Redefining Antibiotics

“Hit hard and hit fast,” German scientist Paul Ehrlich told the International Congress of Medicine in 1913. Ehrlich’s battle strategy wasn’t for war but for combatting bacterial infections, developing “magic bullets” to kill bacteria. A therapia sterilisans magna he also called his concept: a drug that—with one single dose—could kill all of the microorganisms inside a person. Ehrlich’s ideas, even before the discovery of penicillin, were representative of how scientists at the time aimed to treat infections: the bacteria must die.

In the following decades, the idea of killing bacteria to treat infections remained at the backbone of microbiology, although new antibiotics could be divided into two classes: bacteriocidal drugs that killed bacteria and bacteriostatic ones that stopped them from reproducing. The lines between the two were often blurry, but that didn’t matter: antibiotics from both classes, by the middle of the century, were poised to eliminate bacterial infections as a health threat.

“The time has come to close the book on infectious disease,” said Surgeon General William Stewart in 1967. “We have basically wiped out infection in the United States.”

In retrospect, although, it’s not surprising to most microbiologists that this claim was soon proved wrong. If there’s one sure thing about bacteria, it’s that they evolve. So when a population of bacteria is exposed to a drug that kills them, the natural response of the community is to evolve. The bacteria that survive—due to a chance mutation—lead to a new, resistant strain. Some bacteria start producing molecules that directly shut off drugs; others build up heightened defenses that keep drugs from getting into the bacterial cells or immediately eject them out; and some alter the structures of key molecules so that drugs no longer bind to them.

“This mess we’re in with antibiotic resistance is not an accident,” says Spellberg. “It’s a predictable outcome of the way we have designed and used antibiotics.”

However, if Ehrlich’s “hit hard and hit fast” approach to fighting antibiotics isn’t working, there are other battle tactics that might: know your enemy, use deception, cut communication, and be sneaky.

“The idea is that maybe we shouldn’t be trying to kill them in the first place. Maybe we should treat infections without killing the bug; by passively starving it, by disarming it, or directly blocking the host targets,” says Spellberg.

Spellberg and other scientists and clinicians hypothesize that by developing drugs that let antibiotics remain alive, but render them incapable of causing disease in humans, the evolutionary pressure for them to evolve resistance will lessen. In his own laboratory, he’s tested such drugs on 20 successive generations of bacteria without seeing drug-resistant mutations.

“It’s a fundamental break from a hundred years of how people have thought about treating infections,” he says. “This isn’t just about new classes of antibiotics, it’s about redefining what an antibiotic is.”

Crowd Control

Alone, a single cell of Pseudonoma aeruginosa—the bacterium blamed for many hospital-acquired infections—can’t cause much damage to the human body. In fact, the bacteria won’t even produce virulence factors, the compounds that make it pathogenic to humans, if it doesn’t sense neighbors. However, add a few thousand other cells of P. aeruginosa, and suddenly the bacteria aren’t lone warriors; they’re a team. When they sense the presence of unique signaling molecules produced by their allies, the cells start making those virulence factors, ramping up to cause an infection.

“Bacteria evolved multicellularity,” says Bassler. “They’re so small that the only way“This mess we're in with antibiotic resistance is not an accident.”they get any bang for their buck is by coordinating activities with each other.”

Bassler and other scientists have spent the last two decades trying to decode the molecules that P. aeruginosa and other bacteria use to send each other messages, alerting neighbors to their presence through so-called “quorum sensing.” By developing ways to turn off these quorum-sensing pathways in bacteria, scientists including Bassler believe they can treat bacterial infections in humans.

Recently, for example, Bassler’s laboratory found a molecule called chlorolactone that blocked one of the quorum-sensing receptors in Chromobacterium violaceum, a bacterium found in water and soil. She wondered whether the chlorolactone—or similar chemicals—would also work to stop quorum sensing in the more medically important P. aeruginosa. So her group created 30 related molecules and began testing them in the bacteria. One of them, called metabromo-thiolactone (mBTL), bound to a quorum-sensing receptor and stopped the bacteria—even when in a group—from initiating behaviors, such as virulence factor production, that are synchronized through quorum sensing. The data appeared in a PNAS paper in October 2013 (3).

“What was surprising was that shutting down this one receptor really shuts down the whole system,” says Bassler.

With the successful results in isolated bacterial cultures, Bassler’s team began testing whether the drug would work against bacteria that were actively infecting human cells. So next, they cultured the bacteria with human lung cells. Once again, adding mBTL decreased the virulence of the bacteria.

Drugs such as mBTL that target quorum sensing, Bassler says, could be used to treat infections without killing the bacteria—simply by stopping their production of virulence proteins. Such an approach, she says, is less likely to lead to resistance because bacteria will remain alive.

However, Pete Greenberg, a microbiologist at the University of Washington who also studies quorum sensing—and is widely credited with launching the entire field—says such benchtop discoveries must be taken with a grain of salt.

“We seem to move forward a step and then backward a step,” says Greenberg. “The critical animal experiments that have actually been done are few and far between.”

Most compounds that target quorum sensing, he says, have only been tested as preventive measures for infections—and even then, they have only been successful in worms and occasional mice models. For a drug to be useful in treating human disease, it must be able to treat widespread infections inside the body. It’s not clear, Greenberg says, whether quorum sensing is as vital to bacteria once an infection has been established. Data from human patients, for example, have shown that bacteria in established infections naturally accumulate mutations in quorum-sensing genes, suggesting these pathways are no longer important.

“Everybody who gets into this field gets really excited, but then they realize as they learn the details that there’s a lot of fundamental science that we still need to do,” he says.

