DNA Device’s Heredity Scrutinized by U.S.

When the history of the just-dawning genetic revolution is finally written, a clunky-looking machine the size of a sidewalk trash can will play a starring role.

The automated DNA sequencer is letting researchers quickly crack the biochemical code of life, an achievement that could one day turn incurable diseases into treatable ones.

But the machine is at the vortex of a struggle over wealth, fame and, quite possibly, control of the genetic code itself.

The sequencer’s developers say they invented the device without a penny from the federal government, the usual source of funds for such endeavors. Their act of entrepreneurial wizardry, they say, entitles them to sweeping rights over their invention.

But The Times has turned up a paper trail that suggests a quite different story: one in which the developers collected millions in federal funds and failed to provide the government with certain key rights, such as discounts on purchases of the sequencers. Federal officials are now investigating.

The difference between these two versions of events could have a big financial effect on the inventors of the sequencer, the machine’s manufacturer and the California Institute of Technology, where the device was developed.

It could also affect the fortunes of investors who are wagering billions of dollars on claims made by the manufacturer, PE Corp., which has gained widespread recognition in both financial and scientific circles for its pioneering work.

Most important, the dispute could influence who gains control of the human genetic blueprint and all the medical miracles that it is expected to generate: the public or a few drug and biotech companies.

Among the documents examined by The Times:

* A series of federal grants to Caltech scientists in the mid-1980s to devise and improve the machine, including a $2.5-million National Science Foundation grant specifically “to automate DNA sequencing.” If the grants were used in inventing the device, the government would be due some rights to the results. The developers say they had already invented the machine when the federal money started to flow.

* A 1988 licensing agreement in which Caltech gave nearly exclusive authority to use the sequencer technology to a company started by the inventors, something not commonly done at the time when federal funds were involved. PE subsequently purchased the firm and, according to one recent study, captured 92% of the U.S. market for sequencers. The agreement included provisions promising the government the very discounts the law says are due it when federal funds are used.

* Letters and memos that show the machine’s inventors bickered over its creation in ways that some experts say could jeopardize their government-issued patents. The government is studying whether to challenge the patents.

Representatives of Caltech, PE and the researchers credited with the invention all said that no federal funds were used in developing the machine. They also denied there was anything improper about the 1988 licensing agreement and said there were no disagreements among the inventors that threaten their patents.

They acknowledged that the developers may have slipped up by filing over-exuberant grant applications, but described any mistakes as innocent. “Based on everything we know at this time, we believe we handled the invention and patenting of the sequencer in an entirely proper manner,” said Caltech Provost Steven E. Koonin.

Michael Hunkapiller, president of PE Biosystems, the PE affiliate that manufactures the sequencer, said he expects no effect on the company from the federal investigation and newly raised questions about the machine’s invention. “The issue is more with Caltech,” said Hunkapiller, who worked on the sequencer there and is listed in patents as one of its inventors.

Lab Notebooks Are Key Pieces of Evidence

The lead researcher, former Caltech scientist Leroy E. Hood, dismissed the entire episode as a peevish show of rivals’ envy. “Someone out there has a vendetta,” Hood said in a telephone interview. “Someone is trying to make trouble.”

But the matter appears to go beyond personal feuds.

Some of the inventors’ principal contentions do not seem to square with the paper trail. Caltech officials maintain, for example, that the inventors’ laboratory notebooks, which have been submitted under subpoena to federal investigators, prove that the government has no claim to the sequencer because the device was up and running by the time researchers began receiving federal funds.

But a review of the notebooks found that they contain a sustained lament by the researchers about their inability to get the machine to function correctly during the crucial period just before the start of federal funding.

“Very disappointing. What’s the trouble?” read the notebook entry after one early 1985 experiment. “Sob! What’s wrong,” read another.

The developer’s handling of the funding, licensing and patenting of the Caltech machine stands in marked contrast to that of several other fundamental inventions of the genetic revolution.

