A. Maureen Rouhi

C&EN Washington

Every 10 seconds the sound of a bell punctuated the talk of one speaker at a recent briefing on infectious diseases at the National Press Club in Washington, D.C. Consumer advocate Ralph Nader was making the point that a person dies of tuberculosis every 10 seconds somewhere in the world. And he was calling on the U.S. "to take the lead in fighting this scourge."

A male patient with chronic tuberculosis in León, Nicaragua.

After antibiotics to treat tuberculosis became widely available in the 1950s, it was believed that the disease would eventually be eliminated. But Mycobacterium tuberculosis, the organism that causes the disease, has proven to be very resilient. Tuberculosis continues to be a public health threat, to the extent that the World Health Organization (WHO) has declared tuberculosis a global public health emergency. WHO's efforts to control and publicize the epidemic and advocacy by people such as Nader--as well as stories about Russian inmates languishing in overcrowded jails where the disease is efficiently propagated--are helping to focus attention on a problem that needs to be solved globally if people are to be spared locally.

But even before these recent clarion calls, a sharp rise in the incidence of tuberculosis in the U.S. between 1985 and 1992 already had jolted public health officials from a complacency brought about by the previous decades' consistent decline in tuberculosis cases and tuberculosis death rates. The public health systems for dealing with tuberculosis had been allowed to deteriorate. Now an infrastructure to deal with outbreaks is again in place, and research aimed at a better understanding of the disease and development of new therapies is thriving. But eradication of tuberculosis will eventually boil down to new drugs and new vaccines, the sorts of things only drug companies can make. Whether enough companies will join the battle is not clear.

A woman is observed taking her tuberculosis drugs in Dhaka, Bangladesh. [Photos by Jad Davenport/WHO Global Tuberculosis Program]

Rebuilding the public health infrastructure to deal with tuberculosis has come at a steep price, as the experience of New York City shows. A tuberculosis epidemic swept the city during a 43-month period between 1991 and 1994, with the number of cases reaching 4,000. Now everything is under control, but "we spent a billion dollars to rebuild the infrastructure," says Barry N. Kreiswirth, director of the Tuberculosis Center at the nonprofit Public Health Research Institute (PHRI), in New York City.

Kreiswirth says the money went for building negative-air-pressure rooms in a city jail to isolate people with the disease. As is happening in Russia, a lot of the tuberculosis in the New York outbreak was transmitted in jails, he says. Most hospitals also had to install isolation rooms. "The infrastructure that had been there historically had been dismantled, because politicians predicted in the 1970s that tuberculosis would be eradicated from North America by 2010," he adds.

Money also was spent on directly observed therapy, short course--DOTS, as it's called--a public health strategy advocated by WHO that has been adopted by 102 countries, including the U.S. With this strategy, when an infectious case is detected, health and community workers and volunteers are mobilized to observe the patient swallowing the correct dosage of drugs on a daily basis and to document that the patient has been cured.

Taking the lead in maintaining the public health infrastructure that deals with tuberculosis is the Centers for Disease Control & Prevention (CDC), Atlanta. About 5% of the center's resources are used to support state and local tuberculosis control programs, says Thomas M. Shinnick, chief of CDC's Tuberculosis & Mycobacteriological Branch. In 1998, that amounted to $144 million for tuberculosis control. CDC also assists local authorities during outbreaks and conducts epidemiological studies.

Tracking tuberculosis is being enhanced by the use of molecular biology tools. So-called molecular epidemiology using DNA fingerprinting is uncovering information that could not be obtained before and sometimes shattering long-held assumptions about the disease.

[Armando Waak/Pan American Health Organization]

One such assumption has been that most tuberculosis cases in U.S. cities are due to reactivated latent infection, says Peter M. Small, an assistant professor in the Division of Infectious Diseases & Geographic Medicine at Stanford University Medical Center, and principal investigator at the Stanford Center for Tuberculosis Research.

Among people infected with M. tuberculosis, only about 5% will have primary active tuberculosis--that is, they will manifest the disease within a few years after infection. About 90%, although infected, will never have tuberculosis. That condition is called latent tuberculosis. And about 5% of infected people will be fine for decades, then become sick long after they were infected. That condition is called reactivation tuberculosis.

