Ohman Lab Research
Most of the research in our lab is on Pseudomonas aeruginosa, an opportunistic bacterial pathogen that causes a diverse range of infections among patients compromised by underlying disease, chemotherapy, or occupational factors. Such opportunistic infections include acute pneumonia in emphysema patients, chronic pneumonia in chronic obstructive pulmonary disease (COPD) patients and cystic fibrosis patients, bacteremia in burn and cancer patients, as well as many other infections. Due to the innate resistance of P. aeruginosa to a wide range of antimicrobial agents, such infections are life threatening because they are very difficult to treat. P. aeruginosa produces numerous toxins and degradative enzymes as its main virulence factors. The capsule-like exopolysaccharide called alginate is a major virulence factor of P. aeruginosa during pulmonary infection by protecting the bacteria from the host immune response and contributing to biofilm formation. Overproduction of alginate leads to a mucoid phenotype, which is common in strains that cause chronic pulmonary infection. Infections with mucoid P. aeruginosa are virtually impossible to eradicate. A variety of tissue-damaging proteases are also secreted by P. aeruginosa, and several are important virulence factors during pulmonary infections.
I. Alginate and infection. The role of alginate in lung infection is complex and appears to give P. aeruginosa several advantages. Alginate confers anti-phagocytic properties typically associated with bacterial capsules. Alginate-producing bacteria are less susceptible to normal, antibody-dependent bactericidal mechanisms, and immuno-determinants for opsonic antibody are blocked by alginate. Alginate is released to the environment and increases the viscosity of the bronchial secretions in the lung, which interferes with mucociliary airway clearance and impairs movement of phagocytes.
Alginate is an unbranched polysaccharide that is composed of two kinds of uronic acid residues: β-D-mannuronic acid (M), and its C5 epimer, α-L-guluronic acid (G) in β-1,4 linkage (Fig. 1).
Alginate biosynthetic genes are clustered in a large operon (Chitnis and Ohman 1993) referred to here as the algD operon (Fig. 2). The algD promoter in P. aeruginosa requires the alternative sigma factor σ22 to activate its transcription. At least 13 gene products are involved in the synthesis and secretion of alginate. One of our goals is to better understand the molecular mechanisms of the multi-protein complex for alginate polymerization, polymer modification, and transport out of the bacterial cell.
Synthesis of a nucleoside diphospho-sugar precursor of alginate. The main precursor of alginate is the nucleoside diphospho-sugar, GDP-mannuronic acid (Fig. 3). Step 1 in its synthesis is the conversion of fructose 6-phosphate to mannose 6-phosphate by AlgA (phosphomannose isomerase). Step 2 is the production of mannose 1-phosphate by phosphomannose mutase (PMM), an activity found in AlgC. The algC gene is not in the algD operon, and is also involved in the biosynthesis of LPS. Step 3 is the formation of GDP-mannose by AlgA's GDP-mannose pyrophosphorylase (GMP) activity. AlgA is a bifunctional enzyme with guanosine 5'-diphospho-D-mannose pyrophosphorylase (GMP) activity (step 3) as well as PMI activity (step 1). The AlgA sequence shows high homology to GMP enzymes like ManC of E. coli. Step 4 requires AlgD, a GDP-mannose dehydrogenase (GMD), which uses NAD+ as a cofactor to form GDP-mannuronic acid, the primary sugar-nucleotide precursor of alginate.
Alginate polymer formation. The second half of the alginate pathway leads to extracellular polymer formation. The initial polymeric product of the alginate biosynthetic pathway appears to be polymannuronate (poly-M) (Chitnis and Ohman 1990). It is likely that polymerization and secretion across the inner membrane occurs simultaneously; there is no evidence for an isoprenoid lipid carrier. Alg8 bears structural homology to β-glycosyl transferases and thus appears to be the alginate polymerase. Glycosyl transferases catalyze the transfer of sugar residues from an activated donor substrate, often a nucleoside diphospho (NDP) sugar, onto an acceptor molecule, which can be the growing carbohydrate chain. Among the 26 families of NDP-sugar glycosyltransferases, Alg8 belongs to Family 2, which is a large group of processive and inverting glycosyl transferases, which is consistent with the β-1,4 linkages in alginate. A model for the membrane topology of Alg8, a glycosyl transferase (GT), was constructed using PhoA fusions (Fig. 4). This provided evidence for a large cytoplasmic loop containing the active domains predicted for β-GTs like Alg8 and 5 transmembrane (TM) domains, the first of which resembles a cleavable signal peptide. The C-terminal TM domain of Alg8 was critical for the polymerization reaction in vivo. Alanine substitution mutagenesis showed that all of the predicted active site residues in the widely spaced D, DxD, D, LxxRW motif were required for polymerization activity in vivo, and two of these substitutions also affected Alg8 protein stability as shown by Western blot (Oglesby, Jain et al. 2008).