Stanford and UC San Francisco, for example, acknowledged their use of federal funds in devising a gene-copying technique that has proved as crucial to the revolution as the sequencer. After receiving patents for the work, the universities issued inexpensive licenses to all comers--hundreds in total.

First Drafts of Genetic Code Expected Soon

PE, on the other hand, has issued only a handful of licenses for the machine and has sued some of the recipients for patent infringement.

In addition, the company started Celera Genomics, an affiliate that is using the sequencer machines in an all-out effort to be the first to decipher the entire human genetic code and claim property rights to it. The firm’s chief rival, the publicly led Human Genome Project, which also uses PE sequencers, is making its findings immediately and freely available to the public.

Both efforts expect to complete first drafts of the genetic code next month.

If the developers can defend their version of events, they will not only have protected themselves from legal trouble, but they will also have struck an important blow for the view that only the private sector has the necessary drive to unlock the medical wonders of the genetic code.

Celera Genomics has certainly shown staggering drive. It has set 300 PE-made sequencers and some of the world’s fastest computers to work 24 hours a day in an effort to be the first to map the genetic code. By contrast, the Human Genome Project moved painfully slowly before Celera’s arrival forced it to step up its pace.

Celera’s colorful president, J. Craig Venter, has used every forum he can find--from scientific colloquium to congressional hearing--to argue that only the lure of profits will assure that the new genetic knowledge gets translated into innovative drugs and medical treatments. He has been unrelenting in his criticism of the public project as sluggish and insensitive to the need to develop new cures quickly.

“We believe that entrepreneurial efforts such as Celera are the best way to progress,” Venter said at a recent congressional subcommittee hearing. Quoting his company’s marketing slogan, he added: “Speed matters. Discovery can’t wait.”

If the developers’ version of events fails to hold up, however, the results could also be far-reaching.

In addition to possible financial penalties, the developers’ patents on the sequencing technology could be vulnerable to challenge by rivals or the government itself. “Fights between inventors can provide the opening wedge that competitors need to go after patents,” said Robert P. Merges, co-director of the Center on Law and Technology at UC Berkeley.

More broadly, evidence that the sequencer’s invention was not the entrepreneurial tour de force its developers have portrayed could undermine the arguments of Venter and other executives that the nation should entrust the genetic code’s development to the private sector.

And it could breathe new life into the warnings of public officials like Dr. Harold Varmus, the former director of the National Institutes of Health, and Dr. Francis S. Collins, the leader of the Human Genome Project, that private ownership of the code would harm research, stall drug development and force patients with hereditary diseases to wait still longer for help.

“While establishing a monopoly on commercial uses of the human genome sequence may be in Celera’s business interests,” Varmus, Collins and others recently wrote the company, “it is not in the best interest of science or the general public.”

Federal officials, led by the Department of Health and Human Services inspector general’s office, are investigating whether the government overpaid millions of dollars to purchase the machines, the latest version of which sells for $300,000. Under a 1980 law, the government is owed a discount on the products of federally funded research. Until recently, the government and government-funded scientists were the largest purchasers of the devices, having bought about 6,000 of them over the last decade.

The story of the sequencer begins two decades ago in the Caltech labs of biochemist Leroy Hood.

Hood is a Big Name in Science, a man at ease at the crossroads of research, power and celebrity. He has traveled to Africa with Bill Gates, appeared on “Larry King Live” with former junk-bond mogul Michael Milken and Gen. Norman Schwarzkopf, hobnobbed with heads of government and Nobel laureates.

He nearly won a Nobel Prize of his own for his work on the immune system. By the early 1980s, his new ambition was to invent a group of machines, especially the DNA sequencer, that would, in his words, “revolutionize biology in the same way [atom smashers] revolutionized physics.”

The state of scientific play as Hood began his work was that DNA had been discovered to be the language in which all living things are instructed how to grow and change. Scientists had figured out its basic structure: two tightly coiled chains of four distinct chemical building blocks, represented by the letters A, T, G and C, extending nearly endlessly.