[University of Stellenbosch photo]

By analyzing DNA fingerprints of bacteria from people with tuberculosis in San Francisco, researchers at the Stanford center have shown that 30% of the cases are due to recently transmitted disease. "This finding changed the focus of tuberculosis control in San Francisco," Small says. "It pointed to a new need to identify and interrupt transmission. That need would not have been apparent otherwise, because we had assumed that tuberculosis occurred mostly in people who had been infected in the distant past."

Investigators at Johns Hopkins School of Hygiene & Public Health and School of Medicine, the Baltimore City Health Department, and the Maryland Department of Health & Mental Hygiene recently reached a similar conclusion following a study to determine, through DNA fingerprinting, the patterns of tuberculosis transmission in Baltimore. Until the late 1970s, the city had one of the highest rates of tuberculosis among cities in the U.S. An aggressive DOTS program introduced in 1978 has turned the city around, reducing tuberculosis rates by 58% between 1981 and 1996.

The study found matchings of M. tuberculosis strains, indicating recent transmissions, says one of the investigators, William R. Bishai, an assistant professor in the Center for Tuberculosis Research at Johns Hopkins School of Hygiene & Public Health.

"Even if you do DOTS and tuberculosis control by the book, there are still pockets of recent transmission that continue to occur," Bishai says. "That finding suggests that even with the most effective tool that we have now, eradicating tuberculosis may not be achievable."

DNA fingerprinting also helped establish "the exquisite susceptibility to tuberculosis of people with HIV," the virus that causes AIDS, says Small. That finding emphasized the importance of identifying infectious cases of tuberculosis among such patients and treating them promptly.

The New York City epidemic revealed how devastating the HIV-tuberculosis combination can be. It was bad enough that the tuberculosis outbreak involved a highly multi-drug-resistant (MDR) strain. This strain, called strain W, is resistant to all the first-line drugs--isoniazid, rifampin, ethambutol, and pyrazinamide--as well as to one second-line drug, kanamycin. (An MDR strain is one that is resistant to at least isoniazid and rifampin.)

"It was also bad luck that the tuberculosis wards in many hospitals happened to be juxtaposed to the HIV wards," Kreiswirth explains. With very weak immune systems, the people with HIV who were infected with strain W developed a form of the disease that could be treated only with the more toxic second-line agents. "That's why the outbreak had such a high mortality," he says.

At the PHRI Tuberculosis Center, Kreiswirth and others also are using DNA fingerprinting to understand how tuberculosis spreads globally. One of their projects is DNA fingerprinting of isolates from Russia. "We're seeing commonalities among strains from Russia and Asia, which suggest that a large family of tuberculosis strains evolved from the Asia-Russia region," Kreiswirth says. "When you start looking at strains, you can start understanding the history of the migration of peoples. Because people bring their tuberculosis strains with them."

Research back on track

Patrick J. Brennan, a chemist by training and a professor of microbiology at Colorado State University, Fort Collins, has been engaged in mycobacterial research since his postdoctoral work in Dublin, Ireland, in 1965. He has seen the crest and troughs of support for tuberculosis research. Now is "a wonderful time to be working in this area," he says. "Anybody with a good idea in tuberculosis research is now funded. It's probably one of the most active and most dynamic research areas in all of infectious diseases."

Funding for research has increased from the doldrum years of the 1970s and '80s, when very little attention was paid to tuberculosis. According to Ann M. Ginsberg, program officer for tuberculosis, leprosy, and other mycobacterial diseases at the National Institutes of Health's National Institute of Allergy & Infectious Diseases, funding from NIAID alone was $37 million in 1998, compared with only $4 million in 1991. And NIAID's $37 million is only part of about $60 million that NIH as a whole made available in 1998 for tuberculosis.

Sidebar: Tuberculosis facts and figures

The bulk of the basic work supported by NIAID is directed to understanding the biology and molecular genetics ofM. tuberculosis, host immune responses, and bacteria-host interactions; developing new vaccines, diagnostics, and drugs; and improving molecular tools for epidemiological research. NIAID is funding more directly applied research too, through contract programs.

One of these is the Tuberculosis Research Materials & Vaccine Testing Contract at Colorado State. According to John T. Belisle, an assistant professor of microbiology and one of the program's three principal investigators, the contract began in 1992 at the height of the tuberculosis crisis in the U.S.

As NIH geared up to bolster tuberculosis research, it also realized that it was difficult to attract enough talent because of the problems associated with the growing and handling of the organism and that it needed to provide support for obtaining bacteria and bacterial products, Belisle explains.

