Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Cloning and characterization of the oprF gene for protein F from Pseudomonas aeruginosa Woodruff, Wendy Anne 1988

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1988_A1 W66_8.pdf [ 7.08MB ]
JSON: 831-1.0098352.json
JSON-LD: 831-1.0098352-ld.json
RDF/XML (Pretty): 831-1.0098352-rdf.xml
RDF/JSON: 831-1.0098352-rdf.json
Turtle: 831-1.0098352-turtle.txt
N-Triples: 831-1.0098352-rdf-ntriples.txt
Original Record: 831-1.0098352-source.json
Full Text

Full Text

CLONING A N D CHARACTERIZATION OF THE oprF GENE FOR PROTEIN F FROM Pseudomonas aeruginosa  by  Wendy Anne Woodruff B.Sc. (Microbiology), University of Alberta, 1982  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in The Faculty of Graduate Studies Department of Microbiology  We accept this thesis as conforming to the required standard  The University of British Columbia August 1988 © Wendy Anne Woodruff, 1988  In  presenting  degree freely  at  this  the  available  copying  of  department publication  of  in  partial  fulfilment  University  of  British  Columbia,  for  this or  thesis  reference  thesis by  this  for  his thesis  and  scholarly  or  her  for  of  Microbiology  The University of British C o l u m b i a 1956 Main Mall Vancouver, C a n a d a V6T 1Y3 Date  September 21,  1988  I  I further  purposes  gain  the  shall  requirements  agree  that  agree  may  representatives.  financial  permission.  Department  study.  of  be  It not  that  the  be  Library  an  advanced  shall  permission for  granted  is  for  by  understood allowed  the  make  extensive  head  that  without  it  of  copying my  my or  written  ii.  ABSTRACT The oprF gene encoding porin protein F from Pseudomonas aeruginosa was cloned onto a cosmid vector into Escherichia coli. Protein F was expressed in large amounts in E. coli and retained its heat- and reduction-modifiable and immunological characteristics. The cloned oprF gene product was purified from E. coli and characterized with respect to pore-forming ability in black lipid bilayers. Small channels, with an average single channel conductance of approximately 0.4 nS, were observed. A similar small channel size was observed for native protein F. The oprF sequences were used as a DNA-DNA hybridization probe with chromosomal D N A from the 17 IATS (International Antigen Typing Scheme) strains of P. aeruginosa, 52 clinical isolates and the non-aeruginosa Pseudomonads. Conservation of oprF sequences was observed among all the P. aeruginosa strains and to a lesser extent among the non-aeruginosa strains of the P. fluorescens rRNA homology group. Insertion mutations in the oprF gene were created in vivo by Tn2 mutagenesis of the cloned gene in E. coli and in vitro by insertion of the streptomycin-encoding Q fragment into the cloned gene, followed by transfer of the mutated protein F gene back into P. aeruginosa and homologous recombination with the chromosome. The oprF mutants were characterized by gel electrophoresis and immunoblotting, and it was shown that the mutants had lost protein F. The P. aeruginosa oprF mutants were characterized with respect to growth rates, antibiotic permeability and cell surface hydrophobicity. The results of these studies indicated that major alterations in the cell surface had occurred and that the cells were unable to grow in a non-defined liquid medium without added electrolytes. Marginal differences were observed in MICs (minimum inhibitory concentrations) of hydrophilic antibiotics for the oprF mutants compared with their protein Fsufficient parents. The putative roles of protein F in antibiotic permeability and general  iii. outer membrane permeability are discussed. Evidence for extensive homologies between protein F, the OmpA protein of E. coli and PHI of Neisseria gonorrhoeae are presented. A role for protein F in prophylactic anti-Pseudomonas therapy, as a target for vaccine development, is proposed.  iv.  TABLE OF CONTENTS Page ABSTRACT  ii.  TABLE OF CONTENTS List of Tables List of Figures ACKNOWLEDGEMENTS DEDICATION INTRODUCTION A. General Remarks  iv. viii. ix. xi. xii. 1 1  B. Clinical Significance and Antibiotic Resistance of P. aeruginosa  1  C. The Gram-negative Outer Membrane  2  D. Porins  6  E. Porins and Antibiotic Resistance  7  F. Protein F of P. aeruginosa  8  G. Aims of and Rationale for this Study  10  MATERIALS A N D METHODS A. Bacterial strains, bacteriophages and plasmids  11  B. Media and growth conditions  11  C. D N A techniques and manipulations 1. Plasmid isolation  15  2. Chromosome isolation  15  3. Tri-parental conjugation  15  4. Construction of the gene bank  16  D. Techniques for characterization of protein F 1. Black lipid bilayer analysis  16  T A B L E O F C O N T E N T S (Continued)  Page 2. Antibiotic mimimal inhibitory concentrations (MIC) determinations  16  3. N P N uptake assays  17  4. Growth curve experiments  17  5. LPS preparation and electrophoresis  17  6. Alkaline phosphatase assays  17  7. Bacteriophage susceptibility testing  18  8. Isolation of specific D N A fragments from gels . . . .  18  9. Southern blots and DNA hybridization  18  10. Transposon mutagenesis and gene replacement. . .  19  E. Monoclonal antibodies and antibody procedures 1. Monoclonal antibodies  21  2. Antibody procedures a. Colony immunoblotting  21  b. Western blots  21  F. Protein procedures 1. Outer membrane preparation  22  2. Cell envelope preparation  22  3. SDS-PAGE  22  4. Purification of protein F  23  RESULTS Chapter 1. Cloning the oprF gene from P. aeruginosa A. Introduction: Other outer membrane protein genes cloned B. Screening the gene bank  24 27  vi. T A B L E O F C O N T E N T S (Continued)  Page C. Characterization of the cloned gene product D. Subcloning strategies and restriction mapping  27 29  E. Characterization of novel gene products from pWW4, pWW5 and pWW12  31  F. Conservation of oprF among the Pseudomonaceae  35  G. Summary  38  Chapter 2. Insertion mutagenesis of the oprF gene in vivo and in vitro  A. Introduction  40  B. Transposon mutagenesis of the oprF gene in E. coli (pWW13) and gene replacement in P. aeruginosa  41  C. Q mutagenesis of the oprF gene  43  D. Characterization of the P. aeruginosa oprF mutants  45  E. Summary  50  Chapter 3. Functional studies of purified protein F and the P. aeruginosa oprF strains  A. Introduction  51  B. Growth rate determinations  52  C. Miscellaneous uptake studies 1. N-phenylnaphthylamine uptake  56  2. Phosphate uptake  58  D. Antibiotic uptake analysis: MICs  58  E. Black lipid bilayer studies  60  F. Summary  66  vii. T A B L E O F C O N T E N T S (Continued)  Page DISCUSSION A. Protein F: Cloning, mutagenesis and conservation  68  B. Protein F: Antibiotic resistance  71  C. Protein F: Channel size  74  D. Protein F: Relationship with OmpA  78  E. Protein F: Clinical relevance  83  REFERENCES  85  viii.  LIST OF TABLES Table  Title  Page  I  Bacterial strains and bacteriophages  12  II  Plasmids  14  HI  Hybridization of Pseudomonas species D N A with pWW2200  IV  37  Doubling times for P. aeruginosa H103 and its derivatives H608 oprFr.Tnl and H636 oprF::Q in broth culture during the logarithmic phase of growth at 37°C  V  53  Uptake of N P N by strain HI03 and its protein F-deficient derivatives H608 oprFv.Tnl and H636 oprF::Q  VI  Geometric mean MIC determinations for the oprFy.Tnl H608 and the wild-type strain H103 (RPl::Tn2)  VII  57  61  Mean MICs for strain H103 and its protein F-deficient Q. insertion mutant H636  62  ix.  LIST OF FIGURES Figure  Title  1.  Schematic representation of the cell envelope of a gram-negative bacterial cell  2.  4  SDS-PAGE profiles of outer membrane preparations of E. coli containing oprF clones  3.  Page  26  SDS-PAGE profiles of outer membrane preparations of £. coli and P. aeruginosa showing the 2-mercaptoethanolmodifiability of protein F  4.  28  Western immunoblots of outer membrane proteins of P. aeruginosa HI 03 and E. coli strains containing the truncated oprF encoding plasmids  30  5.  Subcloning strategy for oprF  32  6.  Western immunoblot analysis of outer membrane proteins of E. coli strains containing pHN4, pWW12, pWW13 and pWW2200  7.  Diagram of the restriction site heterogeneity of the oprF gene among the 17 serotypes of P. aeruginosa  8.  34  36  Restriction map of the oprF gene showing the locations of the insertion sites of Tnl and Q in the oprF insertion mutants  9.  42  Strategy for the construction of pWW2500 for Q mutagenesis of the oprF gene  44  X.  LIST O F F I G U R E S (Continued)  Figure 10.  Title  Page  SDS-PAGE profiles of cell envelope proteins of the protein F-deficient P. aeruginosa strains and their protein F-sufficient parents  11.  46  SDS-PAGE profile of silver-stained lipopolysaccharide (LPS) from the protein F-deficient P. aeruginosa mutants and their protein F-sufficient parents  12.  47  Autoradiograph of a Southern blot hybridization of a radiolabeled oprF probe to chromosomal D N A digests of P. aeruginosa HI03, H636 and H637  13.  Effect of NaCl on the logarithmic growth rate of H103 and H636 grown in PP2 broth  14.  49  55  SDS-PAGE profile of cell envelope preparations of P. aeruginosa H608 oprF::Tn2 grown in minimal medium with varying amounts of inorganic phosphate  15.  SDS-PAGE profiles of the purified protein F used for black lipid bilayer analysis  16.  65  Schematic representation of the speculative disulfide rearrangement model  18.  64  Histogram of the conductance increases observed for purified protein F in black lipid bilayer membranes  17.  59  75  Comparison of the amino acid homology between protein F from P. aeruginosa, OmpA from E. coli, and PHI from N. gonorrhoeae  80  xi. Acknowledgements  The financial assistance of the Canadian Cystic Fibrosis Foundation and R.E.W. Hancock is gratefully acknowledged. The author would like to acknowledge the friendship, guidance, support and training of her supervisor, R.E.W. Hancock, the members of his lab 1983 through 1988, the members of her supervisory committee, the members of the Department of Microbiology, the University of British Columbia, Barbara H. Iglewski, the members of her lab in 1984 and 1985, and the members of the Department of Microbiology and Immunology at Oregon Health Sciences University, Portland, Oregon. In particular, the assistance of Jan Treit, Carrie Hirsch and Frank Tufaro in the word processing of this manuscript was invaluable. The love and support of many family members and personal friends is humbly acknowledged.  xii. DEDICATION  This thesis is lovingly dedicated to the multitude of people whose love, support, humor, discipline and nurturing have enabled me to accomplish this task and to the glory of the Creator who in His infinite wisdom has given me the privilege of a unique, first-hand view of a small part of His creation.  1  INTRODUCTION A. General Remarks Pseudomonas aeruginosa is a gram negative, non-fermentative, nonmotile, rod-shaped bacterium. P. aeruginosa can be isolated from a variety of environments ranging from soil, food and standing water, through to medical instruments and disinfectants. This organism is metabolically diverse, nutritionally undemanding and amply endowed with enzymes, toxins, pigments, slime layers and adhesins, all of which help facilitate its ability to live and replicate on animate and inanimate moist surfaces. It is therefore not surprising that P. aeruginosa is an opportunistic pathogen. Infection by P. aeruginosa requires either local impairment of the patient's anti-infection defenses (i.e., damage to skin and mucous membranes) by physical stress, diseases, accidents, medical treatment or by general impairment of the immune system. True to its opportunistic nature, P. aeruginosa is largely a nosocomial pathogen which establishes infections in specific subgroups of patients. These include immunocompromised patients, especially cancer or transplant patients, burn victims, patients with eye infections, diabetics, intravenous drug users, premature infants and patients with indwelling medical apparati such as catheters or shunts. P. aeruginosa is most commonly isolated from patients with pneumonia or respiratory tract infections. The association of P. aeruginosa with cystic fibrosis has long been established and chronic, long term infection with this organism often leads to pulmonary failure and death in cystic fibrosis patients.  B. Clinical Significance and Antibiotic Resistance of P. aeruginosa P. aeruginosa is not considered to be a regular constituent of human microflora. Extensive studies of stool samples and ear-nose-throat swabs showed that 3% or less of the healthy adult population carried P. aeruginosa (Botzenhart & Ruden, 1987). This finding coupled with the opportunistic,  2  non-invasive nature of P. aeruginosa infections makes it difficult to understand the clinical incidence of this pathogen. For American hospitals in 1980, Escherichia coli was the most frequent isolate followed by Staphylococcus aureus, P. aeruginosa and Klebsiella sp. E. coli remains  the most frequent hospital acquired pathogen (approximately 18% of total isolates); however, by 1984 P. aeruginosa had moved to second place, accounting for 11.4% of all nosocomial isolates (Horan, et al, 1986). In one study it was reported that 43% of patients hospitalized for more than fifteen days became carriers of P. aeruginosa (Bodey, et al., 1983). Another study reported that P. aeruginosa was responsible for 20% of gram negative bacteremic infections but that in such infections, the incidence of mortality was 84% (Young, et al, 1982). The severity of and high mortality rate caused by P. aeruginosa infections is due in part to this organism's ability to resist antimicrobial therapy. Pseudomonas is significantly more resistant to antibiotics when compared with other gram negative organisms. This is generally true for all antibiotics but particularly so for (3-lactam compounds. More than fifty penicillins, cephalosporins and other B-lactam antibiotics are used clinically and the minimal inhibitory concentration (MIC) required to kill P. aeruginosa, in vitro, is eight to one hundred times that required to inhibit E. coli (Rolinson, 1986). This phenomenon is attributed in part to low outer membrane permeability (Angus, et al, 1982; Godfrey & Bryan, 1987). Experiments designed to quantify whole cell permeability estimate that the outer membrane permeability of P. aeruginosa is twelve-fold (Nicas & Hancock, 1983) to one hundred-fold (Yoshimura & Nikaido, 1982) lower than that of E. coli. Since the target sites of most antibiotics are inside the outer membrane, this structure constitutes a formidable barrier to antibiotic therapy and control of infection.  C. The Gram Negative Outer Membrane The outer membrane of a gram negative bacterium consists of an asymmetric lipid bilayer interspersed with protein molecules. Associated  3  with this membrane is a layer of peptidoglycan inside the periplasmic space which separates the outer membrane from the inner (cytoplasmic) membrane. These three layers are collectively referred to as the cell envelope and are shown in a schematic diagram in Figure 1. The outer membrane is the cell structure which is in contact with the external environment. Accordingly, it contains several unique types of molecules. The outer leaflet of the lipid bilayer contains few of the phospholipids normally encountered in cytoplasmic membranes and has for its major lipid component, lipopolysaccharide (LPS). LPS is composed of a lipid A core moiety embedded in the bilayer, a hydrophilic core oligosaccharide and a highly variable O-antigenic side chain of repeating units of unusual sugars which protrudes into the external medium. Only 10-25% of LPS molecules in a given strain have an O-antigenic side chain (Nikaido & Hancock, 1986). LPS is highly antigenic. The inner leaflet of the outer membrane contains phospholipids, predominantly phosphatidylethanolamine (Nikaido & Hancock, 1986) and a molecule called lipoprotein which is non-covalently associated with the peptidoglycan layer (Mizuno & Kageyama, 1979). The  outer membrane lipid bilayer contains a large number of  proteins, but a few are present in large amounts and seem to predominate. Accordingly, these are the most well characterized. Under normal growth conditions there seem to be three categories of proteins in enteric bacteria: porins, which form transmembrane channels; heat-modifiable proteins, which have reduced electrophoretic mobility after heating in sodium dodecyl sulfate (SDS) and the murein lipoprotein. Under special conditions such as phosphate, iron or magnesium depletion or in the presence of high salt, the synthesis of new proteins is induced and they appear in the outer membrane. As well, there is a phospholipase activity associated with the outer membrane of E. coli (Nikaido & Hancock, 1986) which may indicate an enzymatic function for some outer membrane proteins. In P. aeruginosa nine major outer membrane proteins have been described (Nikaido & Hancock, 1986). Proteins D2, E, F, G, HI, H2 and I are expressed under normal laboratory growth conditions. Under these conditions, proteins F, H2 and I predominate. Proteins I and H2 have similar characteristics to the enteric  4  Figure 1. Schematic representation of the cell envelope of a gram negative bacterial cell. The outer membrane is a bilayer containing lipopolysaccharide (LPS) molecules (shown with the long sugar side chains and the crosshatched lipid A portion), phospholipids, lipoprotein (L), globular proteins and porin molecules (P). Separating the inner and outer membranes is the periplasmic space which contains a layer of crosslinked peptidoglycan. The inner membrane is a bilayer of phospholipids interspersed with proteins which may span the membrane but do not form constantly open channels.  5  Braun's lipoprotein and peptidoglycan-associated lipoprotein respectively. Proteins D l , F and P have pore-forming activity and proteins D l , P and HI are induced in the presence of glucose or the absence of phosphate and magnesium ions, respectively. Protein F is a heat-modifiable protein (Mizuno & Kageyama, 1979; Hancock & Carey, 1979). The contribution of the outer membrane, particularly the outer membrane proteins, to the life of a bacterial cell, has been the focus of intensive research for the past fifteen years (for reviews: Nikaido & Vaara, 1985; Nikaido & Hancock, 1986; Hancock, 1987). It has been shown that the cell envelope contributes to maintenance of cell shape, general permeability, antibiotic resistance and evasion of host defenses. Each of these functions, in whole or in part, has been attributed to outer membrane proteins. The role of lipoprotein and LPS is largely considered to be in maintaining the structural integrity and barrier properties, respectively, of the outer membrane. Little is known about the interactions between lipids and proteins or their functional interdependence. Some evidence suggests that LPS is essential for the poreforming activity of the porins (Schindler & Rosenbusch, 1981), but this has been disputed (Nikaido & Rosenberg, 1981; Nikaido & Rosenberg, 1983; Hancock, 1987). The heat-modifiable proteins are thought to play a structural role and together with the lipoprotein may be responsible for the shape of the cell (Sonntag, et al, 1978). The heat-modifiable (OmpA-like) proteins have been shown to function as phage and colicin receptors (Lugtenberg & van Alphen, 1983) and to be necessary in stabilizing mating aggregates during F-pilus mediated conjugation (Manning & Achtman, 1979). Porin proteins have a unique function in the outer membrane. They form transmembrane hydrophilic channels, allowing the passive diffusion of small solutes into the cell. The diameter of the porin channel determines the size of molecule which can enter the cell, therefore, porins are responsible for the selective permeability characteristics of the outer membrane. This class of molecules will be discussed in detail below.  6  D. Porins The outer membrane is a selective permeability barrier which allows hydrophilic molecules of a certain size to enter the cell but excludes larger hydrophilic molecules and hydrophobic molecules. The exclusion limit of a given outer membrane appears to be a function of the most extensively characterized outer membrane proteins, the porins. Porins constitute a significant proportion of the outer membrane proteins, being present in 10^ to 10^ copies per cell. These molecules function as water-filled transmembrane channels or pores and serve as receptors for colicins and phage. Since the term porin was coined in 1976 (Nakae, 1976), forty-four channel-forming molecules have been identified and characterized from thirty-two species representing ten separate families of gram negative bacteria (Hancock, 1987a). The best characterized porins are Omp F, Omp C, Pho E and Lam B of E. coli. The genes for these and several other enteric porins have been cloned and sequenced (Tommassen, et al., 1982; Mizuno, et al, 1983; Overbeeke, et al, 1983). In general, porins have a monomeric molecular weight of 28,000 to 48,000, a high percentage of P-sheet structure and a strong affinity to LPS; are resistant to denaturation by SDS at low temperatures; are non-covalently associated with the peptidoglycan layer in the periplasm and are functionally active as trimers (Nikaido & Vaara, 1985). Many of the enteric porin proteins are immuno- logically crossreactive (Hofstra & Dankert, 1979). A variety of methodologies exist to study the pore-forming ability of proteins in model membranes. These are based on the formation of porincontaining liposomes or black lipid bilayer membranes. Using these assay systems, the channel diameters and exclusion limits have been calculated for many porins and range from 0.6 nm for protein P of P. aeruginosa to 2.3 nm for the 32 kiloDalton (kD) protein of Spirochaeta aurantia and 600 Daltons for enteric porins to 6000 Daltons for protein F from P. aeruginosa, respectively (Hancock, 1987b). Despite significant differences in these model membrane systems, it has been shown by Hancock (1987b) that the estimated porin diameter from both black lipid bilayer studies and liposome studies is very similar. Most porins form general permeability channels for small  7  hydrophilic molecules but may be cation or anion selective (Nikaido, Rosenberg & Foulds, 1983). Some early investigators of E. coli porins in model membranes showed that application of high voltage (>100 mV) across a planar phospholipid bilayer containing porins could cause the pores to close (Schindler & Rosenbusch, 1978). Later investigators have failed to find any physiological or experimental data to support voltage dependent channel closure and it has been suggested that such high voltages are nonphysiological (Nikaido & Vaara, 1985). There is some evidence with Neisseria gonorrhoeae porins that a membrane potential is required for activity but this may be related to the ability of these molecules to spontaneously transfer from the gonococcal outer membrane into eukaryotic membranes (Blake & Gotschlich, 1987).  