He’s still optimistic about the approach—particularly for nonhuman applications such as infection prevention in plants and livestock—but thinks the basic research needs to continue before claims of new, revolutionary antibiotics will be realistic.

Slow Starvation

In the past, antibiotics were often developed by screening libraries of chemical compounds for their ability to kill bacteria. Once a chemical was identified as potentially bactericidal, it could be further studied. Now, in many cases, the process happens in the reverse: once a scientist understands a process that’s key to bacteria functioning—like quorum sensing—then they develop ways to block it.

At Vanderbilt University, microbiologist Eric Skaar is taking that approach when it comes to the metabolism of metals by bacteria. Bacteria—as well as all living things—rely on metals including iron, zinc, calcium, and manganese to carry out basic cellular processes.

“If you want to kill or weaken a living thing, one of the simplest ways to do that is to take away its food,” says Skaar.

In 2008, Skaar surveyed all of the proteins that the human body produces at sites of infection. One protein that was up-regulated in comparison with the rest of the body was called calprotectin. Skaar’s team revealed that calprotectin was attempting to fight infections by binding to manganese and zinc, keeping the metals away from bacteria. He’s spent the 5 years since the discovery delving further into the details of how bacteria use metal and how drugs could be designed to keep them from getting metal.

“There’s no question that if you keep metal away from a test tube of bacteria, the bacteria won’t be able to grow and will dramatically change their gene expression,” Skaar says. “And bacteria will probably not evolve to not need metal at all because it’s so vital.”

When calprotectin lowers the availability of manganese and zinc to bacteria, Skaar’s laboratory showed in 2011, the bacteria don’t immediately die, but become more susceptible to natural regulation by the human immune system (4). In early 2013, Skaar published new data in PNAS on how calprotectin binds to manganese (5). The molecular details, he says, could help scientists design new ways to keep the metal away from bacteria. Now, he’s completing a new genetic survey, determining which pathways within bacteria rely on metals.

“This can tell us which proteins are required for growth and those themselves may be good drug targets,” Skaar explains.

Efforts related to nutrient sequestration—the idea of blocking bacterial growth by limiting metals—are targeted not only at developing molecules to hoard metals but at making drugs that block metal receptors within the bacteria.

“What we need to understand next is what the front line in this battle for metal is,” says Skaar. “When we identify a factor that can bind a metal, we then identify a bacterial protein that can take that metal away, and then there are host proteins that combat that bacterial protein.”

Like many studies on quorum sensing, work on nutrient sequestration has largely been conducted on isolated cells. The leap to humans has yet to be made.

Getting to Know the Enemy

Among bacteria that cause disease in humans, some fall into the group of Gram-positive bacteria—which have fewer, but thicker, outer layers surrounding each bacterial cell—or Gram-negative bacteria, which have an extra outer membrane. Generally, infections from Gram-negative bacteria are more drug resistant and harder to develop new drugs for.

“The main problem with Gram-negative bacteria is that antibiotics can’t get access to the inside of the cell,” says Harvard Medical School microbiologist Stephen Lory. “The Gram-negatives have this double membrane and from a chemical point of view, that is a challenge to penetrate.”

Moreover, he says, Gram-negative bacteria including P. aeruginosa, which Lory focuses his research on, often contain extra pumps designed to expel compounds that make it“If you talk to most experts who practice medicine, or to pharmaceutical companies or biotech, most of these kinds of new drugs aren't even on their radar.”through the membranes. Therefore, designing drugs that target genes or proteins inside the cells is extra challenging. However, that doesn’t worry Lory.

“My solution to this permeability problem is to completely ignore it,” he says. “Just target proteins on the surface or outside of the cell instead.”

Some of these outside proteins include those that other people have also homed in on: quorum-sensing molecules and pumps that take in nutrients. However, Lory’s targeting something else: the very machinery that the cell uses to build its outer membrane. Mammalian cells don’t use the same proteins, he says, so drugs targeting this pathway would be specific to Gram-negative bacteria.

However, his approach to finding proteins to target—and what led him to this pathway—has been a broad search strategy. In 2009, Lory’s laboratory studied all 262 proteins that they found on the outer membrane of P. aeruginosa and found 12 that are actively recognized by the human immune system, suggesting their usefulness as targets for vaccines (6). More recently, Lory has turned from such a protein-based screening to a genetic screen. By inserting transposons—mobile bits of DNA—into 30,000 random spots in the P. aeruginosa genome, he and his colleagues could figure out which genes were vital to the survival of the bacteria. The final list had 636 genes (7).

“Of those essential proteins, most of them function in the cytoplasm, so you end up with that same problem of getting through the membranes,” says Lory. However, with a targeted effort at matching up accessible proteins with those that are essential, he thinks there are drug targets yet to be discovered.

Scientists like Lory, Skaar, and Bassler are getting to know bacteria better through basic fundamental microbiology, genetics, and biochemistry studies. Each, if they wanted to, can recreate a Fleming-like breakthrough in a Petri dish: when they add their drug of choice—whether it targets quorum sensing, hoards away metals, or blocks membrane formation—colonies of bacteria will die, stop growing, or stop producing virulence factors. However, although Fleming’s observation led to a revolutionary drug, not every compound that effects bacteria on the laboratory bench will work in humans.

“If you talk to most experts who practice medicine, or to pharmaceutical companies or biotech, most of these kinds of new drugs aren’t even on their radar,” says Spellberg. “We’re not at that inflection point yet.”

Moving forward, he says, must involve a focus on basic research and a better understanding of bacteria, but also a push to move potential compounds toward animal and human trials.

“Science has become increasingly difficult. When you run through chemical libraries, the low-hanging fruit have already been plucked. We need to challenge the way we’ve done things.”