"That's how the contract for research materials originated," Belisle says. Through this program, researchers around the world have access to high-quality standardized research reagents, from whole cells to various products derived from M. tuberculosis, including DNA, proteins, lipids, and antibodies. Research associates working with Belisle prepare the materials in a biosafety level-3 facility.

In 1997, vaccine testing was added to the contract. Ian M. Orme, a professor of microbiology at Colorado State and principal investigator for the vaccine-testing component of the contract, says he suggested the service. Very few places are equipped with the biosafety level-3 facilities required to test tuberculosis vaccines on animals, he says. Before NIH made this service available, people with ideas but with no means to test them "just had to talk nicely to me" and the few others who had the facilities and were willing to test for others on the side, he explains. Now anyone in the world with a vaccine that passes an NIH screening committee's scrutiny can get the material tested in the Colorado facility.

Another component of this contract program is development of skin-test reagents, for which Brennan is the principal investigator.

Another NIAID contract is the Tuberculosis Prevention & Control Research Unit at Case Western Reserve University, Cleveland, directed by Jerrold J. Ellner, a professor of medicine and pathology. The unit applies advances in basic immunology and microbiology to control of tuberculosis in endemic areas. Through epidemiological research and clinical trials in places as different as Brazil, Uganda, and New York City, the multidisciplinary program--involving 10 institutions and about 50 investigators--hopes to identify correlates of protective immunity as well as microbiological markers that can be used to guide the development of new drugs and vaccines.

Genome sequence

Money, materials, and facilities are in place. Last year, another weapon in the war against tuberculosis came into play--the complete genome sequence of M. tuberculosis [Nature, 393, 537 (1998)]. Armed with the organism's blueprint for life, researchers now can attack the disease with more firepower than ever before.

This milestone in tuberculosis research comes more than a century after March 24, 1882, when the German microbiologist Robert Koch announced that the disease is caused by rod-shaped bacteria that can be seen with a microscope when stained in a special way. Since then, tuberculosis has claimed at least 200 million lives, while scientists have been struggling to explain why M. tuberculosis is such a successful pathogen.

The most obvious reason is that people get infected simply by breathing. Infection can occur anywhere.

In Buenos Aires, for example, bus travel could be responsible for up to 30% of the new infections, according to Carlos Castillo-Chavez, a professor of biomathematics at Cornell University. Working with researchers at the University of Belgrano, Buenos Aires, he has been developing mathematical models to estimate the effect of public transportation on transmission of tuberculosis.

Castillo-Chavez believes their findings will apply also to other places where people use public transport extensively. But more important, he says, is that infection also occurs in unexpected places such as airplanes, bars, schools, and churches. "Tuberculosis is widely viewed as a disease of the poor and homeless people, but studies show it happens in all sorts of places."

Sidebar: A coat of elaborate construction

So it's very easy for M. tuberculosis to penetrate its host. What it does after it secures a foothold in the lung is a puzzle that researchers are still trying to solve. The genome sequence is changing how that work gets done, at the very least speeding it up.

Unlike other infectious agents such asStaphylococcus aureus, M. tuberculosis does not produce a toxin. Its pathogenicity arises from strategies it has evolved to survive in its host, including a slow growth cycle (dividing every 24 hours, compared with every 20 minutes for Escherichia coli), a complex cell envelope, the ability to colonize macrophages, and the ability to remain quiescent and then reactivate decades later. These characteristics not only complicate diagnosis and treatment, they limit the pace of research.

Consider the effort that has gone into elucidating the cell wall of M. tuberculosis. Brennan and his collaborators have made major contributions to the chemical definition of this structure, how it is held together, and how it is synthesized. To do that, over a period of many years, they have had to figure out pathways, develop assays, isolate and purify enzymes, sequence genes, and express them in E. coli.

"That's not necessary nowadays," says Brennan. "We can predict the actual enzymes from the genome and just go ahead and express them. Everything is so much easier, so much more efficient, so much faster now."

John D. McKinney, an assistant professor and head of the Laboratory of Infection Biology at Rockefeller University, New York City, compares the effort and resources he mustered as a postdoctoral researcher in trying to clone genes encoding enzymes that are potential drug targets to what's required now.

