Granted, when word emerged some weeks ago that MJ Research was legally challenging the validity of Caltech and Applied Biosystems' four-color sequencing patents, it hardly elicited a gasp of shock from the industry. People have debated for years which of the names listed on the patents — Lee Hood, Lloyd Smith, brothers Mike and Tim Hunkapiller, and Kip Connell — actually belong there. Similarly, there have been murmurings for some time that Henry Huang, a little-known virologist at Washington University who used to work in Hood's Caltech lab, was the real inventor.
Conference chatter doesn't amount to much, but MJ's team thinks its lawsuit just might be the stone David slings at Goliath. At stake for Caltech, Applied Biosystems, Applera, and Celera, all named as defendants: the automated four-color, fluorescent DNA sequencing patent estate owned by Caltech and licensed to ABI. It is arguably the cornerstone of ABI's sequencing business. In addition to revenue from its own sequencers and their reagents — from the ancient ABI 373 to the just-released 3730 — ABI collects licensing fees from companies that use the patented process, such as Amersham Biosciences and MJ for their respective MegaBACE and BaseStation.
If MJ succeeds in overturning the patents, it would open the doors for anyone to use four-color fluorescence for sequencing — and could potentially raise the floodgates for lawsuits against ABI from companies that have had to pay for the license all these years.
MJ is charging that the sequencing patents are invalid because they intentionally omit a key inventor, Huang, and that the technology was developed using federal funds, entitling the government to full ownership of the patents or a royalty-free license. In addition, MJ alleges that the defendants deliberately overcharged the government, and that they monopolized the sequencing market.
Invoking a provision of the 1863 False Claims Act that allows individuals or companies to sue on behalf of the US, MJ is seeking damages of hundreds of millions of dollars in overcharges to the government. And if lawyers can convince a jury that ABI knowingly charged too much, damages are automatically tripled in penalty. MJ is also looking to have the patents invalidated and the IP turned over to the public domain.
Patents Under Siege
This lawsuit is not the first time these patents have been scrutinized. Indeed, MJ prompted a full-scale government investigation as early as 1998 when it brought the matter to the attention of the US Department of Justice. "MJ actually had no intention of bringing suit themselves," says the company's attorney Allen Foster.
Months passed and MJ saw no progress, so it filed its own suit, which was kept under seal until the DOJ, HHS, and NSF wrapped up their joint investigation this spring. The DOJ's decision to unseal the case and not get immediately involved was hailed as a victory by both camps: ABI and Tim Hunkapiller contend that if there were any merit to the case, the DOJ would have exercised its option to take over the suit. MJ reps say that if the case had no merits, the DOJ would have shut it down. The government will monitor the case and reserves the right to step in and take over at any time.
That MJ, a small instrumentation company based near Boston, has taken the burden of what promises to be a years-long legal struggle is something of a peculiarity. MJ's Foster says federal funds paid to make the sequencer, and the public has been cheated of the rewards of its investment. "It's wrong," Foster says in his chummy but no-nonsense Southern drawl, "and that's why MJ is going to do this."
Indeed, the government is eyeing quite a windfall if MJ wins the civil case. But to believe that MJ, which stands to get 15 to 30 percent of any earnings from the suit, is driven by pure civic-mindedness would be naive. There's little love lost between MJ, headed by cofounding brothers Mike and John Finney, and ABI, also known for its sibling force: president Mike Hunkapiller and his brother, much-relied-on consultant Tim. "MJ Research has been … fighting ABI on things all the way back to PCR for years," asserts Tim Hunkapiller, 49. He says the lawsuit most likely is a negotiating stunt engineered to get MJ its own license to the valuable patents. MJ's Foster refutes that charge, pointing out the license MJ acquired when it bought ABI licensee GeneSys Technologies, now MJ GeneWorks.
MJ stoked the anti-Applera flames again when it set up nonprofit shell GCorp in 2000 to launch the annual Advances in Genome Biology and Technology conference, known as the "anti-Celera, anti-Venter" haven for scientists rebeling against TIGR's GSAC. Neither of the Finneys returned calls seeking comment.
