relationship have been limited. The similarity of Lp(a) to low-density lipoprotein and plasminogen provides an enticing link between the processes of atherosclerosis and thrombosis, although a clear demonstration of this association in vivo has not been provided. Clearly, Lp(a) is a risk factor for both atherothrombotic and purely thrombotic events; a plethora of mechanisms to explain these clinical findings has been provided by both in vitro studies as well as animal models for Lp(a).     

Autonomous functions of structural domains on human tissue-type plasminogen activator.

Transfected mouse Ltk- cells were employed for transient expression of recombinant human tissue-type plasminogen activator (t-PA; EC 3.4.21.31) or of recombinant-t-PA deletion proteins, encoded by SV40-pBR322-derived t-PA cDNA plasmids. The t-PA cDNA deletion mutants have two features in common, i.e., cDNA programming the signal peptide and the coding region for the light chain. Consequently, recombinant t-PA mutant proteins are efficiently secreted and display plasminogen activator activity. The gene encoding the amino-terminal heavy chain [an array of structural domains homologous to other plasma proteins (finger, epidermal growth factor, and kringle domains)] was mutated using restriction endonucleases to delete one or more structural domains. The stimulatory effect of fibrinogen fragments on the plasminogen activator activity of t-PA was demonstrated to be mediated by the kringle K2 domain and to a lesser extent by the finger/epidermal growth factor region but not by the kringle K1 domain. These data correlate well with the fibrin-binding properties of the recombinant t-PA deletion proteins, indicating that the stimulation of the activity by fibrinogen fragments is based on aligning the substrate plasminogen and t-PA on the fibrin matrix. Our results support the evolutionary concept of exon shuffling, arranging structural domains that constitute autonomous functions of the protein.     

Structural mimicry of a native protein by a minimized binding domain

The affinity between molecules depends both on the nature and presentation of the contacts. Here, we observe coupling of functional and structural elements when a protein binding domain is evolved to a smaller functional mimic. Previously, a 38-residue form of the 59-residue B-domain of protein A, termed Z38, was selected by phage display. Z38 contains 13 mutations and binds IgG only 10-fold weaker than the native B-domain. We present the solution structure of Z38 and show that it adopts a tertiary structure remarkably similar to that observed for the first two helices of B-domain in the B-domain/Fc complex [Deisenhofer, J. (1981) Biochemistry 20, 2361–2370], although it is significantly less stable. Based on this structure, we have improved on Z38 by designing a 34-residue disulfide-bonded variant (Z34C) that has dramatically enhanced stability and binds IgG with 9-fold higher affinity. The improved stability of Z34C led to NMR spectra with much greater chemical shift dispersion, resulting in a more precisely determined structure. Z34C, like Z38, has a structure virtually identical to the equivalent region from native protein A domains. The well-defined hydrophobic core of Z34C reveals key structural features that have evolved in this small, functional domain. Thus, the stabilized two-helix peptide, about half the size and having one-third of the remaining residues altered, accurately mimics both the structure and function of the native domain.     

A conserved domain is present in different families of vesicular fusion proteins: A new superfamily

We have analyzed conserved domains in t-SNAREs [soluble N-ethylmaleimide-sensitive factor (NSF) attachment protein (SNAP) receptors in the target membrane], proteins that are believed to be involved in the fusion of transport vesicles with their target membrane. By using a sensitive computer method, the generalized profile method, we were able to identify a new homology domain that is common in the two protein families previously identified to act as t-SNAREs, the syntaxin and SNAP-25 (synaptosome-associated protein of 25 kDa) families, which therefore constitute a new superfamily. This homology domain of approximately 60 amino acids is predicted to form a coiled-coil structure. The significance of this homology domain could be demonstrated by a partial suppression of the coiled-coil properties of the domain profile. In proteins belonging to the syntaxin family, a single homology domain is located near the transmembrane domain, whereas the members of the SNAP-25 family possess two homology domains. This domain was also identified in several proteins that have been implicated in vesicular transport but do not belong to any of the t-SNARE protein families. Several new yeast, nematode, and mammalian proteins were identified that belong to the new superfamily. The evolutionary conservation of the SNARE coiled-coil homology domain suggests that this domain has a similar function in different membrane fusion proteins.     

