Retracing Steps to Find New Antibiotics

When linezolid (Zyvox) received federal approval in early 2000, it was the first completely new antibiotic compound to reach the pharmaceutical market in 35 years. The synthetic compound even proved effective against methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant Enterococcus faecium (VRE) bacteria, for which no other line of defense existed. Its creator, New Jersey-based Pharmacia, sounded confident that few people would become resistant to the drug. It was not to be. Within months, patients infected with MRSA and VRE were not responding to linezolid.1,2 Such quick resistance epitomizes the dilemma of antibiotics development: Drugs cost hundreds of millions of dollars and take at least a decade to develop, and then become increasingly less effective. (see 'Renewing the Fight Against Bacteria,'). In the past, a classic screening approach, which tests in vitro the inhibitory effect of synthetic and natural compounds or extracts, led to the discovery of many current antibacterial drugs. Linezolid, for example, was developed in this way: thousands of compounds were reviewed to find one that kills bacteria. Today, this method yields few, novel, promising structures. The related method of developing second and third generation drugs based on existing pharmaceuticals is no longer considered an option because cross resistance reduces the effectiveness of macrolides (based on erythromycin), rifapentine (based on penicillin), and carbapenems (based on imipenim). Some pharmaceutical companies are turning to new approaches to find anti-biotics that will elude resistance. Many are mining the bacteria genomes mapped in recent years, exercising a target-based approach that screens for enzyme inhibitors that attack critical metabolic pathways never targeted by antibiotics. The process is still young, and pharmaceutical companies report trouble turning in vitro enzyme inhibitors into broad-spectrum antibacterial drugs. Some researchers, however, are returning to old ideas, once considered as possibilities but ignored when classic screening reliably yielded new compounds. Scientific advances and new technology have revealed at least two of these ideas, bacteriophages and antimicrobial peptides, as promising tools for fighting disease bacteria. Other research is opening possibilities for new antibiotics created from novel approaches, including cyclic peptides, that invade bacteria and then form nanotubes, which poke holes in the membrane (see '21st Century Antibiotics,'). Manipulating Nature At the University of California, San Diego, a research team led by dermatologist Richard Gallo and infectious disease specialist Victor Nizet has demonstrated that antimicrobial peptides in mammals act as naturally occurring antibiotics in vivo that can kill disease bacteria.3 The team's work with these peptides, called cathelicidins (caths), which are found in mammalian skin, lungs, intestines and white blood cells, suggested that the peptides provide a first line of defense against invading bacteria and keep infections in check until the immune system can mount a full-blown attack. "We observed that when laboratory mice were infected in their skin with strep, the skin cells and infiltrating white blood responded by producing a large amount of cath antimicrobial peptide," Nizet says. This suggested that the mouse and the strep bacteria could serve as a model to determine how important caths are to immune defense. Caths are one of a dozen or so known peptides with antimicrobial properties in humans. First discovered in the 1980s, they were essentially ignored by pharmaceutical companies primarily because they were superfluous as long as marketed antibiotics were doing the job. Concerns also existed about the therapeutic use of protein molecules and difficulties in large-scale production and purification. Today, these peptides are being reconsidered as an alternative therapeutic approach. Advances in genetic techniques have made antimicrobial peptides more accessible to study, too. Gallo's lab created a knockout mouse of an antimicrobial peptide. The Gallo and Nizet teams found that this knockout failed to produce a cath called CRAMP, which is homologous to the human LL37. When infected with strep bacteria, the mice lacking the caths developed fast-growing necrotic skin lesions that were larger and more persistent than those produced in normal mice. In companion studies, strep engineered to resist caths produced more severe infections in normal mice. Together, the research showed that the specific bacterial-killing cath activity was important to skin defense against infection; they were acting as natural antibiotics. Besides uncovering