Johns Hopkins Magazine
Johns Hopkins Magazine Current Issue Past Issues Search Get In Touch
  Sweet Persistence

Inventor Gil Levin has patented a natural sugar that doesn't add calories or decay teeth -- an innovation that grew out of his earlier experiments on Mars and promises to be huge. So when will tagatose hit grocery stores near you?

By Dale Keiger
Photos by Bill Denison

Gilbert V. Levin tosses a piece of candy across his desk. Wrapped in gold foil, the confection is a nugget of chocolate, dark and sweet. The sugar that sweetens it came not from cane or sugar beets but from a dairy byproduct: whey. That's not all that's distinctive about it. Unlike ordinary sugar, it will not further bulge a bulging waistline. Nor will it decay teeth. It strengthens beneficial bacteria in the gastrointestinal track. Diabetics can eat it; not only will it do no harm, but if early clinical studies are borne out, it may actually alleviate symptoms of type 2 diabetes. Unlike non-nutritive sweeteners like aspartame, cyclamates, or saccharin, it occurs naturally, in minute quantities, in yogurt, soy products, even breast milk. Cooks can bake with it and use it to caramelize.

It's called D-tagatose, and the only thing you can't do with it that you can do with sugar is buy some in a grocery store. Gil Levin hopes to rectify that by next summer. The company he founded in 1967, Spherix, holds patents on D-tagatose. You'd think that a corporation with a patented natural sugar that doesn't harm the body would need massive vaults to store its profits, or at least would be trading at a premium on the New York Stock Exchange in anticipation of riches. Spherix, however, has never been hugely profitable and has struggled for more than a decade to put D-tagatose on store shelves. Its stock price was trading last August at around $4.50 per share, down from a 52-week high of $11.77. By early October, it had rebounded a bit to $6.71.

Levin, who holds three degrees from Hopkins, including a PhD, and who served three years on the university's board of trustees, comes across as patient, avuncular, self-effacing, good- humored. He smiles easily and seems to possess a mild personality. But his career demonstrates that he's also stubborn. For 26 years, he has asserted that an experiment of his invention on the 1976 Viking Mars landers found living microorganisms in the planet's red soil. NASA, and pretty much the rest of the scientific community, begged to differ when he first made his claim, and for years after. But Levin has never wavered, and he's gaining support from other scientists and from data generated by more recent Mars probes. Back in 1991, he forecast D-tagatose would be on store shelves in five years. Now, he believes, 2003 finally will see the introduction of Gaio-tagatose, as the sugar will be called. He just won't give up, not on Mars, not on his sugar, not on his company.

Life in Martian soil? Levin designed an experiment aboard the 1976 Viking landers aimed at finding out.
Photo courtesy Spherix/NASA
"It takes persistence," he says.

Now 78 years old, Levin is still CEO of Spherix, which was named Biospherics Research Inc. when he founded it in 1967. He seems in robust health, with remarkable recall of details from 60 years in the past. He often looks bemused as he discusses his life, seated in his modest office at Spherix's headquarters in Beltsville, Maryland. His office shelves hold more books than you'll find in the usual executive suite. A lot of them are about Mars. One is his Hopkins dissertation, "Metabolic Uptake of Phosphorus by Sewage Organisms."

"I like to say that I went to Hopkins for 22 years," Levin says. He entered the university in 1941 as a civil engineering major. When World War II intervened, he went to sea in 1944 as a radio operator in the U.S. Merchant Marine. "I'd never been farther than New York City or York, Pennsylvania," he says. He returned from the war in 1946 and completed his bachelor's degree the following year. He stayed at Hopkins to earn a master's in sanitary engineering in '48, then came back in 1960 to put in the three years he needed for a PhD in environmental engineering.

His work with sugar and his experiments on Mars are intertwined, and that story begins in 1952. Four years out of Hopkins and working as a public health engineer for the District of Columbia, Levin was fooling around with a better way to detect microorganisms. The standard method was to immerse a test sample in a nutrient broth that would culture any present microorganisms. The metabolic processes of those microorganisms would produce bubbles of carbon dioxide. Spot the bubbles, and you'd spotted the microorganisms. Levin thought the process was too slow. "Why wait to see bubbles?" he says. "If the carbon dioxide were radioactive, we could detect it much sooner." Levin reasoned that if he put carbon-14 (14C) in the nutrient broth, microorganisms would respire carbon dioxide labeled with radioactivity, detectable by a Geiger counter well before the first appearance of bubbles.