Crystal structure of truncated human apolipoprotein A-I suggests a lipid-bound conformation

The structure of truncated human apolipoprotein A-I (apo A-I), the major protein component of high density lipoprotein, has been determined at 4-Å resolution. The crystals comprise residues 44–243 (exon 4) of apo A-I, a fragment that binds to lipid similarly to intact apo A-I and that retains the lipid-bound conformation even in the absence of lipid. The molecule consists almost entirely of a pseudo-continuous, amphipathic α-helix that is punctuated by kinks at regularly spaced proline residues; it adopts a shape similar to a horseshoe of dimensions 125 × 80 × 40 Å. Four molecules in the asymmetric unit associate via their hydrophobic faces to form an antiparallel four-helix bundle with an elliptical ring shape. Based on this structure, we propose a model for the structure of apo A-I bound to high density lipoprotein.     

INAUGURAL ARTICLE by a Recently Elected Academy Member:Structural features of the kringle domain determine the intracellular degradation of under-γ-carboxylated prothrombin: Studies of chimeric rat/human prothrombin

Vitamin K antagonists such as warfarin inhibit the vitamin K-dependent γ-glutamyl carboxylation during protein processing and block the secretion of under-γ-carboxylated prothrombin (FII) in the rat but not in the human or bovine. Under-γ-carboxylated prothrombin is also secreted from warfarin-treated human (HepG2) cell cultures but is degraded in the endoplasmic reticulum in warfarin-treated rat (H-35) cell cultures. This differential response to warfarin has been shown to be determined by the structural difference in the proteins rather than by the origin of the cell line. When recombinant rat prothrombin (rFII) and human prothrombin (hFII) were expressed in a transformed human kidney cell line (HEK293), secretion of rFII but not hFII was drastically decreased in response to warfarin. To determine the structural signal required for this differential response, chimeric cDNAs with the propeptide/Gla domains, kringle domain, and serine protease domain exchanged between rFII and hFII were generated (FII_RHH and FII_HRR, FII_RRH and FII_HHR, FII_RHR and FII_HRH) and expressed in both warfarin-treated HEK293 cells and HepG2 cells. The presence of the hFII kringle domain changed the stability of rFII to that of hFII, and the rFII kringle domain changed the stability of hFII to that of rFII. The kringle domain therefore is critical in determining the metabolic fate of under-γ-carboxylated prothrombin precursors during processing. Prothrombin contains two kringle structures, and expression of additional rFII/hFII chimeras (FII_HrhH and FII_HhrH, FII_RrhR, and FII_RhrR) was used to determine that the first of the two kringles plays a more important role in the recognition process.     

Impact of apolipoprotein(a) on in vitro angiogenesis

Angiostatin, which consists of the kringle I-IV domains of plasminogen and which is secreted into urine, is an efficient inhibitor of angiogenesis and tumor growth. Because N-terminal apolipoprotein(a) [apo(a)] fragments, which also contain several types of kringle IV domains, are found in urine as well, we evaluated the potential angiostatic properties of these urinary apo(a) fragments and of a recombinant form of apo(a) [r-apo(a)]. We used human microvascular endothelial cell (hMVEC)-based in vitro assays of tube formation in 3-dimensional fibrin matrixes. Purified urinary apo(a) fragments or r-apo(a) inhibited the basic fibroblast growth factor/tumor necrosis factor-alpha-induced formation of capillary-like structures. At concentrations varying from 0.2 to 10 microgram/mL, urinary apo(a) fragments inhibited tube formation by as much as 70%, whereas there was complete inhibition by r-apo(a). The highest concentrations of both inhibitors also reduced urokinase plasminogen activator production of basic fibroblast growth factor-induced hMVEC proliferation. The inhibitors had no effect on plasminogen activator inhibitor-1 expression. If our in vitro model for angiogenesis is valid for the in vivo situation as well, our data point toward the possibility that apo(a) may also be physiologically operative in modulating angiogenesis, as the concentration of free apo(a) found in humans exceeds that tested herein.     