The two enzymes, isocitrate lyase and malate synthase, are involved in a pathway called the glyoxylate shunt. Cloning and sequencing the isocitrate lyase gene from M. tuberculosis before the genome sequence came out took several months. For malate synthase, "we simply scanned the genome sequence, and we were able to clone it in a day," McKinney says.

The genome sequence also revealed how lucky McKinney was to choose the pathway involving these two enzymes. McKinney is interested in the persistence of M. tuberculosis, which seems to be related to its ability to live on fatty acids with the help of two pathways, the glyoxylate shunt and the -oxidation pathway. Testing the hypothesis requires mutants in which genes of either pathway are knocked out. Because the glyoxylate shunt does not exist in vertebrates, McKinney decided to target that pathway first.

As it turns out, the genome sequence shows an enormous redundancy in the genes of the -oxidation pathway. "We could not have foretold that," McKinney says. "Had we begun by knocking out any of the genes that had so many copies, we wouldn't have seen the phenotype we were after. We would be groaning now as we look at the sequence."

The preponderance of enzymes for lipid metabolism is one of the most remarkable revelations of the genome sequence, says Clifton E. Barry III, a chemist and chief of the Tuberculosis Research Section at NIAID's Laboratory of Host Defenses. His group was involved in annotating the sequence, getting it in pieces as it was being worked out by researchers in Europe.

"For the longest time," he says, "I kept thinking that I had things wrong or that [the European researchers] had duplicated things, because there were too many enzymes involved in lipid biosynthesis." But it wasn't a mistake. M. tuberculosis produces about 250 distinct enzymes involved in fatty acid metabolism, compared with only about 50 for E. coli.

"In some ways you might have predicted such abundance," Barry says. "I like to say that M. tuberculosis is the chemist's bug. It produces lots of fancy natural products and lipids. But this kind of complexity was considerably more than I thought."

As a result of this finding, an idea that's now getting much attention is thatM. tuberculosis probably lives primarily on lipids in its host, Barry says. In terms of drug design, "we should be hitting lipid degradative enzymes instead of lipid biosynthetic enzymes, because the degraded lipids are what they live on."

The genome sequence has spawned other hypotheses. One seeks to explain why about 10% of the genome is devoted to two very repetitive protein families of acidic glycine-rich proteins. The biological functions of these proteins are unknown, but researchers believe they have immunological relevance, either by giving the organism an enormous ability for camouflage by allowing it to easily modify the antigens it presents to the host immune response or by interfering with the host's ability to process antigens.

Another genome-derived hypothesis is that M. tuberculosis does not have an imperative need for oxygen to survive, as has been widely assumed.

"If you look at any standard microbiology text, M. tuberculosis is described as an obligate aerobe," Barry says. "But the genome shows that it has all the genes for microaerophilic [limited oxygen] or anaerobic metabolism. They may, in fact, live most of their lives microaerophilically, but the way we grow them in the lab is not like that at all. We grow them aerobically on glucose, even though there are plenty of suggestions in the old literature that the artificial lab medium has no bearing on how M. tuberculosis grows in vivo."

The genome sequence is also leading scientists in new directions as they begin to comprehend the biology of M. tuberculosis from its genetic blueprint.

"When we finished the work," Barry tells C&EN, "I went around and told everybody: 'We should stop the experimental work for six months while we think about what this information means. Just sit in front of your computer and just think about it.' That's how profoundly it is changing research."

And Valerie Mizrahi, the head of the Molecular Biology Unit at the South African Institute for Medical Research, Johannesburg, says, "As a scientist working in a developing country, I feel that the genome sequence has leveled the playing field, because ideas rather than high-throughput technical capacity are now the most important criteria for future success."

What was published last year is the genome sequence of the H37Rv strain ofM. tuberculosis. The multi-institutional and primarily European effort was supported by Wellcome Trust, a British charitable institution. H37Rv is a well-characterized strain that was isolated in 1905 and that has been used extensively in research. It has full virulence in animal models, but there are doubts whether it actually causes disease in humans.

Also soon to be published is the genome sequence of the strain called CDC1551. This project is being undertaken by the Institute for Genetic Research (TIGR), Rockville, Md., with funding from NIH. According to CDC's Shinnick, this strain causes classical tuberculosis in humans and is transmitted more efficiently than other strains involved in outbreaks. The sequence is complete and much of it can be accessed at TIGR's web site (http://www.tigr.org/tdb/mdb/mdb.html#progress).