He says he brought his idea to one of the Hopkins engineering faculty, who told him it would never work. But he then showed it to Abel Wolman '13 (Eng '15), the legendary Hopkins engineering professor widely credited with making drinking water safe for the world. Wolman, who in the 1940s had convinced Levin to study sanitary engineering, said, "Why don't you write a proposal?"

Levin replied, "What's a proposal?"

Wolman explained, and Levin wrote one. He submitted it to the U.S. Atomic Energy Commission, which dawdled but eventually approved funding that Levin describes as a tenth of what he'd asked for. Nevertheless, he proceeded (early evidence of that stubbornness), and in three months had a 14C detection process that not only worked but was indeed much faster than the old method.

Six years later, Levin's wife, Karen, who was then a reporter for Newsweek and later became a vice president of Biospherics, took him to a Christmas party. Over martinis, he met NASA administrator Keith Glennan and asked him if the agency was interested in searching for life on Mars. Glennan replied that NASA had just hired Clark Randt, an MD, to direct its new biology division. Levin contacted Randt and proposed equipping a Mars landing craft with an experiment that would test Martian soil with a 14C nutrient broth that could detect any microorganisms that were based on water and carbon. NASA liked the idea and began funding development of the experiment and necessary instruments.

Levin remains convinced his experiment found life on Mars -- despite contradictory data from NASA. "With a small company," he says, "your results are not taken as seriously." Again, Levin needed patience. It would be 16 years before Viking I landed on Mars, on July 20, 1976. Viking II followed on September 3. Both carried a "labeled-release" or "LR" experiment designed by Levin and Biospherics. Each Viking used a robotic shovel to scoop the Martian dirt. The dirt was dumped into a box, from which .5 cubic centimeters were directed to one of a quartet of incubation chambers mounted on a carousel. The lander then sealed the chamber, enclosing the soil and trapping a sample of Martian atmosphere. For 24 hours, a radiation detector monitored the captured air sample to establish a baseline radiation level. Then the liquid nutrient (a soup of sodium formate, calcium glycolate, glycine, D-alanine, L-alanine, sodium D-lactate, and sodium L-lactate, each labeled with 14C) was injected into the soil, and for eight days the radiation level of the tiny amount of air trapped in the cells was measured every four minutes for the first two hours, every 16 minutes thereafter, for eight days. Then a second injection of nutrient was followed by eight more days of monitoring. At the completion of the cycle, the chamber would be purged, the carousel would rotate, and the experiment repeated using a fresh chamber.

Were there microorganisms in Martian soil, living according to the same biochemical rules as life on Earth, they would intake 14C from the broth and respire radioactivity into the air trapped inside the incubation chambers. If radioactive gas evolved, a duplicate soil sample would be placed in a fresh chamber, heated sufficiently to kill any microorganisms, then injected with more radioactive nutrients. If no radioactive gas evolved from the heated sample, that would confirm that the first response had been from living microorganisms. Sure enough, Viking's instruments detected radioactive gas coming from the nutrient-dosed samples; for the heated soil, the instruments detected no radiation. Levin and Patricia A. Straat, then a senior research biochemist at Biospherics and co-experimenter on the Viking LR project, published several papers, including an interim report in the December 17, 1976, issue of Science: "Tests at both landing sites provide remarkably similar evolution of radioactive gas upon addition of a radioactive nutrient to the Mars sample. The 'active' agent in the Mars sample is stable to 18°C, but is substantially inactivated by heat treatment for 3 hours at 50°C and completely inactivated at 160°C, as would be anticipated if the active response were caused by microorganisms."

Levin says, "My experiment satisfied all requirements for detecting microorganisms on Mars." But NASA had contradictory data from a second experiment on the landers, a molecular analysis that tested Martian soil samples with a gas chromatograph-mass spectrometer (GC-MS) that failed to detect any organic compounds. No organic compounds meant no life, at least no life as it's known on Earth. Eventually the agency concluded that Levin's experiments had not found microorganisms, but had been fooled, perhaps by something exotic in Martian soil chemistry that created a false positive.

Levin never bought the idea that chemistry could account for the change in data when the soil samples were heated; to his mind, only biology could explain that. But he and Straat exercised caution in the journal BioSystems, where they wrote, "The Viking Labeled Release Experiment has produced evidence for life on Mars. However, non-terrestrial soil chemistry may be mimicking a biological response. All hypotheses require further study before a conclusion can be reached."

