The Greenpoint Men’s Shelter for the homeless is as grim as its name suggests: a crumbling yellow-brick pile in a desolate corner of northwest Brooklyn. Even in normal times its residents hate being here, with robbery, assault, and abuse their everyday companions. But one day last winter they had something worse to worry about than poverty, homelessness, and the squalor of the city’s stark and overburdened institutional charity. They were terrorized by a 32-year-old fellow resident whom we’ll call Walter.

Walter wasn’t violent; he was ambling placidly around the premises brandishing a can of soda, apparently harmless, even companionable. According to Gregory Usenbor, a courtly and soft-spoken man who works for New York City’s public health agency,

Walter triggered the alarm by appearing at the shelter bizarrely swathed in a surgical mask. But, Usenbor recalled, every time he wanted a sip of soda he’d pull the mask down. Everywhere he went the shelter clients--even the caseworkers--were running away from him. They were running from that lowered mask and the legendary killer it was releasing into the room: Walter’s highly contagious case of tuberculosis.

TB is back, and with a vengeance. And as Walter’s case illustrates, its return is a mirror of everything that went wrong in American health care during the 1980s. Walter has been homeless for the last five years. He is an IV drug user with a history of multiple sexual partners. He also has AIDS, a predisposing cause of TB, and a factor, Usenbor estimates, in 30 to 40 percent of the TB cases he sees. Damage to Walter’s immune system by the human immunodeficiency virus probably paved the way for the TB he already had when he first came to the attention of the city’s Bureau of Tuberculosis Control, a year before his masked appearance at the Greenpoint shelter.

The TB bureau first became involved in Walter’s case when he was admitted, coughing, to a Brooklyn hospital. After being discharged, he went off medication and wound up--sick again--at Woodhull Hospital. He’s stubborn, says Usenbor, who has the thankless job of overseeing treatment for the city’s burgeoning population of homeless people with TB. They couldn’t get him to stay put at Woodhull; he was wandering from one room to another day and night. But he was always telling me he knew what he was doing. He kept telling me he knew a lot about TB--how to prevent it, how to take care of himself. He doesn’t think straight sometimes. Eventually Walter left Woodhull against medical advice. Usenbor tried to get him into an outpatient program at the city’s Bedford-Stuyvesant chest clinic, but Walter vanished into the streets, going without treatment until Usenbor ran him down at the shelter months later.

Multiply Walter a few thousand times and you begin to grasp the dimensions of the problem. Take people in straitened conditions, lives rendered chaotic by poverty, drug abuse, or mental illness, many with damaged immune systems. All these factors conspire to make them easy prey to the bacillus; they also make it devilishly hard to ensure that they’ll persist in taking a combination of drugs for the six months needed to bring the infection under control. To control most TB, all you need to do is swallow two or three different oral antibiotics once a day for at least six months. Your cough generally vanishes within weeks, your lassitude lifts, and--vital from the public health vantage point--you can no longer spread the disease. Simple, as long as you’re decently housed and fed, your life is under control, you have a place to keep your prescription and a way to refill it--all nearly impossible if you’re like Walter.

And plenty of people are. Last year there were 3,673 new TB cases in New York City, up more than 140 percent from 1980. There may be as many as a million infected people in the city. Nationally there has been an 18 percent increase since 1985. Worldwide 1.7 billion people are probably infected--that’s one-third of the world’s population. Not all are currently sick; there are around 10 million cases with active symptoms, and 3 million deaths occur a year--a total that, so far at least, dwarfs mortality from AIDS.

To make matters worse, public health investigators have begun to detect more and more drug-resistant cases, some defying all 11 of the known effective anti-TB drugs. In one recent New York City survey, a third of TB cases had bacilli resistant to one drug, and nearly a fifth resisted two: namely, isoniazid and rifampin, the two most widely used remedies and hitherto the most effective.

Less than a decade ago this dire scenario would have seemed like fantasy. For some 40 years, ever since antibiotics effective against the disease first became available, the number of new cases had been dropping, from more than 84,000 in 1953 down to an all-time low of 22,000 in 1984. Emboldened by such numbers, the U.S. Department of Health and Human Services began to formulate a plan to wipe out TB by the first decade of the twenty-first century. Ironically, while the agency was planning, the case rate abruptly started zooming upward again, beginning in 1985.

Scientists sheepishly admit that in the 40 years during which TB appeared to be in retreat they fell asleep at the switch while the bacillus bided its time. Research on Mycobacterium tuberculosis, the organism that causes TB, simply crumbled. The bacillus is difficult to work with. And dangerous: unlike AIDS, TB is easy to catch, spreading casually through the air via a cough, a sneeze, or even a sentence spoken aloud at average volume. And in the 1960s and 1970s research on TB was unfashionable. Cancer and heart disease were the magnet diseases, attracting funds and ambitious researchers. Drug companies, by and large, blithely stopped working on new TB remedies.