E. Porins and Antibiotic Resistance As stated earlier, the outer membrane of gram negative bacteria is a permeability barrier to all molecules which gain access to the cell. Antibiotics are no exception. Three recognized pathways for antibiotic uptake exist in gram negative bacteria: the hydrophilic pathway, the self-promoted uptake pathway and the hydrophobic pathway (Moore, et ah, 1987). It should be noted that other, alternative routes of entry may exist (Nayler, 1987), and that a particular compound may be able to permeate the outer membrane by more than one pathway (Yamaguchi, et ah, 1985). Because porins are water-filled transmembrane channels which allow the diffusion of hydrophilic molecules into the cell it seemed reasonable that these molecules would serve as a major uptake route for hydrophilic antibiotics, particularly pMactams. Studies involving porin-deficient or porin-altered mutants of E. coli, Salmonella typhimurium  and P. aeruginosa have shown that there is a positive  correlation between loss of, or change in, a porin protein and an increase in the minimum inhibitory concentration of (3-lactams (Godfrey & Bryan, 1987; Nikaido, et al, 1977). This phenomenon seems to be dependent on the size and physicochemical characteristics (particularly charge) of the antibiotic (Nayler, 1987; Curtis, et al, 1985) but leaves little room for doubt that porins  8  may be involved in P-lactam uptake. Recently, it has been proposed that porins alone are not responsible for P-lactam uptake, but that porin-mediated, low outer membrane permeability acts in synergy with periplasmic Plactamases to effect resistance (Vu & Nikaido, 1985; Hancock & Woodruff, 1988).  F. Protein F of P. aeruginosa Protein F from P. aeruginosa has been purified and characterized by several groups (Hancock, et al, 1979; Benz & Hancock, 1981; Mizuno & Kageyama, 1979; Yoshimura, et al, 1983). In each case, a single protein of approximately 35,000 D has been isolated and found to have pore-forming ability. Protein F was shown to decrease in apparent mass, M , after heating r  from approximately 35 kD to approximately 40 kD and thus is classified as a heat-modifiable protein. A similar decrease in mobility was also observed when protein F was treated with 2-mercaptoethanol (Hancock & Carey, 1979) or trichloroacetic acid (Mizuno & Kageyama, 1979). Amino acid analysis showed that protein F contained four cysteine residues (Mizuno & Kageyama, 1979). Some evidence suggests these cysteines form two intrachain disulfide bonds (Hancock & Carey, 1979) and reduction of these bonds could account for the shift in electrophoretic mobility. Besides the single cysteine residue at the N H 2 terminus of all enteric lipoproteins (Yu, 1987), the only other outer membrane proteins known to contain cysteine residues are the OmpA proteins of Enterobacteriaceae (Braun & Cole, 1984), PHI of Neisseria gonorrhoeae (Gotschlich, et al.,1987) and M O M P of Chlamydia trachomatis  which in the EB form is extensively polymerized by interchain disulfide bonds (Bavoil, et al, 1984). It has been postulated that the disulfide bonds are involved in channel formation in protein F (Moore, et al, 1987). Circular dichroism studies showed that protein F was rich in p-sheet structure (Mizuno & Kageyama, 1979). Protein F has been the source of controversy in the scientific literature. One of these controversies involves the oligomeric structure of protein F. Most other porin molecules characterized can be isolated as trimers and are  9  functional as pores only in this configuration (Hancock, 1987b). In contrast, protein F has rarely been isolated in a multimeric configuration and most pore-forming activity has been associated with these monomers. The only other bacterial protein shown to have pore-forming monomers is from Rhodopseudomonas sphaeroides (Weckesser, et al., 1984). Yoshimura, etfl/.(1983),showed that in contrast to E. coli Omp  F, SDS-purified protein F  could not be crosslinked by dimethylsuberimidate and sedimentation studies by analytical centrifugation showed only monomers, never oligomers. Angus and Hancock (1983) showed that 40% of Triton X-100 purified protein F could form dimers when crosslinked by dithio-bis-succininidyl propionate and 20% could form trimers. These authors cite the unavailability of appropriately spaced and oriented amine groups to interact with the crosslinker for the difficulties in demonstrating protein F trimers. In addition, the difference in the detergents used during purification may result in a conformational change in protein F (see Discussion) which may  include  the ability of this protein to associate into multimers. Such a proposal is speculative. The size of the protein F channel has also been the subject of controversy. Preliminary studies of vesicles containing outer membrane proteins from P. aeruginosa showed that this bacterium had an outer membrane exclusion limit of approximately 6000 Daltons, ten-fold higher than that of outer membranes from enteric bacteria (Hancock, et al., 1979). Purified protein F was then incorporated into black lipid bilayer membranes and shown to form large channels with a single-channel conductance of 5 nS in 1M KC1, a value approximately two and one half times that of Omp  F  (Benz & Hancock, 1981). These data were paradoxical. As stated earlier, the outer membrane of P. aeruginosa was known from in vivo studies of permeability to be much less permeable than that of E. coli yet it was observed to have porins more than twice as large. To account for this, Benz and Hancock (1981) proposed that most of the protein F channels existed in a closed state, that is, only a few molecules formed large channels. This explanation seemed plausible based on the observation that 5 ng/mL of protein F had to be added to the black lipid bilayer membrane to observe channels at the same frequency as a 0.1 ng/mL solution of Omp  F from  10  E. coli (Benz, et al, 1982). Other investigators have been unable to demonstrate the existence of such large channels in studies of permeability in vivo. Caulcott, et al, (1984), measured the efflux of radiolabeled solutes from plasmolyzed cells of P. aeruginosa and concluded that disaccharides (molecular weight, approximately 350) were the largest sugars which could penetrate the outer membrane. Yoneyama and Nakae (1986) estimated the exclusion limit of the P. aeruginosa outer membrane to be larger than hexose (molecular weight, 180) but smaller than an uncharged disaccharide by measuring the penetration of sugars into plasmolyzed cells. Thus, the findings of Caulcott, et al. and Yoneyama and Nakae, that the exclusion limit of P. aeruginosa is 180 to 350 Daltons, contrast to those of Benz and Hancock, who reported the exclusion limit to be 6000 + 3000 Daltons. Resolution of this conflict is critical to understanding the role of protein F in permeability but particularly the role of protein F in antibiotic uptake as most hydrophilic antibiotics are larger than 350 Daltons (Nayler, 1987; Rolinson, 1986).  G. Aims of and Rationale for this Study In light of the clinical significance of P. aeruginosa infections resulting from the extreme resistance of this organism to antibiotic therapy, it was decided to attempt to resolve some of the controversy regarding the permeability of the outer membrane in the hope that clinically and pharmaceutically relevant data would be obtained. The first approach was to clone the gene for protein F from the chromosome of P. aeruginosa into E. coli. This would allow the comparative influence on permeability of nonprotein F components of the Pseudomonas outer membrane to be examined as the cloned gene product could be characterized in the E. coli background. The second approach was to construct defined mutants of the protein F gene and to look at the effect of loss of protein F on the permeability of the P. aeruginosa outer membrane.  11  MATERIALS AND METHODS A. Bacterial strains, bacteriophages and plasmids. The strains of Eschericha coli and Pseudomonas used in this study and their bacteriophages are listed in Table I. In addition, the seventeen International Antigen Typing Scheme (IATS) prototype strains of P. aeruginosa, A T C C numbers 33348-33364, and fifty-six clinical isolates of P. aeruginosa were used. The clinical isolates were the generous gift of Suzanne Steinbach (Boston University School of Medicine, Boston Massachusetts) and included forty cystic fibrosis sputum isolates, one urine isolate, six sepsis isolates and nine blood isolates. The plasmids used in this study are listed in Table II. B. Media and growth conditions In general, bacteria were grown in rich media broth cultures of 37oC with rigorous aeration. Some strains of non-aeruginosa Pseudomonads were grown at 30oC. The rich media used were LB (1% tryptone, 0.5% yeast extract, 1% NaCl, p H 7.0), TY (0.8% tryptone, 0.5% yeast extract, 0.5 % NaCl p H 7.0), PP2 (1% proteose peptone no. 2) or Mueller-Hinton broth. Media were solidified with 2% Bacto-agar. All media components were from Difco Laboratories, Detroit, Michigan. Minimal media were Vogel-Bonner medium (Maniatis, et al, 1982), BM2 (Gilleland, et al, 1974) supplemented with 0.2% (v/v) succinate for Pseudomonas or 1% (v/v) glucose for E. coli or phosphate deficient medium (Hancock, et al, 1982). Cultures of E. coli JF733 were routinely grown on LB with 300 mM  NaCl  and 0.1% (w/v) glucose to suppress induction of OmpF and LamB. Porin deficient strains of P. aeruginosa were grown in Mueller-Hinton broth or PP2 with 200 mM NaCl. Antimicrobial agents were used in selective media at the following concentrations for E. coli: tetracycline at 25 (ig/mL; ampicillin at 50 (ig/mL; for P. aeruginosa: tetracycline at 100 |ig/mL, streptomycin at 500 |ig/mL, carbenicillin at 300 (ig/mL and mercuric  12  Table I: Bacterial Strains and Bacteriophages Strain  Characteristics  Reference  E. coli K-12  HB101  hsdR hsdM recA13 ara-14 thi-1 proA2 lacYl  Maniatis et al, 1982  galK2 mtl-1 xyl-5 supE 44, X-  Maniatis et al, 1982  MM294  pro thi end A hsdW  DH5cc  endAl hsdR17 supE44 thi-1 recAl gyrA96 A {argF-lacZYA)U\69  JF733  relAl  BRL  08O dlac ZAM15  aroA ilv met his purE41 pro cyc-1 xyll lacY29  J. Foulds, 1979  cpsL77 tsx63 ompA ompC LE392  hsdR514 supE44 sup58 lacYl galK2 galT22 metBl  Maniatisef al, 1982  trpR55 AT2453  thi-1 lys relA spoTl  Bukhari & Taylor, 1971  P. aeruginosa P A O PAOl Cm  H376  PAO1840 met-9020 leu-9005 hex-9001  H457  PA0222 met-28 trp-6 \ys-\2 his-4 ilv-226 R68.45  H576  H103 oprP::Tn501  H608  H103  H636  H103 oprFv.Q.  This study  H637  H376 oprF::Q.  This study  H638  H575 oprF::Q.  This study  H670  H457 oprF::Q  This study  H103(RPI)  H103 (RPI Tcr Kmr Apr [Tnl])  Nicas & Hancock, 1983  PCC23  PA053 oprF*  Godfrey, et al. 1984  r  prototroph; wild type reference strain  Hancock & Carey, 1979  H103  oprF::Tnl-3  Poole & Hancock, 1986 This study  Table I: Bacterial Strains and Bacteriophages (Continued)  Other Pseudomonaceae A T C C number H289  P. fluorescens  13525  H291  P. putida  12633  H348  P.  H358  P.  H359  P. cepacia  25416  H360  P.  chloraphis  9446  H361  P.  maltophilia  13637  H362  P.  pseudomallei  23343  H363  P.  solanacearum  11696  H364  P. stutzeri  17588  H365  P. syringae  19310  P.  9355  C351  anguilloseptica aureofaciens  acidovorans  Azotobacter  vinelandii  13985  13705  Bacteriophages  F116L  P. aeruginosa transducing phage  Krishnapillai, 1971  XcI857  c/857  Maniatis et al, 1982  14  Table II: Plasmids  Name  Characteristics/Origin  Reference  pLAFRl  IncPlTc rZx Xcos  Friedman et al.  pCP13  IncPl Tcr Kirf rlx Xcos  Darzins et al.  pRK404  IncPl (RK2 replicon) Tcr  Ditta et al.  r  plac and the multicloning site of pUC8 Jorgensen et al.  pRZ102  ColEl::Tn5 mob K m  R68.45  IncPl Tcr Cbr Kmr Trar Cma+  Haas & Holloway  pRK2013  ColEl-Tra(RP2)+ Kmr  Figurski & Helinski  pUC18  ColEl Apr  Yanisch-Perron et al.  RSF1010::Tnl  IncQSmrCbr  Ohman et al.  RSF1010::Tn5  IncQSmr Kmr  Ohman et al.  RSF1010::Tn501  IncQSmr Hgr  Ohman et al.  RSF1010::Tn7  IncQ Smr  Ohman et al.  pHN4  pLAFRl+ 25 kb insert of P.aeruginosa D N A oprF+  This study  pWW4  pCP13 + 4.7 kb insert of non-colinear P. aeruginosa  This study  r  D N A coding for a truncated protein F molecule pWW5  pUC8 + 2.0 kb insert of pWW4 which  This study  codes for a truncated protein F molecule pWW12  p L A F R l + 11.2 kb EcoRl fragment of pHN4  This study  including an altered oprF gene. pLAFr + 11-kb EcoRl fragment of pHN4 opr F +  This study  pWW13 oprFv.Tnl  This study  pWW2200  pRK404 + 2.4 kb Psfl insert from pWW13, oprF+  This study  pWW2300  pUC18 + 1.4 kb Pstl-Sall insert from pWW220  This study  pWW2300 + Q. fragment ligated into  This study  pWW13 pWW13::Tnl-3  pWW2300£2  the Smal site of oprF,  oprF-.iQ.  pRK404 + 3.5 kb Pvull fragment pWW2400  of pWW2300Q,oprF::Q pRZ102 + 2.5 kb Sail fragment of  pWW2500  This study  pWW2400,  oprF-.-.Q.  This study  When using the pUC or pRK404 plasmids, plates were spread with 50JIL of dimethyl formamide containing 50 ng of 5-bromo-4-chloroindoyl-P-D-galactoside. C. D N A techniques and manipulations 1. Plasmid isolation Large scale quantities of plasmid DNA were isolated by an alkaline lysis method (Maniatis, et al, 1982) followed by centrifugation in ethidium-bromide-cesium  chloride gradients. Alternatively, plasmid  D N A was isolated by a boiling procedure followed by PEG precipitation (Sadhu and Gedman, 1988) after which gradient centrifugation was not required. Small scale isolation of plasmid D N A was done by the rapid boiling method of Crouse, et al. (1983). Routine screening of transformants for inserts was by the "slot-lysis" procedure of Sekar (1987). 2. Chromosomal D N A isolation Chromosomal D N A was isolated from 10 mL overnight cultures by the procedure of Meade, et al., (1982), resuspended in sterile distilled water and quantitated by measurement of the absorption at 260 nm. 3. Tri-parental conjugation Overnight cultures of the donor plasmid-containing strain and the helper plasmid strain E. coli MM294 (pRK2013), were grown in broth at 37°C with shaking. The recipient Pseudomonas strain was grown overnight at 42©C with shaking. Aliquots of the donor, helper and recipient strains (0.5-2 mL, depending on the density of the cultures) were mixed, diluted into 5 volumes fresh broth and incubated 60 min at room temperature. The cell mixture was then filtered onto a 0.45 um membrane. The filter was placed cell side up on a freshly made non-selective agar plate and incubated 6-24 h at 30°C. Finally, cells were washed off the filter into sterile saline, serially diluted, spread onto selective plates and incubated at 30°C for up to 48h.  16  4. Construction of the gene bank The gene bank was constructed following the procedure described by Goldberg and Ohman (1984). Chromosomal DNA, isolated from HI 03, was partially digested with EcoRl and size fractionated by centrifugation through linear 10-40% sucrose gradients. Fragments of 15-25 kb were ligated into the EcoRl site of the cosmid pLAFRl. The ligated molecules were packaged in vitro into phage lambda particles and then transfected into E. coli HB101. The transfectants were plated onto LB agar plates containing 25 (ig/mL tetracycline. Colonies were scraped into broth supplemented with 10% (v/v) glycerol and stored at -70oC. Individual colonies from this bank were screened using the colony immunoblot procedure with monoclonal antibodies specific for protein F. D. Techniques for characterization 1. Black lipid bilayer analysis The techniques, methods and instrumentation used for this procedure have been described in detail elsewhere (Woodruff, et al, 1986). For these analyses the protein was diluted 10 thousand-fold in 2% Triton X-100, the aqueous solution in the chamber was 1 M KCl, the bilayer was formed from 1.5% oxidized cholesterol in n-decane, and the applied voltage was 50 mV. 2. Antibiotic minimal inhibitory concentrations (MIC) determinations MICs were determined by the agar dilution method. Approximately ten thousand cells grown to log phase in Mueller-Hinton broth were spotted onto Mueller-Hinton agar plates containing doubling two-fold dilutions of antibiotics. The antibiotics used were gifts from the manufacturers and were used according to their recommendations. MIC endpoints were determined after 24 h of incubation at 37oC. Each MIC analysis was performed in triplicate and each experiment was repeated at least five times on separate days.  3. NPN uptake assays The uptake of N-phenylnaphthylamine (NPN) by intact bacterial cells was measured exactly as described by Loh, et al, (1984). 4. Growth rate measurements The effect of salts and sugars on the growth rate of the protein F deficient strains and their wildtype parent was assessed as follows: cultures of stationary phase bacteria grown in PP2 broth plus 200 mM  NaCl  were diluted forty-fold into Klett flasks of fresh PP2 broth containing 0 to 400 mM  KCl, NaCl, MgCl2, CaCl2, sucrose, glucose or succinate. Cultures  were incubated at 37o C with vigorous shaking. Cell density was determined using a Klett-Summerson photometer with a green filter at hourly intervals for an 8-12 hour period. 5. LPS preparation and electrophoresis The procedure followed was a modification of that of Hitchcock and Brown, (1983) for LPS staining of protease-digested whole cell lysates. Bacteria were streaked onto agar plates and incubated 18 h at 37oC. Loopfuls of cells were suspended in sterile saline to an optical density of approximately 0.4 at 600 nm. Cells from a 1.5 mL aliquot were centrifuged, resuspended in lysing buffer (2% (v/v) SDS, 4% (v/v) 2mercaptoethanol, 10% (v/v) glycerol, 1M TrisCl pH 6.8) and heated ten min at 100°C. Fifty micrograms of proteinase K were added and the mixtures were incubated 2 h at 60oC then sonicated briefly and an aliquot loaded directly onto a 15% acrylamide, 0.5% SDS gel (Peterson et al., 1985). After electrophoresis the LPS was visualized by periodate treatment and silver staining (Tsai & Frasch, 1982) and photographed. 6. Alkaline phosphatase assays The hydrolysis of p-nitrophenyl phosphate (pNPP) by alkaline phosphatase of a whole cell bacterial extract was assayed as follows: 5 mL of log phase culture were broken in the French press. An aliquot of the cell extract was mixed with a solution of pNPP in a spectrophotometer cuvette and the increase in absorbancy over time at 410 nm was monitored  18  on a chart recorder. The rate of hydrolysis of pNPP was calculated from the slope of the curve on the chart recorder tracing. 7. Bacteriophage susceptibility testing The oprF mutant strains and their parents were tested for sensitivity to thirty Pseudomonas-speciiic bacteriophages from the lab phage collection (Nicas & Hancock, 1980). Diluted phages (108 pfu/mL) were spotted from a multisyringe applicator onto a lawn of bacteria on an agar plate. The plates were incubated overnight at 37C and phage susceptibility, as determined by plaque formation, was scored. 8. Isolation of specific DNA  fragments from gels  Many different techniques were attempted but the most satisfactory was the band interception onto DEAE-cellulose paper (Schleicher and Schuell Inc., Keene N.H.)  method of Winberg and  Hammarskjold (1980). The protocol followed was that included by Schleicher and Schuell with the DEAE paper. 9. Southern blots and DNA Fragments of DNA,  hybridization techniques which had been digested with restriction  enzymes, were separated on agarose gels by electrophoresis in 89 mM  Tris-  89 mM  DNA  borate-5 mM  EDTA pH8.0 buffer. After electrophoresis, the  was transferred to Zeta-probe membranes (Bio-Rad Laboratories, Richmond, CA) by capillary transfer in a 0.4 M N a O H solution exactly as described in the manufacturer's instructions. DNA  probes were prepared  by isolating the desired fragment from an agarose gel, then radioactively labelling by DNA  elongation using the Klenow fragment of  DNA  polymerase I in the presence of 32p-dATP, hexameric random primers (Pharmacia, Dorval, Quebec), dCTP, dGTP and dTTP (Feinberg & Vogelstein, 1983). Unincorporated nucleotides were removed by passing the labelling reaction mixture over an Elutip-d column (Schleicher and Schuell, Keene, N.H.) as per manufacturer's instructions. Labelled probes (approximately 108 cpm/|ig DNA)  were hybridized to DNA  bound to Zeta-  probe membranes following the Bio-Rad protocol for standard  19  hybridizations. For high stringency experiments, the hybridizations took place at 68°C; for low stringency experiments the hybridization was at 58°C and dextran sulfate was added to the hybridization solution to a final concentration of 10% (w/v). Conditions for washing the Zeta-probe and autoradiography were those recommended by the manufacturer. For hybridizations of radiolabeled D N A to bacterial colonies, the probe was prepared as described above. The colonies were hybridized on nitrocellulose circles following the protocol outlined by Maniatis, et ah, (1982). 