Privately, though, Levin felt that he'd found living microorganisms. He saw something more besides contradictory data behind NASA's reluctance to endorse his results: "With a small company, your results are not taken as seriously as if you were a large company or university. My experiment was produced by Biospherics. The GC-MS experiment [that failed to detect organic matter] was produced by MIT."

While all this was happening, Levin had a company to run. He'd begun Biospherics almost on impulse when some lab space came on the market just as he was contemplating a faculty position proffered by Colorado State University. He figured that running his own company would afford more freedom to pursue ideas than he'd ever find working for another lab or university, so he signed a lease for the space. The resulting corporation, which for a brief perilous time after its start-up had been reduced to a few hundred dollars in its bank account, now had products and services and employees and shareholders and the need to find more profits.

Still, when he could, Levin kept after the Mars experiment. In 1981, he and Straat wrote in the journal Icarus that they had tried to replicate the Mars results on Earth using hydrogen peroxide (NASA's main possible explanations for the labeled-release findings) and could not do it. That didn't seem to convince any skeptics.

Twenty-five years after the Viking landers experiments on Mars, scientists found evidence of water ice -- offering more credence than ever to Levin's earlier findings.
Photo courtesy Spherix/NASA

But 15 years later, new discoveries returned the LR experiment to the center of the debate. In August 1996, four scientists (Everett K. Gibson Jr. and David S. McKay of NASA's Johnson Space Center, Kathie L. Thomas-Keprta of Lockheed Martin, and Christopher S. Romanek of the Savannah River Ecology Laboratory) announced in Science that analysis of a meteorite that had come to Earth from Mars -- it was literally a piece of Mars flung to Earth by some sort of major impact -- showed evidence of fossilized life. Four years later, the same scientists plus five new collaborators published a follow-up paper in Precambrian Research that presented more evidence in support of their position. In 2001, McKay, D.M. Warmflash of the Johnson Space Center, and S.J. Clemett of Lockheed Martin wrote a paper that reviewed the meteorite data and said "... a biological interpretation of the Viking LR results [Levin's experiment] can no longer be ruled out."

For another 20 years, critics disagreed, not only because of the GC-MS results. There was also no evidence of surface water on Mars. No water, no life, no matter what Levin's instruments said. Then came photos from the Mars Global Surveyor in 2000, images of hundreds of gullies that certainly looked like the work of flowing water, albeit in the past. More startling were the May 2002 gamma-ray spectrometer readings from the orbiting Mars Odyssey spacecraft that indicated reservoirs of ice just below the Martian surface. Levin had already taken 1997 Mars Pathfinder surface temperature data and created a model in which Martian soil could contain 1 percent moisture; that doesn't sound like much, but he points out that it's enough to support microbial life in Death Valley, California. Joseph Miller, an associate professor of cell and neurobiology at the University of Southern California, adds, "We know there are terrestrial microbes which exist quite happily in ice. And there appears to be a phenomenal amount of water ice on Mars, maybe 20 centimeters below the surface."

As for the GC-MS data, Arthur Lafleur of MIT, one of the engineers who had worked on the GC-MS experiment and co-authored the paper reporting its data, announced in 2001 that a new evaluation had determined that the instrument, though remarkably sensitive, was not sensitive enough to detect a low concentration of microorganisms, even in Earth soil with known microbial content. A paper co-authored by Lafleur stated, "Millions of microorganisms are required to provide sufficient organic matter to evoke a positive response in the GC-MS, whereas [Levin's] LR has consistently detected as few as 10 cells."

Miller of Southern Cal discounts the chemical explanations for Levin's results. "Basically, the [chemical] idea is that there is a highly oxidizing material in the Martian soil, peroxides or superoxides," he says. "When the nutrient medium in the LR experiment came into contact with this stuff, all the amino acids and whatnot were very rapidly oxidized, yielding CO2 without having to invoke biology. The problem is that no one has ever been able to detect this hypothetical material on Mars! Spectroscopic observations from the Hubble space telescope were unable to find any signature for this stuff."

Earlier this year, Miller was quoted in USC Health Magazine: "I think back in 1976, the Viking researchers had an excellent reason to believe they'd discovered life; I'd say it was a good 75 percent certain. Now I'd say it's over 90 percent. And I think there are a lot of biologists who would agree with me."

