Summary of Research Projects

Clostridium difficile is an opportunistic pathogen that causes pseudomembranous colitis (PMC) in patients that are on antibiotics or chemotherapeutics or that have a physical condition that alters the normal microflora in the large intestine. PMC is a common problem among hospital and nursing home patients, and is potentially fatal due to extensive damage to intestinal mucosal cells.

In a healthy body, the normal microflora confers colonization resistance upon the intestine, preventing a large number of potential pathogens from establishing colonies. Antibiotic and anti-tumor chemotherapies destroy this protective mechanism, and often result in the growth of C. difficile and lead to the toxemias associated with growth of the bacterium.

Several virulence factors are associated with diseases caused by C. difficile. Two high molecular weight exotoxins, released during growth of vegetative cells, trigger the damage to the intestinal mucosa. Sporulation and native resistance to several antibiotics are important in allowing the bacterium to survive the initial antimicrobial treatments. Adhesion to mucosal surfaces aids in the establishment of niches in the colon, and growth substrate utilization is apparently a major determinant of colonization resistance. My research focuses primarily on a multifunctional virulence factor, the oligopeptide permease of C. difficile, which underlies both substrate utilization and adhesion properties of the bacterium. Students in my lab are also characterizing one of the native antibiotic resistances of the bacterium, the cycloserine resistance determinant.

C. difficile prefers peptides as a substrate, but peptides are not abundant in the normal colon. The preference for peptides could explain why C. difficile does not colonize the intestines of healthy people. When the normal microbiota is disturbed, peptide concentrations rise in the colon, as do the numbers of C. difficile. By understanding the mechanism that C. difficile uses to metabolize peptides, we will be able to understand in greater detail what controls its growth and, more importantly, how to prevent colonization in patients whose normal flora is impaired due to antimicrobial treatment. Because the intracellular metabolism of peptides is essentially identical to that of amino acids (which C. difficile does not use preferentially as growth substrate), it is highly likely that transport of peptides across the membrane is the important physiological event dictating the growth properties of the pathogen.

The primary mode of peptide transport in bacteria is the multisubunit ATP Binding Cassette (ABC) permease, which uses the hydrolysis of ATP to drive the transport of peptides that are 2 to 8 amino acids in length (Click on image to see Opp animation). The complex is composed of an external ligand binding protein, two hydrophobic membrane-spanning subunits, and two peripheral membrane proteins. The latter contain sites for the binding and hydrolysis of ATP, and are the most conserved of the subunit types. In gram-positive bacteria, the ligand binding protein is a lipoprotein that is covalently attached to the outer leaflet of the cytoplasmic membrane via a thioeither linkage between an N-terminal cysteine residue and a phospholipid molecule. This subunit doubles as an adhesin in several pathogenic bacteria.

Peptide permeases tend to show selectivity with respect ot substrate size, but not sequence. Thus, permeases specific for dipeptides (encoded by dpp operons) can be distinguished from oligopeptide permeases (opp), which transport peptides over a fairly broad range of sizes. Previously characterized di- and oligopeptide permease operons have a variety of gene orders, but in all cases the genes are oriented in the same direction with respect to each other and are all expressed from a single promoter. Thus, we were quite surprised to find two, divergently transcribed operons encoding the opp system of C. difficile. The unusual operon structure raises a number of interesting questions, in particular regarding the manner in which coordinated expression of the operons is achieved, and perhaps regulated. We are using RNase protection assays to study the regulation of transcription of the C. difficile opp genes, and have preliminary evidence that each of the two "mini-operons" has its own distinct regulatory pattern. We are also attempting to purify the ligand binding OppA protein in quantities sufficient for the isolation of antibodies. These antibodies will be useful in characterizing the adhesin function of the subunit.

Cycloserine is a D-alanine analog that inhibits several steps in the bacterial peptidoglycan pathway.

It enjoyed widespread use in the 1950s and 1960s, but since then, resistant bacteria have become more common, and potential toxic side effects have come to light. Cycloserine use as an antibiotic is now largely restricted to tuberculosis treatments. Recently, however, have appeared numerous reports of the cognitive-enhancing properties of cycloserine. This has led to increasing use of the drug for the treatment of neurologic disorders, including Alzheimer's disease, epilepsy, and schizophrenia.

C. difficile and several other clinically significant clostridia are resistant to cycloserine, by mechanisms that are not understood. Since potentially fatal intestinal disorders can result from the use of cycloserine, we have begun an investigation into the molecular basis for cycloserine resistance in the clostridia.