10. Transposon mutagenesis and gene replacement The transposon mutagenesis protocol followed was that of Ohman, et ah, (1985). Plasmid RSF1010::Tnl was transformed into E. coli HB101 (pWW13). Transformants containing both plasmids were infected with bacteriophage ArI857. Plasmid pWW13 was chosen as the target for mutagenesis because it has an 11 kb fragment of P. aeruginosa DNA, including the oprF gene, in a cosmid vector which is too small to be packaged into A, particles unless Tnl has transposed onto pWW13. A A, lysate was prepared from the E. coli HB101 (pWW13, RSF1010::Tnl) and used to transfect E. coli LE392. Transfectants which were tetracycline and carbenicillin resistant and protein F-deficient as determined by colony immunoblots with a protein F-specific monoclonal antibody (MA5-8) were selected for further manipulation. The sites of Tn2 insertion into oprF were ascertained by restriction mapping and Southern blot analysis (using pWW2200 as a probe). The mutated plasmids pWW13 oprF::Tn2 were triparentally conjugated into P. aeruginosa as the first step of gene replacement. The D N A of the transconjugants was packaged in vivo with F116L. The resulting phage lysate was used to infect P. aeruginosa PAOl. Transductants which were carbenicillin resistant and tetracycline sensitive (had lost pWW13 but retained Tn2 ) were screened for the loss of protein F by colony immunoblotting as above. One of the P. aeruginosa oprFv.Tnl mutants was named H608 and further characterized. This procedure was also followed using Tn5, Tn7 and Tn501 in place of Tnl.  The Q fragment is a commercially available D N A cartridge (Amersham Corp. Oakville, Ontario) of 2.0 kb which contains a streptomycin resistance gene flanked on each end by three transcriptional stop sequences. As supplied by the manufacturer, Q has Smal ends. In order to insert Q into oprF, several plasmids needed to be constructed (see Fig. 9). A 1.5 kb Pstl-Sall fragment containing the amino-terminal coding region of oprF was ligated into the Smal site of pUC18 to form pWW2300. This fragment was made blunt-ended by reaction with the Klenow fragment of D N A polymerase I as per Maniatis, et al, (1982). Plasmid pWW2300 was mutagenized by the insertion of Q into the Smal site of the oprF sequences to form pWW2300 Q. The 3.5 kb Pvull fragment of pWW2300 Q was isolated and ligated into the filled-in Hindlll site of pRK404 resulting in pWW2400. This step was useful in generating Sail ends on either side of the oprF::Q so this fragment could then be isolated and ligated into the unique Sail site in the Tn5 sequences of pRZ102. This final construct, pWW2500, was used for gene replacement of oprF in P. aeruginosa with oprFv.Q. .  To effect gene replacement, several features of the pRZ102 sequences of pWW2500 were exploited. A combination of the lack of a suitable replication origin for P. aeruginosa, the presence of a mob site which allowed triparental conjugation from E. coli into P. aeruginosa, and the instability of Tn5 in this host promoted homologous recombination between the oprFv.Q sequences inside the IS50 elements of Tn5 on pWW2500 and the chromosomal oprF in P. aeruginosa. Streptomycin resistant, kanamycin sensitive transconjugants were screened by colony immunoblots procedures to assess the loss of protein F. This procedure was used to construct oprFv.Q. mutants in several different strains of P. aeruginosa (Table I).  E. Monoclonal antibodies and antibody procedures 1. Monoclonal antibodies The production and characterization of the protein F specific monoclonal antibodies used in this study have been described in detail by Mutharia and Hancock, (1983,1985a). Monoclonal antibodies specific for two distinct epitopes on protein F were used. MA5-8 recognizes a surface epitope in the carboxy terminal portion of the molecule; MA4-4 recognizes an epitope resulting from formation of internal disulfide bonds in protein F (it will not bind to protein F which has been treated by 2mercaptoethanol) and cross reacts with protein F-like outer membrane proteins from several other Pseudomonads (Mutharia & Hancock, 1985a). 2. Antibody procedures a. Colony  immunoblotting  Bacterial colonies were transferred from agar plates to nitrocellulose circles by contact. The colonies were lysed by hanging the filters in a chloroform vapor-saturated chamber for 15 min. After removal, the filters were allowed to dry for 5 min then placed in empty petri dishes and covered with blocking buffer [3% bovine serum albumin (BSA) in phosphate buffered saline (PBS) with 10 (ig/mL DNAase I and 40 Hg/mL of lysozyme]. After incubation at 37°C for 1 h on a rocking platform, the filters were washed 3 times in PBS, covered with 1% BSA in PBS containing MA5-8 or MA4-4 (diluted one in one thousand) and incubated 2 h at 37oC. The filters were washed as described then incubated another hour at 37oC with a goat-anti-mouse-immunoglobulin Gperoxidase conjugate (Cappell) and washed again in PBS. The blots were developed using 4-chloro-l-naphthol (0.5 mg/mL in PBS with 10% methanol), a histochemical substrate for peroxidase, b. Western blots Bacterial proteins were separated on SDS-PAGE gels (section F. 3) and electrophoretically transferred to nitrocellulose from the gel using the procedure of Mutharia and Hancock (1983). After transfer, protein F was detected immunologically as described for colony immunoblots,  omitting the chloroform lysis step and the DNase I and lysozyme from the blocking buffer. F. Protein procedures 1. Outer membrane preparation Outer membranes were prepared by the method of Hancock and Carey (1979). Briefly, mid-log phase cells were harvested by centrifugation, treated with lysozyme and DNase, and broken by passage through a French pressure cell. The extract was loaded onto a 70:58:52:18% step sucrose gradient for centrifugation to equilibrium in a rotor with swinging buckets. Two to four bands were visible after centrifugation. The bottom band was collected, diluted with water to reduce the concentration of sucrose and pelleted by ultracentrifugation. This pellet contained outer membrane proteins suitable for electrophoresis or further purification. 2. Cell envelope preparation Cell envelopes, the inner and outer membranes plus peptidoglycan, were prepared as described above except after passage through the French press cell, the lysate was made to 2% Triton X-100 (v/v) final concentration and centrifuged at 45000 rpm in a fixed angle Ti 70.1 rotor (Beckman Instruments. Inc.) for 45 min. The pellet contained the cell envelopes which were resuspended in water and electrophoresed. 3. SDS-PAGE Outer membrane proteins were separated by electrophoresis through 10% or 14% polyacrylamide gels as previously described (Hancock & Carey, 1979). Before electrophoresis, proteins were solubilized in 2% SDS with or without 10% 2-mercaptoethanol for 10 min at 20° or 100°C. After electrophoresis proteins were visualized by staining with Coomassie Brilliant Blue R250 (Bio-Rad) or Western blotting.  23  4. Purification of protein F Protein F was purified from 5 L cultures of P. aeruginosa and E. coli JF733 (pWW2200) grown under conditions described previously (Woodruff, et al, 1986). Outer membranes were prepared on sucrose gradients as described above. After dilution of the sucrose, the membrane pellets were extracted twice with 2% Triton X-100 -10 mM and twice with 2% Triton X-100-10 mM  Tris HC1-10 mM  Tris HC1 p H 8.0 EDTA p H 8.0, and  the pellets were collected by ultracentrifugation at 177,000x g for 45 min after each extraction. After the final extraction the pellet was resuspended in 10 mM  Tris p H 8.0 and treated with 0.1 mg/mL lysozyme for 1 hour at  25oC. Finally SDS and glycerol were added to a final concentration of 2% and 10% respectively and the proteins were electrophoresed on a 10% preparative (3mm) SDS-PAGE. The position of the major protein band was determined by staining narrow gel strips in Coomassie Blue. The band was cut out and the protein was allowed to diffuse from the acrylamide into a small volume of water for 48 h at 4oC.  The purified protein was  electrophoresed on an SDS-PAGE to assess purity and subsequently used in the black lipid bilayer apparatus.  24  RESULTS Chapter 1. C l o n i n g the oprF gene from P. aeruginosa  A. Introduction: Other outer membrane protein genes cloned The genes for outer membrane proteins, both porins and the heatmodifiable OmpA proteins, have been cloned from a variety of bacteria (Cole, et al, 1982; Carbonetti & Sparling, 1987; Freudl & Cole, 1983; Gotschlich, et al, 1986; Henning, et al, 1979; Manning, et al, 1985; Stephens, et al, 1986). Without exception, the outer membrane protein gene products were lethal to their E. coli host if cloned with an intact promoter into a high copy number vector. Accordingly, most genes were cloned in a lysogen system. If plasmid vectors were used, they were generally low copy number and the E. coli host strain was deficient in one or more outer membrane proteins (ie., ompB, ompA).  Through various manipulations, most of the cloned genes have  been expressed in E. coli and have retained their original physical and immunologic properties (Carbonetti & Sparling, 1987; Cole, et al, 1982; Gotschlich, et al, 1987). One group (Braun & Cole, 1984) found that a subclone of OmpA from Serratia marcescens, containing the promoter and approximately 66% of the amino terminal coding sequence, was not lethal but consistently encoded an outer membrane protein of 23,500 Daltons (D) even though a 28,600 D protein was expected from the sequence. This phenomenon had been documented earlier in a study on the export of shortened derivatives of E. coli OmpA where, regardless of the length of sequence deleted from the carboxy terminus, to a maximum of seventy-seven amino acids, a 24,000 D protein was observed (Bremer, et al, 1982; Henning, et al, 1983). The expression of foreign outer membrane protein genes in E. coli has been shown to alter the relative amounts of the porins (Tommassen, et al, 1982) or OmpA (Braun & Cole, 1984), and in one instance caused the disappearance of a major E. coli protein (Carbonetti & Sparling, 1987). It is unknown if these changes resulted from the ability of the cloned gene product to assume the functions of the diminished host proteins. Some of  25  the enteric Omp A proteins could replace the E. coli Omp A protein functionally with regard to serving as phage or colicin receptors and in Fmediated conjugation (Cole, et al, 1982; Freudl & Cole, 1983). The nucleotide sequences of many of these cloned genes have been determined. Tables of both the D N A and amino acid sequence data have been compiled for comparative purposes (Braun & Cole, 1984; Gotschlich, et al, 1987; Mizuno, et al, 1983; Yu, 1987). Significant homology (>60% identity) has been observed between the porins OmpF, OmpC and PhoE at both the D N A and amino acid levels for E. coli but there is no significant homology between these proteins and the LamB porin. Earlier studies, based on various chemical analyses and some sequence data, concluded that there was little evolutionary conservation among the porins within a species or between genera (Lee et al, 1979; Mizuno, et al, 1983; Overbeeke,e£ al, 1983). Sequence homology between PI of N. gonorrhoeae and E. coli porins is less than 23% and there is no homology between these molecules and the porin of Chlamydia trachomatis (Gotschlich, et al, 1987). These findings were in sharp contrast to those of another study using similar techniques which found that the heat-modifiable (OmpA) proteins for all strains of Enterobacteriaceae and some more phylogenetically distant gram negative bacteria were strongly conserved during evolution (Beher, et al, 1980). These conclusions were confirmed when the nucleotide sequence data were compared. The enterobacterial OmpA protein genes have such similar D N A sequences that one group (Cole, et al, 1982) was able to clone the OmpA genes of Shigella dysenteriae, Enterobacter aerogenes and S. marcescens  using DNA-DNA hybridization to a probe from the E. coli OmpA gene (Braun & Cole, 1983; Braun & Cole, 1984; Freudl & Cole, 1983; Yu, 1987). Significant amino acid homology has also been observed between OmpA from E. coli and Pin, a highly conserved heat-modifiable outer membrane protein from N. gonorrhoeae (Gotschlich, et al, 1987), providing further evidence that the heat-modifiable proteins are highly conserved in evolution.  Figure 2. SDS-PAGE profiles of outer membrane preparations of E. coli containing oprF clones . Lane 1, P. aeruginosa HI03; lane 2, E. coli HB101 (pHN4); lane 3, E. coli HB101; lane 4, E. coli HB101 (pWW4); lane 5, E. coli HB101 (pWW5). Molecular weight markers, in thousands (K), are indicated on the right. The position of protein F (F) and the truncated protein F gene product (T) are indicated on the left. The positions of Omp F, Omp C, and Omp A in the E. coli strains are indicated.  27  B. Screening the gene bank The p L A F R l cosmid gene bank of P. aeruginosa P A O l D N A in E. coli HB101  was plated for single colonies and screened for the production  of protein F antigen. Of approximately three thousand five hundred colonies screened, five reacted with monoclonal antibodies specific for protein F. One of these five colonies was arbitrarily selected for characterization and further manipulations. The plasmid in this E. coli HB101  clone was designated  pHN4. C. Characterization of the cloned gene product The reaction, with protein F-specific monoclonal antibodies, of E. coli HB101  colonies harboring cosmids containing P. aeruginosa D N A indicated  that protein F was expressed in E. coli. Native protein F is found in the outer membrane of P. aeruginosa as a major protein which interacts with MA4-4 and MA5-8 and undergoes a shift in electrophoretic mobility from 35 kiloDaltons (kD) to 41 kD on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) after heating or treatment with 2-mercaptoethanol (Hancock & Carey, 1979; Mutharia & Hancock, 1983). Outer membranes were isolated from E. coli HB101  (pHN4) and JF733  (pHN4), a porin-deficient strain of E. coli. SDS-PAGE profiles of these outer membranes are shown in Figures 2 and 3, lanes 4 and 7. In each profile, a protein, having the same molecular weight as native protein F (Fig. 2, lane 1; Fig. 3, lanes 2 & 5), was present and it was not found in outer membrane profiles of E. coli strains without pHN4 (Fig. 2, lane 3 and Fig. 3, lanes 3 & 6). In E. coli JF733, the plasmid-encoded protein was the predominant outer membrane protein. In the porin-sufficient strain, E. coli HB101  containing  pHN4, there was no detectable change in expression of the major outer membrane proteins OmpF, OmpC or OmpA (Fig. 2, lanes 2 and 3). This new protein band was identified as protein F by its interaction with the protein F-specific monoclonal antibodies MA4-4 and MA5-8 (Fig. 4 A & B, lanes 5 and Fig. 6 A & B, lanes 2). In each Western blot, a single major protein band reacted with antibody. A faint reaction occurred with a slightly  1 2  3 U  5 6 7  Figure 3. SDS-PAGE profiles of outer membrane preparations of E. coli and P. aeruginosa showing the 2-mercaptoethanol modifiability of protein F. Samples in lanes 2, 3 and 4 were treated with 5% 2-mercaptoethanol before leading onto the gel. Lanes 2 and 5, P. aeruginosa HI03; lanes 3 and 6, E. coli JF733; lanes 4 and 7, E. co/i JF733 (pHN4). The 2-mercaptoethanol modified (F*) and unmodified (F) forms of protein F are indicated. Molecular weight markers (lane 1) were bovine serum albumin (66,000 D), ovalbumin (45,000 D), glycerol-3-phosphate dehydrogenase (36,000 D), carbonic anhydrase (29,000 D), trypsinogen (24,000 D), soybean trypsin inhibitor (20,000 D) and oc-lactalbumin (14,400 D).  29  higher molecular weight protein in the E. coli (pHN4) lanes with both MA4-4 and MA5-8.  This phenomenon is often observed in outer  membrane preparations from P. aeruginosa and is presumed to be the result of partial heat modification (Hancock & Carey, 1979). In E. coli HB101 (pHN4), (Fig. 4 B, lane 5), a lower molecular weight band was observed which interacted with only MA4-4. This band was presumably a proteolytic breakdown product of protein F. Proteolysis of gene products from outer membrane protein genes cloned in E. coli has previously been reported (Braun & Cole, 1984). Differential reactivity of enzyme-generated cleavage fragments with MA4-4 and MA5-8 has been observed previously (Mutharia & Hancock, 1985a). The effect of 2-mercaptoethanol on the cloned oprF gene product was also examined. In both E. coli HB101  (pHN4) and E. coli JF733 (pHN4), the  protein F band increased its apparent molecular weight after heating in 2-mercaptoethanol. This phenomenon was difficult to observe in E. coli HB101  (pHN4) because the 2-mercaptoethanol modified band comigrated  with OmpA and was obscured. In E. coli JF733 (pHN4) the effect of 2-mercaptoethanol on protein F was easily observed (Fig. 3, lanes 4 & 7) and was identical to the shift in molecular weight observed for protein F in P. aeruginosa (Fig. 3, lanes 2 & 5). D. Subcloning strategies and restriction mapping Cosmid, pHN4, was an unwieldy 45 kb. The first subcloning experiments were aimed at simply reducing the size of the insert in pHN4. Purified pHN4 D N A was digested with a restriction enzyme, religated and transformed into competent E. coli cells. Protein F antigen-expressing transformants were detected using MA5-8 in the colony immunoblot. Transformants reacting with MA5-8 were isolated when pHN4 was digested with Xhol and EcoRl. The Xhol lineage of subclones ended with pWW4and pWW5 (Fig. 5). These plasmids encoded a truncated gene product (Fig. 2, lanes 4 & 5; Fig. 4 A & B, lanes 3 & 4). Further characterization of the D N A of pWW4 and pWW5 by restriction mapping and Southern blotting with pWW2200 D N A led to the conclusion  30  A  B 1  F-  2  3  4  5  1  2  3  4  5  —  T-  Figure 4. Western immunoblots of outer membrane proteins of P. aeruginosa H103 and E. coli strains containing the truncated oprF encoding plasmids. The proteins in panel A were reacted with MA5-8; the proteins in panel B were reacted with MA4-4. Lanes 1, P. aeruginosa HI 03; lanes 2, E. coli HB101 (pCP13); lanes 3, E. coli HB101 (pWW4); lanes 4, E. coli HB101 (pWW5); lanes 5, E. coli HB101 (pHN4). The position of protein F is indicated (F) and that of the truncated gene product is labelled T.  31  that the insert D N A in pWW4 and pWW5 was not colinear on the P. aeruginosa chromosome and the exact length of oprF D N A sequences or the position at which sequences were no longer colinear would be determined only by D N A sequence analysis. The EcoRl line of subclones (Fig. 5) included pWW13 and pWW2200. Both these plasmids encoded a gene product in the outer membranes of their host strains identical to those of E. coli (pHN4) and P. aeruginosa with respect to molecular weight and monoclonal antibody reactivity (Fig. 6 A &B, lanes 2, 4, 5 & 6). The smallest oprF containing subclone, pWW2200, was a 2.4 kb PstI fragment (Fig. 5) in pRK404 and contained unique Kpnl, Smal and Sail sites inside the coding sequence. Repeated attempts were made to ligate the 2.4 kb PstI fragment into a small high copy number vector but transformants containing such constructs were never recovered. This phenomenon has been observed for other outer membrane protein genes (see Results, Chapter 1A ) and by another group working on protein F. Duchene, et al, (1988) successfully sequenced oprF by making overlapping subclones, but were unable to clone the entire gene into a single vector. Their failure presumably was a result of the fact that they used high copy number vectors instead of low copy number, broad host range vectors such as pRK404. The 2.4 kb PstI fragment from pWW2200 was radiolabeled and hybridized to genomic digests of P. aeruginosa PAOl. In digests with enzymes known to cut outside oprF sequences, a single band hybridized to the PstI probe. From these data it was concluded that there is a single copy of oprF in the P. aeruginosa chromosome. This observation is consistent with the demonstration that the structural genes for the enteric porins and heatmodifiable outer membrane protein are present as single copy genes despite the large amounts of protein products synthesized from these genes (Sato & Yura, 1979; Tommassen & Lugtenberg, 1981). E. Characterization of novel gene products from pWW4, pWW5 and pWW12 During the subcloning experiments, several subclones were generated which produced truncated or slightly aberrant protein F gene products. Two  32  Figure 5. Subcloning strategy for oprF. The cosmid clone isolated from the gene bank, pHN4, is shown as a circle. The thick line represents vector sequences. Subclones pWW12 and pWW13 were 11.3 kilobase (kb) and 11.0 kb fragments of pHN4, respectively, containing oprF. Subclone pWW2200 was a 2.4 kb Pst I fragment of pWW13 and contained the entire oprF gene. Subclones pWW4 (4.7 kb) and pWW5 (2.0 kb) produced a truncated gene product which reacted with monocloned antibodies against protein F. The protein F gene sequences, oprF are shown as a thick black box in all the subclones. Sequences of pWW4 and pWW5, which are non-colinear in the P. aeruginosa chromosome, are shown as cross-hatched boxes. A 1.0 kb size reference marker is given for the linear fragments. The restriction enzyme sites indicated by single letter codes are: E, EcoRl; K, Kpnl; M, Smal; P, Pstl; S, Sail. The question mark denotes the possible presence of a Sail site.  33  of these, pWW4 and pWW5, encode a protein of approximately 24,000 D (Fig. 2, lanes 4 &5) which interacted with MA5-8 but not with MA4-4 (Fig. 4 A & B, lanes 3 & 4). The truncated protein was transported to the outer membrane but was not 2-mercaptoethanol modifiable and did not form channels when analyzed in black lipid bilayers. The presence of this protein in the outer membrane of E. coli HB101 caused a significant decrease in the expression of proteins OmpF and OmpC but no noticeable difference in OmpA (Fig. 2). It should be noted that E. coli cells containing plasmids encoding both the entire and truncated protein F molecules were fractionated into inner and outer membrane, periplasmic and cytoplasmic fractions, electrophoresed and blotted to determine the location of protein F. In both cases, the protein F was predominantely found in the outer membrane, only traces of protein were detected in the inner membrane and these may represent contamination. The significance of this observation with respect to transport, assembly or organization of the outer membrane is unknown. Although the exact nature of these plasmids at the D N A level is unknown, a similar phenomenon has been observed with subclones of OmpA from both E. coli and S. marcescens as noted earlier (see Results, Chapter 1A). Another novel protein was encoded by pWW12. This plasmid was isolated during the subcloning of EcoRl fragments of pHN4. Plasmid pWW12 had an 11.3 kb EcoRl fragment and was similar to pWW13 (Fig. 5). At the D N A level, the difference between pWW12 and pWW13 was the length of the insert and consequently the size of the PstI fragments. The EcoRl insert in pWW12 was 11.3 kb, 0.3 kb longer than that in pWW13. However, the PstI fragment containing oprF was 2.1 kb in pWW12, 0.3 kb shorter than that in pWW13. The sizes of restriction fragments upstream of oprF and in the amino terminal coding sequences of pWW12 were identical to those of pWW13. In addition to differences in the size and positions of some of the D N A restriction fragments, the gene products encoded by oprF in pWW12 and pWW13 were different (Fig. 6 A & B, lanes 3 & 4). Protein F from pWW12 was slightly smaller than the native protein and the cloned gene products of pHN4, pWW13 and pWW2200 (Fig. 6 A, lane 3). It did not react with MA5-8  A  B 1 2 3 4 5 6  1 2 3 4 5 6  Figure 6. Western immunoblot analysis of outer membrane proteins of E. coli strains containing pHN4, pWW12, pWW13 and pWW2200. Cell envelope proteins were separated by electrophoresis, transferred to nitrocellulose and reacted with MA5-8 (panel A) and MA4-4 (panel B). Lanes 1, E. coli JF733; lanes 2, E. coli JF733 (prIN4); lanes 3, E. coli MM294 (pWW12); lanes 4, E. coli MM294 (pWW13); lanes 5, E. coli JF733 (pWW2200) and lanes 6, P. aeruginosa H103.  