David W. Deamer, a professor of chemistry and biochemistry at the University of California-Santa Cruz, is not among them: "[Scientists] use weight of evidence to judge results in terms of plausibility, the rule being that 'extraordinary claims require extraordinary evidence.' In my judgment, the labeled-release experiment did not provide sufficient weight of evidence to conclude that microbial life exists on the Martian surface. A chemical reaction involving Martian minerals is a more plausible explanation. Earth has had over 3 billion years of evolution to fill every available niche, yet has never managed to invade the dry valleys of Antarctica, which are the closest terrestrial analogue of the Martian surface. I would be skeptical that microbial life somehow found a way to live in the even harsher Martian surface in such numbers that they could exhibit a high metabolic rate as soon as they were provided with water and nutrients."

Levin remains undeterred. He says, "It is now more difficult to propose a sterile Mars than a live one."

Early clinical trials show that diabetics who use tagatose daily have no increase in blood glucose or insulin. Furthermore, obese patients lost weight. When Levin was thinking about his Mars lander experiment back in the 1960s, he was mindful that ever since Louis Pasteur, scientists have known that various molecules can be described as "left-handed" or "right-handed." Stereoisomers -- left- or right-handed versions of the same molecule -- contain identical constituent elements, behave the same chemically, yet exhibit different properties in biological reactions because of their mirror-image structures. A common case is the molecule limonene. Limonene exists in the peel of both oranges and lemons. But in oranges, it's left-handed and makes grated orange peel smell like an orange. The right-handed limonene found in lemons is made of the same atoms, but smells like lemons because of its different structure. Mint and caraway both contain carvone, but caraway carvone is right-handed and mint carvone is left-handed. So caraway smells like caraway and mint like mint.

Levin knew that as life evolved on Earth, it evolved to use only right-handed carbohydrates and left-handed amino acids. What if Martian life were the opposite? If it had evolved to use southpaw carbs, Levin's experiment would have to test for those, too. Researching this problem, he found a paper that described left-handed sugar as tasting bitter. For some reason -- Levin smiles as he remembers this -- he simply had to find out for himself. So he ordered a quantity of leftie sugar, dabbed a fingertip in the crystals, and tasted them. They weren't bitter at all. They were sweet.

Slow-growing Biospherics needed a breakthrough product, and Levin the CEO immediately thought sweetener. He knew that chemically a left-handed sugar would behave the same as the common right-handed stereoisomer, when used in cooking, for example. But biologically the body would be unable to process it because humans had evolved to use only D-sugars -- D for "dextrose," or right-handed. He began testing L-glucose, L-fructose, L-gulose. Some worked as sweeteners but would be prohibitively expensive. When the baked-goods company Entenmann's inquired about one of his left-handed sugars, Levin had to tell them, "Well, right now it costs $3,000 a pound."

The last sugar to test was L-tagatose. In 1988, when Levin got the first batch from the lab he had assigned to make it, he found that something had gone wrong. His sample was not L-tagatose but D-tagatose, the right-handed variety. Because the test was ready to go, Levin ran it anyway with the right-handed isomer, and to his surprise found that in the bodies of lab rats D-tagatose behaved much the same as a left-handed sugar. Were you to eat some of it, your body would absorb about 25 percent into the bloodstream (you would absorb virtually none of a left-handed sugar) but then would expend more calories to process it than are contained in the sugar. Thus, no surplus calories, no weight gain. (Levin speculates that D-tagatose is so rare in nature that humans never evolved enzymes to process it efficiently.)

D-tagatose does not naturally occur in abundance, but it can be made by a process that starts with whey, which dairy product companies dispose of in abundance. That meant Levin's company might be able to produce its new sugar for a retail price of around $2 a pound -- provided the federal Food and Drug Administration would allow its use as a food product. Biospherics began assembling the data it would need to satisfy regulators, farming out the testing to ensure objectivity. That proved expensive. Biospherics was still shuffling along as a company, profitable thanks mostly to an information services division, but not deep in the pockets. In 1992, Levin began looking for a partner.

Four years passed before he found one: MD Foods, a Danish dairy company that produced a lot of whey as a byproduct of its manufacturing. Biospherics licensed the tagatose manufacturing rights to MD Foods in 1996, with the understanding, Levin says, that the Danes would bring a product to consumer shelves in 1998.

The FDA in 1995 had created a fast-track alternative to the standard food-additive petition. Food products could be evaluated by a panel of FDA-acknowledged experts who could declare the additives GRAS -- Generally Regarded As Safe. Achieve GRAS status and the FDA would allow you to sell your product to the public. To Levin's dismay, his partner, MD Foods, turned out to be an unexpectedly conservative operation. The Danish corporation insisted on a level of testing that went well beyond the requirements for GRAS. The targeted 1998 came and went, as did 1999, as did 2000. MD Foods merged with a Swedish dairy company, Arla Foods. Still no consumer product. Finally, in April 2001, Arla secured GRAS status for D-tagatose.