The scientists had learned that the exact sequence of letters determined whether a person had blue eyes or cystic fibrosis, long legs or a predisposition to cancer. They had even come up with some methods of reading the sequence. But the methods were slow, error-prone and, because they involved radioactivity, dangerous.

As he did with every other challenge, Hood gathered enormous firepower to his side. His Caltech lab was one of the biggest in the nation. Records show that during the 1980s, Hood’s Pasadena lab won at least $30 million in grants from the government’s two big funding agencies, the National Institutes of Health and the National Science Foundation. His lab staff swelled to more than 100 and included the unusual team that he assembled to come up with the sequencer.

Henry Huang was a molecular biologist who had done a stint as a computer programmer at the Jet Propulsion Laboratory. Lloyd Smith, a fencer in college who would later earn a black belt in taekwondo, had gotten a doctorate in the arcane field of biophysics. Michael Hunkapiller was a biochemist who became Hood’s chief aide at Caltech and who later became a PE executive. His brother, Tim, was a computer expert in the midst of earning a doctorate in biology.

As recounted in their own writings immediately after the invention, and as described in several books, each member of the team made a distinctive contribution to the device while Hood sought money to support their work.

Huang tried to make a machine that read the A’s, Ts, Gs and Cs of the genetic code directly, but he couldn’t get it to distinguish the letters. The Hunkapillers hit upon marking each letter with a different colored dye. Smith realized that the dyes had to be fluorescent and devised a way of sticking them to strands of DNA.

The device would eventually include a mechanism for sorting the strands, a laser for making the dyes glow, an optical reader for spotting the colors as they passed by and a computer for deciding whether the colors were distinct enough to call each letter an A, T, G or C.

As the scientists struggled to fit the pieces together, Hood scrambled to come up with research funds from the government, corporations and foundations. When he met with rejection from Washington and only a grudging response from others, he started his own company, Applied Biosystems Inc., or ABI.

“I wrote several grants trying to get the sequencer done. They all failed,” Hood told The Times last year. “I went to 19 different companies--19 out of 19 said no. That’s when I went and started Applied Biosystems so we could finally get these things commercialized.”

The patents that were finally issued on the sequencing technology carried the names of Hood, Smith, the Hunkapillers and an ABI researcher, Charles Connell. ABI clinched near-exclusive control over the device in the 1988 licensing agreement with Caltech.

The machine was such a smash that it helped persuade the government to launch the Human Genome Project in 1990 in order to decipher the entire genetic code. ABI swept the market for sequencers, selling to both the government and drug companies.

If the inventors, the company and the university had their way, the sequencer story would end at this point--with their having completed a successful piece of science and being in near-complete command of the sequencing business.

But instead, questions remain about the government’s role in paying for the crucial research--and about who should get credit for inventing the breakthrough DNA sequencer.

Developers Say U.S. Didn’t Contribute

The developers--Caltech, PE and the inventors themselves--assert that the government contributed nothing toward inventing the device and first making it work. But grant applications, memos and correspondence--all of which have been subpoenaed by the federal government--suggest the opposite.

Most notable is a multimillion- dollar National Science Foundation grant approved in the mid-1980s.

The application for the NSF grant, filed by Hood and his colleagues in September 1984 and revised the following May, said the researchers were seeking $1.8 million (later raised to almost $2.5 million) “to automate DNA sequencing.”

The application included a timetable for producing “the first reliable operating prototype of the DNA sequenator,” as the machine was then called. It requested money for equipment needed to build a “prototype . . . breadboard,” a term scientists use to describe a very early-stage machine.

Hood and his team began work under the grant Sept. 1, 1985. They submitted a stream of annual progress reports saying that the machine had been invented and improved during the course of the grant. They wrote scientific papers crediting the grant with providing some of the money they were using. At a June 11, 1986, news conference, they unveiled the device and thanked the NSF for helping with its development.

They also received a host of other grants from the NSF to advance the sequencer cause. Among them were a 1987 grant for $600,000, in part to improve the machine, and a 1988 grant for at least $13 million to create a Center for Molecular Biotechnology at Caltech that was devoted in large measure to the sequencer.