35  but did react with MA4-4 (Fig. 6 B, lane 3). While the exact nature of the D N A rearrangement could be determined by D N A sequence analysis, which may be useful in defining amino acid residues involved in epitope formation, the simplest explanation for these observations is that a 0.3 kb D N A insertion occured at the carboxy-terminal encoding end of oprF which contained a PstI cleavage site and truncated the carboxy terminus of protein F, thus destroying the antigenic epitope recognized by MA5-8. Such rearrangements have been proposed to occur spontaneously within the surface-exposed amino terminal regions of the highly conserved, heatmodifiable proteins as a means of maintaining sequence polymorphism between genera, but have not been observed in the carboxy terminal domains (Braun & Cole, 1984). F. Conservation of oprF among the Pseudomonaceae Chromosomal D N A was isolated from the seventeen I ATS serotypes of P. aeruginosa and fifty-six P. aeruginosa clinical isolates. The D N A was digested with restriction enzymes, electrophoresed, transferred to nylon membrane and hybridized with the radioactively labelled PstI fragment of pWW2200. Among the P. aeruginosa strains, the hybridization patterns were surprisingly consistent. With the exception of serotype 12 and one clinical isolate, the hybridization patterns were identical for the remaining seventy-one strains for a region of about 5 kb around and including oprF (Fig. 7). The differences for serotype 12 are shown in Figure 7 and involve the Sail and Kpril fragments. Among the other serotypes, there was some heterogeneity within the Kpril sites several kilobases downstream of oprF. The clinical isolate with the unique hybridization pattern was a sepsis isolate and had the same hybridization pattern for KpnI-digested D N A as the serotype 12 type strain. Unfortunately, serotyping data on the clinical isolates was not available. It is interesting to note that in PstI digests of genomic D N A from all seventy-three P. aeruginosa strains, a single 2.4 kb fragment hybridized with the probe of pWW2200. This confirmed the previous conclusion that oprF is present in a single copy in the P. aeruginosa genome. This pattern of tight  Serotypes 1-11,13-17  P  K  KP  K  -M-  ~r S  Serotype  K M S  S  K ©  P  MS  S  K'  12  (R)  P  (K)  1 kb  Figure 7. Diagram of the restriction site heterogeneity of the oprF gene among the 17 serotypes of P. aeruginosa. The oprF restriction sites were strongly conserved for all 17 serotypes except serotype 12 which has the restriction map shown. The non-conserved sites are circled. The one letter codes for restriction enzyme sites are: K, Kpnl; M, Smal; P, PstI; and S, Sail. Several of the other serotypes, 5, 6, 8 and 17, showed some heterogeneity among their Kpnl sites. For these serotypes, the most downstream Kpnl site, K', was 1.4 kb to the right of that in the other strains. Regardless, this Kpnl site was significantly outside the coding region of oprF, which is represented by a thick black rectangle.  Table III: Hybridization of Pseudomonas species D N A with pWW2200  Strain  Chromosome D N A Pst I digests  fragment  H103  P.  H298  P. fluorescens*  4(S)  H348  P.  3.0,1.6 (S)  H358  P. aureofaciens  -  H359  P. cepacia  -  H360  P. chloraphis*  2.8(S)  H361  P. maltophilia  -  H362  P.  pseudomallei  -  H363  P.  solanacearum  -  H364  P. stutzeri*  3.0,1.6 (S)  H365  P. syringae*  3.4 (S)  H398  P. putida*  3.4 (W)  P.  acidovorans  -  A.  vinelandii*  C351  aeruginosa*  Size of hybridizing  anguilloseptica*  2.4  5(S)  *react with M A 1-6 specific for protein H2 (lipoprotein) and belong to the P. fluorescens taxonomic group based on rRNA homology. (DeVos & DeLey, 1983; Mutharia, et al., 1985). S=strong; W=weak interaction with the probe.  conservation was not observed for the exotoxin A genes of P. aeruginosa (Ogle, et ah, 1987) but has been observed for the omp A genes of the Enterobacteriaceae and PHI of N. gonorrhoeae (Gotschlich, et al,  1987;  Nikaido & Vaara, 1985). The interaction of the pWW2200 probe with genomic DNA  of non-  aeruginosa Pseudomonas strains was also examined using less stringent hybridization conditions. The results of these experiments are presented in Table III. Homology, as determined by hybridization of radiolabeled oprF sequences to genomic digests, was observed for P. fluorescens, P. anguilloseptica, P. chloraphis, P. stutzeri, Azotobacter vinelandii  and  P. putida. The hybridization signal was strong for all these strains except P. putida. Two bands hybridized with the oprF probe from PstI digests from P. anguilloseptica and P. stutzeri. When DNA  of these strains was  digested with other enzymes and probed with pWW2200, more than one band hybridized. It is unknown whether there are multiple copies of an oprF-like gene in these strains or simply restriction sites for the enzymes used within the oprF-like sequences. The interaction of outer membrane proteins from these Pseudomonas strains with P. aeruginosa specific antibodies has been determined (Mutharia & Hancock, 1983; Mutharia & Hancock, 1985b). None of these strains interact with protein F-specific MA5-8 but outer membrane proteins of P. syringae and P. putida bind to MA4-4. Interestingly, all the non-aeruginosa Pseudomonas strains containing DNA hybridizing to the oprF probe DNA,  sequences  also interacted with MA1-6, a  monoclonal antibody specific for protein H2, the P. aeruginosa lipoprotein (Mutharia & Hancock, 1985c). G. Summary The oprF gene of P. aeruginosa, which encodes outer membrane protein F, was cloned into E. coli using a cosmid vector. The gene product of oprF was observed in the outer membrane of E. coli and exhibited the same electrophoretic mobility as the native protein. The interaction of the cloned gene with 2-mercaptoethanol and protein F-specific monoclonal antibodies  39  MA4-4 and MA5-8 was identical to that of the native protein. In the course of subcloning, several clones were isolated which expressed novel gene products. One of these clones, pWW5, encoded a 24,000 D protein which interacted with MA5-8, but not MA4-4. Another subclone, pWW12, produced a protein slightly smaller than native protein F which interacted with MA4^4, but not MA5-8. The oprF gene was subcloned as a 2.4 kb PstI fragment. This fragment was radiolabeled and used in hybridization studies with digests of chromosomal DNA. For PstI digested chromosomal D N A from P. aeruginosa strains PAOl, the seventeen IATS serotyping strains and fifty-six clinical isolates, a single fragment of 2.4 kb hybridized to the oprF probe. This indicated that there is a single copy of oprF in the chromosome and that this gene is highly conserved. Differences in the sizes of fragments within the oprF gene were observed in one serotype and one clinical isolate when their D N A was digested with Sail or KpnI. Differences in the sizes of downstream KpnI fragments were observed for several serotypes. Homology was observed when oprF was used to probe chromosome D N A from nonaeruginosa Pseudontonas strains under less stringent conditions, indicating that there is conservation of oprF-like sequences between species as well as within species of Pseudomonas.  40  Chapter 2. Insertion mutagenesis of the oprF gene in vivo and in vitro  A. Introduction The study of outer membrane protein structure and function has been greatly facilitated by the use of strains deficient or altered in these molecules. Traditionally, these mutants have been isolated by selecting for spontaneous phage or antibiotic resistance (Bassford, et al, 1977; Curtis, et al, 1985; Harder, et al, 1981; Henning, et al, 1978) or following chemical mutagenesis (Hrebenda & Heleszko, 1985; Manoil & Rosenbusch, 1982; Nicas & Hancock, 1983). The inherent difficulty in working with these mutants is that the exact nature of the mutation is unknown (ie., point mutation, insertion, deletion, inversion, duplication) and extensive mapping is required to determine if the lesion is in a structural or regulatory gene. More recently, transposon mutants have been used to study the structure and function of E. coli porins (Benson & Decloux, 1985; Jaffe, et al, 1982; Manoil, 1983; Misra & Benson, 1988) and the function of the phosphate induced protein P of P. aeruginosa (Poole & Hancock, 1986a). While Poole and Hancock successfully isolated P. aeruginosa oprF mutants, they observed that transposon mutagenesis in P. aeruginosa is seriously hampered by the tendency of the transposon delivery suicide plasmids to form stable whole plasmid insertion cointegrates. Strains of P. aeruginosa lacking protein F have been isolated following chemical mutagenesis by two groups (Gotoh, et al, 1987; Nicas & Hancock, 1983). In both studies, difficulties were encountered in obtaining protein F-deficient mutants. Only 0.2 - 0.6% of colonies surviving the mutagenesis treatment were protein F-deficient and the frequency of reversion was very high among these strains (Nicas & Hancock, 1983). One of these mutants was characterized with respect to its permeability to nitrocefin, a chromogenic P-lactam which, if able to penetrate the outer membrane, is hydrolysed by periplasmic P-lactamases and changes color (Zimmerman & Rosselet, 1977). Compared with the protein F-sufficient parent, the mutant  41  was approximately six-fold less permeable to nitrocefin (Nicas & Hancock, 1983). These authors concluded, based on these data, that protein F functions as a pore-forming protein in vivo. These results of Nicas and Hancock could not be confirmed in this study as their protein F-deficient strain had completely reverted to a protein F-sufficient phenotype. Later attempts to re-isolate P. aeruginosa oprF mutants by using specific antibody screening of transposon insertion mutants, following mutagenesis with the Tn501 -containing suicide vector pMTlOOO, met with failure (L. Chan & R.E.W. Hancock, personal communication). Because of the difficulties encountered by these previous investigators in isolating P. aeruginosa oprF strains and the instability of the original mutants, it was decided to construct more stable mutants. This was accomplished by genetic manipulation of the cloned gene in E. coli followed by gene replacement in P. aeruginosa, and is described in the following section.  B. Transposon mutagenesis of the oprF gene in E. coli (pWW13) and gene replacement in P. aeruginosa Using the technique described previously in Materials and Methods, insertion mutants of pWW13 containing Tn2, Tn5, Tn502 and Tn7 were isolated. E. coli (pWW13::Tn) colonies were screened for their loss of reactivity with protein F-specific monoclonal antibody indicating that the transposon had inserted in the oprF gene. Plasmids which were oprF::Tnl, oprF::Tn5 and oprF::Tn7, as ascertained by their loss of antigenicity, were isolated. Restriction enzyme analysis confirmed that these isolates had transposon inserts which mapped in the oprF gene. Interestingly, no oprF::Tn501 strains were isolated even though over three thousand E. coli (pWW13::Tn502) colonies were screened for loss of ability to express protein F. Presumably the oprF gene lacks a suitable site for Tn502 insertion as previous investigators were unable to isolate oprF::501 mutants in P. aeruginosa (see Results Chapter 2A). The pWW13 oprF::Tn plasmids were conjugated into P. aeruginosa. In order to facilitate homologous  n  V  P  M S  L  vfvvvvv^wvvvu  •  •  P 1  2-4  kb  opr F  Figure 8. Restriction map of the oprF gene showing the locations of the insertion sites of T n l and Q in the oprF insertion mutants. The Tnl insertion sites are indicated by triangles; the CI fragment was inserted into the Smal (M) site as shown. Strain H608 has Tnl inserted in the position indicated by the left triangle. P=Pstl; S=SalI.  43  recombination with the chromosome, lysates of phage F116L were made on the P. aeruginosa (pWW13 oprF::Tn) transconjugants and used to transfect P. aeruginosa PAOl. Transductants were selected which had lost the antibiotic marker of pWW13, had retained the transposon marker and did not react with the protein F-specific antisera, indicating that gene replacement by homologous recombination had occurred. Gene replacement was achieved only with the pWW13 oprF::Tnl plasmids. Two strains of P. aeruginosa oprF::Tnl were constructed. One of these contains Tnl inserted about two hundred base pairs into the coding sequence of the oprF structural gene; the other Tnl inserted three to four hundred base pairs further downstream (Fig. 8). The strain with the upstream Tnl insertion, designated H608, was selected for further characterization.  C.  Q mutagenesis of the oprF gene Despite attempts to isolate oprF transposon mutants with markers for  non-fi-lactam antibiotics, only Tnl insertion mutants were recovered. T n l codes for a P-lactamase, therefore functional studies of the role of protein F in P-lactam uptake using H608 were not feasible. Aminoglycoside antibiotics enter bacteria by the self-promoted uptake pathway (Moore, et al, 1987), a non-porin route; therefore, it was desirable to construct insertion mutants using aminoglycoside resistance markers like kanamycin or streptomycin. In vivo mutant construction, using Tn5 or Tn7, which code for kanamycin and streptomycin resistance, respectively, was unsuccessful (see Results Chapter 2B), so the commercially available D N A fragment Q was used to construct an oprF insertion mutant in vitro. The Q fragment encodes resistance to streptomycin and spectinomycin and contains three transcriptional stop signals at both ends. CI was ligated into the Smal site of oprF (Fig. 8) and then a fragment containing oprF::Cl was ligated into a gene replacement vector, pRZ102, by the procedure outlined in Figure 9. The final construct, pWW2500, was conjugated into P. aeruginosa.  Transconjugants  were screened for gain or loss of the appropriate markers and loss of protein F antigens. Mutants derived from homologous recombination and  44  Figure 9. Strategy for the construction of pWW2500 for Q mutagenesis of the oprF gene. A 1.4 kb Pstl-Sall fragment containing the amino-terminal half of oprF (pWW2300) was insertionally mutagenized in vitro by the insertion of the Q cartridge into the Smal site. The 3.5 kb Pvull fragment of this plasmid, pWW2300Q, was purified and ligated into the blunt-ended Hindm site of pRK404 to make pWW2400. pWW2400 was digested with Sail and the 3.4 kb Sail fragment was ligated into the Sail site of pRZ102 to make pWW2500. Symbols: ZZZI, oprF DNA; • • , Q. DNA; EI3, Tn5 DNA. The restriction sites indicated by single letter codes are as follows: H, Hindm.; M, Smal; P, PstI; S, Sail; V, Pvull.  Other phenotypes indicated:  Kn, kanamycin resistance; Tc, tetracycline resistance; Ap, ampicillin resistance; mob, mobilization site for conjugation.  45  replacement of the oprF gene with oprFv.Q, were obtained in a variety ofP. aeruginosa strains including mutants in other outer membrane proteins at a frequency of 60-100% per streptomycin resistant transconjugate. One of the P. aeruginosa P A O l oprF::Q, isolates was arbitrarily selected for further characterization and was designated H636.  D.  Characterization of the P. aeruginosa oprF mutants Cell envelopes were isolated from both the P. aeruginosa protein F-  deficient Tnl and Q. mutants. SDS-PAGE profiles of cell envelope preparations from the protein F-deficient mutants and their protein Fsufficient parents in a variety of genetic backgrounds are shown in Figure 10. In each mutant, the cell envelope profile was almost identical to that of its parent, except for the conspicuous absence of protein F. There was no consistent substantial increase in the size or relative amount of any other protein indicating that the cells did not compensate for the loss of protein F by substituting another outer membrane protein (Fig. 10, lanes 1, 2 & 3). For the strains in Figure 10, lanes 4 through 9, the cells were grown in PP2 broth for the parents or PP2 broth supplemented with 200 mM  NaCl for the mutants  (lanes 5, 7 & 9). In the cell envelope profiles of the mutants there was a protein of approximately 20,000 D not observed in the profiles of the parents (lanes 4, 6 & 8). The induction of a protein of similar size was also observed in outer membrane preparations of protein F-sufficient P. aeruginosa P A O l grown in the presence of 50 mM  NaCl or greater (Lydia Chan, unpublished  data) and was, therefore, thought to be a response to growth conditions, not protein F deficiency, in these strains. Accordingly, the cell envelopes of H608 and H636 (Fig. 10, lanes 2 & 3) were isolated from strains grown in MuellerHinton broth without added salt in which media they did not contain the 20,000 D protein. There was no interaction of proteins in the cell envelopes of the P. aeruginosa oprF mutants with protein F-specific monoclonal antibodies. The LPS of the strains shown in Figure 10 was examined by silverstained SDS-PAGE of whole cell preparations treated with lysozyme and  Figure 10. SDS-PAGE profiles of cell envelope proteins of the protein F deficient P. aeruginosa strains and their protein F sufficient parents. Lane 1, HI03 (PAOl, F+); lane 2, H608 (PAOl oprFv.Tnl); lane 3, H636 (PAOl oprF::Q); lane 4, H376 (PAO1840, F+); lane 5, H637 (PAO1840 oprFv.Q); lane 6, H457 (PA022 (R68.45), F+); lane 7, H670 (PA022 (R68.45) oprFv.Cl); lane 8, H576 (PAO oprP::Tn501, F+); lane 9, H638 (PAO oprP::Tn501 oprFv.Q). The position of protein F is indicated on the left. The unmarked arrow indicates the salt-induced protein of approximately 20,000 D present in lanes 5, 7 and 9 (see text). Molecular weights in kD are shown on the right.  Figure 11. SDS-PAGE profile of silver-stained lipopolysaccharide (LPS) from the protein F-deficient P. aeruginosa mutants and their protein F-sufficient parents. Lane 1, H103; lane 2, H608 (H103 oprF:;Tnl); lane 3, H636 (H103 oprFr.Q); lane 4, H376; lane 5, H637 (H376 oprF::Q); lane 6, H457; lane 7, H670 (H457 oprF::Q); lane 8, H576 (oprP::Tn502); lane 9, H638 (H576 oprP::Tn501 oprFwQ).  48  proteinase K (Fig. 11). No major changes were seen between the parents and the mutants. For some strains, there appeared to be a slight loss in higher molecular weight bands in the mutants (Fig. 11, lanes 1-5, 8 & 9). A similar phenomenon was observed for the absA antibiotic-supersusceptible mutants of P. aeruginosa (Angus et ah, 1987). Unfortunately, this technique would not detect alterations in saccharide composition but did indicate that the mutants did not lose their O-side chains or acquire any gross structural abnormalities in their LPS. This observation was confirmed by phage susceptibility testing. Strains H608 (oprF::Tnl) and H636 (oprFv.Q) had phage sensitivity profiles identical to those of HI 03, their protein F-sufficient parent, using thirty phage specific for LPS, non-specified outer membrane protein or pilin receptors. To ensure that the expected DNA  rearrangements had occurred during  gene replacement (ie., homologous recombination at the oprF gene), chromosomal DNA  was isolated from the mutant strains and analyzed by  Southern blot hybridization to a radiolabeled probe of the oprF gene. The hybridization patterns of DNA  from the oprFv.Q. mutants H636 and H637 and  their parent strain HI 03 are shown in Figure 12. In the PstI digest of HI 03, the radiolabeled oprF probe hybridized to a fragment about 2.4 kb; in the PstI digests of H636 and H637, the oprF probe hybridized to a 4.4 kb fragment. This was the expected result if the oprF-.