A membrane topology model for Alg44 was also developed by a PhoA fusion analysis, and this showed a central TM domain and predicted an N-terminal membrane anchor (Fig. 5). An N-terminal PilZ domain in Alg44 for c-di-GMP binding, which is required for alginate synthesis, was localized to the cytoplasmic loop. The long periplasmic C-terminus of Alg44 contains a region similar to membrane fusion proteins (MFPs) of multi-drug efflux systems, which predicts the possibility of its interaction with another protein in this compartment. A Western blot analysis showed reduced levels of AlgE (the OM secretin) in the outer membrane of the alg44 mutant, whereas expression of Alg44 in trans restored AlgE. C-terminal truncations of Alg44 as small as 24 amino acids blocked alginate polymerization in vivo, indicating a critical role for this MFP domain. These studies suggest that Alg44 may act as a co-polymerase in concert with Alg8, the major GT, and that both inner membrane proteins are required in vivo for the polymerization reaction leading to alginate production (Oglesby, Jain et al. 2008).
Mutants defective in one of several periplasmic proteins (AlgKGX) for alginate transport release alginate fragments due to the activity of an alginate lyase (AlgL) in the periplasm, which cleaves the newly formed polymers (Jain and Ohman 2005). However, mutants defective in Alg8 or Alg44 do not secrete polymer or alginate fragments, suggesting that both of these membrane proteins are required for the polymerization reaction.
The algK gene encodes a 52-kDa protein that has a signal peptide characteristic of lipoproteins with an N-terminal cysteine on the mature protein. AlgK-PhoA fusion proteins are active, indicating that AlgK is periplasmic but probably membrane attached. Mutants with a nonpolar ΔalgK mutation are nonmucoid, indicating that AlgK is essential for polymer formation. AlgK– mutants secrete high concentrations of low-molecular weight uronic acids (Jain and Ohman 1998), which are fragments of polymerized alginate. In collaboration with structural biologist, Dr. Lynne Howell (Hospital for Sick Children & University of Toronto), we recently solved the structure of AlgK to 2.5Å using the Se-SAD technique (Keiski, Harwich et al. 2009). AlgK contains 22 α-helices that pack into a right-handed superhelix. The body of the superhelix is composed of 21 anti-parallel α-helices with residues 55-384 forming 9.5 tetratricopeptide (TPR)-like repeats, which are known to be protein-protein interaction motifs.
We also identified algL, which encodes an alginate-degrading lyase, and mapped its location to the algD operon (Schiller, Monday et al. 1993). AlgL has a signal sequence and localizes to the periplasm. Alginate lyases (alginases) are enzymes produced by a wide range of organisms that catalyze the degradation alginate polymers into unsaturated oligosaccharides, often as carbon sources. Although alginate-producing bacteria (Pseudomonas and Azotobacter) have a periplasmic-localized alginate lyase, they are unable to use alginate as a carbon source. Why an alginate degrading enzyme would be coexpressed with the biosynthetic enzymes was mysterious. To better understand the role of AlgL, we examined the effect of a nonpolar ΔalgL mutation in mucoid P. aeruginosa FRD1. Interestingly, such mutants could not be constructed, suggesting that it was a lethal mutation. However, by genetically replacing the algD operon's promoter with a trc promoter under IPTG control, an ΔalgL mutant could be constructed when the operon was silent. However, when algD operon expression was activated with IPTG, the bacteria lysed because the periplasm filled with alginate and the cells burst within a few hours (Jain and Ohman 2005). Thus, AlgL appears to be bi-functional with (a) a role in the periplasmic transport apparatus (i.e., ‘scaffold’) in mucoid P. aeruginosa like AlgG and AlgK, and (b) as a lyase that digests periplasmic alginates that remain after spontaneous disassembly of the scaffold. This study led to the development of a model for the alginate transport complex in the periplasm (Fig. 6), which explains the secretion of small alginate fragments when the scaffold is missing one of the components (Jain and Ohman 2005). Determining which proteins directly interact is currently being pursued
Polymer modification by C5-epimerization.