We have isolated a gene from C. difficile that confers high levels of resistance to cycloserine upon E. coli transformants (see dcs m/s). The gene product, a glutaminase-like protein, has been purified as a His-tagged fusion protein. The 27 kDa enzyme has been shown not to deaminate cycloserine, and most likely catalyzes a transamination reaction that cleaves the ring structure of the molecule.

 We are currently pursuing a structural characterization of the "cycloserinase" protein, in an effort to describe the features of the enzyme that allow for recognition and modification of cycloserine. It is hoped that these studies will be applicable to other cycloserine-binding proteins, such as the NMDA receptors that are modulated by cycloserine.

C. difficile dUTPase:

A screen for genes involved in pH homeostasis and protection from Na⁺ toxicity resulted in the isolation of a C. difficile gene that encodes a member of the family of dUTPases (see dut m/s). These enzymes are essential for viability of prokaryotes and eukaryotes, in that they maintain and control the critical levels of dUMP and dUTP in the cell, particularly during rapid growth. The dUTPases of tumor cells and several animal viruses have been targeted for chemotherapy, and we plan to pursue the design of drugs aimed at the clostridial enzyme as well.

3. difficile Nhe Na⁺/H⁺ exchanger:

Located adjacent to the opp locus of C. difficile is a gene that would encode a homologue of the Nhe-type Na⁺/H⁺ exchangers, which, until recently, were thought to be strictly eukaryotic. The function of the bacterial Nhe antiporters is unknown. The E. coli Nhe homologue was revealed only by whole genome studies. In fact, the nhe gene was not selected by the exhaustive screens for Na⁺/H⁺ antiporters, carried out by Padan, Krulwich and colleagues, which led to the isolation of the nha and cha antiporters of E. coli (see reprints, below).

In a collaboration with Etana Padan, we are beginning an investigation of the C. difficile and E. coli Nhe proteins. We are using PCR to amplify the nhe genes on fragments that can be cloned into His-tag expression vectors. We hope to be able to overexpress and purify the proteins in active forms that can be reconstituted and characterized.

In addition to improving our understanding of the basic physiological processes, studies of the bacterial Nhe proteins will add to our knowledge of the eukaryotic exchangers. The latter enzymes are of enormous importance in ion and pH homeostasis in eukaryotes, and are among the very first processes to be turned on during mitogenic transformations. The bacterial homologues should provide a simpler system for the analysis of the basic biochemistry of the transport proteins. In addition, because the clostridial enzyme may contribute to virulence in allowing the bacterium to survive the pH fluctuations that occur in the alimentary canal, it is an important topic of research.

We are assessing the use of purines as growth substrate by attempting to generate knock-out mutant strains lacking the gene for a major enzyme in the purinolytic pathway. The C. difficile xanthine dehydrogenase gene has been cloned and partially sequenced, and the appropriate suicide vectors have been constructed.

Extremely alkaliphilic bacteria synthesize ATP via a bioenergetic mechanism that apparently requires a) direct interactions between electron transport proteins and the ATP synthase, and b) a pH-sensor that regulates the interactions. Unusual amino acid sequence motifs, perhaps associated with these functions, have been found in the H+-translocating a and c subunits of the ATP synthases of two unrelated alkaliphilic Bacillus species, but not in other organisms. We are continuing our efforts to assess the importance of the unusual sequence features with respect to the adaptation to growth at high pH in general and to the mechanism of H⁺-translocation in specific. We are examining other extreme alkaliphiles, including non-Bacillus species, for the presence of these or other unusual motifs in their ATP synthases

At present, we have isolated four additional alkaliphilic Bacillus strains. The isolates, distinguishable based on colony and cell morphology, growth pH optima, and sporulation patterns were purified from pH 10.5 enrichment cultures initiated with local soil samples. Chromosomal DNA from one isolate (UA005) was used as template in PCR amplification of the region of the atp operon that encodes the sequence motifs of the a subunit of F_O-ATPase. The deduced amino acid sequence encoded by the cloned PCR product contains the a subunit lysine-218, indicative of the alkaliphile F_O motif. The entire UA005 a subunit domain is equally related to the B. alcalophilus and B. firmus OF4 domains. Preliminary 16S rDNA sequence comparisons support the placement of UA005 in the sixth Bacillus rRNA group.