Cytohesin-1, a cytosolic guanine nucleotide-exchange protein for ADP-ribosylation factor

Cytohesin-1, a protein abundant in cells of the immune system, has been proposed to be a human homolog of the Saccharomyces cerevisiae Sec7 gene product, which is crucial in protein transport. More recently, the same protein has been reported to be a regulatory factor for the αLβ2 integrin in lymphocytes. Overexpression of human or yeast ADP-ribosylation factor (ARF) genes rescues yeast with Sec7 defects, restoring secretory pathway function. ARFs, 20-kDa guanine nucleotide-binding proteins initially identified by their ability to stimulate cholera toxin ADP-ribosyltransferase activity and now recognized as critical components in intracellular vesicular transport, exist in an inactive cytosolic form with GDP bound (ARF-GDP). Interaction with a guanine nucleotide-exchange protein (GEP) accelerates exchange of GDP for GTP, producing the active ARF-GTP. Both soluble and particulate GEPs have been described. To define better the interaction between ARF and Sec7-related proteins, effects of cytohesin-1, synthesized in Escherichia coli, on ARF activity were evaluated. Cytohesin-1 enhanced binding of ³⁵S-labeled guanosine 5′-[γ-thio]triphosphate [³⁵S]GTP[γS] or [³H]GDP to ARF purified from bovine brain (i.e., it appeared to function as an ARF-GEP). Addition of cytohesin-1 to ARF3 with [³⁵S]GTP[γS] bound, accelerated [³⁵S]GTP[γS] release to a similar degree in the presence of unlabeled GDP or GTP[γS] and to a lesser degree with GDP[βS]; release was negligible without added nucleotide. Cytohesin-1 also increased ARF1 binding to a Golgi fraction, but its effect was not inhibited by brefeldin A (BFA), a drug that reversibly inhibits Golgi function. In this regard, it differs from a recently reported BFA-sensitive ARF-GEP that contains a Sec7 domain.     

A cdc5+ homolog of a higher plant, Arabidopsis thaliana

We cloned and characterized a cDNA corresponding to a cdc5⁺ homolog of the higher plant, Arabidopsis thaliana. The cDNA, named AtCDC5 cDNA, encodes a polypeptide of 844 amino acid residues. The amino acid sequence of N-terminal one-fourth region of the predicted protein bears significant similarity to that of Schizosaccharomyces pombe Cdc5 and Myb-related proteins. Overexpression of the AtCDC5 cDNA in S. pombe cells is able to complement the growth defective phenotype of a cdc5 temperature-sensitive mutant. These results indicate that the AtCDC5 gene is a plant counterpart of S. pombe cdc5⁺. This is the first report of a cdc5⁺-like gene in a multicellular organism. We also demonstrated that a recombinant AtCDC5 protein possesses a sequence specific DNA binding activity (CTCAGCG) and the AtCDC5 gene is expressed extensively in shoot and root meristems. In addition, we cloned a PCR fragment corresponding to the DNA binding domain of human Cdc5-like protein. These results strongly suggest that Cdc5-like protein exists in all eukaryotes and may function in cell cycle regulation.     

Identification of the prooxidant site of human ceruloplasmin: A model for oxidative damage by copper bound to protein surfaces

Free transition metal ions oxidize lipids and lipoproteins in vitro; however, recent evidence suggests that free metal ion-independent mechanisms are more likely in vivo. We have shown previously that human ceruloplasmin (Cp), a serum protein containing seven Cu atoms, induces low density lipoprotein oxidation in vitro and that the activity depends on the presence of a single, chelatable Cu atom. We here use biochemical and molecular approaches to determine the site responsible for Cp prooxidant activity. Experiments with the His-specific reagent diethylpyrocarbonate (DEPC) showed that one or more His residues was specifically required. Quantitative [¹⁴C]DEPC binding studies indicated the importance of a single His residue because only one was exposed upon removal of the prooxidant Cu. Plasmin digestion of [¹⁴C]DEPC-treated Cp (and N-terminal sequence analysis of the fragments) showed that the critical His was in a 17-kDa region containing four His residues in the second major sequence homology domain of Cp. A full length human Cp cDNA was modified by site-directed mutagenesis to give His-to-Ala substitutions at each of the four positions and was transfected into COS-7 cells, and low density lipoprotein oxidation was measured. The prooxidant site was localized to a region containing His⁴²⁶ because Cp_H426A almost completely lacked prooxidant activity whereas the other mutants expressed normal activity. These observations support the hypothesis that Cu bound at specific sites on protein surfaces can cause oxidative damage to macromolecules in their environment. Cp may serve as a model protein for understanding mechanisms of oxidant damage by copper-containing (or -binding) proteins such as Cu, Zn superoxide dismutase, and amyloid precursor protein.     