The proportion and distribution of β-D-mannuronic acid (M) and its C-5 epimer, α-L-guluronic acid (G), affect the properties of the polysaccharide. Instead of a repeating subunit structure, the monomers in P. aeruginosa alginate are arranged as random block-structures of poly-M or poly-MG, but not poly-GG. Alginates rich in G form strong, brittle gels, whereas those rich in M are weaker but more flexible. The gene algG encodes a polymer-level C-5 alginate epimerase, which converts M to G (Chitnis and Ohman 1990). AlgG localizes to the periplasm, suggesting that its substrate is poly-M. We developed an in vitro epimerase assay where recombinant AlgG incubated with poly-M was shown to produce poly-MG. The epimerization reaction is sensitive to acetylation, suggesting a link between AlgG's C5-epimerase activity and acetylation (Franklin, Chitnis et al. 1994). Like an ΔalgK mutant, an ΔalgG mutant of P. aeruginosa is nonmucoid but secretes dialyzable uronic acids, which are breakdown products of alginate by alginate lyase AlgL as shown by NMR (Jain, Franklin et al. 2003). We collaborated on a similar study of alginate producing P. fluorescens, which showed the same results (Gimmestad, Sletta et al. 2003). Thus, AlgG and AlgK protect new polymers from degradation by AlgL during transport through the periplasm and suggest that they are part of a multi-protein complex for translocation of alginate to the outer membrane.
Polymer modification by acetylation.
P. aeruginosa alginate is also modified by the addition of acetyl groups to the O-2 and/or O-3 positions of Mannuronates. We identified genes involved in alginate acetylation, called algF, algI and algJ (Franklin and Ohman 1993; Franklin and Ohman 1996). Interestingly, algI,J,F are not required for alginate polymerization or transport, but the alginate produced lacks O-acetyl groups (Franklin and Ohman 2002). Using alkaline phosphatase (AP) fusions to localize the proteins, AlgF and AlgJ were localized to the periplasm. The signal peptide was cleaved from AlgF but not from AlgJ, which remained tethered to the membrane. AlgI is an inner membrane protein with 7 transmembrane domains. These 3 proteins play a role in the transfer of acetyl groups, possibly from acetyl-coenzyme A ([H3C-OC-]-S-CoA) to M residues (Franklin and Ohman 2002) (Fig. 7).
The O acetylation of alginate plays an important role in the formation of microcolonies and biofilms (Nivens, Ohman et al. 2001). When strain FRD1 (a mucoid clinical isolate) and a nonmucoid derivative were grown in flow-cell biofilms, FRD1 produced microcolonies 6X thicker in depth than those of the nonmucoid strain, which formed a densely packed biofilm. Interestingly, an algJ mutant, which overproduced alginate lacking O-acetyl groups, could not even produce a biofilm. Thus, although alginate is not not required for P. aeruginosa biofilm development, it plays a vital role in the the formation of thick biofilms, but only if it is O-acetylated (Nivens, Ohman et al. 2001).
Alginate O-acetylation is also critical for the resistance of P. aeruginosa to opsonic phagocytosis (Pier, Coleman et al. 2001). Avoiding host defenses is required for the establishment and maintenance of chronic lung infections. When mucoid clinical isolate FRD1 (WT) was treated with complement+anti-alginate antibodies+human peripheral blood leukocytes (HPBLs), it was still nearly 100% resistant. However, an isogenic algJ mutant (FRD1153 algJ-), which appears just as mucoid, was nearly totally killed. Thus, mucoid Pseudomonas is highly resistant to antibody-dependent and -independent opsonic killing as long as alginate is acetylated (Pier, Coleman et al. 2001). Future studies are in progress on the O acetylation of alginate because it is so critical in the resistance of P. aeruginosa to host immune effectors.
Regulation of alginate biosynthesis.
Most enzymes for alginate biosynthesis are encoded by the algD operon (Chitnis and Ohman 1993), which is controlled by a complex hierarchy of regulators encoded by genes around its circular chromosome (Fig. 8). The alternative sigma factor σ22 encoded by algT (also called algU) is essential for alginate production (Flynn and Ohman 1988; Flynn and Ohman 1988). σ22 shows high sequence similarity (66% identity) to σE of E. coli, an ECF sigma factor that responds to envelope stress (DeVries and Ohman 1994). From a plasmid clone (6-8 copies/cell), algT activates alginate production in otherwise nonmucoid strains of P. aeruginosa and several other Pseudomonas species (Goldberg, Gorman et al. 1993). The algT gene is in an operon with 4 other genes called mucA-mucB-mucC-mucD. Mutations in mucA are typically responsible for the mucoid conversion observed in clinical isolates. MucA is an inner membrane protein with one transmembrane domain and appears to be the main regulator to post-translationally control σ22 activity (Mathee, McPherson et al. 1997). MucA acts as an anti-sigma factor that binds and sequesters σ22, thus affecting its ability to transcribe in vitro. MucB(AlgN) is also a negative regulator (Goldberg, Gorman et al. 1993) and a periplasmic protein that may bind the periplasmic domain of MucA (Mathee, McPherson et al. 1997). In addition, 3 regulators bind the algD promoter encoded by algB, algR and amrZ (Jain and Ohman 2004).