What researchers do know about the bacillus simply underlines the mistake of ignoring it. Its cell wall, for example, is a tough, waxy, nearly impermeable barrier almost unique among bacteria.The earliest antibiotics, beginning with penicillin in 1940, had no effect whatever on it. It wasn’t until 1947, after a long and painful search, that researchers finally gave physicians an effective drug for their patients: streptomycin. While several more drugs came along in the late 1940s and the 1950s, some have bad side effects, and there are very few backup weapons in the medical arsenal if the frontline antibiotics fail.

Traditionally we think of TB as a lung ailment, inseparable from a hacking cough and pulmonary hemorrhaging à la Camille. The bacteria attack patches of tissue in the lungs, killing cells and making breathing difficult. Sometimes the attack extends to adjacent blood vessels, and patients cough up blood. Eventually the lungs become too riddled with infection to function, and the patient suffocates.

But while these are the obvious symptoms, in recent years those few researchers who have stuck the course and kept working on TB have come to realize that it’s really--like AIDS--a disease of the immune system. Thanks at least in part to its armor, M. tuberculosis is able not just to survive but to multiply in (of all things) the macrophages--the white blood cells whose normal job is to ingest and destroy foreign bodies. Much like the AIDS virus, M. tuberculosis actually hides in these immune system cells, hitching a ride with them to infiltrate the body, even using them as breeding grounds in which to multiply as it bides its time for a later all- out assault. Once it breaks out of these macrophages, their destruction begins to downgrade the immune system.

Then there’s the matter of how the bacillus multiplies. Common laboratory bacteria double their numbers every 20 minutes. The TB bacillus takes 22 hours to do the same thing, and that poky metabolism lends it a survival advantage. Anti-TB drugs appear to work only when the bacillus is active. But, being sluggish creatures, TB bacilli can remain dormant--and hence unresponsive to antibiotics--for months. That’s why patients have to stay on drugs for at least six months, with the bacillus, like the tortoise in its race with the hare, always ready to sneak into the lead if they’re careless.

Finally, there’s drug resistance. In someone being treated for tuberculosis with isoniazid, the drug of choice, about one in every million bacteria undergoes a spontaneous mutation that renders it resistant. That’s not a new problem: within months during the first streptomycin trials in the 1940s, it was apparent that the bacillus had an uncanny ability to parry medical thrusts, mutating its way around any drug launched at it. As new anti-tuberculosis drugs became available, combination therapy solved this problem: if you took two or more drugs simultaneously, the second would polish off any bacilli resistant to the first. But then the stresses of the 1980s created a critical mass of socially afflicted individuals vulnerable to tuberculosis and marginal enough to make it hard for them to complete treatment. That’s given tuberculosis a new foothold, and the cycle Walter illustrates--taking drugs just until the symptoms go away, thus leaving resistant bacilli to regroup and multiply--has clearly encouraged the development of resistant strains.

What can be done? Some steps are obvious: find all the people who have TB, make sure they take their medication, and be especially vigilant about those with drug-resistant disease strains. As Thomas Frieden, director of New York’s Bureau of Tuberculosis Control, says, We have the means to stop the epidemic. All we have to do is apply what we already know. But what we don’t know is enormous, and what isn’t being done is downright frightening. Example: The only practical test for TB infection is an early-twentieth-century antique--a pinprick under the arm with a bacillus-derived protein. You have to wait two or three days for a skin reaction, and the reading of the results can be maddeningly ambiguous. In addition, despite the upsurge in new cases and the gravity of future prospects, there still aren’t enough people with the disease in this country to make development worthwhile to drug companies. It’s not a big market, says Barry Bloom of the Howard Hughes Medical Institute at the Albert Einstein College of Medicine in New York, with understated irony--a small number of people of whom a lot are on public finance anyway. It’s not attractive from a business perspective.

I really think the scientific community is to be blamed, adds biochemist Patrick Brennan, whose lab at Colorado State University is now at work on tuberculosis.

He cites himself as an example. In the early 1960s I’d worked intensively on TB drug resistance but decided to get out of it because I thought TB was cured. There was still plenty, especially in developing countries, but either we weren’t aware of it or the figures weren’t available. In the last few years, scientific attention and funding began to drift back into TB research, but the field remains a poor relation to cancer and AIDS. In 1992 the budget of the National Institute of Allergy and Infectious Diseases called for a total of $5.1 million for all research on drug-resistant TB. One medium-size biology lab can have a bigger budget than that. (Officials later added another $4.3 million by funneling it in from other programs.) For fiscal 1993 the NIAID proposed a further $4.1 million increase.