Whatever the motives, MJ's case could be stronger than ABI's official no-comment and Tim Hunkapiller's boiled-down assessment that "there's no issue here" would indicate. Bookcases-full of documents gathered by the DOJ and by MJ's legal team suggest that, if nothing else, the battle will be fiercely fought.
Almost the only thing Henry Huang and Tim Hunkapiller agree on about those early '80s days at Hood's Caltech lab is that they were close friends. They also each claim to be principal inventor of the automated sequencer.
After joining in 1977, Huang spent much of his time in Lee Hood's lab trying to design an automated DNA sequencer. He experimented with various techniques but never built a functional instrument before leaving for a teaching position at WashU. Even today, Huang, 52, doesn't use an automated sequencer — he still analyzes DNA "the old-fashioned, radioactive way. The cost of doing those hundreds of sequencing reactions on an ABI machine would be too expensive," he says.
By the time Caltech and ABI announced their working sequencer in June 1986, Huang assumed it was completely different from his efforts and didn't even review the patent until MJ's lawyers asked him to a year and a half ago. The main points of the patent — optical detection process, fluorescent dyes, four colors in one lane, software to capture data and convert it to sequence — "are the key concepts that I had come up with pretty much before the summer of 1982," Huang says. "It looks awfully like what I had drawn [then]."
(Huang is not involved in the lawsuit. If his name were added to the patent now, he does stand to collect future royalties. But if MJ wins a jackpot from the suit, Huang won't get a dime of it.)
During much of the time he worked on the sequencer, Huang and his "best buddy" and fellow night owl Tim Hunkapiller would go out almost every night to the Salt Shaker, an eatery near campus, for a late snack while they talked over their lab work. Huang says he also talked with Lee Hood and lab chief Mike Hunkapiller about the project, particularly about using fluorescent dyes. According to an ABI spokeswoman, Mike Hunkapiller says Huang never worked with fluorescent detection. Hunkapiller declined to comment.
Huang claims that when he left, "none of the people involved [at the lab] had any expertise in sequencing DNA nor had thought about it much." But less than a month and a half after Huang's departure, the Hunkapillers and Smith signed an invention disclosure, witnessed by Hood, for an automated DNA sequencer.
The disclosure was dated October 1, 1982. But according to MJ's Foster, "the defendants, the purported inventors of this patent, have not produced one single piece of paper which would indicate that they did any work on this invention prior to October 1, 1982." Huang has handed over reams of correspondence, lab notebooks, and documents of his work on the sequencer. The defendants' papers from after the invention disclosure have also been examined, and Tim Hunkapiller says numerous reviews have satisfied investigators that the invention is rightfully theirs.
Given the schedules of the Hunkapillers and Smith in the weeks following Huang's move to WashU, "the best we could figure was that there was a 10-day period that they could've invented it," Huang says. That's 10 days for people who weren't working on sequencing to invent the instrument without ever writing anything down, notes the MJ team.
The charge doesn't faze Tim Hunkapiller, who says of course it looks suspicious to Huang, who didn't know that he and Mike were working on the idea. Hunkapiller says he thought of the concept for the sequencer after reading an article in the defunct magazine High Technology. "I don't blame Henry for not knowing that," he says. "I never discussed it with Henry. I discussed it with my brother." He says he and his brother worked together on figuring out the elements before bringing in Lloyd Smith to handle the chemistries. Huang's project, says Hunkapiller, did not use four-color sequencing, never worked, and did not play a role in what would become the ABI sequencer.
The invention disclosure was apparently shelved for a year before it was filed in late November 1983. Huang calls the delay "a curiosity" while MJ counts it a strong indicator that it was less than legit.
Even without debating Huang's account, the listed inventors have credibility problems of their own stemming from an apparent inability to tell the same invention history. "Lloyd and Lee said a lot of self-aggrandizing and sloppy things very early on," Tim Hunkapiller says, "and that has confused this thing for 20 years." Lee Hood, 63, did not return calls seeking comment, and Lloyd Smith declined to discuss matters of inventorship or other aspects that might be under litigation with the MJ suit.