Novel mechanism for carbamoyl-phosphate synthetase: A nucleotide switch for functionally equivalent domains

Carbamoyl-phosphate synthetases (CPSs) utilize two molecules of ATP at two internally duplicated domains, B and C. Domains B and C have recently been shown to be structurally [Thoden, J. B., Holden, H. M., Wesenberg, G., Raushel, F. M. & Rayment, I. (1997) Biochemistry 36, 6305–6316] and functionally [Guy, H. I. & Evans, D. R. (1996) J. Biol. Chem. 271, 13762–13769] equivalent. We have carried out a site-directed mutagenic analysis that is consistent with ATP binding to a palmate motif rather than to a Walker A/B motif in domains B and C. To accommodate our present findings, as well as the other recent findings of structural and functional equivalence, we are proposing a novel mechanism for CPS. In this mechanism utilization of ATP bound to domain C is coupled to carbamoyl-phosphate synthesis at domain B via a nucleotide switch, with the energy of ATP hydrolysis at domain C allowing domain B to cycle between two alternative conformations.     

Thrombospondin interaction with plasminogen. Evidence for binding to a specific region of the kringle structure of plasminogen

Platelet thrombospondin interacts with plasminogen in a specific and saturable manner. Thrombospondin was found to specifically bind to plasminogen and the nonenzyme chain of plasmin. Preincubation of 125I- labeled thrombospondin with 30 mmol/L lysine was without effect in the binding of thrombospondin to immobilized plasminogen; preincubation of 125I-labeled plasminogen with 30 mmol/L lysine, on the other hand, significantly reduced the binding of plasminogen to immobilized thrombospondin, suggesting that the interaction of thrombospondin with plasminogen is not the direct result of the lysine binding sites of plasminogen. Arginine and benzamidine, ligands known to specifically bind to the kringle 5 domain of plasminogen, blocked the binding of thrombospondin to plasminogen. Limited elastase proteolysis of plasminogen and plasmin resulted in the generation of two distinct thrombospondin binding domains, one of which was retained on lysine- agarose. The isolation and amino-terminal analysis of these domains following elastase proteolysis of plasminogen identified them, respectively, as a domain containing kringle structures 4 and 5 and plasmin and the other domain consisting of kringle 5-plasmin. A 16- residue synthetic peptide, which represents the amino acids linking kringle 4 to kringle 5 (residues 435-450 of native plasminogen), was without effect in either binding to thrombospondin or blocking the binding of thrombospondin to plasminogen. Plasminogen, therefore, possesses a single thrombospondin interactive site that is independent of, but influenced by, the lysine binding site containing kringle structures and most likely is located within the kringle 5 domain.     

Characterization by yeast artificial chromosome cloning of the linked apolipoprotein(a) and plasminogen genes and identification of the apolipoprotein(a) 5'' flanking region.

The apoprotein(a) [apo(a)] gene encodes a protein component of the circulating lipoprotein(a) [Lp(a)]. The apo(a) gene is highly homologous to the plasminogen gene. It encodes one of the most polymorphic human proteins, due to variability in the number of repetitions of structures called kringles. In addition, Lp(a) levels vary among individuals by more than two orders of magnitude, the high levels being highly correlated with predisposition to early atherosclerotic disease. To better understand the genetics and function of the apo(a) gene, we have cloned in yeast artificial chromosome vectors DNA fragments comprising the linked apo(a) and plasminogen genes and other members of the plasminogen family. By a combination of pulsed-field gel electrophoresis and genome walking experiments, we have identified the 5' portion and flanking regions of the apo(a) gene.     

Positionally cloned human disease genes: Patterns of evolutionary conservation and functional motifs

Positional cloning has already produced the sequences of more than 70 human genes associated with specific diseases. In addition to their medical importance, these genes are of interest as a set of human genes isolated solely on the basis of the phenotypic effect of the respective mutations. We analyzed the protein sequences encoded by the positionally cloned disease genes using an iterative strategy combining several sensitive computer methods. Comparisons to complete sequence databases and to separate databases of nematode, yeast, and bacterial proteins showed that for most of the disease gene products, statistically significant sequence similarities are detectable in each of the model organisms. Only the nematode genome encodes apparent orthologs with conserved domain architecture for the majority of the disease genes. In yeast and bacterial homologs, domain organization is typically not conserved, and sequence similarity is limited to individual domains. Generally, human genes complement mutations only in orthologous yeast genes. Most of the positionally cloned genes encode large proteins with several globular and nonglobular domains, the functions of some or all of which are not known. We detected conserved domains and motifs not described previously in a number of proteins encoded by disease genes and predicted functions for some of them. These predictions include an ATP-binding domain in the product of hereditary nonpolyposis colon cancer gene (a MutL homolog), which is conserved in the HS90 family of chaperone proteins, type II DNA topoisomerases, and histidine kinases, and a nuclease domain homologous to bacterial RNase D and the 3′-5′ exonuclease domain of DNA polymerase I in the Werner syndrome gene product.     