-.Q. sequences from pWW2500 had homologously recombined by a double crossover event into the chromosome, as oprF is contained on a 2.4 kb Psrl fragment and was mutagenized by insertion of the 2.0 kb Smal Q fragment, making a 4.4 kb Psfl fragment in the chromosome of the mutants. Accordingly, Smal digests of the parents and the mutants had identical hybridization patterns, as expected, because the Q fragment would be excised by Smal. Attempts were made to conjugate pWW13 into the protein F-deficient strain H608 to show that a plasmid containing oprF could restore the protein F phenotype in the mutant strain. Six separate experiments were set up using the oprFv.Q strain H608 as a conjugation recipient and in each experiment no transconjugates were recovered. Control experiments in which pLARFl was conjugated into protein F-sufficient recipients worked normally. Possible explanations for these observations are discussed below (see Discussion).  H103  H636  H637  PS  PS  PS  Figure 12. Autoradiograph of a Southern blot hybridization of a radiolabelled oprF probe to chromosomal DNA digests of P. aeruginosa H103, H636 and H637. The DNA was digested with Psfl (P) or Smal (S). The oprF probe was the 2.4 kb PstI fragment of pWW2200. The size of molecular weight markers (1.0 kb ladder) is shown on the left in kilobase pairs (kb).  50  E. Summary P. aeruginosa oprF insertion mutants were constructed in vivo by T n l mutagenesis of the cloned gene in E. coli and in vitro by insertion of Q, a streptomycin-resistance encoding DNA fragment, into the cloned gene, followed by transfer of the mutated protein F gene back into P. aeruginosa. Homologous recombination between the mutant genes, oprF: :Tn2 and oprF::Q, resulted after bacteriophage or plasmid D N A containing these mutant genes was introduced into P. aeruginosa strains. oprFv.Q, mutants were constructed in vivo in P. aeruginosa strains already carrying various mutations. SDS-PAGE analysis of the oprF mutants demonstrated that these strains had lost protein F. The LPS of the protein F-deficient mutants was not grossly altered from that of the protein F-sufficient parents as determined by SDS-PAGE of the LPS and phage susceptibility testing using LPS specific phage. Southern blot analysis of chromosomal DNA isolated from the mutants, and probed with the oprF gene, showed that the expected double crossover event had occurred and that there was a single chromosomal oprF gene with a Tnl or Q insert. Attempts to conjugate the cloned oprF gene back into protein F-deficient P. aeruginosa strains were unsuccessful.  51  Chapter 3. Functional studies of purified protein F and the P. aeruginosa oprF strains  A. Introduction Functional characterization of outer membranes and their proteins has been frustrated by the lack of assay systems and the inherent difficulties of studying a complex integral structure in a living organism which can adjust to, adapt to and compensate for both internal and external changes. The experimental techniques employed to study outer membrane proteins can be divided into two groups: experiments designed to look at function in the intact cell, usually a mutant, and those designed to study purified proteins. In general, studies of purified non-porin membrane proteins have been confined to defining structural rather than functional parameters. Purified porin proteins can be studied in model membrane systems. Four basic systems exist to show pore-forming ability in vitro: 1) radioisotope efflux using proteoliposomes; 2) substrate uptake into enzyme-containing proteoliposomes; 3) liposome swelling and 4) black lipid bilayer conductance assays. These systems plus the biophysical equations and mathematical assumptions used to analyze the data have been described in detail by Hancock (1987a). In vivo assays of permeability include MIC determinations to assess differential antibiotic uptake between porin mutants and their parent strains (Nikaido & Vaara, 1985), permeation or efflux of radiolabeled sugars by plasmolyzed whole cells (Caulcott, et al, 1984; Yoneyama and Nakae, 1986) and spectrophotometric measurement of the rate of hydrolysis of a chromogenic hydrophilic molecule cleaved by a periplasmic enzyme (Zimmerman & Rosselet, 1977). Several of these techniques have been criticized for design flaws (Hancock, 1987; Nikaido & Vaara, 1985). The most widely accepted technique for assessing in vivo permeability is that of Zimmerman and Rosselet (Nikaido & Vaara, 1985),which measures the penetration and hydrolysis of a chromogenic p-lactam antibiotic. Definition of the function of the non-porin proteins, lipoproteins and heat-modifiable proteins, has traditionally involved comparisons of strains  52  deficient in these proteins with their protein-sufficient parents. Such studies have included determination of: changes in growth rate and substrate requirements (Manning, et al, 1977; Sonntag, et al, 1978); increases in MIC (Harder, et al, 1981; Hrebenda & Heleszko, 1985; Komatsu, et al, 1981; Nikaido, et al, 1983; Yamaguchi, et al, 1985); sensitivity to detergents, chelating agents, acridine dyes and hydrophobic agents (Nikaido & Vaara, 1985; Nurminen, et al, 1976; Schweizer, et al, 1976); changes in sensitivity to bacteriophage and colicins (Henning, et al, 1978; Manoil, 1983); loss of ability to function in conjugation (Achtman, et al, 1978; Manoil & Rosenbusch, 1982) and electron microscopy to look for changes in cell shape (Sonntag, et al, 1978), blebbing (Yem & Wu, 1978) or pilin production (Manning, et al, 1977). Purified protein F and the protein F-deficient strains were characterized using some of these techniques in an attempt to clarify the function of protein F in the outer membrane of P. aeruginosa and to resolve some of the controversy concerning the size of protein F channels.  B. Growth rate determinations Preliminary studies to determine growth rates of P. aeruginosa HI03 and its oprF derivatives H608 oprF::Tn2 and H636 oprF::Q were performed in a very rich medium, Mueller-Hinton broth (MHB), and in a less rich, nondefined medium, PP2. Similar doubling times were observed for HI 03 (51 min), H608 (54 min) and H636 (58 min) on MHB. However, growth of H608 in PP2 was significantly slower than that of HI 03, and H636 would not grow on PP2 (Table IV). The growth rate of the previously isolated protein Fdeficient chemical mutant had been substantially enhanced by adding 50 mM NaCl to PP2 (Nicas & Hancock, 1983), therefore, it was not surprising that when NaCl was added to the PP2, the doubling times of H608 and H636 approximated those of HI03 (Table IV). The effects of other salts and sugars, observed to stabilize growth of outer membrane protein mutants in E. coli (Manning, et al, 1977; Sonntag, et al, 1978), was determined for the exponential growth of H608 and  Table IV: Doubling times for P. aeruginosa H103 and its derivatives H608 oprF::Tnl and H636 oprF::Q in broth culture during the logarithmic phase of growth at 37<>C Doubling Time (in minutes) Growth Medium  H103  PP2a PP2 + KC1  2  62  89  94  100 mM  58  65  75  mM  55  59  67  50 mM  47  80  142  100 mM  49  61  126  mM  47  62  66  15 mM  54  99  30 mM  52  94  61  68  267  200  mM  70  76  107  400  mM  70  76  101  100  mM  66  80  103  200  mM  72  66  73  400  mM  75  92  91  58  60  56  51  54  58  PP2+succinate 200 mM MHBb a  oprF::Q,  50 mM  PP2+glucose 100 mM  PP2+sucrose  oprF::Tnl  128  200 PP2+MgCl  H636  55  200 PP2+NaCl  H608  Proteose peptone #2 broth (Difco) at 1% w/v;  ^Mueller Hinton broth (Difco)  denotes no observed doubling time  54  H636 (Table IV). Comparable doubling times were observed for H103 and its oprF mutant derivatives H608 and H636 on MHB, PP2 plus 200 mM succinate, PP2 plus 200 mM KCl, PP2 plus 200 mM NaCl and PP2 plus 200mM sucrose. Excluding these exceptions, the growth rate of H636 was significantlyslower than that of HI 03 or H608. This was particularly surprising for cultures grown in PP2 plus lOOmM glucose (doubling times of HI 03, H608 and H636 were 61.0, 68.3 and 266.6 min, respectively) because the glucose-inducible porin protein D l should facilitate comparable levels of uptake in all three strains. In contrast to H608 oprF::Tnl which grew slowly, H636 oprFv.Q. did not grow at all on PP2 or PP2 plus MgCl (Table IV) The large differences 2  between doubling times of H608 and H636 for most of the substrates examined are somewhat puzzling as both are insertion mutants within a two hundred base pair region of the amino terminal coding region of the oprF gene (Fig. 8) and neither expressed protein F, as determined by Coomassie blue staining of cell envelopes fractionated by SDS-PAGE (Fig. 10). It is possible that the two different mutations induce different adaptative responses in P. aeruginosa, Autolysis of cells of H608 occurred if the cultures were grown longer than 18-20 h or if just left on the bench overnight. H608 and H636 also died when repeatedly frozen (-70C) and thawed, even in the presence of glycerol or dimethyl sulfoxide. Similar observations have been made in mutants of E. coli lacking OmpF, OmpC and OmpA (Schweizer, et al, 1976). The minimum NaCl concentration required to support growth of P. aeruginosa oprFv.Q. strain H636 was determined. A n overnight culture grown in PP2 with 200 mM NaCl was diluted into fresh PP2 broth containing concentrations of NaCl from 0 to 50 mM. Addition of 50 mM NaCl to cultures of HI03 slightly enhanced the growth rate (Fig. 13, panel A). Addition of greater than 10 mM NaCl to PP2 broth was required for a culture of H636 to grow exponentially (Fig. 13, panel B). The lag phase of H636 cultures supplemented with 20, 30 or 40 mM NaCl was twice that of a culture grown on 50 mM NaCl. Even so, the growth rate of H636 grown on PP2 with 50 mM NaCl was significantly slower than that of HI 03, its protein Fsufficient parent, grown on PP2 with or without salt (Fig. 13, panels A & B; Table IV) or of H636 grown in PP2 plus 200 mM NaCl (Table IV). The flask of  Figure 13. Effect of NaCl on the logarithmic growth rate of H103 and H636 grown in PP2 broth. Cultures of P. aeruginosa HI 03 (panel A) and H636 oprFv.Q. (panel B) were set up in PP2 broth supplemented with NaCl at the concentrations shown. For H103 grown with 10, 20,30 and 40 mM NaCl, the growth curves were between those of 0 mM  and 50 mM  omitted from this Figure for clarity (panel A).  and have been  56 PP2 with no salt inoculated with H636 maintained an almost constant Klett reading for the 8 h of the experiment (Fig. 13, panel B) indicating that the cells did not lyse. This was confirmed by phase contrast light microscopy. Whole cells were observed in the no salt H636 flask even though no detectable growth had occurred. Gross abnormalities such as swollen or clumped cells were not observed and the cells remained motile. C. Miscellaneous uptake studies 1. N-phenylnaphthylamine uptake The observation that the loss of protein F in the oprF mutants was not compensated for by an increase in an existing protein or the synthesis of a new outer membrane protein seemed to indicate that a major change in the cell surface had occurred which might alter the hydrophobicity of the membrane. Hydrophobic permeability can be determined using the fluorescent probe N-phenylnaphthylamine (NPN) (Angus, et al, 1987; Loh, et al, 1984). The responses of H103, H608 and H636 to N P N are shown in Table V. The oprF mutants H608 and H636 took up three to five-fold more N P N than did their protein F-sufficient parent H103 (Table V). This was 3050% of the maximum N P N uptake possible by the cells observed after treatment with EDTA had disrupted their membranes exposing the maximal number of N P N binding sites. Similar increases in hydrophobic permeability have been observed in strains of E. coli missing OmpC (Siden & Boman, 1983) and omp A and ompB mutants (Schweizer, et al, 1976). P. aeruginosa strain H608 oprF::Tnl strain H636 oprF::Q.  was permeable to 60% more N P N than  This observation is somewhat surprising, based on the  similarities between these mutants genetically (Fig. 8), and may reflect a difference in the cell surfaces of these mutants not detected during characterization studies of the cell envelope proteins (Fig. 10) and LPS (Fig. 11). Such a difference may possibly account for some of the results observed in the growth studies.  57  Table V: Uptake of NPN by strain H103 and its protein F-deficient derivatives H608 oprFnTnl and H636 oprFv.Q,  Relative fluorescence (% of maximum) untreated cells  EDTA-treated cells  rI103 oprF+  0.95 ± 0.00 (10.5%)  9.05 + 0.05 (100%)  H608 oprFv.Tnl  5.43 ± 0.46 (50.9%)  10.67 ± 0.17 (100%)  H636 oprFv.Q  2.94 ± 0.06 (28.8%)  10.20 ± 0.69 (100%)  58  2. Phosphate uptake During the course of experiments to study phosphate uptake and regulation, a student in our laboratory observed that in P. aeruginosa H608 oprF: :Tnl the porin induced under low levels of inorganic phosphate, protein P, was induced at phosphate levels five times higher than those of its parental protein F-sufficient strain H103 (Michael Sauve, personal communication). This preliminary observation was confirmed (Fig. 14). In H608, protein P was induced when concentrations of 0.2, 0.6 and 1.0 mM inorganic phosphate were added to the phosphate deficient minimal medium (Fig. 14, lanes 2 through 6). In contrast, protein P is induced in H103 only when the external phosphate concentration is 0.2 mM  or lower (Hancock, et  al., 1982). These observations suggested that, under phosphate sufficient conditions, protein F may function as a pore for inorganic phosphate. Attempts were made to further investigate this phenomenon by measuring the kinetics of uptake of radioactive phosphate by the method of Poole and Hancock (1984) and the levels of induction of periplasmic alkaline phosphatase, a marker enzyme for the induction of the phosphate regulated operon which includes protein P (Poole & Hancock, 1984). The results of these experiments were negative. Significant differences were not observed between protein F-sufficient parent and protein F-deficient mutant strains judged by these experiments, a result which may reflect the complexities of regulation of the high and low affinity phosphate uptake systems which, conceivably, could function independently of phosphate permeability through the outer membrane. It remains unclear why protein P was induced at higher concentrations in oprF strains than in wildtype cells. D. Antibiotic uptake analysis: MICs Agar dilution MICs, to compare antibiotic sensitivities of the oprF mutants H608 and H636 to their parent HI 03, were set up in Mueller-Hinton agar because similar doubling times were observed for these strains in this medium (Table IV). Strain H608 had T n l inserted in the oprF gene. This transposon encoded a (3-lactamase, therefore, it was not possible to assess  59  Figure 14. SDS-PAGE profiles of cell envelope preparations of P. aeruginosa H608 oprF::Tn2 grown in minimal medium with increasing amounts of inorganic phosphate. Lane 1, outer membranes of P. aeruginosa HI03 grown on 0.2 mM  phosphate to show induction of protein P in a protein F  sufficient strain. Lanes 2 through 6 are cell envelope preparations of H608 grown in phosphate deficient medium supplemented with inorganic phosphate as follows: lane 2, 0.2 mM lane 4,1.0 mM  phosphate; lane 3, 0.6 mM  phosphate; lane 5, 3.0 mM  phosphate;  phosphate; lane 6,10.0 mM  phosphate. The positions of proteins F and P are indicated. The size of molecular weight markers is shown in kD on the right.  60  MICs for antibiotics susceptible to hydrolysis with this enzyme. MICs for p-lactam antibiotics resistant to P-lactamase hydrolysis could be determined using strain H103 (RP1) as a control. Plasmid RP1 contained Tnl and expressed similar levels of P-lactamase compared to H608, as assessed by the hydrolysis of nitrocefin, a chromogenic P-lactam (Table VI). Small increases in MIC were observed for H608 compared with HI 03 (RP1) on the p-lactamase stable P-lactams aztreonam and ceftazidime (Table VI). No significant differences in resistance were observed for the quinolone antibiotic norfloxacin or the aminoglycosides tobramycin and gentamicin, compounds thought to use non-porin routes of entry into P. aeruginosa (Hancock, et al, 1981; Moore, et al, 1987) or a combination of porin and non-porin pathways in E. coli (Chapman & Georgopapadakou, 1988). Similar patterns of resistance were observed when the MICs for HI 03 and the protein F-deficient-Q-mutagenized strain H636 were compared (Table VII). Small increases in resistance, 2.5 to 3.2-fold, were observed for H636 on the P-lactams cefpirome, cefotaxime and cefepime. Marginal, less than twofold, increases were observed in the MICs of other P-lactams with the exception of imipenem (Table VII), a carbapenem thought to permeate P. aeruginosa by a non-protein F pathway (Buscher, et al, 1987; Quinn, et al, 1986). For some of the P—lactams in Tables VI and VII, the increases in MIC for the oprF strains compared with their parent strains were statistically significant (P<0.05 by Fisher's exact test) but in each case the difference in MIC was three-fold or less. As observed for the oprF::Tnl mutant, resistance to the quinolone norfloxacin was unaffected by the oprFv.Q. mutation.  E. Black lipid bilayer studies Protein F was purified from P. aeruginosa HI03 and from E. coli JF733 (pWW2200) by selective solubilization, followed by SDS-PAGE and passive elution. SDS-PAGE analysis of the purified proteins revealed that these proteins were essentially free from outer membrane contaminants (Fig. 15). The faint higher molecular weight band in the purified protein F  Table VI: Geometric mean MIC determinations for the oprF::Tnl mutant H608 and the wild-type strain H103 (RPl::Tnl)  MIC (|ig/mL)a of Strain  (3-lactamase  Azfd  Cef  Nor  Tob  Gen  1.7  1.0  0.5  2.0  2.0  4.0b  3.4b  0.8  0.7  2.8  levelc  H103 (RPl::Tnl) H608 (oprF: :Tnl)  2.2 1.5  aGeometric means of four or more independent results. bSignificantly different from H103 (RPl::Tnl). Using Fisher's exact test, (P=0.05). For all other antibiotics tested, the H608 MICs were not significantly different from those of the control strain. Cnmol of nitrocefin hydrolysed/min/mg of cells dAzt=Aztreonam, Cef=Ceftazidime, Nor=Norfloxacin, Tob=Tobramycin and Gen=Gentamicin  62  Table VII: Mean MICs for strain H103 and its protein F-deficient Q insertion mutant H636  Geometric mean MIC (|ig/mL)a for:  MIC ratios for porin-deficient and porin-sufficient strains  Antibiotic H103 oprF+  H636 oprF::Q  P. aeruginosa^  E. colic  Cefpirome  1.0  3.0d  3.0  -  Cefotaxime  4.0  12.7  3.2  2  Cefepime  0.8  2.0  2.5  -  Aztreonam  1.4  2.4  1.7  4  Carbenicillin  16  28d  1.8  8  Ceftazidime  1.0  1.6  1.6  2  Piperacillin  1.7  2.3  1.4  2  Cefpir amide  2.0  2.5  1.3  -  Cefsulodin  1.6  2.0  1.3  -  Imipenem  3.2  2.5  0.8  -  Norfloxacin  3.0  2.6  0.9  -  a  Geometric means of three to five determinations  bRatio of strains H636 and H103 c  Data from (Curtis, et al., 1985, Jaffe, et al., 1982 Jorgensen, et al., 1979 and Komatsu, et al., 1981) -, No information available.  dp=0.05 by Fisher's exact test for MICs of H636 compared with H103.  63  preparations (Fig. 15, lanes 2 & 4) was the heat-modified form of protein F (Hancock & Carey, 1979); the lowest molecular weight band was lysozyme, added during the final stages of purification to free the protein from peptidoglycan. Lysozyme did not interfere with and indeed assisted the analysis of these proteins in artificial membranes (Parr, T., and R.E.W. Hancock, unpublished data). Addition of purified protein F to the 1 M K C l solution bathing the black lipid bilayer membrane resulted in stepwise increases in electrical conductance which were observed on the oscilloscope of the apparatus and recorded on the chart recorder. Discreet increases in the conductance represented the incorporation of single channel-forming units of protein F into the membrane. By tabulating the number and size of the discreet conductance increases from the chart recorder tracing, the probability histograms and average single channel conductance for a particular protein can be calculated. These data, for protein F purified from P. aeruginosa HI03 and E. coli JF733 (pWW2200), are shown in Figure 16. For both preparations, predominantly small channels were observed at similar frequencies. The average single channel conductance of native protein F was determined to be 0.41 nS and that of the cloned protein F was 0.38 nS. In the native protein F preparation, some larger channels were observed (5.2% of all channels observed were larger than 1.2 nS).  In one experiment 15% of the channels  observed were between 1.2 and 4.0 nS, and of these, slightly more than half (52%), were between 1.2 and 2.0 nS. For the cloned protein F preparation, the largest channels observed had a conductance of 3.4 nS. Channels larger than 1.2 nS were observed at a frequency of 6.8% overall, a result similar to that observed for the native protein F. Sixty-three percent of the larger channels observed were between 1.2 and 2.0 nS. Nonetheless, for protein F isolated from its native P. aeruginosa background, and also from the cloned oprF gene in E. coli, the vast majority of channels observed were less than 0.6 nS. These observations are in contrast to those published earlier, (Benz & Hancock, 1981), in which the average channel conductance of purified protein F was reported to be 5 nS.  