Indeed, when Robert Cook-Deegan set out to include a chapter on the sequencer in his book The Gene Wars: Science, Politics, and the Human Genome, he couldn't get a consensus story from the principals on the patent. Almost every other chapter in the book needed about three revisions; the sequencing chapter took more than 15 iterations because the inventors kept objecting to each other's accounts.
For instance, Lloyd Smith wrote an extensive letter to Cook-Deegan in March 1991 detailing the steps of the invention and concluding, "Neither Tim nor Mike played any role in any of this work." He adds, "The entire implementation of the project, from before day one, was done by myself."
A sticking point remains for Huang: he went back to visit the Hood lab a couple of times in the years after he left, and he says nobody told him about the sequencer invention. He checked in with his successor Smith on his old project; though they didn't speak in detail, "my assumption was that it was in very capable hands and good enough," Huang says. But according to Smith's letter to Cook-Deegan, he declared the project unfeasible and shut it down almost immediately after Huang's departure.
Another angle MJ will bring up is the Husimi connection. A professor at Japan's Saitama University, Yuzuru Husimi was trying to build an automated sequencer at the same time as Huang. He and his students came up with a prototype, and he presented the findings at a September 1982 conference in Aspen, Colo., which Tim Hunkapiller attended. MJ's lawyers believe that Hunkapiller went to Husimi's talk, the only one in its time slot, where he would have heard Husimi say the instrument could be used for automated DNA sequencing.
If the connection is true, Husimi's work should have been listed on the patent as prior art — and knowing failure to do so could invalidate the patent, says uninvolved biotech patent lawyer Breffni Baggot. Husimi's name does not appear on the original patent, issued in 1992. Tim Hunkapiller says that he had never heard of Husimi until just weeks ago when news of the lawsuit came out. But Husimi's work is cited on a follow-on patent for automated sequencing that names Hunkapiller as one of the inventors.
No matter which story is true, the MJ team will have to show that Huang was intentionally excluded from the patent, a possibility that Caltech lawyer Jim Asperger says has been carefully investigated and never borne out. As Hunkapiller says, "Had Henry been an appropriate person on the patent, it would've been no issue. We were all friends." As the principals don't even own the patent — they signed over rights to Caltech — what would be the motive for leaving Huang off?
That's the easy part, according to Baggot. Leaving Huang off would have meant getting bigger shares of the 15 percent royalty afforded by Caltech to the inventors — by MJ's estimates, each of the principals has gotten more than $1 million from the patent. It also meant sharing less of the credit: a common enough reason people are left off patents all the time, Baggot says.
But MJ proposes that the plot is thicker than just leaving Huang's name off for higher royalties or more fame. Its argument relies on the Bayh-Dole Act, which went into effect mid-1981 and allowed universities to retain rights to products of government-funded research. Before that, the law said the government owned any technology invented with federal funds.
Huang relied for his sequencing work on federally funded grants that started as early as 1979 — a fact that is the foundation for MJ's argument that, because his work came before Bayh-Dole, the sequencer has technically belonged to the government all along. Foster contends that Huang was "a pre-Bayh-Dole-Act inventor. Even if there are other inventors [afterward], that vests ownership of the patents in the government."
And that would have been the primary motive for not crediting Huang as an inventor, according to Foster. "They knew that Henry invented the sequencer before the Bayh-Dole Act and therefore the government owned the whole thing," he says. "They didn't want Henry's name anywhere close to it. … By the time they filed that invention disclosure, they were very aware that this was going to be big."
Rubbish, says Tim Hunkapiller: it's ludicrous to think he and the other inventors could have predicted the Human Genome Project or anything else that made the sequencing market as lucrative as it is today.
Even if Huang's involvement can't be proven, the MJ lawyers plan to show that the Hunkapillers, Smith, and Hood used extensive federal funds to build the sequencer. These funds would be covered by Bayh-Dole so Caltech would still be the legitimate patent owner, but the law entitles the government to use the technology without paying a royalty. That alone could be worth millions if proven in court.