Drug targets

At present, the best therapy for tuberculosis is several drugs taken over a period of at least six months. However, many tuberculosis patients won't adhere to such a long-term, multidrug course of treatment. With poor compliance, treatment fails, leading to relapses or, worse, evolution of MDR tuberculosis.

Sidebar: Long therapy makes patients drop out

In the search for new drugs, "therapy that would cure tuberculosis in two months would be a major leap forward," says Ken Duncan, a research biologist at Glaxo Wellcome Research & Development, Stevenage, England. "The only way to make that step forward is by understanding more about the fundamental biology of the organism."

In the meantime, the growing number of cases of MDR tuberculosis worldwide is of utmost concern. As even more strains evolve that are not susceptible to the drugs available now, tuberculosis could quickly become incurable.

"We're trying to solve the MDR tuberculosis problem with innovative drug targeting," says Colorado State's Brennan. Specifically, his research, which is funded largely by the National Cooperative Drug Discovery Group within NIH, is targeting the bacterium's unusual cell wall.

Brennan's work on the biosynthesis of the M. tuberculosis cell wall has led to assays and screens to test drug candidates. His group is collaborating with the pharmaceutical company SmithKline Beecham, which has converted some of its assays into high-throughput screens for combinatorial libraries and has come up with a number of hits that it is pursuing.

"We're also taking a targeted approach, by synthesizing transition-state analogs and competitive inhibitors," Brennan says. At the same time, his group has a program to test the hits in SmithKline Beecham's screens against live M. tuberculosis cells, as well as in animal and macrophage models.

At NIAID, research in Barry's lab also has focused on the cell wall, particularly the cell wall lipids called mycolic acids. These extremely long fatty acids form a broad family of more than 500 closely related structures and comprise about 30% of the dry weight of M. tuberculosis, Barry says.

Recently, Barry and coworkers at Baylor College of Medicine, Houston, identified an enzyme in mycolic acid biosynthesis that is targeted by isoniazid, the most widely used antituberculosis drug. A spin-off from that work has been a relatively fast assay for activity against mycolic acid synthesis.

"We observed that when we treat M. tuberculosis with the right concentration of isoniazid, synthesis of certain proteins is upregulated," Barry says. "We've taken the promoter that's driving expression of those genes and fused it to a luciferase. So now we have a strain of M. tuberculosis that produces light when treated with isoniazid or anything that inhibits mycolic acid synthesis."

This reporter strain gives a response within 30 minutes of treatment, compared with three to four weeks for regular strains in a classical growth-inhibitory assay, because of the extremely slow growth of M. tuberculosis. In collaboration with the Princeton, N.J.-based company Pharmacopeia, Barry's lab is now using a high-throughput version of the assay to screen Pharmacopeia's drug discovery library for potential antituberculosis agents.

Meanwhile, a different track has developed in Barry's lab since the genome sequence became known. "We've been struggling with how to integrate that information into drug development," Barry says. The solution seems to be to find a "general paradigm" for how to move to the next generation of chemotherapy against an organism, starting with a drug that works (or an active lead compound) and the genome of the target.

Following this direction, Barry's group is harnessing DNA microarray technology for drug development. Instead of screening one target against millions of potentially inhibitory molecules, the whole spectrum of possible targets offered by the genome is tested with one compound at a time.

"We look for something that's upregulated and design a reporter strain, as we did for isoniazid," Barry explains. "Then we can make a library of compounds based on the original and screen it very quickly and efficiently. The advantage is that you'll be screening against specific targets but still in the context of the whole cell instead of a single enzyme. Lots of brilliant inhibitors of M. tuberculosis enzymes are out there, but in general they don't work against the whole organism, in part because of the elaborate cell wall that acts as a barrier for many compounds."

This whole-genome approach to drug discovery jibes with the notion that defeating M. tuberculosis might be more feasible by interference of pathways other than those involved in cell wall biosynthesis. Researchers believe the mechanisms by which the organism interacts with its host also could lead to potential vaccine and drug targets. But those mechanisms are still poorly understood.

When M. tuberculosis enters the lung, it is engulfed by a macrophage. Like any other foreign particle, the organism goes into a compartment of the macrophage called the phagosome. Ordinarily, the phagosome fuses with another compartment in the macrophage, called the lysosome, which contains chemicals and enzymes that degrade the contents of the phagosome. This fusion does not occur with M. tuberculosis, and it survives inside a cocoon in the very cell that should have spelled its doom.