Figure 15. SDS-PAGE profiles of the purified protein F used for black lipid bilayer analysis. Lane 1, cell envelopes of P. aeruginosa H103; lane 2, protein F purified from P. aeruginosa HI03; lane 3, cell envelopes of E. coli JF733 (pWW2200); lane 4, protein F purified from E. coli JF733 (pWW2200); lane 5, cell envelopes of P. aeruginosa PCC23; lane 6, protein F purified from P. aeruginosa PCC23. The protein F band is marked. The lowest molecular weight band in lane 4 is lysozyme added during purification to free the protein from peptidoglycan. The molecular weights of the protein markers are indicated on the right in kD.  03  n=750 x=0-41 nS  02 0-1 0 c  3  03  ^  0-2  B n=1268 x=0-38 nS  01 > o  0-1  tr 0 0-2  04  0-6  a8  10  1-2  0 3 n=343 02  x=077 nS  0-1  W Single  0-8  1-2  Channel  1-6  20  24  Conductance (nS)  Figure 16. Histogram of the conductance increases observed for purified protein F in black lipid bilayer membranes. Nanogram per millilitre concentrations of the purified protein were used. The number of conductance increase events observed, n, is shown for protein F purified from P. aeruginosa P A O l (panel A), from E. coli (pWW2200) (panelB) and from P. aeruginosa PCC23 (panel C). The average single channel conductance, x, is given in nanoSiemens (nS) a unit of conductance equal to lO"^ ohms" ; the standard deviations are 0.16 (panel A), 0.24 (panel B), and 1  0.52 (panel C). Measurements were made in 1M KC1 with an applied voltage of 50 mV. It should be noted that the scale of the x-axis in panel C is half that of the x-axes in panels A and B.  66  Protein F from P. aeruginosa PCC23, the oprF* mutant, was purified and examined in black lipid bilayer (Fig. 15 & 16). No difference in the mobility of protein F, purified from PCC23 was observed (Fig. 15, lanes 5 & 6) as compared with either native or cloned protein F (Fig. 15, lanes 1-4). This is in contrast to the observations of Godfrey and Bryan (1987). In a preliminary black lipid bilayer experiment, the average single channel conductance of protein F from PCC23 was observed to be 0.77 nS (Fig. 17, panel C), almost twice that of native or cloned protein F. It should be noted, nonetheless, that the most frequently observed channel in PCC23 was 0.4 nS.  F. Summary The protein F-deficient P. aeruginosa mutants H608 oprFv.Tnl and H636 oprFv.Q. were analyzed for functional differences compared with their protein F-sufficient parent HI 03 with respect to growth rates, uptake of the hydrophobic probe NPN, inorganic phosphate and antibiotic resistance. Protein F, purified from P. aeruginosa and E. coli clones harboring plasmids with oprF, was analyzed for pore-forming ability in black lipid bilayers. Major differences in growth rate, assessed by doubling time, were observed between HI 03, H608 and H636 on the non-defined medium PP2 (Table IV). The doubling time of H608 was 2.5 times longer than that of HI 03 on PP2 and H636 would not grow at all during the time course of the experiment. These effects could be reversed (for H636) or modulated if the PP2 was supplemented with NaCl, KCl, glucose, sucrose or succinate. MgCl2 had little effect on growth rate. Similar doubling times were observed for all three strains in  MHB.  The uptake of a hydrophobic fluorescent probe, NPN, was determined. Intact cells of the oprF mutants accumulated three to five times more NPN than did HI 03 indicating a major alteration in outer membrane hydrophobicity and/or composition (Table V). Loss of protein F from the outer membrane of P. aeruginosa was correlated with increase in the concentration of inorganic phosphate which would cause induction of the phosphate specific porin, protein P (Fig. 14). This observation implies a role  67  for protein F in phosphate uptake under phosphate sufficient conditions. Analysis of antibiotic resistance by MIC assays showed small to marginal increases in resistance for the protein F-deficient mutants compared to their protein F-sufficient parent for ten P-lactams (Tables VI and VII). These increases were never more than three-fold. Analysis of purified protein F from both P. aeruginosa and E. coli (pWW2200) in the black lipid bilayer apparatus showed similar results: the average single channel conductances observed were 0.41 nS and 0.38 nS, respectively (Fig. 16). In both cases the most frequently observed channels had conductances of less than 0.4 nS. Approximately 5% of the channels observed were larger than 1.2 nS but no channels larger than 4 nS were observed. Protein F, purified from the oprF* mutant P. aeruginosa PCC23, had an average single channel conductance of 0.77 nS, but the most frequently observed channel was 0.4 nS (Fig. 16).  68  DISCUSSION  The aims of this study were to clone the structural gene, oprF, encoding outer membrane protein F from P. aeruginosa and to construct stable oprF insertion mutants. It was hoped that characterization of protein F and of mutants lacking this protein would help to elucidate the role of protein F and promote an understanding of the mechanisms of hydrophilic antibiotic resistance in P. aeruginosa. The techniques for and results of experiments designed to clone oprF and to construct and characterize the P. aeruginosa oprF mutants have been presented in the preceeding pages. The following discussion will attempt to interpret these results and observations in light of existing information in the literature and to assess their potential clinical relevance. A. Protein F:  Cloning, mutagenesis and conservation  The cloning of the oprF gene into the cosmid vector pLAFRl and identification of clones using protein F-specific antibodies was straightforward, because protein F was expressed in E. coli under its own promoter. Difficulties were encountered when attempts were made to subclone oprF into high copy number vectors, a phenomenon observed by other investigators who cloned outer membrane proteins (Freudl & Cole, 1983; Carbonetti & Sparling, 1987; Stephens et al., 1985). During the subcloning of oprF, several plasmids were generated which encoded novel forms of protein F. While the exact nature of these plasmids is unknown, it is interesting to note the similarities between the truncated protein encoded by pWW4 and pWW5 and the cloned OmpA proteins of E. coli and S. marcescens. It has been observed that the carboxy-terminal coding region of omp A could be deleted and that regardless of how many basepairs had been deleted, up to a defined maximum, the gene product would be proteolytically degraded to approximately 24,000 Daltons before translocation to the outer membrane (Bremer et ah, 1982; Braun and Cole, 1984). A similar phenomenon was observed for the gene products  of pWW4 and pWW5 in E. coli (Fig. 2, lanes 4 and 5; Woodruff et al, 1986). The reasons for this apparent proteolytic degradation or the implications of these observations with regard to assembly and translocation of membrane proteins and structural or functional domains of OmpA and protein F are unknown. Hybridization experiments using oprF sequences to probe chromosome D N A of P. aeruginosa and other Pseudomonas strains showed that restriction enzyme sites and fragment sizes, as well as sequences, of the protein F gene were tightly conserved both within a species and between species. Mutharia et al. (1985) found that protein Fspecific monoclonal antibodies only reacted with outer membranes of P. aeruginosa, P. putida and P. syringae.  In contrast, sufficient  conservation of oprF D N A sequences to allow hybridization under fairly stringent conditions, was observed for 52 clinical isolates and nine Pseudomonas strains and Azotobacter vinelandii from the P. fluorescens rRNA homology taxonomic group (De Vos and De Ley, 1983). Interestingly, Mutharia et al. did observe antigenic conservation of epitopes of the P. aeruginosa lipoprotein, H2, among all the strains of this group. Interspecies and intergenic conservation has been observed for outer membrane proteins of Enterobacteriaceae and Neisseria sp., particularly for the heat-modifiable proteins (Hofstra & Dankert, 1979; Hofstra & Dankert, 1980; Cole et al, 1982; Gotschlich et al, 1987b), and for protein P of P. aeruginosa (Poole and Hancock, 1986b). It should be noted that antigenic conservation or even peptide mapping data do not necessarily correlate with significant DNA-DNA hybridization (Lee et al, 1979; Beher et al, 1980; Mizuno et al, 1983). Characterization of chromosomal restriction fragments hybridizing to radiolabeled gene probes from the exotoxin A gene of P. aeruginosa showed that this genetic region had sufficient sequence heterogeneity that toxA gene probes could be used as epidemiological markers for typing clinical isolates (Ogle et al, 1987). Comparison of the lack of conservation of these P. aeruginosa virulence factors with the extreme conservation of protein F, also a virulence factor, raises interesting questions regarding the evolutionary origins of these virulence proteins and their relative  significance to the bacterial cell. To determine the function of a particular protein, it has been traditional to isolate mutant strains lacking that protein and to compare the mutant with its wildtype parent. This is particularly true for the study of antibiotic resistance in gram negative bacteria because decreased permeability of antibiotics was correlated with a decrease in amounts of outer membrane proteins in early studies of these proteins (Bavoil & Nikaido, 1977; Harder et al, 1981; Komatsu et al, 1981; Jaffe et al, 1982). Unfortunately, most of the studies of antibiotic resistance have been performed using spontaneous outer membrane protein mutants selected for their ability to grow on high levels of |3-lactams or following chemical mutagenesis. In such strains, multiple lesions may be present, or the mutation that causes the cell to alter its outer membrane proteins may have pleiotropic effects not immediately apparent or completely masked. Attempts were made to circumvent these problems, when studying the role of protein F in antibiotic resistance, by constructing insertion mutants of oprF in P. aeruginosa in vivo and in vitro.  A major disadvantage with working with outer membrane proteins is that these proteins represent a significant component of the outer membrane. Protein F constitutes about 15% of the outer membrane proteins and is present at an estimated 200, 000 copies per cell (Benz & Hancock, 1981; Angus et al, 1982; Mutharia & Hancock, 1983). Removal of such an abundant protein from the outer membrane of oprF mutants created a change in the cell surface. This conclusion is based on the observation of 3-to-5-fold increases in uptake of the hydrophobic probe, NPN, by these mutants compared with their protein F-sufficient parents. Because the outer membrane profiles of the oprF mutants were seemingly identical to those of their parents except for the conspicuous absence of protein F, the gaps left by the loss of protein F were likely filled in by lipids or lipid-like molecules. This explanation is supported by data from E. coli mutants lacking OmpF and OmpC or OmpF, OmpC and OmpA proteins which have been shown to contain increased amounts of phospholipids compared with their protein sufficient parents (van Alphen et al, 1977; Schweizer et al, 1976). For the triple mutant, the  phospholipid to protein ratio was shown to increase from 0.96 in the parent to 2.2 in the mutant (Schweizer et ah, 1976). The amounts of lipopolysaccharide and lipoprotein remained the same in both strains. Changes in outer membrane proteins may be accompanied by alterations in LPS. This phenomenon has been observed for S. typhimurium (Nikaido & Vaara, 1985), E. coli (Havekes et ah, 1976;  van  Alphen et ah, 1977; Hiruma et ah, 1984) and P. aeruginosa (Godfrey & Bryan, 1987). In each of these mutants, the P-lactam antibiotic resistance of the mutant compared with its parent was significantly increased, in some instances by as much as 300-fold. Hiruma et ah, (1984) concluded that LPS, via its polysaccharide chain region, constitutes a selective barrier against the permeation of P-lactams proportional to the hydrophobicity of the molecule. In contrast, Godfrey and Bryan (1987) showed that, although the level of the LPS O-side chain sugar fucosamine was increased in the protein F altered mutant PCC23, permeability of Plactams in liposomes containing this purified protein F* was not affected by the presence or absence of LPS. The LPS of the oprF mutants was not grossly altered as determined by silver staining of LPS gels (Fig. 11). It should be noted that this technique would not detect changes in the distribution of LPS or in sugar residues. Therefore, mutations within the structural gene of an outer membrane protein, such as oprF, may  have  undetected pleiotropic effects on other membrane components. Despite these inherent difficulties and potential aberrations in comparing protein-deficient mutants with protein-sufficient parent strains, with respect to antibiotic resistance, a contributing role for an outer membrane protein in antibiotic resistance may be determined by comparative MIC analysis. B. Protein F: Antibiotic resistance A role for protein F in the unusually high antibiotic resistance of P. aeruginosa was postulated on the basis of the observation of large, 5 nS, channels observed in black lipid bilayers reconstituted with preparations of outer membrane proteins or purified protein F (Benz & Hancock, 1981).  72  Such channels would easily allow antibiotics to permeate the outer membrane. However, these large channels represented only 1% of the protein F molecules, therefore despite the potential ability of antibiotics to go through these channels, Benz and Hancock proposed that the probability of such an event was low enough to account for the observed antibiotic resistance in this organism. The first evidence for a contribution of protein F to antibiotic resistance was the observation of significant decreases in the permeability of nitrocefin, a p-lactam, in a chemically mutagenized protein F-deficient P. aeruginosa strain (Nicas & Hancock, 1983). MIC determinations for clinically relevant antibiotics using this strain were not reported due to its high reversion frequency to protein F-sufficiency. More recent evidence has been reported by Godfrey and Bryan (1987) who isolated a chemically mutagenized pleiotropic mutant, PCC23, of P. aeruginosa, altered in protein F, which was resistant to more than 300 times the levels of some P-lactams than its parent strain. It should be noted that this mutant was selected based on its resistance to ten-fold higher levels of ticarcillin compared with its parent strain. The construction of stable insertion mutants in the structural gene for protein F described in this study afforded an opportunity to examine the role of protein F in antibiotic resistance. The mutant oprF P. aeruginosa strains are isogenic with their parents except for a single defined lesion (Fig. 12). It was therefore surprising, particularly in light of previous data, that small or marginal differences in antibiotic resistance (determined by MIC) were observed for both oprFv.Tnl and oprF-Q P. aeruginosa strains compared with their parents (Tables VI and VII). It was found that similar results had been obtained for spontaneous E. coli mutants lacking OmpF and, or OmpC (Table VII), particularly for more recently synthesized cephalosporins (Harder et ah, 1981; Curtis et ah, 1985; Yamaguchi et ah, 1985). These observations have caused several investigators to question the proposed role of porins as exclusive routes for P-lactam penetration through the outer membrane (Curtis et ah, 1985; Yamaguchi et ah, 1985; Nayler, 1987).  In discussing the MIC results presented in this study, for the P. aeruginosa oprF mutants, two issues must be addressed. Firstly, are porin proteins the major routes of uptake of hydrophilic antibiotics and secondly, is protein F significantly involved in P-lactam uptake in P. aeruginosa?  An increasingly large body of data is accumulating which suggests that hydrophilic antibiotic permeation into gram negative bacteria is a multifactorial phenomenon. A role for LPS alteration in P-lactam resistance has been demonstrated (Godfrey et al, 1984; Hiruma et al, 1984; Angus et al, 1987). Recently, Hancock and Woodruff (1988) proposed that the outer membrane acts in synergy with P-lactamases to resist the action of P-lactams. This proposal is supported by the early observations of Harder et al, (1981) that porin-deficient mutants selected on the basis of their resistance to P-lactams had MIC profiles 4-fold or greater than porin-deficient mutants selected on the basis of resistance to bacteriophages known to use porins as receptors. The observed increases in MICs for porin-deficient strains suggest that porins are involved in hydrophilic antibiotic permeability but it is doubtful they constitute the only means of entry for hydrophilic molecules into gram negative bacteria. The conclusions of Godfrey and Bryan (1987), with respect to the putative significant role of the protein F-altered (F*) mutant in antibiotic permeability, PCC23, must be re-examined in light of the above observations and proposals.  This mutant had significantly altered LPS  compared with its parent strain (Godfrey et al, 1984). Major changes in P-lactamases or penicillin-binding proteins were not observed (Godfrey et al, 1984). Godfrey and Bryan showed that the conformation of protein F was dependent on the method used to purify the protein. These investigators showed that protein F, purified by electroelution, was almost completely in the F* configuration, whereas protein F, purified by column chromatography in Triton X-100, was divided between the two configurations (Godfrey & Bryan, 1987). In SDS-PAGE of outer membrane preparations from PCC23, the predominant band was the native, unaltered form. The F* band was barely visible (Godfrey et al, 1984). A  similar observation was made for protein F purified by passive elution from SDS-PAGE in this study. Negligible amounts of the protein F purified from PCC23 in this study were in the F* position (Fig. 15, lane 6), and the putative F* band may simply be protein F that was partially heat modified during preparation of the samples for electrophoresis (Hancock & Carey, 1979). It is possible that for this particular chemical mutant, PCC23, the difference between the F and F* configurations of protein F may reflect an artifact created during purification. The  conformational  change did not result in the formation of 5 nS channels (Fig. 16) and despite the increased average single channel conductance, the most frequently observed channels were 0.4 nS, the size observed for the unaltered native and cloned protein F preparations. Indeed, Godfrey and Bryan provided no definitive proof for the relationship between antibiotic resistance and protein F alterations. It would be interesting to clone and sequence the putative oprF* gene of PCC23, and to move it into a wildtype genetic background, in order to assess the effect of this mutation on antibiotic resistance independently of the observed LPS alterations.  C. Protein F: Channel size The controversy surrounding the channel size and exclusion limit of protein F was presented earlier (see Introduction). To recapitulate, Benz and Hancock (1981) proposed that protein F forms large 5 nS channels, but few of these channels (<1%) exist in an open, functional configuration. Caulcott et al. (1984) and Yoneyama and Nakae (1986) proposed that the exclusion limit of P. aeruginosa was such that the outer membrane was permeable to disaccharides and monosaccharides respectively. Once oprF had been cloned, and it was observed that protein F was expressed in E. coli, it was of interest to purify the cloned protein and examine it's pore-forming ability in black lipid bilayers. Preliminary studies of the cloned gene product were performed by T. R. Parr. Parr  75  Figure 17. Schematic representation of the speculative disulfide rearrangement model. A cross-section through the channel, parallel to the membrane, is shown. This model proposes a way in which the proposed channel heterogeneity could be generated based on rearrangement of the four cysteine residues of protein F into two different cystine disulfide bond pairs. The pairing in panel A would result in the rare, large channel configuration; that in panel B, the more common small channel configuration (from Moore, et al., 1987).  76  observed that protein F from both P. aeruginosa HI 03 and E. coli JF733 (pHN4) formed large channels of 2-5 nS and reported that in some experiments up to 65% of the channels observed had a conductance of greater than 4nS (Woodruff et al, 1986). When the black lipid bilayer apparatus instrumentation was set for higher sensitivity, very small channels of 0.34 nS to 0.38 nS were observed at a greatly increased frequency compared with the large channels (Woodruff et al, 1986).  The  two channel sizes were observed in protein F from both E. coli and P. aeruginosa, therefore this phenomenon was unlikely an artifact but a real characteristic of protein F. It was concluded that protein F formed a heterogeneous channel.  