σ22 belongs to the extra-cytoplasmic function (ECF) subfamily of sigma factors, which often control stress responses. A well characterized ECF sigma is E. coli σE, which is co-transcribed with RseAB (which resemble MucAB) and together form an envelope-stress response system. An σE-RseAB complex sequesters the sigma factor until envelope stress (e.g., misfolded envelope proteins) signals DegS protease and YaeL/RseP protease to sequentially degrade the anti-sigma factor RseA in a process called regulated intramembrane proteolysis (RIP) to de-represses sigma factor activity. Given the potential similarity between σE-RseAB and σ22-MucAB, we asked whether σ22 in P. aeruginosa could be part of an extracytoplasmic stress response system. To identify stress stimuli that might cause transcriptional induction of the algD promoter (PalgD), we developed a plate bioassay, using a plasmid-encoded PalgD-cat, to provide chloramphenicol-resistance as a marker for PalgD activation. This revealed that antibiotics attacking peptidoglycan synthesis (e.g., D-cycloserine) strongly induce PalgD transcription in wild-type P. aeruginosa PAO1 (Wood, Leech et al. 2006). Such induction required known transcriptional regulators (i.e., σ22, AlgB and AlgR) and also AlgW protease, a homologue of DegS protease. A transcriptome analysis using microarrays verified that sublethal levels of D-cycloserine strongly activated expression of genes in the algD operon (Wood, Leech et al. 2006).
MucA is the anti-sigma factor of σ22, and we have recently shown that MucB interacts with MucA as part of this sigma-sequestration complex (Wood and Ohman 2009). We hypothesized that AlgW protease is activated by cell wall stress to cleave MucA and release σ22. To test this, we showed that when strain PAO1 was exposed to D-cycloserine, MucA was degraded within just 10 min and σ22 was activated, all of which was AlgW dependent (Wood and Ohman 2009). Also, a mutant analysis showed that YaeL protease cleaves MucA, but only after cleavage by AlgW. Another protease called MucD was shown to degrade the periplasmic cell wall stress signals. Microarray analyses were used to identify the genes of the σ22 regulon, which included gene products that contribute to recovery from cell wall stress (Wood and Ohman 2009). A model for two PalgD activation mechanisms is shown in Figure 11.
II. Proteases involved in P. aeruginosa Pathogenesis
Much of the pathogenesis of P. aeruginosa has been attributed to its ability to secrete a variety of toxic and degradative enzymes that include: Proteases (e.g., elastase, staphylolysin, aminopeptidase, alkaline protease, lysine-specific protease [also known as protease IV, PrpL, and LysC endopeptidase]) and Exotoxins (e.g., toxin A, phospholipase C1 & 2; lipase; exoenzymes S,T). P. aeruginosa has a large arsenal of other virulence factors that include biofilm formation, exopolysaccharides, type III secretion, adherence mechanisms and a complex quorum system to control virulence factor expression. Pathogenesis is often multifactorial, and the site of infection can affect which combination of virulence factors are important contributors to disease. We are studying 3 metalloproteases and their propeptides.
Elastase is encoded by the lasB gene (PA3724 at www.pseudomonas.com), and its expression is under the control of the LasRI and RhlRI quorum sensing systems. It is the most active and abundant of all the proteases produced by P. aeruginosa and is a classic virulence factor. Elastase accumulates in most nutrient growth media and is produced by P. aeruginosa during pulmonary infections where it damages pulmonary tissue. Elastase is described in the MEROPS database of proteases (http://merops.sanger.ac.uk/) as a member of the M4 family of neutral zinc-metalloproteases. The prototype protease of this large homologous group is Thermolysin of Bacillus thermoproteolyticus, and so elastase is also called Pseudolysin (Kessler and Ohman 2004). Although Elastase is found in the culture supernatant as a 33-kDa protein, the lasB gene actually encodes a 53.6-kDa precursor called pre-proElastase. The domain located between the signal peptide and the mature domain is an 18-kDa propeptide. The signal peptide is removed during transfer across the inner membrane. ProElastase -> Elastase processing occurs in the periplasm and is autocatalytic, so a substitution at H420 (= H223 in mature Elastase) blocks proteolytic activity and thus autoprocessing (McIver, Kessler et al. 1991). After the propeptide is cleaved from mature in the periplasm, they still form a non-covalent complex as shown by our long-time collaborator, Dr. Efrat Kessler, Tel-Aviv University. (Kessler and Safrin 1988).