As the bacterium multiplies, the numbers eventually become so overwhelming that the macrophage dies, releasing bacteria to be taken up by more macrophages. At two to three weeks into the infection, enough macrophages are infected that they start recruiting other macrophages and T-cells to the site. The congregation of cells around the infected macrophage forms a granuloma.

In 90% of cases, the bacteria become latent or inactive but remain for the rest of the person's life. Often the infected person is unaware of the war raging in the lung. In about 5% of cases, however, the people don't get a protective immune response. The bacteria keep multiplying and start spreading to other parts of the lung, producing primary active tuberculosis within months of the initial infection. In another 5% of cases, the protective immune response that induced the latent state somehow disappears decades later. Now roused, the bacteria begin to proliferate, causing reactivation tuberculosis.

Very little is known about the latent organism or what causes it to reactivate, yet understanding it is critical in the control of tuberculosis, Bishai says. At Johns Hopkins, he is using molecular biology tools to try to understand latency with the goal of developing drugs and vaccines that would help to control it.

"One-third of the world is infected. If in 1999 a potent vaccine or drug were to be introduced, we would still have about 2 billion people running around who could get reactivation tuberculosis anytime in the remainder of their lives," Bishai explains. "To really achieve control of tuberculosis in the foreseeable future, we have to come up with better strategies for controlling latent tuberculosis."

It is not clear whether latency means that the organism is simply hiding or that it is dormant. Prolonged treatment with antibiotics decreases the risk of reactivation, suggesting that the organism is metabolically active even though it is not proliferating, Bishai says.

Colorado State's Orme cites other evidence for metabolic activity. One is that when placed in animals without immunity, these supposedly dormant bacteria grow almost instantaneously, without going through the expected lag phase. "That doesn't look like dormant tuberculosis to me," he says.

As Rockefeller University's McKinney notes, there's a sharp contrast in the courses of infection by acute pathogens and by M. tuberculosis. "If you infect an animal with an acute infectious agent, the organism replicates very rapidly for the first couple of weeks. Once the host immune responses kick in, bacterial numbers rapidly decline as the agent is eliminated," he says.

Tuberculosis is very different. In its host, M. tuberculosis displays the same exponential growth within the first two weeks but at a much slower rate, because it divides so slowly. But when host immune defenses become engaged, bacterial numbers in the lung remain steady instead of dropping.

"So from two to three weeks onward, the organism is no longer growing and dividing," McKinney says. "It is not surprising that, at that point, it's very difficult to treat with drugs that target dividing cells. The host appears quite healthy, and yet the disease develops slowly but inexorably over the ensuing months."

Both Orme and McKinney believe the disease progresses because the very presence of bacteria induces a nonstop immune response.

The infection elicits "an immunosurveillance-type mechanism that is continually looking for it and continually squashing and sitting on it to keep it under control," Orme says. The immune response is always on, McKinney adds, unleashing reactions that are meant to protect the host but end up damaging host tissues.

"We'd like to understand the functions required by the organism for stationary-phase persistence in the host," McKinney says. "Targeting those functions with new drugs could allow us to cure tuberculosis more rapidly."

Very little also is known about what triggers reactivation of the latent bacteria. "What we know from human epidemiological studies is that immunosuppressive conditions predispose to reactivation, and so HIV infection, administration of steroid immunosuppressive drugs, taking cancer chemotherapy, and old age are all risk factors," Bishai says.

Vaccines and diagnostic tools

Researchers are also hoping to mobilize new vaccines and new diagnostic tools in the battle against tuberculosis.

"When one thinks of the global situation, where you have billions of people infected and billions at risk of becoming infected," says Shinnick, it is clear why the world needs a better vaccine against tuberculosis. "We don't have the resources to treat everybody, but everyone can get a one-shot vaccine."

The current vaccine--called BCG for bacille Calmette-Guérin--"has been around for nearly 70 years," says Orme. "It's very safe and very cheap to produce. But the more people have looked at clinical trials, the more they realize that it's not as effective as one would hope." Although it works well in children, it is not effective in adults.