The vast majority, (>99%), of the protein F  molecules, previously considered to be closed and non-functional were now  proposed to form the small, previously undetected, channels. Subsequently, a speculative model to explain the channel  heterogeneity of protein F was proposed (Moore et al, 1987). This model is schematically represented in Figure 17 and has been named the disulfide rearrangement model The basis of this model is the differential pairing of the four cysteine residues of protein F to form two disulfide bonds. The most thermodynamically stable arrangement of the cysteines into disulfides would result in the formation of the small, mostfrequently observed channels. Different pairing of the cysteines would result in the formation of the rare larger channels. The disulfide rearrangement model represents an attractive hypothesis for the preliminary observations of Parr and is in good agreement with the proposals of Benz and Hancock. In the current study, attempts were made to confirm the preliminary observations of Parr and to demonstrate pore function for protein F in vivo. Purified protein F from P. aeruginosa and E. coli JF733 (pWW2200) was examined in the black lipid bilyer system. In agreement with the results of Parr (Woodruff et al, 1986), a variety of channel sizes were observed, but the average single channel conductance was 0.41 nS for native protein F and 0.38 nS for cloned protein F. In contrast to Parr's findings, only 5% of the channels observed were larger than 1.2 nS and none were larger than 4 nS. Significantly more channel events were  77  monitored in the current study than in the preliminary study. The black lipid bilayer data showed that protein F formed pores. The results of this study indicate that the channels formed are very small. When the black lipid bilayer data are combined with the MIC results, the possibility must be entertained that the actual size of protein F is represented by the small, antibiotic-impermeable, 0.4 nS channel. Results of attempts to show large channel-forming ability for protein F in vivo were disappointing. The results of comparative growth rate studies of oprF strains with their wildtype parents showed that loss of protein F had a significant impact on the doubling time of cultures grown in PP2, a phenomenon which could be modified by the addition of salts or sugars. A similar observation has been made for E. coli strains defective in OmpA (Manning et al, 1977) and omp A, Ipp (lipoprotein) mutants (Sonntag et al., 1978). The influence of medium osmolarity on the regulation of the ompB locus, which alters the ratio of OmpF to OmpC in the outer membrane, has been well documented (Lugtenberg et al, 1976; van Alphen & Lugtenberg, 1977; Ozawa & Mizushima, 1983). These data imply a role for outer membrane proteins in maintaining the osmotic integrity of the cell, a role which may be simply structural or may require pore-forming functional ability. The minimum size of such a pore could be very small, permeable only to ions and small hydrophilic molecules. The observation that protein P, the phosphate induced porin in P. aeruginosa was induced at higher levels of external phosphate in the oprF: :Tn2 mutants than in HI 03, suggested a role for protein F in the permeability of inorganic phosphate. However, this observation may be unrelated to the loss of protein F per se. The accumulating evidence supports the conclusion that the small channels observed in the artificial membrane system represent the actual size of the protein F channels. The large pores, consistently observed only in early preparations of protein F, purified by gel sieving in Triton X-100, (R.E.W. Hancock, personal communication), may represent a contaminating porin, present in the outer membrane in very small amounts (Hancock, 1987b; Woodruff & Hancock, 1988). Alternatively, the obser-  vation of large conductance increases in the black lipid bilayers may be the result of protein aggregates or the simultaneous entry of multiple porin molecules (Hancock, 1987b). The former hypothesis is more appealing for the following reasons. Firstly, the preparations of protein F used in the original studies were known to be contaminated with small amounts (visible on SDS-PAGE) of other outer membrane proteins (Hancock et al, 1979; Benz & Hancock, 1981). Such contamination was less likely in later preparations due to a change in purification methodology and the isolation of cloned protein F from E. coli (Woodruff et al, 1986). Secondly, observations that P. aeruginosa can grow on a variety of large compounds as sole carbon sources, such as pentamethionine, whereas E. coli can not (Becker & Naider, 1974; Miller & Becker, 1978), provides evidence that the proposed large pore exists. Thirdly, on a strictly empirical level, it would seem advantageous to a bacterium as metabolically diverse as P. aeruginosa, to have a large repertoire of genes encoding porins of different sizes which could be induced under appropriate conditions. The ability to selectively induce larger porins (already observed for growth on glucose and phosphate), without requiring the constant presence of such pores, would confer a selective impermeability advantage to P. aeruginosa in the presence of undesirable compounds such as antibiotics. Such a proposal is strictly speculative. D. Protein F: Relationship with OmpA Protein F is an integral component of the outer membrane of P. aeruginosa. This conclusion is implicit in the observations that the oprF gene is apparently conserved between P. aeruginosa strains and within the P. fluorescens rRNA homology group (Table III; Fig. 7) and the difficulties in isolating oprF mutants (Nicas & Hancock, 1983; Gotoh et al, 1987) . Recently, the oprF gene from P. aeruginosa was sequenced by colleagues who used pWW5, isolated in this study, to clone fragments of the oprF gene (Duchene et ah, 1988). These authors compared the protein F sequence to the recorded sequences of other proteins.  Surprisingly, similarity was found between protein F and the enterobacterial heat-modifiable OmpA proteins, not the porin proteins. The other OmpA-like non-enterobacterial protein sequenced is PHI from N. gonorrhoeae (Gotschlich et al., 1986b). When the amino acid sequences of Omp A from E. coli, PHI from N. gonorrhoeae and protein F from P. aeruginosa were compared, more regions of homology were observed than were originally reported by Duchene et al. (1988) (Fig. 18). It is thought that the OmpA protein consists of two domains, the amino terminal 180 residues repeatedly traverse the outer membrane while the carboxy terminal residues form a domain on the periplasmic side of the membrane (Bremer et al., 1982; Morona et al., 1984) No homology exists between the amino terminal domains of the protein F, OmpA, and PIII proteins. Not unexpectedly, these regions are also the least conserved among the enterobacterial OmpA proteins and have been shown to form the recognition sites for phage and colicins (Morona et al, 1984). Extensive regions of homology between the carboxy terminal domains of Omp A, PIII and protein F is illustrated in Figure 18. It is very interesting that these regions of homology begin around the proline-rich region found in all three proteins. This region consists of proline residues alternating with alanine, valine or glutamate residues. A similar proline-rich region has been observed for the M protein of Streptococcus pyogenes (Hollingshead et al, 1986) and protein A of Staphylococcus aureus (Guss et al, 1984). This proline-rich region is thought to be the site where membrane-attached proteins pass through the peptidoglycan layer of gram positive cell walls (Hollingshead et al, 1986) and may have an analogous function in the OmpA-like proteins. To align the sequences, gaps of 27 residues in the sequences of OmpA and PIII and 11 and 15 residues in the sequences of OmpA and protein F have been introduced. These gaps accomodate the residues between cysteine residues. As discussed earlier (see Introduction), cysteine residues are rare in outer membrane proteins and are not generally found in porins. PITJ and protein F each have four cysteine  20  30  proOmpA  10 10 50 60 MKKTAIAIAVALAGFATVAQAAPKDNTWYTGAKLGWSQYHDTGFINNNGPTHENQLGAGA  proOprF  MK LXNTLGWIGSLVAASAMNAFAQGQNSVEIEAF GK AYFTDSVRNMK NADLY  20  10 proOmpA  * 30  40  50  70 80 90 100 110 120 FGGyQVNPYVGFEMGYDWLGRMPYKGSVENGAyKAQGVQLTAKLGYPITDDLDiyTRLGG  proOprF GGSIGYFLTDDVELALSYGEYHDVRGTYETGNKKVHGNLTSLDAIYHFGTPGVGLRPYVS 60 70 80 90 100 110  proOmpA  130 140 150 160 170 180 MWRADTKSNVYGKNHDTGVSPVFAGGVEYAITPEIATRLEYQWTNNIGDAHTIGTRPDN  proOprF  AGLAHQNITNlNSDSQGRQQMTMANIGAGLKYYFTENFFAKASLDGQYGLFiCRDNGHQGE  proPIII  MTKQLKLSALFVALLASGTAVAGEASVQGYTVSGQSNETVRNNYGECWKNAYFDKA 40 50  proOmpA proOprF proPIII  10  20 *  30  200  210  190 GMLSLGVSYRFGQGEAAPWAPAPAP  APEVQ  WMAGLGVGFNFGGSKAAPAPEPVADVCSDSDNDGVCDNVDKCPDTPANVTDDANGCPAVA SQGRVBSGDAVAVPEPEPAPVAYYEQ 60  220  70  230  ,  80  APQYV  260  90  270  proOmpA  240 250 TEHFTI£SDVIfKFNKATIXPEGQAAIJ)QLYSQLSNLDPKDGSVVVLGZTDRIGSDAYNQ  proOprF  EW-RVQIDVKFDFDKSKVKENSYADIKNLADFMKQ—YPSTSTTVEGHTDSVGTDAYNQ  proPIII  DETISLSAKTLFGFDKDSLRAEAQDNLKYLAQRLSR—TNYQSYRVEGHTDFMGSBRYNQ 100 110 120 140  proOmpA proOprF proPIII  proOmpA proOprF proPIII  130  280 290 300 310 GLSERRAQSWDYLISK-GIPADKISARGMGESNPVTGNTC  DNVKQRAA  KLSERRANAVRDVLVNEYGVEGGRVNAVGYGESRPVADNATAEG— ALS ER R AYVYANNLVSN-GVP ASRISAYGLGESQAQMTQYCQAEVAKLGAK ASK AKKR EA 150 160 170 180 190 200 330 340 LIDCLAPDRRVEIEVKGIKDWTQPQA RAINRRVEAEVEAEAK LIACIEPDRRVDVKIRSIYTRQVYPARNHHQH 210 220 230  Figure 18. Comparison of the amino acid homology between protein F from P. aeruginosa, OmpA from E. coli and Pin from N. gonorrhoeae. Single letter codes for the amino acids are used. Double dots indicate identities, single dots indicate conservative substitutions. is the start of the mature protein. Residues of PHI homologous to OmpA are underlined. Data were compiled from Duchene, et al, (1988) and Gotschlich, et al., (1987). The conservative substitutions are based on a weighting value of > or = to 8 according to the table for conservative substitutions from Doolittle (1986). A  residues, whereas the enteric OmpA proteins have only two cysteines. The position of the additional cysteine residues, presumed to form a disulfide loop (Gotschlich et al, 1987b; Moore et al, 1987), is noteworthy. In PIII this loop is situated directly upstream of the proline-rich region whereas, in protein F, all four cysteines (presumably forming two disulfide loops) are immediately downstream of the proline-rich region. Therefore, as discussed above, the position of the disulfide loops is likely periplasmic, not submerged in the bilayer. The second pair of cysteines in PIII and the only pair in OmpA are located close to the carboxy terminus. The residues between the cysteines, putatively contained within the disulfide loops, show little similarity among the three proteins. Similarities between the genes for PIII, OmpA and protein F were observed in codon usage (Braun & Cole, 1984; Gotschlich et al, 1987b; Duchene et al, 1988). This was unexpected as the %G+C values for P. aeruginosa (67%) differ markedly from those of E. coli (50-51%) and N. gonorrhoeae (49.5%) (Cowan, 1974; Doudoroff & Palleroni, 1974; Reyn, 1974). In addition to the observed partial amino acid sequence homologies, protein F shares the following characteristics with PIII and OmpA: they have similar molecular weights; strong peptidoglycan and LPS association; substantial 6-sheet structure; are heat modifiable; are resistant to protease degradation in intact cells; and are strongly antigenically conserved within and between species (Hofstra & Dankert, 1980; Nikaido & Vaara, 1985; Gotschlich et al, 1987b; Woodruff & Hancock, 1988). A comparison of functional properties between OmpA, PLTI and protein F is hampered by the absence of information on PIII mutants. E. coli omp A mutants have been fairly extensively characterized and some striking similarities between these strains and the P. aeruginosa oprF mutants exist. As cited earlier, ompA mutants of E. coli had decreased growth rates compared with wildtype strains (Manning et al, 1977) and omp A Ipp double mutants grew poorly in a non-defined medium but normally if the medium was supplemented with an electrolyte such as NaCl or KC1 (Sonntag et al, 1978). A similar effect was not observed for  ompB Ipp mutants missing OmpF and OmpC. Interestingly, the omp A Ipp mutants adopted a spherical morphology which implied a structural role for OmpA. A decrease in growth rate, modified by addition of salt was also observed for P. aeruginosa oprFwTnl and oprFv.Q. strains (Table IV). If these data imply a functional role for OmpA in ion uptake or osmoregulation, the correlation for a similar function for protein F is clear. A role for OmpA in stabilizing mating pair formation during conjugation is well established (Havekes et al, 1976; Achtman et al, 1978; Manoil & Rosenbusch, 1982). A similar lack of ability to engage in conjugation was observed for the oprFv.Tnl mutant in this study. Repeated attempts to conjugate oprF containing plasmids into this strain met with failure (Results, Chapter 2D). Several explanations for this observation are possible. The most obvious is that there is a problem with copy number. The broad host range plasmids used for cloning protein F are present in 5 to 8 copies per host chromosome equivalent (Figurski & Helinski, 1979). It has been observed that multiple copies of outer membrane protein genes are lethal to the host cell (Results Chapter 1A) and this may also be the situation when the recipient cell is the native bacterium. Alternatively, based on the structural homology between protein F and OmpA, a role for protein F in conjugation must not be overlooked. A functional role for pore formation by OmpA has been hypothesized based on the observation that ompA mutants showed overall reduced transport rates for radioactive glutamine and proline (Manning et al., 1977). The methodologies used in these studies have been criticized by Nikaido (1979) who later reported a similar transport defect for ompA mutants belonging to two separate lines (Nikaido & Vaara, 1985). One of the criticisms cited against a pore-forming role for OmpA was the observation that an omp A mutant of S. typhimurium showed unaltered permeability to cephaloridine compared with its parent strain (Nikaido, 1979). It should be pointed out that if OmpA, by analogy to protein F, forms small channels, it is likely these channels would be permeable to amino acids, but impermeable to a B-lactam as large as  cephaloridine (molecular weight, 415: Yoshimura & Hikaido, 1985). The likelihood that OmpA and PLTJ form small channels, in the outer membranes of E. coli and N. gonorrhoeae respectively, is significant based on the in vivo data of the omp A mutants and structural homologies observed between these proteins and protein F. While such a possibility has far-reaching implications, in terms of membrane permeability, antibiotic therapy, and osmotic regulation of gene expression (Higgins et al., 1988), it should be recognized that minimal amino acid homology was observed between the aminoterminal membrane-spanning domains of OmpA, PIII and protein F.  E. Protein F: Clinical relevance The earlier observation that protein F may have a marginal role in antibiotic permeability through the outer membrane of P. aeruginosa has interesting implications for clinical therapies. The impetus for cloning and gene and structurally characterizing a 5 nS channel with respect to the design of effective theraputic agents is obvious. However, if protein F is not involved in the formation of large channels, as the results of this study indicate, is it still a potential target molecule for anti-Pseudomonas therapy? The answer is definitely affirmative. Aside from designing very small antibiotics, able to permeate the small channels of protein F, or determining how to alter outer membrane permeability via protein F, an obvious tactic would be to construct an anti-Pseudomonas vaccine based on protein F. Several features of protein F make it an ideal vaccine candidate. Protein F is surface exposed, present in high copy numbers in the outer membrane, and very tightly conserved within P. aeruginosa strains and among the nine strains of the P. fluorescens rRNA homology group. Several anti-Pseudomonal vaccines are commercially available (Mutharia et al, 1985). These are based on LPS antigens and tend to be highly toxic, selectively active for specific serotypes and produce shortlived responses. A vaccine based on an outer membrane protein may  eliminate these problems. A precedent for effective outer membrane protein based vaccines has been established. A vaccine containing outer membrane proteins from N. meningitidis was protective, non-toxic and non-pyrogenic in both humans and animals (Zollinger et al, 1978). The existence of techniques to purify antigenically unaltered protein F, essentially free from contaminating LPS (Parr et al, 1985) and from an E. coli background (Woodruff et al, 1986), would negate the toxicities and serotype specificities found in LPS vaccines. The evidence presented in this study shows that protein F from P. aeruginosa forms very small 0.4 nS channels in the outer membrane of this organism. Such channels would probably be impermeable to antibiotics. Thus, these small channels likely contribute to the low rate of permeation of antibiotics across the outer membrane and consequently the intrinsic high antibiotic resistance observed in this organism both clinically and in the lab. Despite its contribution to the antibiotic resistance of P. aeruginosa, protein F may play a key role in the future in an effective anti-Pseudomonas therapy. A vaccine based on protein F, an ideal candidate for several of its intrinsic properties, would provide prophylactic therapy against infection by P. aeruginosa for high risk groups, significantly lessening the threat of this often fatal opportunistic pathogen.  85  REFERENCES  Achtman, M., S. Schwunchow, R. Helmuth, G. Morelli and P. A. Manning. 1978 Cell-cell interactions in conjugating Escerichia coli Con- mutants and stabilization of mating aggregates. Molec. Gen Genet. 164: 171-183. Angus, B.L., J. Fyfe and R.E.W. Hancock. 1987. Mapping and characterization of two mutations to antibiotic super-susceptibility in Pseudomonas aeruginosa. J. Gen. Microbiol. 133: 2905-2914. Angus, B.L., and R.E.W. Hancock. 1983. Outer membrane proteins F, P, D l of Pseudomonas aeruginosa and PhoE of Escherichia coli:  Chemical crosslinking to reveal native oligomers. J. Bacteriol. 155: 1042-1051. Bassford, P.J., D.L. Diedrich, C.L. Schnaitman and P. Reeves. 1977. Outer membrane proteins of Escherichia coli. VI. Porin alteration in bacteriophage resistant mutants. J. Bacteriol. 131: 608-622. Bavoil, P., and H. Nikaido. 1977. Pleiotropic transport mutants of Escherichia coli lack porin, a major outer membrane protein. Molec. Gen. Genet. 158: 23-33. Bavoil, P., A. Ohlin and J. Schacter. 1984. Role of disulfide bonding in outer membrane structure and permeability in Chlamydia trachomatis. Infect. Immun. 44: 479-485. Becker, J.M., and F. Naider. 1974. Stereospecificity of tripeptide utilization in a methionine auxotroph of Escherichia coli K-12. J. Bacteriol. 120: 191-196. Beher, M.G., C A . Schnaitman and A.P. Pugsley. 1980. Major heat  86  modifiable outer membrane protein in gram negative bacteria: comparison with the OmpA protein of Escherichia coli. J. Bacteriol. 143: 906-913. Benson, S.A., and A. Decloux. 1985. Isolation and characterization of outer membrane permeability mutants in Escherichia coli K-12. J. Bacteriol. 161: 361-367. Benz, R., and R.E.W. Hancock. 1981. Properties of the large ionpermeable pores fromed from protein F of Pseudomonas aeruginosa in lipid bilayer membranes. Biochim. Biophys. Acta 646: 298-308. Benz, R., R.E.W. Hancock and T. Nakae. 1982. Porins from gram negative bacteria in lipid bilayer membranes, p. 123-134 In R. Antolini, A. Gliozzi and A. Gorio (ed.), Transport in Biomembranes: Model Systems and Reconstitution.  Raven, New  York. Bodey, G.P., R. Bolivar, V. Fainstein and L. Jadeja. 1983. Infections caused by Pseudomonas aeruginosa. Rev. Infect. Dis. 5: 279-313. Braun, G. and S.T. Cole. 1983 Molecular characterization of the gene coding for major outer membane protein Omp A from Enterobacter aerogenes. Eur. J. Biocem. 137: 495-500. Braun, G., and S.T. Cole. 1984. D N A sequence analysis of the Serratia marcescens ompA gene: Implications for the organization of an enterobacterial outer membrane protein. Mol. Gen. Genet. 195: 321-328. Bremer, E., S.T. Cole, I. Hindennach, U. Henning, E. Bech, C. Kurz and H. Schaller. 1982. Export of a protein into the outer membrane of Escherichia coli K-12 stable incorporation of the OmpA protein  87  requires less than 193 amino-terminal amino-acid residues. Eur. J. Biochem. 122: 223-231. Bukhari, A.I., and A.L. Taylor. 1971. Mutants of Escherichia coli with a growth requirement for either lysine or pyridoxine. J. Bacteriol. 105: 988-998. Buscher, K-H., W. Cullmann, W. Dick and W. Opferkuch. 1987. Imipenem resistance in Pseudomonas aeruginosa resulting from diminished expression of an outer membrane protein. Antimicrob. Agents Chemother. 31: 703-708. Carbonetti, N.H., and P.F. Sparling. 1987. Molecular cloning and characterization of the structural gene for protein I, the major outer membrane protein of Neisseria gonorrhoeae. Proc. Natl. Acad. Sci. USA. 84: 9084-9088. Caulcott, C.A., M.R.W. Brown and I. Gonda. 1984. Evidence for small pores in the outer membrane of Pseudomonas aeruginosa.  FEMS  Microbiol. Lett. 21: 119-123. Chapman, J.S., and N.H. Georgopapadakou. 1988. Route of quinolone permeation in Escherichia coli. Antimicrob. Agents Chemother. 32: 438-442. Cole, S.T., I. Sonntag and U. Henning. 1982. Cloning and expression in Escherichia coli K-12 of the genes for major outer membrane protein OmpA from Shigella dysenteriae, Enterobacter aerogenes  and Serratia marcescens. J. Bacteriol. 149: 145-150. Cowan, S.T., 1974. Gram-negative facultatively anaerobic rods. p. 290351. In R. E. Buchanan and N.J. Palleroni (ed.) Bergey's manual of determinative bacteriology. The Williams & Wilkins Company., Baltimore.  88  Crouse, G.F., A. Fischaud and H. Lehrach. 1983. A n integrated and simplified approach to cloning into plasmids and single-stranded phages. Methods Enzymol. 101: 78-89. Curtis, N.A.C., R.L. Eisenstadt, K.A. Turner and A.J. White. 1985. Porinmediated cephalosporin resistance in Escherichia coli K-12. J. Antimicrob. Chemother. 15: 642-644. Darzins, A., and A.M. Chakrabarty. 