The prion model for [URE3] of yeast: Spontaneous generation and requirements for propagation

The genetic properties of the non-Mendelian element, [URE3], suggest that it is a prion (infectious protein) form of Ure2p, a mediator of nitrogen regulation in Saccharomyces cerevisiae. Into a ure2Δ strain (necessarily lacking [URE3]), we introduced a plasmid overproducing Ure2p. This induced the frequent “spontaneous generation” of [URE3], with properties identical to the original [URE3]. Altering the translational frame only in the prion-inducing domain of URE2 shows that it is Ure2 protein (and not URE2 RNA) that induces appearance of [URE3]. The proteinase K-resistance of Ure2p is unique to [URE3] strains and is not seen in nitrogen regulation of normal strains. The prion-inducing domain of Ure2p (residues 1–65) can propagate [URE3] in the absence of the C-terminal part of the molecule. In contrast, the C-terminal part of Ure2p cannot be converted to the prion (inactive) form without the prion-inducing domain covalently attached. These experiments support the prion model for [URE3] and extend our understanding of its propagation.     

Crystal structure of Lyme disease antigen outer surface protein A complexed with an Fab

OspA (outer surface protein A) is an abundant immunogenic lipoprotein of the Lyme disease spirochete Borrelia burgdorferi. The crystal structure of a soluble recombinant form of OspA was solved in a complex with the Fab fragment of mouse monoclonal antibody 184.1 and refined to a resolution of 1.9 Å. OspA has a repetitive antiparallel β topology with an unusual nonglobular region of “freestanding” sheet connecting globular N- and C-terminal domains. Arrays of residues with alternating charges are a predominant feature of the folding pattern in the nonglobular region. The 184.1 epitope overlaps with a well conserved surface in the N-terminal domain, and a hydrophobic cavity buried in a positively charged cleft in the C-terminal domain is a potential binding site for an unknown ligand. An exposed variable region on the C-terminal domain of OspA is predicted to be an important factor in the worldwide effectiveness of OspA-based vaccines.     

Isolation and characterization of the human apolipoprotein A-I gene.

We have recently shown that an inherited polymorphism occurring in the human apolipoprotein A-I (apo A-I) gene is related to decreased high density lipoprotein and apo A-I levels in the plasma of two patients with severe premature atherosclerosis. Analysis of the molecular basis of this polymorphism and its possible effects on apo A-I gene expression requires direct comparison of both normal and polymorphic apo A-I alleles. Here we report the isolation and characterization of the normal human apo A-I gene and we show that the gene is interrupted by three intervening sequences, IVS-1, IVS-2, and IVS-3, occurring in the 5' noncoding region of apo A-I mRNA, the mRNA sequence coding for the signal peptide of apo A-I, and the sequence coding for the mature protein, respectively. In addition, the nucleotide sequence analysis of the apo A-I gene allowed determination of the complete amino acid sequence of the primary translation product of apo A-I mRNA. This amino acid sequence consists of 267 residues including a 24-residue-long amino-terminal extension (preprosegment). Finally, we show that the apo A-I gene contains six 66-base-pair-long tandemly repeated DNA segments, which suggests that the gene may have evolved by intragenic duplication events.     

Inorganic polyphosphate and the induction of rpoS expression

Inorganic polyphosphate [poly(P)] levels in Escherichia coli were reduced to barely detectable concentrations by expression of the plasmid-borne gene for a potent yeast exopolyphosphatase [poly(P)ase]. As a consequence, resistance to H₂O₂ was greatly diminished, particularly in katG (catalase HPI) mutants, implying a major role for the other catalase, the stationary-phase KatE (HPII), which is rpoS dependent. Resistance was restored to wild-type levels by complementation with plasmids expressing ppk, the gene for PPK [the polyphosphate kinase that generates poly(P)]. Induction of expression of both katE and rpoS (the stationary-phase σ factor) was prevented in cells in which the poly(P)ase was overproduced. Inasmuch as this inhibition by poly(P)ase did not affect the levels of the stringent-response guanosine nucleotides (pppGpp and ppGpp) and in view of the capacity of additional rpoS expression to suppress the poly(P)ase inhibition of katE expression, a role is proposed for poly(P) in inducing the expression of rpoS.