Orme is involved in the search for a better vaccine from the secreted proteins of M. tuberculosis. His research has shown that they protect mice and guinea pigs against tuberculosis to different degrees. "We're trying to sift through a pool of about 200 proteins to come up with a sort of top 20 that could be the basis of a vaccine," he says.

Other approaches include plasmid DNA vector-based vaccines and recombinant and mutant BCG vaccines. The facility that Orme directs at Colorado State has been testing candidates based on various approaches from Germany, England, South Africa, and the U.S.

The proteins of M. tuberculosis are also of interest to Colorado State's Belisle and Suman Laal, an assistant professor of pathology at New York University. The two are working jointly to develop an inexpensive test for on-the-spot diagnosis of tuberculosis that a technician can perform and evaluate. They're looking at the full complement of proteins produced by M. tuberculosis to find the right ones to use in a diagnostic test.

Laal says the ideal proteins for such a test are those that will distinguish tuberculosis from other lung infections and detect it at an early stage. Current diagnostic tools--such as smear tests, bacterial cultures, and chest X-rays--are not very helpful, because they detect the disease only when it's already fairly late in its progression.

The tuberculosis skin test is also available, but it is not an indicator of active disease. All sorts of people will give a positive skin test: those who have had tuberculosis and have been cured, those who have been infected but do not have active disease, and those who have been vaccinated with BCG. "We're looking for markers of active infection," says Laal.

Another element of the tapestry of tuberculosis control is identification of so-called surrogate markers, or indicators, of drug efficacy and protective immunity. These tools are required if drug and vaccine candidates are to move rapidly through clinical trials.

Because of the slow course of the disease, there are tremendous problems in clinical testing of new vaccines or drugs for tuberculosis. For example, Ellner notes that clinical trials for the BCG vaccine involved a quarter of a million people followed for 20 years. "It's not really feasible to repeat studies of that kind," he says. "When new vaccines or drugs are in clinical trials we'd like to have a quick read of their efficacy."

The search for surrogate markers is itself a huge undertaking. For example, in the tuberculosis research unit Ellner directs at Case Western Reserve, the task involves studying the immune responses before, during, and after treatment of large populations of patients in Kampala, Uganda, as well as the immune responses of those with whom the patients have been in contact.

Making a difference

Although times may be good for tuberculosis research in academic and government labs, the pharmaceutical industry appears uninterested. Many of the researchers who spoke to C&EN mentioned that one obstacle in the battle against tuberculosis is the unwillingness of most drug companies to commit significant resources to it.

Sidebar: It could happen to you

Most companies will test their compounds for activity against M. tuberculosis "just to show that they have tested them," is how PHRI's Kreiswirth sees the situation in industry. "If they don't find activity, then they can go on with their development and not worry about tuberculosis."

The drug companies "don't see a good market, because most of the people who have the disease are poor," says Ginsberg.

That's not to say that nothing is happening in pharmaceutical companies with regard to tuberculosis. Last year, for example, Hoechst Marion Roussel introduced rifapentine as a new drug for tuberculosis. The structure of rifapentine is almost identical to that of rifampin, a standard antituberculosis drug. And like existing drugs, it must be taken over a period of six months in combination with other drugs. It is not a radical departure from current therapies.

Among companies surveyed by the Pharmaceutical Research & Manufacturers of America (PhRMA) in 1998, only Seattle-based PathoGenesis had any antituberculosis agent in clinical development. However, results of Phase II clinical trials were inconclusive, says Maryellen Thielen, PathoGenesis' director of investor relations. And the company has decided "to put some of [the tuberculosis] work on the back burner."

PathoGenesis' exit from the tuberculosis scene "does not mean that other companies won't try again with new research projects," says PhRMA spokesman Jeff Trewhitt. "We do not know what is in preclinical testing." But for now, it leaves a British company as the only one with a focused push to find new drugs or vaccines for tuberculosis. And what a difference this one is making.

The company is Glaxo Wellcome. Its seriousness about tuberculosis research is embodied in the program called Action TB, an international research initiative launched in 1993 to which Glaxo Wellcome has committed about $32 million for the 10 years through 2003.

Glaxo Wellcome recognized tuberculosis as an unmet medical need, says Glaxo's Duncan, who is also manager of Action TB. "We saw the need for better therapies and vaccines, and we chose to fund some basic research to move the field forward."