In recent interviews, the inventors, PE and Caltech acknowledge the grants but say their existence does not contradict the claim that the machine was invented without federal funds.

According to this account, Hood and the other inventors had already come up with a working machine by the start of the 1985 grant. Hood nevertheless wrote the grant in a way that made it appear that money from it would go for the invention.

He did this in part out of what one person familiar with Caltech’s case termed “human laziness.” He simply lifted the language from previous grant applications even though he and his team had devised the machine in the interim, this source said. He also did it to curry favor with the NSF in hopes of landing future grants--something the sequencer’s developers assert remains a standard grant-writing practice.

Officials familiar with the federal grant process disagree that the tactic is commonplace. Asked whether he had regularly encountered it, Erich Bloch, who headed the NSF during the 1980s and described himself as an admirer of Hood, said, “Of course not.”

“It’s pretty clear to me from reading [the 1985 grant application] that the machine did not exist” at the time the grant started, Bloch said. (Under federal law, lying or concealing facts on a federal grant application is punishable by a fine and up to five years in prison.)

To make their case that the machine was already up and running, the inventors and Caltech have relied heavily on an affidavit their lawyers recently obtained from John C. Wooley, an associate vice chancellor at UC San Diego who headed the NSF unit that approved the 1985 grant.

Wooley said that he visited Hood’s lab in the spring of 1985, some six months before the start of the grant, and was shown “a working prototype of the DNA sequencer and a large volume of preliminary data.” Elsewhere in the affidavit, he described the device as a “fully functional prototype.”

In two lengthy telephone interviews with The Times, Wooley said he had seen the machine actually turning out four-color data to identify the A’s, Ts, Gs and Cs of the genetic code.

Wooley concluded in the affidavit that he was “well aware” before approving the grant that the device was operating. He said he believed that money from the grant would go for improvements “in order to enable the DNA sequencer to be commercialized.” Koonin and other Caltech officials say the affidavit clinches their case that the inventors came up with the machine without federal funds.

However, Wooley’s account does not match the lab notebooks of researchers working on the machine during the period when the former NSF official said he visited. The notebooks show the researchers were able to get parts of the machine to function but that at critical moments the sequencer would not fully operate.

For example, on March 1, 1985, the researchers described their first effort to use all four dyes, and they later recorded the results as “not any good. Strange. . . .” On March 6 they made another attempt but reported that one of the four colors “didn’t show up” and a second was “weak.”

In fact, after a thorough review of key notebooks, The Times could find no sign that the machine ever fully functioned or produced four-color data during the spring or early summer of 1985, when Wooley said he visited the Hood lab and saw the machine operating.

Caltech officials said recently that they had discovered a graph depicting four-color results dated Aug. 28, four days before the start of the NSF grant. But by late last week, they were still looking for the lab notebook entry for an experiment that day.

The 1980 law governing discoveries made with federal grants gives the government certain rights.

One is the right to a discount so that Washington doesn’t pay twice for something, first by funding the research that leads to its invention, then by paying full price to use the product.

Another is the right to influence how federally funded inventions are used. The government may encourage universities that patent the results of their research to issue broad, rather than exclusive, licenses in order to ensure an invention’s widest possible use.

The sequencer’s inventors, PE and Caltech have all argued that neither of these rights amounts to much. Even if federal funds were used in the machine’s invention, they said, the results would be inconsequential.

In particular, they argue that the law is unclear about whether Washington is owed any discount at all. If it is, they contend, it is due only on purchases by the single U.S. agency that funded the work and not to other arms of the government.

But government policy memos and the legislative history of the 1980 law make clear that a discount is owed on any federally funded invention.

The 1988 licensing agreement included provisions that appear to directly contradict the developers’ current position. It said that the government would be owed a discount on all of its purchases of sequencers.

The contract also said that the discount was owed not just to the NSF, which provided the grant to Caltech, but also to “any of [thegovernment’s] departments or agencies, or contractors or subcontractors or other parties for use in connection with programs funded in whole or in part by the Federal Government throughout the world by or on behalf of the government of the United States.”