Caltech's team will rely on a particular point in Bayh-Dole that gives the government its royalty-free rights if federal funding is used to make an instrument, but does not give those rights if the federal funds are only used after the technology is functional. "The invention was reduced to practice without federal funds," Caltech's Asperger says. Hunkapiller says that, in fact, the team's requests for federal funds for the project were denied.
MJ lawyers, though, cite a slew of grants they say were used for the sequencer project and began before the sequencer was functional — which would give the government rights to that license.
Caltech has more than the Bayh-Dole Act to worry about, though. In 1988, ABI and the school signed a licensing agreement through which ABI can use the patent estate and sublicense it to others. The agreement still governs the ABI/Caltech patent relationship, and it expressly states that ABI must give the government or any government-funded party (start with the Human Genome Project to get a sense of how big this could be) a royalty discount above and beyond any volume discount they might be afforded.
Caltech argues that the government has gotten more of a break from ABI than the license agreement's royalty-free discount would have amounted to. ABI "gives them a nine to 10 percent discount — that's by far a bigger discount than any other purchaser from ABI receives," Asperger says.
But the language in the '88 licensing agreement appears just as unbending as MJ in insisting that the government deserved a separate discount beyond the lowered bulk price. ABI has said that about half of its sequencer and reagent revenues come from government or government-funded purchasers — over the years, say MJ lawyers, the money ABI owes the government as a result has racked up to the hundreds of millions. And MJ can sit back and enjoy the trial: as the lawsuit drags on, the potential damages toll rises.
The allegedly illegitimate patents lay the foundation for other major claims in the lawsuit. MJ argues that ABI used the patents to keep an unfair stranglehold on the sector, preventing competitors from building sequencers that would have driven down market prices. Charging the defendants as conspirators, MJ alleges that ABI wielded its licensed patent estate to keep out DuPont, Hitachi, and Li-Cor, among others. According to Tim Hunkapiller, ABI is obligated by its agreement with Caltech to license the patent estate — and usually does so at a royalty rate of a single-digit percent — and that there's no basis to the notion that ABI refused to grant licenses to competitors.
MJ goes so far as to say that ABI gave preferred access to its sequencers to sister company Celera, giving it an advantage in the genome race and a head start on gene patenting. MJ's suit charges that more than 6,500 gene sequences for which Celera has filed patent applications would otherwise have remained in the public domain — part of the relief sought by the suit would invalidate any of these patents should they be granted to Celera. Even TIGR is brought into the fray as a co-conspirator for enabling the monopoly: MJ cites as evidence the 1998 agreement between TIGR and Celera through which TIGR agreed to stop work on the human genome in return for Celera stock.
Monopoly charges could result in the defendants' losing the patent estate altogether — a move that could conceivably devastate ABI and hurt Caltech, which gets a significant amount in royalties from its ownership of the patents
Even if the suit winds up to a settlement, no one's expecting it to be over soon. "It'll be years of this crap," Hunkapiller says. "I'm pretty damn tired of this." The case, already picked over by industry insiders for years, is just getting started. At press time, MJ lawyers expected to serve their complaint to the defendants and officially start the discovery phase of document exchanges, conferences, and interrogations.
Caltech's Asperger isn't looking forward to years in the courtroom, but "we'll vigorously fight it, and we expect to win," he says.
MJ's Foster is just as confident. "I know what their story is, but it just doesn't hold water," he says. "It's my experience that a little blood has to be shed before the defendants do the right thing."