But it still wasn't satisfied. Arla submitted the sugar to the FDA for further evaluation, a step that Levin maintains was unnecessary. In October 2001, the FDA issued a no-objection letter, which is what amounts to approval in this process. Arla now promises to begin shipping the product to grocery stores in the spring of 2003, none too soon for a frustrated Levin, who says, "This has delayed my efforts to deliver my promised rewards to our shareholders, to the point of personal embarrassment." An equally frustrated Spherix -- Biospherics adopted the new name in 2001 -- filed for arbitration last year, seeking compensation and other redress from Arla for what it describes as "an unreasonably long time for its licensee to bring tagatose to market."

Spherix's new pesticide keeps flies at bay (horse breeders love it); basically citric acid, it is safe enough to eat, says Levin. Meanwhile, the sugar has met the FDA GRAS requirement for animal and human drugs, cosmetics, and non-food products such as toothpaste and mouthwash. Spherix retains the rights to those applications, which means no slow-moving Scandinavian partners this time. The University of Maryland has conducted two small clinical trials, administering D-tagatose to patients suffering from type 2 diabetes. Over a 14-month period, the diabetics experienced no increase in blood glucose or insulin as a result of taking tagatose daily with meals. Furthermore, obese patients lost weight, and subjects experienced a blunting of the rise in glucose levels normally triggered by carbohydrates. Levin wants more trials to see if tagatose may have application as treatment for diabetics. "But we're out of dough," he says, contemplating the extraordinary expense of drug testing. "If we get a big pharmaceutical company interested, or a medical school with money, we might do it."

Levin goes to his office every day, monitors new findings on Mars, waits for tagatose to come to a grocery shelf near you, and works to keep his company in business long enough to reap what he anticipates as a major payoff, at last. In 1998, Spherix lost $1.98 million, and another $5.2 million in 1999. The company did manage its best year in 2001 ($567,823 in profits), before the Wall Street bears took a bite out of its share price, but research the company's stock on an Internet site like and you'll find that no one recommends buying it. In a 1991 interview in Warfield's magazine, Levin conceded that he was a better inventor than salesman or corporate manager. One of his company's first products (it grew out of Levin's PhD dissertation) was PhoStrip, an organic process for removing phosphorus and nitrogen from municipal waste water. PhoStrip works, it's relatively cheap, and it reduces the amount of sludge created by water treatment. But Biospherics/ Spherix has always struggled to get anyone to buy it. In the 1991 interview, Levin said, "I'm a lousy salesman."

The company now is optimistic about a new product called FlyCracker. While experimenting with the manufacture of left-handed sugars in 1988, Biospherics workers left a bucket of an intermediate compound sitting in a warehouse over a weekend with the door open. On Monday morning, they came back to find dead flies everywhere. The discovery prompted further research that led to a compound that prevents fly larvae from developing into adult insects. Spread it around a horse barn, feed lot, or a poultry yard, and it works effectively as a pesticide, significantly reducing the fly population. Because the compound is basically citric acid, it is so safe, Levin claims, that you can actually eat it and suffer no harm. Spherix is now in its second year of selling FlyCracker, and Levin says horse breeders love it, but chicken farmers have been slower to give it a try. The product, like tagatose, has generated more hope than profits.

But Gil Levin remains convinced his day is coming. Tagatose will be on store shelves and in a gazillion carts at the check-out, FlyCracker will be killing flies throughout America, PhoStrip will be ridding streams of organic pollutants, and he'll be acknowledged, finally, as the man who found life on Mars. The latter still matters very much to him, both for the credibility it would lend to his company's other scientific efforts and the personal satisfaction it would bring. Spherix's 2001 annual report contains a timeline of recent developments important to the company, and by implication its shareholders. The fourth of five items is, "NASA establishes a Web site to present the Spherix data results from the Viking Mars lander for use by other scientists."

He watches his visitor eat the piece of chocolate, a prototype made for sampling purposes, and notes that the tagatose in it is 92 percent as sweet as sugar. "It takes an expert to tell the difference," he says. "We hope taste will sell."

He smiles. "It takes persistence."

Dale Keiger is a senior writer at Johns Hopkins Magazine.

Return to November 2002 Table of Contents

  The Johns Hopkins Magazine | The Johns Hopkins University | 3003 North Charles Street |
Suite 100 | Baltimore, Maryland 21218 | Phone 410.516.7645 | Fax 410.516.5251