William Jacobs Jr. is one of the mavericks who stayed with TB before the scientific community at large reawakened to its urgency. Jacobs works with Barry Bloom on the molecular genetics of mycobacteria at Albert Einstein. He points out that M. tuberculosis is as tricky to work with as it is to treat, and for some of the same reasons. First of all, since it’s dangerous and so easily transmitted, it requires expensive and cumbersome maximum-security containment facilities: They must be sealed off and outfitted with fail-safe air-exchange systems to keep bacteria from slipping through a ventilator. Everybody who enters or exits must go through elaborate decontamination rituals. Work with the bacillus itself must be confined to aseptic safety cabinets.

Then there’s the bacterium’s slow growth: the sluggish metabolism that gives the bacillus a survival advantage in vivo slows lab experiments to an agonizing crawl. E. coli forms a colony in eight hours, Jacobs says. M. tuberculosis takes three to four weeks. Furthermore, though the TB bacillus looks svelte in an electron micrograph, with its sleek tubular body and luminous golden sheen, when it grows into a colony it forms sticky clumps that make separating particular populations of cells difficult. And the tough cell walls make it extremely hard to get DNA in and out of the cells.

That’s unfortunate, because much of the basic research on TB (as opposed to trial-and-error drug tests) involves the genetics of the bug. Which of its genes produce proteins unique either to M. tuberculosis or to one of its strains? These could be used not only to devise reliable tests for general TB, but also to identify the particular variety a patient has contracted. Which genes code for the antigens that provoke an immune response when the bacterium invades its host? Which genes produce the near- impregnable cell wall, and how? Which help the bacillus accomplish its formidable feat of surviving inside the immune system’s scavenger cells? Which genes make a particular strain unusually virulent; are there any that might weaken it, making it less dangerous?

At the moment such questions remain largely unanswered. It’s a very complicated organism, Brennan says. There aren’t many people who understand the chemistry of the key components responsible for its interaction with the host. We don’t understand how the cell wall is synthesized, where drugs act, or the mechanism of drug resistance.

But scientists have at last begun to make some progress. In 1987 Jacobs hit upon a clever biomechanism, which he called a shuttle phasmid, that opened the bacillus up to recombinant DNA research. Many bacteria come equipped with circular DNA molecules called plasmids, which are completely separate from their chromosomes. Bacteria use plasmids to conjugate--that is, to dock up with each other and exchange genetic material. Biologists have used plasmids’ cell-hopping ability to transfer genetic material from one bug to another. The TB bacillus, however, doesn’t have plasmids. So Jacobs built one, but one that also has characteristics of a phage, or invading virus. Hence the name phasmid.

Jacobs’s invention starts as a plasmidlike ring of experimental DNA in lab workhorses like E. coli. These bacteria multiply and reproduce the rings. Then Jacobs takes one of the rings and inserts it into the fast- growing mycobacterium M. smegmatis, which is easier to work with than M. tuberculosis. Once inside, the phasmid acts like an invading virus, inserting itself into the mycobacterial genome. It then begins making copies of itself, each of which gets coated with a protective protein shell that allows it to leave the cell and burrow right through the cell wall of another, nearby mycobacterium. It heads straight for the bug’s genome, where it inserts itself, delivering its cargo of new, experimental DNA. Jacobs makes sure that the nearby bug is M. tuberculosis. Suddenly gene- transfer experiments with mycobacteria, once impossible, become relatively easy.

In August bacteriologists in London and Paris--using the techniques Jacobs and others pioneered--managed to track down a gene that appears to help the bacillus resist isoniazid. Their results were something of a surprise: they found that the gene (called katG) enables the bacillus to resist isoniazid when it’s deleted from the bug’s DNA. Usually bacteria fight drugs by acquiring a new gene, which often produces a protein that disables the drug. So at this point nobody is sure just how the disappearance of katG armors M. tuberculosis against isoniazid. The scientists speculate that the gene may produce an enzyme that activates isoniazid once it gets inside the bacterium.

Intriguing though this is, it still leaves us far from knowing how to counter drug-resistant TB. But research into the DNA of M. tuberculosis is at least beginning to accelerate, and more discoveries are surely in the offing. The bacillus’s entire DNA sequence is currently being read by Collaborative Research, a biotech company in Waltham, Massachusetts, using a new technology that allows a more rapid read-through of the 3 million to 4 million base pairs that make up the bug’s hitherto cryptic DNA library. We think it’ll take about three years to sequence a genome this size, says Douglas Smith of Collaborative Research. When they’re through, Barry Bloom adds, we’ll know all the bacterium’s DNA sequences and therefore every possible drug target.

But all that lies in the future. The steps basic researchers have begun to take still leave them frustrated and far from eradicating the White Plague, as tuberculosis was once ominously called. We still don’t know enough about the metabolism of the bug; we don’t fully know how any drug works, Bloom laments. We just don’t really know. And until we do, Walter and the thousands like him who are wandering the streets with uncontrolled tuberculosis are time bombs exploding in slow motion.