the cath's potential as an antibiotic, the research also reveals a layer of the immune system that was previously underappreciated, Gallo says. "Primary resistance to infection in the skin was thought to be through systems such as T-cells or B-cells. What we've now shown is that secreted peptides from a number of cell types represent a very early barrier to slow bacteria growth," Gallo says. A big remaining question is how these peptides, which have been around for thousands of years, have maintained their effectiveness against bacteria, Gallo says. Another question, adds Nizet, is why some people encounter potentially pathogenic bacteria and develop disease, while others do not. Finding Peptide Antibiotics Related research at North Carolina State University recently found a new type of peptide antibiotic in the mast cells of fish. The research, funded by Sea Grant, discovered a new site at which fish express peptide antibiotics; the finding is important because it seems to answer the question of the biological function of mast cells—the most common tissue immune cell found in fish and other vertebrates, including humans, Nizet says. These peptides, called piscidins, may also hold promise as templates to develop new therapeutics against bacterial infection. A number of features make piscidins potentially attractive, says lead investigator, immunologist and aquatic medicine specialist Ed Noga. Piscidins are small, simple-structured, linear peptides, and evidence shows that they can function under high salt concentrations. "Most other peptide antibiotics are less promising for a number of reasons, especially toxicity and low activity under physiologic conditions," he says. The next step, he adds, is to identify antibiotic-containing mast cells in other animals, including humans. Georgetown University immunologist Michael Zasloff encourages the renewed pursuit of antimicrobial peptides for antibiotic purposes. Zasloff, in a recent paper, writes that laboratory and clinical studies "confirm emergence of resistance against antimicrobial peptides is less probable than observed for conventional antibiotics," providing incentive to develop both natural and man-made antimicrobial peptides.4 Besides resistance, the other major barrier to new antibiotics is the long and often expensive process involved in making new drugs. But Stanford University genetic engineer Chaitan Khosla has found a way to force bacteria to churn out antibiotics. Khosla and colleagues manipulated the biosynthetic machinery of Escherichia coli in such a way that it manufactured the antibiotic erythromycin.5 Normally, erythromycin is only available through a slow and expensive fermentation process using the soil bacterium Saccharopolyspora erythraea. (It takes S. erythraea 4 hours to double in number while E. coli doubles in 20 minutes.) But this new process is not just a production tool, Khosla notes; it also allows rapid acceleration of the way the biosynthetic pathways can be engineered to make new types of antibiotics. "This means, at least theoretically, that we can make slightly different forms of erythromycin easily," and therefore elude resistance, Khosla says. Now that it has developed a practical way to ferment E. coli to make reagent quantities of erythromycin, Khosla's group is developing practical ways to make atomic-scale manipulation of the drug a highly predictable practice, he says. "This new technology can help reduce development costs because it allows you to make modular-based variations of erythromycin, many of which might have been impractical to make by synthetic approaches alone. And then it reduces cost at the level of manufacturing because fewer steps are required to make quantities of the drug," Khosla says. He estimates that it will be about five more years before marketable drugs will be based on this technology. Jennifer Fisher Wilson (jfwilson@snip.net)is a contributing editor. References 1. S. Tsiodras et al., "Linezolid resistance in a clinical isolate of Staphylococcus aureus," Lancet, 358:207, July 21, 2001. 2. R. Gonzales et al., "Infections due to vancomycin-resistant Enterococcus faecium resistant to linezolid," Lancet, 357:1179, April 14, 2001. 3. V. Nizet et al., "Innate antimicrobial peptide protects the skin from invasive bacterial infection," Nature, 414:454-7, November 22, 2001. 4. M. Zasloff, "Antimicrobial peptides of multicellular organisms," Nature, 415:389-95, Jan. 24, 2002. 