Cast of Characters: (x) Claims to be principal inventor of the automated DNA sequencer
Charles "Kip" Connell An ABI engineer still with the company, he was added under protest to the sequencer patent for his work with lasers for dye detection
John Finney An engineer from Harvard, he cofounded MJ Research, known for its thermocyclers and now the BaseStation, in 1987
Mike Finney John's brother, a biologist with his PhD from MIT, he was the other cofounder of MJ Research
Lee Hood Joined Caltech faculty in 1970, co-founded ABI to commercialize the protein sequencer in 1981; his lab became known as a hotbed for technology innovation
Henry Huang (x) A Hong Kong native who worked in Hood's lab from 1977 to 1982; he was the first to seriously pursue an automated DNA sequencer there before leaving to go into virology at Washington University
Mike Hunkapiller Chief technician of Hood's lab, he helped develop the protein and DNA sequencers; joined ABI, where he's now president, in 1983
Tim Hunkapiller (x) A computer scientist by training, Mike's brother worked in Hood's lab; now heads up his own biotech consulting firm, Discovery Biosciences, and still consults for ABI
Yuzuru Husimi A professor at Saitama University in Japan, his lab worked on an automated sequencer the same time Huang did in Hood's lab
Lloyd Smith (x) He joined Hood's lab in 1982 and was to pick up the sequencing project when Huang left. Now at University of Wisconsin, Smith started up his own company, Third Wave Technologies, which ABI tried to acquire in 2000
The lawsuit could be nearly as confusing as the issues behind it. In short:
MJ has sued ABI, Applera, Caltech, and Celera for 24 counts: violation of the False Claims Act and the Bayh-Dole Act, monopolization, conspiracy, breach of contract, and patent invalidity, among others.
If MJ can prove Henry Huang invented the sequencer: The technology should have belonged to the government under pre-Bayh-Dole law. Caltech never had rights to the patents, ABI never had rights to the license. MJ's asking that the government reclaim ownership of the patents and name Huang an inventor.
If MJ can show that people on the patent used federal funds: The government should have gotten a royalty-free license to the technology. Under the False Claims Act, MJ is arguing that every single invoice submitted to government parties for sequencers or reagents is fraudulent, each subject to a $10,000 penalty, and also is seeking repayment for the overcharges.
1977: Henry Huang joins Lee Hood's lab; Tim Hunkapiller enrolls at Caltech, later joins Hood's lab
August 1979: Huang and Hood apply for and receive first NIH grant for automated DNA sequencer project
1981: Lee Hood helps found Applied Biosystems to commercialize protein sequencer, developed with Mike Hunkapiller
July 1981: Bayh-Dole Act goes into effect, giving universities the right to keep patents on federally funded research
April 1982: Lloyd Smith joins Hood's lab
Late August 1982: Henry Huang leaves Hood's lab for WashU position
September 1982: Tim Hunkapiller attends Aspen, Colo., conference where Husimi talks about four-color automated DNA sequencer idea
October 1, 1982: Smith and Hunkapillers sign invention disclosure for automated DNA sequencing machine; Hood signs as witness; no prior art listed
July 1983: Mike Hunkapiller joins ABI; Hood, Smith, and Tim Hunkapiller would later sign on as consultants
November 28, 1983: Invention disclosure filed
January 16, 1984: Caltech files first patent applications for automated DNA sequencer; Hood listed as inventor
September 1, 1985: NSF grant awarded to Hood's lab for development of the sequencer; Hood and others claim that the sequencer was built and functioning when the grant kicked in
1986: False Claims Act amended to make it easier for people to sue on behalf of the government. Since then, the provision has returned almost $5 billion to the US
June 11, 1986: Caltech and NSF host press conference to announce working "sequenator"
October 1987: Caltech files amended patent naming Charles Connell as inventor
January 14, 1988: Caltech, ABI sign licensing agreement giving ABI a license to the patent estate and the right to grant sublicenses to other parties
October 15, 1990: Caltech files continuation patent that supersedes previous sequencer patents
December 15, 1992: Main patent approved for Caltech, due to expire in 2009
October 13, 1998: Secondary sequencing patent granted to Caltech, due to expire in 2015
1998-1999: MJ Research prompts DOJ investigation, saying ABI and others have cheated government of hundreds of millions on sequencer developed with federal funds
September 2000: MJ Research files lawsuit under seal naming ABI, Applera, Caltech, and Celera as defendants
March 2002: DOJ officially declines case, reserving right to intervene at any time
May 2002: With documents unsealed, MJ Research has 120 days to serve defendants and enter discovery phase of lawsuit