1984. Cloning of genes controlling alginate biosynthesis from a mucoid cystic fibrosis isolate of Pseudomonas aeruginosa. J. Bacteriol. 159:  9-18.  De Vos, P., and J. De Ley. 1983. Intra- and inter-genic similarities of Pseudomonas and Xanthomonas ribosomal ribonucleic acid cistrons. Int. J. Syst. Bacteriol. 33: 487-509. Ditta, G., T. Schmidhauser, E. Yakobsen, P. Lu, X-W. Liang, D. Finlay, D. Guiney and D.R. Helinski. 1985. Plasmids relating to the broad host range vector pRK290, useful for gene cloning and for monitoring gene expression. Plasmid. 13: 149-153. Doolittle, R.F., 1986. Of urfs and orfs. A primer on how to analyze derived amino acid sequences. University Science Books Mill Valley, California. Doudoroff, M., and N.J. Palleroni. 1974. Gram negative rods and cocci, p. 217-243. In R.E. Buchanan and N.E. Gibbons (ed.) Bergey's manual of determinative bacteriology. The Williams and Wilkins Company., Baltimore. Duchene, M., A. Schweizer, F. Lottspeich, G. Krauss, M. Marget, K. Vogel, B-U. von Specht and H. Domdey. 1988. Sequence and transcriptional start site of the Pseudomonas aeruginosa outer membrane porin protein F gene. J. Bacteriol. 170: 155-162.  89  Figurski, D., and D.R. Helinski. 1979. Replication of an origin-containing derivative of plasmid RK2 dependent on a plasmid function provided in trans. Proc. Natl. Acad. Sci. USA.  76: 1648-1652.  Feinberg, A.P., and B. Vogelstein. 1983. A technique for radiolabelling DNA  restriction endonuclease fragments to a high specific activity.  Anal. Biochem. 132:  6-13.  Foulds, J., and T. Chai. 1979. Isolation and characterization of isogenic E. coli strains with alterations in the level of one or more major outer membrane proteins. Can. J. Microbiol. 25: 423-427. Freudl, R., and S.T. Cole. 1983. Cloning and molecular characterization of the omp A gene from Salmonella typhimurium.  Eur. J.  Biochem. 134: 497-502. Friedman, A.M.,  S.R. Long, S.E. Brown, W.J. Buikema and I.M. Ausubel.  1982. Construction of a broad host range cosmid cloning vector and its use in the genetic analysis of Rhizobium mutants. Gene 18: 289-296. Godfrey, A.J., and L.E. Bryan. 1987. Penetration of (3-lactams through Pseudomonas aeruginosa porin channels. Antimicrob. Agents Chemother. 31_: 1216-1221. Godfrey, A.J., L. Hatleled and L.E. Bryan. 1984. Correlation between lipopolysaccharide structure and permeability resistance in (3lactam-resistant Pseudomonas aeruginosa. Antimicrob. Agents Chemother. 26: 181-186. Goldberg, J.B., and D.E. Ohman. 1984. Cloning and expression in Pseudomonas aeruginosa of a gene involved in the production of alginate. J. Bacteriol. 158: 1115-1121.  Gotoh, N., H. Wakeke, T. Nishino and T. Tanino. 1987. Rapid extraction method for detection of outer membrane protein F of Pseudomonas aeruginosa. J. Micro. Methods. 6: 265-271. Gotschlich, E.C., M.S. Blake, J.M. Koomey, M. Seiff and A. Derman. 1986. Cloning of the structural genes of three H8 antigens and of protein III of Neisseria gonorrhoeae. J. Exp. Med. 164: 868-881.  Gotschlich, E.C., M. Seiff and M.S. Blake. 1987. The D N A sequence of the structural gene of gonococcal protein III and the glanking region containing a repetitive sequence. J. Exp. Med. 165: 471-482. Guss, B., M. Uhlen, B. Nilsson, M. Lindsay, J. Sjoquist, and J. Sjodaht. 1984. Region X, the cell-wall-attachment part of staphylococcal protein A. Eur. J. Biochem. 138: 413-420. Haas, M., J.M. Becker and R.V. Miller. 1981. Peptidase activity in the inner membrane of Pseudomonas aeruginosa.  Biochim. Biophys.  Acta 643: 256-260. Haas, D., and B.W. Holloway. 1978. Chromosome mobilization by the R plasmid R68.45: a tool in Pseudomonas genetics. Mol. Gen. Genet. 158: 229-237. Hancock, R.E.W. 1987a. Model membrane studies of porin function. In M. Inouye (ed.),  Bacterial outer membranes as model systems.  John Wiley & Sons, Inc. New York. Hancock, R.E.W. 1987b. Role of porins in outer membrane permeability. J. Bacteriol. 169: 929-933. Hancock, R.E.W., and A.M. Carey. 1979. Outer membrane of Pseudomonas aeruginosa:  heat and 2-mercaptoethanol  modifiable proteins. J. Bacteriol. 140: 902-910.  91  Hancock, R.E.W., G.M. Decad and H. Nikaido. 1979. Identification of the protein producing transmembrane diffusion pores in the outer membrane of Pseudomonas aeruginosa P A O l .  Biochim. Biophys.  Acta 554: 323-331. Hancock, R.E.W., K. Poole and R. Benz. 1982. Outer membrane protein P of Pseudomonas aeruginosa: regulation by phosphate-deficiency and  formation of small anion-specific channels in lipid bilayer  membranes. J. Bacteriol. 150: 730-738. Hancock, R.E.W., V. Raffle and T.I. Nicas. 1981. Involvement of the outer membrane in gentamicin and streptomycin uptake and killing in Pseudomonas aeruginosa.  Antimicrob. Agents  Chemother. 19: 777-785. Hancock, R.E.W., and W.A. Woodruff. 1988. Roles of porin and P-lactamase in P-lactam resistance of Pseudomonas aeruginosa. Rev. Infect. Dis. 10: 770-775. Hancock, R.E.W., E.A. Worobec, K. Poole, and R. Benz. 1987. Phosphatebinding site of Pseudomonas aeruginosa outer membrane protein p. 186-190. In A. Torriani-Gorini, F. G. Rothman, S. Silver, A. Wright, and E. Yagil (ed.), Phosphate metabolism and cellular regulation in microorganisms. American Society for Microbiology, Washington, D.C. Harder, K.J., H. Nikaido and M. Matgsuhashi. 1981. Mutants of Escherichia coli that are resistant to certain beta-lactam compounds lack the ompF porin. Antimicrob. Agents Chemother. 20: 549-552. Havekes, L.M., B.J.J. Lugtenberg and W.P.M. Hoekstra. 1976. Conjugation deficient ddE. coliKll F- mutants with heptose-less lipopolysaccharide. Molec. Gen. Genet. 146: 43-50.  Henning, U., H-D. Royer, R.M. Teather, I. Hindennach and CP. Hollenberg. 1979. Cloning of the structural gene (ompA) for an integral outer membrane protein of Escherchia coli K-12. Proc. Natl. Acad. Sci. USA. 76: 4360-4364. Henning, U., I. Sonntag and I. Hindennach. 1978. Mutants (ompA) affecting a major outer membrane protein of Escherichia coli K12. Eur. J. Biochem. 92: 491-498. Higgins, C.F., C.J. Dorman, D.A. Stirling, L. Waddell, I.R. Booth, G. May and E. Bremer. 1988. A physiological role for D N A supercoiling in the osmotic regulation of gene expression in S. typhimurium and E. coli. Cell 52: 69-584. Hiruma, R., A. Yamaguchi and T. Sawai. 1984. The effect of lipopolysaccharide on lipid bilayer permeability of P-lactam antibiotics. FEBS Lett. 170: 268-272. Hitchcock, P.J., and T.M. Brown. 1983. Morphological heterogeneity among Salmonella lipopolysaccharide chemotypes in silverstained polyacrylamide gels. J. Bacteriol. 154: 269-277. Hofstra, H., and J. Dankert. 1979. Antigenic cross reactivity of major outer membrane proteins in Enterobacteriaceae species. J. Gen. Microbiol. I l l : 293-302. Hofstra, H., and J. Dankert. 1980. Major outer membrane proteins: common antigens in enterobacteriaceae species. J. Gen. Microbiol. 119: 123-131. Hollingshead, S.K., V.A. Fischetti and J.R. Scott. 1986. Complete nucleotide sequence of type 6 M protein of the group A Streptococcus. J. Biol. Chem. 261: 1677-1686.  Horan, T.C, J.W. White, W.R. Jarvis, T.G. Emori, D.H. Culver, V.P. Munn, C. Thornsberry, D.R. Olson and D.F. Hughes. 1986. Nosocomial infections surveillance. Morbid. Mortal. Wkly. Rep. 35: suppl. 1:17-29. Hrebenda, J., and H. Heleszko. 1985. Mutations affecting the resistance of Escherichia coli K-12 to B-lactam antibiotics. J. Antimicrob. Chemother. 16: 149-156. Jaffe, A., Y.A. Chabbert and O. Semonin. 1982. Role of porin proteins OmpF and OmpC in the permeation of B-lactams. Antimicrob. Agents Chemother. 22: 942-948. Jorgensen, R.A., S.J. Rothstein and W.S. Reznikoff. 1979. A restriction enzyme cleavage map of Tn5 and location of a region encoding neomycin resistance. Mol. Gen. Genet. 177: 65-72. Komatsu, Y., K. Murakami and T. Nishikawa. 1981. Penetration of moxalactam into its target proteins in Escherichia coli K-12: comparison of a highly moxalactam resistant mutant with its parent strain. Antimicrob. Agents Chemother. 20: 613-619. Krishnapillai, V. 1971. A novel transducing phage. Its role recognition of a possible new host-controlled modification system in Pseudomonas aeruginosa. Mol. Gen. Genet. 114: 134-143. Lee, D.R., C A . Schnaitman and A.P. Pugsley. 1979. Chemical heterogeneity of major outer membrane pore proteins of Escherichia coli. J. Bacteriol. 138: 861-870. Loh, B., C Grant and R.E.W. Hancock. 1984. Use of the fluorescent probe 1-N-phenylnaphthylamine to study the interactions of aminoglycoside antibiotics with the outer membrane of P. aeruginosa. Antimicrob. Agents Chemother. 26: 546-551.  Lugtenberg, B., R. Peters, H. Bernheimer and W. Berendsen. 1976. Influence of cultural conditions and mutations on the composition of the outer membrane proteins of Escherichia coli. Molec. Gen. Genet. 147:251-262. Lugtenberg, B., and L. van Alphen. 1983. Molecular architecture and functioning of the outer membrane of Escherichia coli and other gram negative bacteria. Biochim. Biophys. Acta. 737: 52-115. Maniatis, R., E.F. Fritsch and J. Sambrook. 1982. Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York. Manning, P.A., A.P. Pugsley and P. Reeves. 1977. Defective growth functions in mutants of Escherichia coli K-12 lacking a major outer membrane protein. J. Mol. Biol. 116: 285-300. Manning, P.A., and M. Achtman. 1979. Biogenesis and functions, p. 409448. In M. Inouye (ed.) Bacterial outer membranes. John Wiley & Sons, New York. Manning, P.A., E.J. Bartowsky, D.I. Leavesly, J.A. Hackett and  M.W.  Heuzenroeder. 1985. Molecular cloning using immune sera of a 22 kDal outer membrane protein of Vibrio cholerae. Gene. 34: 95-103. Manoil, C. 1983. A genetic approach to defining the sites of interaction of a membrane protein with different external agents. J. Mol. Biol. 169: 507-519. Manoil, C, and J.P. Rosenbusch. 1982. Conjugation-deficient mutants of Escherichia coli distinguish classes of functions of the outer membrane OmpA protein. Mol. Gen. Genet. 187: 148-156.  95  Meade, H.M., S.R. Long, G.B. Ruvkin, S.E. Brown and R.M. Ausubel. 1982. Physical and genetic characterization of symbiotic and auxotrophic mutants of Rhizobium meliloti induced by transposon Tn5 mutagenesis. J. Bacteriol. 149: 114-122. Messing, J., and J. Vieria. 1982. A new pair of M13 vectors for selecting either D N A strand of double-digest restriction fragments. Gene. 19: 269-276. Miller, R.V., and J.M. Becker. 1978. Peptide utilization in P. aeruginosa: evidence for membrane-associated peptidase. J. Bacteriol. 133: 165-171. Misra, R. and S.A. Benson. 1988. Isolation and characterization of Omp  C  porin mutants with altered pore properties. J. Bacteriol. 170: 528533. Mizuno, T., M-Y. Chou and M. Inouye. 1983. A comparative study on the genes for three porins of the Escherichia coli outer membrane. J. Biol. Chem. 258: 6932-6940. Mizuno, T., and M. Kageyama. 1978. Separation and characterization of the outer membrane of Pseudomonas aeruginosa J. Biochem. (Tokyo) 86: 991-1000. Mizuno, T., and M. Kageyama. 1979. Isolation and characterization of major outer membrane proteins of Pseudomonas aeruginosa to strain PAO with special reference to peptidoglycan-associated protein. J. Biochem. 86: 979-989. Moore, R.A., W.A. Woodruff and R.E.W. Hancock. 1987. Antibiotic uptake pathways across the outer membrane of Pseudomonas aeruginosa. Antibiot. Chemother. 39: 172-181.  Morona, R., M. Klose and U. Henning. 1984. Escherichia coli K-12 outer membrane protein (Omp A) as a bacteriophage receptor: analysis of mutant genes expressing altered proteins. J. Bacteriol. 159: 570-578. Movva, N.R., K. Nakamura and M. Inouye. 1980. Gene structure of the Omp  A protein, a major surface protein of Escherichia coli  required for cell-cell interaction. J. Mol. Biol. 143: 317-328. Mutharia, L.M., and R.E.W. Hancock. 1983. Surface localization of Pseudomonas aeruginosa outer membrane porin protein F by using monoclonal antibodies. Infect. Immun. 42: 1027-1033. Mutharia, L.M., and R.E.W. Hancock. 1985a. Characterization of two surface localized antigenic sites on porin protein F of Pseudomonas aeruginosa. Can. J. Microbiol. 31_: 381-386. Mutharia, L.M., and R.E.W. Hancock. 1985b. Monoclonal antibody for an outer membrane lipoprotein of the Pseudomonas fluorescens group of the family Pseuomoadaceae. Int. J. of Syst. Bacteriol. 35: 530-532. Mutharia, L.M., J.S. Lam, and R.E.W. Hancock. 1985c. Use of monoclonal antibodies in the study of common antigens of gram negative bacteria, p. 131-157. In A.J.L. Macario and E.C. Macario (ed.) Monoclonal antibodies against bacteria. Academic Press, Inc., Orlando, Florida. Nayler, J.H.C. 1987. Review. Resistance to B-lactams in gram negative bacteria: relative contributions of B-lactamase and permeability limitations. J. Antimicrob. Chemother. 19: 713-732. Nicas, T.I., and R.E.W. Hancock. 1980. Outer membrane protein HI of Pseudomonas aeruginosa:  involvement in adaptative and  mutational resistance to ethylenediaminetetraacetate, polymyxin B  and gentamicin. J. Bacteriol. 143: 872-878. Nicas, T.I., and R.E.W. Hancock. 1983. Pseudomonas aeruginosa outer membrane permeability: isolation of a porin F-deficient mutant. J. Bacteriol. 153: 281-285. Nikaido, H. 1979. Nonspecific transport through the outer membrane, p. 361-407. In M. Inouye (ed.) Bacterial outer membranes. John Wiley & Sons, Inc., New York. Nikaido, H., and E.Y. Rosenberg. 1981. Effect of solute size on diffusion rates through the transmembrane pores of the outer membrane of Escherichia coli. J. Gen. Physiol. 77: 121-135. Nikaido, H., and E.Y. Rosenberg. 1983. Porin channels in Escherichia coli: studies with liposomes reconstituted from purified proteins. J. Bacteriol. 153: 241-252. Nikaido, H., E.Y. Rosenberg and J Foulds. 1983. Porin channels in Escherichia coli: studies with p-lactams in intact cell. J. Bacteriol. 153: 232-240. Nikaido, H., and M. Vaara. 1985. Molecular basis of bacterial outer membrane permeability. Microbiol. Rev. 49: 1-32. Nikaido, H., and R.E.W. Hancock. 1986. Outer membrane permeability of Pseudomonas aeruginosa, p. 145-193. In J.R. Kotatch (ed.), The bacteria: a treatise on structure and function, vol. 10. Academic Press, Inc., New York. Nurminen, M., K. Lounatman, M. Sarvas, P.H. Makela and T. Nakae. 1976. Bacteriophage-resistant mutants of Salmonella typhimurium deficient in two major outer membrane proteins. J. Bacteriol. 127: 941-955.  98  Ogle, J.W., J.M. Janda, D.E. Woods, and M.L. Vasil. 1987. Characterization and use of a D N A probe as an epidemiological marker for Pseudomonas aeruginosa. J. Infect. Dis. 155: 119-126. Ohman, D.E., M.A. West, J.L. Flynn, and J.B. Goldberg. 1985. Method for gene replacement in Pseudomonas aeruginosa used in construction of recA mutants: recA-independent instability of alginate production. J. Bacteriol. 162: 1068-1074. Overbeeke, N., H. Bergmans, F. van Mansfeld and B. Lugtenberg. 1983. Complete nucleotide sequence of ddddphoE, the structural gene for the phosphate limitaion inducible outer membrane pore protein of Escherichia coli K-12. J. Mol. Biol. 163: 513-532. Ozawa, Y., and S. Mizushima. 1983. Regulation of outer membrane porin protein synthesis in Escherichia coli K-12: ompF regulates the expression of ompC. J. Bacteriol. 154: 669-675. Parr, T.R., K. Poole, G.W.K. Crockford, and R.E.W. Hancock. 1985. Lipopolysaccharide-free Escherichia coli Omp  F and  Pseudomonas aeruginosa protein P porins are functionally active in lipid bilayer membranes. J. Bacteriol. 165: 523-526. Peterson, A.A., R.E.W. Hancock, and E.J. McGroarty. 1985. Binding of polycationic antibiotics and polyamines to lipopolysaccharides of Pseudomonas aeruginosa. J. Bacteriol. 164: 1256-1261. Poole, K., and R.E.W. Hancock. 1984. Phosphate transport in Pseudomonas aeruginosa:  involvement of a periplasmic  phosphate-binding protein. Eur. J. Biochem. 144: 607-612. Poole, K., and R.E.W. Hancock. 1986a. Isolation of a Tn501 insertion mutant lacking porin protein P of Pseudomonas aeruginosa. Mol. Gen. Genet. 202: 403-409.  Poole, K., and R.E.W. Hancock 1986b. Phosphate-starvation-induced outer membrane proteins of members of the families Enterobacteriaceae and Pseudomonaceae demonstration of  immunological cross-reactivity with as antiserum specific for porin protein P of Pseudomonas aeruginosa. J. Bacteriol. 165: 987-993. Quinn, J.P., E.J. Dudek, C A . DiVincenzo, D.A. Lucks and S.A. Lerner. 1986. Emergence of resistance to imipenem during therapy for Pseudomonas aeruginosa infections. J. Infect. Dis. 154: 289-294. Reyn, S.T. 1974. Gram negative cocci and coccobaccilli. p. 427-436. In R.E. Buchanan and N.E. Gibbons (ed.), Bergey's manual of determinative bacteriology. The Williams & Wilkins Company, Baltimore. Rolinson, G.N., 1986. 6-lactam antibiotics. J. Antimicrob. Chemother. 17: 5-36. Sadhu, C, and L. Gedman. 1988. A procedure for the preparation of RNA-free plasmid DNA. Bio techniques 6: 11-12. Sato, J., and T. Yura. 1979. Chromosomal location and expression for the structural gene for major outer membrane protein la of Escherichia coli K-12 and of the homologous gene of Salmonella  typhimurium.  J. Bacteriol. 139: 468-477.  Schindler, H., and J.P. Rosenbusch. 1981. Matrix protein in planar membranes: clusters of channels in a native environment and their functional reassembly. Proc. Natl. Acad. Sci. USA 78: 2302-2306. Schweizer, M., H. Schwarz, I. Sonntag, and U. Henning. 1976. Mutational change of membrane architecture. Mutants of Escherichia coli K12 missing major proteins of the outer cell envelope membrane. Biochim. Biophys. Acta. 448: 474-491.  100  Sekar, V. 1987. Rapid screening procedure for the identification of recombinant clones. Biotechniques 5: 11-13. Siden, I., and H. Boman. 1983. Escherichia coli mutants with an altered sensitivity to cecropin D. J. Bacteriol. 154: 170-176. Sonntag, I., H. Schwarz, Y. Hirota, and U. Henning. 1978. Cell envelope and shape of Escherichia coli: multiple mutants missing the outer membrane lipoprotein and other major outer membrane proteins. J. Bacteriol. 136: 280-285. Stephens, R.S., C-C. Kuo, G. Newport, and N. Agabian. 1985. Molecular cloning and expression of Chlamydia trachomatis major outer membrane protein antigens in Escherichia coli. Infect. Immun. 47: 713-718. Stephens, R.S., G. Mullenbach, R. Sanchez-Pescador and N. Agabian. 1986. Sequence analysis of the major outer membrane protein gene from Chlamydia trachomatis serovar L2. J. Bacteriol. 168: 12771282. Tommassen, J., and B. Lugtenberg. 1981. Localization of phoE, the structural gene for outer membrane protein e in Escherichia coli K12. J. Bacteriol. 147: 118-123. Tommassen, J., P. Overduin, B. Lugtenberg and H. Bergmans. 1982 Cloning of pho E, the structural gene for the Escherichia coli phosphate limitation-inducible outer membrane pore protein. J. Bacteriol. 149: 668-672. Tsai, C-M., and C.E. Frasch. 1982. A sensitive stain for detecting lipopolysaccharides in polyacrylamide gels. Anal. Biochem. 119: 115-119.  101  van Alphen, W., and B. Lugtenberg. 1977. Influence of osmolarity on the growth medium on the outer membrane protein pattern of Escherichia coli. J. Bacteriol. 131: 623-630. van Alphen, L., B. Lugtenberg, R. van Boxtel and K. Verhoef. 1977. Architecture of the outer membrane of Escherichia coli.  Biochim.  Biophys. Acta. 466: 257-268. Vu, H., and H. Nikaido. 1985. Mechanism of resistance of a B-lactamase constitutive Enterobacter cloacae strain to the third generation Blactams: Role of B-lactam hydrolysis. Antimicrob. Agents Chemother. 27: 393-398. Winberg, G., and M-L. Hammarskjold. 1980. Isolation of D N A from agarose gels using DEAE-paper. Application to restriction site mapping of adenovirus type 16 DNA. Nucl. Acids Res. 8: 253-264. Woodruff, W.A., and R.E.W. Hancock. 1988. Construction and characterization of Pseudomonas aeruginosa protein F-deficient mutants after in vitro and in vivo insertion mutagenesis of the cloned gene. J. Bacteriol. 170: 2592-2598. Woodruff, W.A., T.R. Parr, R.E.W. Hancock, L.F. Hanne, T.I. Nicas and B.H. Iglewski. 1986. Expression in Escherichia coli and function of Pseudomonas aeruginosa outer membrane porin protein F. J. Bacteriol. 167: 473-479. Yamaguchi, A., N. Tomiyama, R. Hiruma and T. Sawai. 1985. Difference in pathway of Escherichia coli outer membrane permeation between penicillins and cephalosporins. FEBS. Lett. 181: 143-148. Yanisch-Perron, C, J. Vieria and J. Messing. 1985. Improved M l 3 phage cloning vectors and host strains: Nucleotide sequences of the M13mpl8 and pUC19 vectors. Gene. 33: 103-119.  102  Yem, D.W., and H.C. Wu.  1978. Physiological characterization of an  Escherichia coli mutant altered in the structure of murein lipoprotein. J. Bacteriol. 133: 1419-1426. Yoneyama, H., and T. Nakae. 1986. A small diffusion pore in the outer membrane of Pseudomonas aeruginosa.  Eur. J. Biochem. 157:  33-38. Yoshimura, F., and H. Nikaido. 1985. Diffusion of B-lactam antibiotics through the porin channels of Escherichia coli K-12. Antimicrob. Agents Chemother. 27: 84-92. Yoshimura, F., L.S. Zalman and H. Nikaido. 1983. Purification and properties of Pseudomonas aeruginosa porin. J. Biol. Chem. 258: 2308-2314. Young, L.S., P. Stevens and B. Kaijser. 1982. Gram negative pathogens in septicaemic infections. Scand. J. Infect. Dis. Suppl. 31: 78-94. Yu, F., 1987. Appendix: DNA and amino acid sequences of outer membrane proteins and lipoproteins. In: M. Inouye (ed.), Bacterial outer membranes as model systems. John Wiley & Sons, Inc New York. Zimmerman, W., and A. Rosselet. 1977. Function of the outer membrane of Escherichia coli as a permeability barrier to betalactam antibiotics. Antimicrob. Agents Chemother. 12: 368-376. Zollinger, W.D., R.E. Mandrell, P. Altieri, S. Berman, J. Lowenthal and M. S. Artenstein. 1978. Safety and immunogenicity of a Neiserria meningitidis type 2 protein vaccine in animals and humans. J. Infect. Dis. 137: 728-739.  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items