5. B.A. Pfeifer et al., "Biosynthesis of complex polyketides in a metabolically engineered strain of E. coli," Science 291:1790-2, March 2, 2001. Bacteriophages: The Comeback Kids IN FOCUS | Nicole Johnston In an era of 'superbugs' that are resistant to some of the most powerful antibiotics, scientists are searching for new approaches to kill bacteria. Now, a team of researchers at Rockefeller University (New York, NY) has discovered a highly potent alternative to antbiotics that viruses, known as bacteriophages, have already been using for millions of years to burst open and destroy bacterial cells. Bacteriophages (or phages) are viruses that only infect bacteria, and cannot infect humans or cause disease. Nearly all bacteria play host to a specific kind of phage. Since these phages are highly discriminating, they only invade particular bacterial species. But when it is time to burst their way out of their host and infect neighboring bacterial cells, phages use a rapid-fire approach, killing bacteria in the process. Taking advantage of this strategy, Rockefeller University scientists, led by microbiologist Vincent A. Fischetti, isolated the phage-manufactured weapons, known as lytic enzymes or "lysins," that viruses use to destroy bacteria. These enzymes are so lethal to their target bacterial species, that bacteria are destroyed almost instantly, says Fischetti. The Rockefeller team isolated lysins that are specific for either group A streptococci or Streptococcus pneumoniae, and found that, in vitro, tiny quantities could kill at least 10 million bacterial cells in five seconds. In mice heavily colonized with either organism, the same amount of corresponding lysin eradicated the organism from the nose and throat, the normal reservoir for these organisms.1,2 So far, they haven't been able to find any bacteria resistant to the enzymes. "Resistance appears to be a rare event," Fischetti says. "Phage do not want to get trapped inside a bacterium, so they devised a system to get around resistance." Furthermore, an adverse immune response is not anticipated given the tiny quantity of enzyme required to eradicate the target organism from mucous membranes. The team's work has been well received. "Vincent Fischetti has pioneered the use of phage enzymes into human therapeutic settings," says microbiologist Alexander Sulakvelidze, University of Maryland, and vice president for research and development at Intralytix, Inc., a Baltimore-based biotechnology company specializing in phage therapeutic development. Fischetti says he believes that these enzymes won't replace antibiotics but could prevent infection from occurring in the first place. "The organism you carry is usually the organism that will ultimately cause infection," he says. "No technology was available to control them until now. You had to wait for infection to occur before treatment." This would reduce the need for antibiotics and, thus, the incidence of antibiotic resistance. And since nearly all bacteria play host to a corresponding phage, enzymes could be developed for nearly all pathogens. These enzymes also could be used as a powerful adjunct to antibiotics, digesting their way through impenetrable biofilms so that antibiotics can get in and do their work, he says. And given their discriminating target specificity, Fischetti says, they could be used for the development of rapid diagnostic tests for infectious organisms. The highly stable enzymes could be administered in a nasal spray or in a lyophilized form. Lysins wouldn't cause the adverse side effects associated with antibiotic therapy; such effects stem from the broad-spectrum activity of antibiotics. "These phage enzymes may be nature's 'smart bombs' in that they target only the disease bacteria without any collateral damage; that is, affecting the surrounding normal bacteria," explains Fischetti. Phage lytic enzymes have been used in the laboratory for years, but their therapeutic potential has been largely overlooked until now. Microbiologist Ry Young, Texas A&M University in College Station, Texas, says it is about time. "I think it's going to be a really terrific strategy... and I like it because it calls attention to phage biology that has been inappropriately ignored for too long," he says. "Natural biocontrol of bacteria is by their predators and their predators are bacteriophage." References 1. D. Nelson et al., "Prevention and elimination of upper respiratory colonization of mice by group A streptococci by using a bacteriophage lytic enzyme," Proceedings of the National Academy of Sciences, 98:4107-12, 2001. 2. J.M. Loeffler et al., "Rapid killing of Streptococcus pneumoniae with a bacteriophage cell wall hydrolase," Science, 294:2170-2, 2001. 3. V.A. Fischetti, "Phage antibacterials make a comeback," Nature Biotechnology, 19:734-5, 2001. Nicole Johnston (johnstnj@mcmaster.ca) is a freelance writer in Hamilton, Ontario.

Retracing Steps to Find New Antibiotics

Interested in reading more?

Become a Member of