Koonin, the Caltech provost, said that under the developers’ current, narrow definition of their obligations, the discounts due to the U.S. could total at most $400,000.

Using the license agreement’s original broader terms and recent industry studies, The Times estimated that the amount of overcharges in the last decade alone could be $25 million or more. That would represent about half of the royalties that Caltech collected on all inventions made at the school during the period, and Koonin conceded that its loss would be a substantial blow.

Legal Advice Is Debated

In interviews in recent weeks, Caltech and PE officials asserted that the provision’s inclusion in the 12-year-old agreement had been a mistake, the result of bad legal advice, and that it has been removed. They said that the language was removed late last year or early this year, just as the investigation of the sequencer was heating up.

The 1988 licensing agreement was interesting for another reason: The government’s general practice at the time was to encourage broad, nonexclusive licensing of federally funded inventions.

The agreement suggests how influential Washington was on this point; Caltech appears to have adopted the government’s position as its own.

The agreement requires ABI, the start-up company that PE subsequently bought, to “make [sequencing technology] available to third parties on reasonable terms and conditions.” ABI promised to make its “best efforts” to find others to use the device.

But records indicate that ABI and then PE issued licenses in only about a half dozen cases. By contrast, Stanford and UC San Francisco have issued more than 300 licenses for their genetic copying technology.

Another unresolved issue is potentially more threatening to the developers. That is the matter of exactly who was involved in inventing the machine.

The inventors, PE and Caltech all assert that the answer is the five men named in the patents on the sequencing technology: Smith, Hood, the Hunkapillers and Connell. But the answer is far less clear in personal letters, memos and other documents examined by The Times.

The conflict arises in documents obtained from Robert Cook-Deegan, an official of the National Academy of Sciences, who sought to prepare an account of the sequencer machine’s invention during the late 1980s for Congress and subsequently wrote a book, “The Gene Wars.”

Cook-Deegan repeatedly interviewed Hood and the other inventors, exchanged letters with them and offered them drafts of his account. His papers, now on file at Georgetown University, show that he could not get the inventors to agree on crucial aspects. Independent experts say such disagreements can leave patents open to attack.

As part of the group’s patent application, for example, each inventor had to swear that the others had contributed to the invention. But in a letter to Cook-Deegan dated March 2, 1991, Smith denied that the Hunkapillers had had anything to do with what was supposed to have been their chief contribution--coming up with the idea of using four colors to mark the DNA building blocks.

Instead, Smith said, the credit was at least partially due Huang.

“The four-dye idea had been around a long time and was not conceived by Tim or Mike,” Smith wrote.

In a recent telephone interview, Huang, now a professor at Washington University in St. Louis, said he had been studying the idea of using four colors as early as 1980, two years before the Hunkapillers. Documents indicate that Huang ordered several dyes as early as the fall of 1981. Huang was not included on the patent.

Asked why he had not pursued the issue of his omission from the patent, Huang said that he cared more about the science than about matters of ownership and money.

“If I opened the door for someone else to go through, that’s enough. You have to decide where you get your gratification.”

Under patent law, applicants must name everyone involved in a new idea or device and must describe the state of knowledge at the time of an invention.

If someone is left off inadvertently or information is omitted unintentionally, patent experts said, the problem is easily corrected. But for intentional omissions, the consequences can be dire.

Said Cleveland State University law professor Michael H. Davis, coauthor of the standard textbook on intellectual property: “If you play fast and loose with the patent system, the remedy is the invalidity of the patent.”

(BEGIN TEXT OF INFOBOX / INFOGRAPHIC)

At the Heart of a Revolution:

the DNA Sequencer

Publicly funded research centers and a private company are racing to complete a working draft of the human genetic code this year. A more detailed, finished version, listing in order some 3 billion chemical building blocks or letters that make up the entire genome, should be complete by 2003. None of this would have been possible without the development of highly automated technology, pioneered by scientists at Caltech in the 1980s.

*