The propeptide (Pro) of Elastase functions as an intramolecular chaperone, which is essential for the folding of Elastase in the periplasm (McIver, Kessler et al. 1995). The propeptide also acts as a potent inhibitor of Elastase activity, which is another common feature of propeptides. The periplasmic Elastase + propeptide complex is then an acceptable substrate for secretion by the Xcp type II secretion apparatus. (Kessler and Safrin 1988) (Fig. 12). The entire propeptide+enzyme complex is translocated across the outer membrane by the “piston-like” Xcp type-II secretion apparatus (Kessler, Safrin et al. 1997). Dissociation of the complex and Elastase-mediated degradation of the propeptide take place extracellularly. This pathway involves non-covalent interactions between the propeptide and mature domain, which are poorly understood. When lasB is expressed in E. coli, mature-active Elastase (33-kDa) is readily found in the periplasmic fraction but not in the supernatant because E. coli does not have an appropriate type II secretion apparatus (McIver, Kessler et al. 1995).
Staphylolysin (LasA protease) also has elastase activity and it's gene was discovered during our search for mutants defective in elastin degradation (Ohman, Cryz et al. 1980). The lasA gene encoding Staphylolysin was cloned by genetic complementation, and defined chromosomal lasA mutants have been constructed and characterized (Goldberg and Ohman 1987). Staphylolysin (aka, LasA protease) is classified as an M23a protease in the MEROPS database of proteases. The name Staphylolysin comes from the ability of this protease to rapidly lyse Staphylococcus aureus cells. Staphylolysin dramatically increases the elastin-degrading activity of Elastase and other proteases including human neutrophil elastase (Kessler, Safrin et al. 1997), and so Staphylolysin and Elastase work synergistically to enhance protein degradation and thus necrosis. We found that the preferred cleavage sites of Staphylolysin in elastin are Gly-Ala peptide bonds within Gly-Gly-Ala sequences surrounded by apolar sequences; Gly-Phe and Gly-Gly provide additional sites of cleavage in elastin (Kessler, Safrin et al. 1997). In 1996 we published the sequence of lasA which acknowledged its precursor structure and large propeptide (Gustin, Kessler et al. 1996). This showed that the lasA gene encodes a 45.6-kDa pre-proStaphylolysin, which is more than twice the size of the final, 20-kDa mature Staphylolysin. A classic N-terminal signal peptide was observed, and the propeptide-mature cleavage site was determined by N-terminal sequencing, which established its pre-pro-mature structure. The propeptide is actually longer than the mature protease, suggesting the importance of this propeptide. ProStaphylolysin is secreted into the culture medium by the Xcp type II secretion apparatus (Kessler, Safrin et al. 1998). ProStaphylolysin requires another enzyme for the removal of its propeptide, and processing is readily observed in the presence of extracellular P. aeruginosa proteases (Gustin, Kessler et al. 1996).
Aminopeptidase is encoded by pepB (PA2939) and not yet well studied. It is a Zn-metallo protease of the M28a family of proteases. It is an exopeptidase that removes amino acids from the N-termini of polypeptides in contrast to endopeptidases (e.g., Elastase and Staphylolysin) that cleave peptide bonds within polypeptide chains. Many bacterial species produce broad-specificity aminopeptidases that permit utilization of peptides supplied in the growth medium or in vivo. We have demonstrated that P. aeruginosa secretes an aminopeptidase that releases amino acids from a broad range of substrates (Cahan, Axelrad et al. 2001). Aminopeptidase sequentially removes amino acids from peptides that would be generated in vivo by other Pseudomonal proteases, which would allow amino acids to be taken up and metabolized. Thus, Aminopeptidase most likely provides a mechanism of nutrient acquisition from the host and so promotes the growth and proliferation of the organisms at sites of infection. The precursor of Aminopeptidase shows a typical Pre-Propeptide-Mature domain structure, where the 24 aa signal peptide (Pre) is removed when the protein is secreted as an inactive 58-kDa ProAminopeptidase.
* Citatations above are from this laboratory and can be found on our Publications webpage.