Through Action TB, Glaxo Wellcome is supporting research by various groups in England, Canada, the U.S., South Africa, and the Gambia. What has emerged is a partnership between industry and academia that is bringing together researchers in the various disciplines--chemistry, biochemistry, molecular biology, immunology, and microbiology--that must be brought to bear on the problem.

In addition, Glaxo Wellcome has "brought some of the discipline of a pharmaceutical company to the problem so that researchers are thinking about what we actually need out there," Duncan says. "It's not good enough to simply say we need more drugs," he explains. "We've identified the properties that would be desirable in a new drug and considered the factors that would make it a big step forward in therapy."

Action TB has been crucial to researchers in South Africa, according to Paul D. van Helden, a professor at the Medical Research Council Centre for Molecular & Cellular Biology at the University of Stellenbosch Medical School, Cape Town.

Van Helden leads a multidisciplinary team that's applying various approaches to identify surrogate markers of patient response to therapy. Through a long-term study using modern tools ranging from DNA fingerprinting to geographic information systems, the team is also studying how the disease spreads in an area where the high incidence is not associated with HIV infection. Clinicians, immunologists, molecular biologists, and epidemiologists, as well as sociologists, geographers, and anthropologists, make up the team.

"Much of this work would not have been possible without the support from Action TB," van Helden tells C&EN. "We have not had adequate funding from local sources."

Nulda Beyers, one of the clinicians in the team, adds that because the work is being funded well, "we are in a position to study many variables," including social and demographic factors, attitudes of patients toward disease and treatment, host immunological factors, and bacterial genetic factors.

Collaboration extends beyond national boundaries, as exemplified by the research of Mizrahi at the South African Institute for Medical Research. There, she is trying to identify new drug and vaccine targets through gene knockouts. Research in her lab is done with the cooperation of Action TB researchers at the London School of Hygiene & Tropical Medicine and at Glaxo Wellcome Research & Development. She's also adopting techniques for random mutation that have been developed by a U.S. group associated with Action TB.

"We've succeeded in developing a reliable method for doing targeted gene knockouts in M. tuberculosis that has sufficient general applicability to allow us to inactivate any nonessential gene within a short period," Mizrahi tells C&EN. "The success of this research has hinged upon a highly productive collaboration with other Action TB groups."

But it's not only by providing research funds and promoting cooperative endeavors that Action TB is making such an impact. The program also has a progressiveness that comes like a breath of fresh air.

For example, McKinney thinks that at NIH "there's an enormous amount of conservatism, and basically they want everything to have been done before they'll give you the money for a line of research."

At the moment, McKinney is establishing his lab at Rockefeller. In seeking funds to launch his independent research, he got a quick response from Action TB, which, he says, awarded him a grant one month after he submitted a solicited proposal. With four other collaborators, McKinney is also seeking NIH funding for a tuberculosis program. He says they submitted a proposal in January, and "it's my understanding that there won't even be any meeting to discuss the proposal until this summer."

And according to Mizrahi, Action TB gave "due consideration to the fact that tuberculosis research is slow and that a long-term vision is required." This vision, she adds, "has enabled us to establish methods and techniques during the first few years that are only now paying dividends."

Action TB's goal for 2003 is to have at least one drug in the early stages of development and a backup, if possible, says Duncan. The backup would be either a molecule with a different mode of action or a molecule against the same target but from a different chemical series. "We also hope to have understood the immune response better to have some vaccine candidates at least at the stage of having evaluated them in animal models," he adds.

As to whether poor patients in developing countries will be able to afford new therapies, Duncan says: "To a great extent, you can't decide how you're going to deliver a drug or a vaccine to the patients until you've got it. But we have to think in terms of the potential Third World market. That limits the types of molecules you may consider. You have to keep the chemistry simple and keep the starting point inexpensive so that the final drug is going to be relatively cheap to produce."

With research picking up, there is increasing hope that tuberculosis will be eliminated. But as Ellner puts it, "We can't really be sure that, despite all the basic research, pharmaceutical companies will see the opportunity to go ahead with the new vaccines and the new drugs that are necessary to eradicate the disease. The scientific and public health communities must provide the advocacy that our patients cannot."

The world can be sure with one pharmaceutical company at least: "If we come up with an agent that's substantially better than what we have today, [Glaxo Wellcome] will find a way of making that available," Duncan says.

[Previous Story][Next Story]

Chemical & Engineering News
Copyright © 1999 American Chemical Society