navigating the patent maze

Announcing …
…the latest newsletter of the Cambia Intellectual Property Resource. http://www.cambiaip.org/Newsletters/Newsletter/news_Q1_2004.pdf

In this issue:

Who Owns the Rights to Agrobacterium-Mediated Transformed Plants?
The longest-running patent dispute in plant biotechnology between the European Patent Office and inventors from the Max Planck Society ended in November 2003.

A Patent Application is Not the Same as a Patent
Explains the difference between a published patent application and a granted patent and how to easily identify which kind of document you are looking at.

The US Patent and Trademark Office (USPTO) now has an on-line searchable database of assignments for U.S. patents and published patent applications. Assignment data for pending (and unpublished) and abandoned applications are not available.

A tour of the search page (found here) shows the typical bare-bones style as their search page for patents. Searchable terms include:
Reel/Frame Number: (location of data on tape; casual users aren’t likely to know this)
Patent Number: (e.g. 6345127)
Publication Number: (refers to patent applications e.g. 200300012456 — must contain 11 digits)
Assignor Name: (usually the inventor name)
Assignor Index: (same as above — see below for difference between index and name)
Assignee Name: (usually a company or university)
Assignee Index: (same as above)
Assignor/Assignee Name: (this means the name could be either assignor or assignee)

Do not be fooled (like I was) by the underlining of the input terms, thinking that means they link to help. The help file, such as it is, is found by clicking on the identically underlined “online help”, which is at the top of the page. “Help” failed to help me understand the difference between ‘name’ and ‘index’.

After some trial and error, I deduced that ‘name’ means just that; the results contain all entries with the inputted name or begin with the inputted name. (Example: ‘camb’ will return ‘cambia’, ‘cambridge’, etc.) ‘Index’ returns an alphabetical list of names contained in the entire dataset, with the inputted name at the top of the list. (Example: ‘cambia’ returns ‘cambia’ ‘cambber’ ‘camden’ etc.) Clear as mud yet?

Remember that the assignee of the patent rights is only the owner of the rights. The assignee may license away, give away, or sell any or all of the rights it owns. If you want to know who actually controls the patent rights, this database will not inform you. While ownership is public knowledge, the transaction between the owner and the recipient of the rights is often held in secret. When the transfer of rights is done by contract, the privacy of the contract is respected by law.

Sometimes it is possible though to find out who has control of the patent rights by searching news articles, press releases, or by asking the owner (assignee). Good luck.

With fanfare, the US Patent and Trademark Office announced an “action plan” for biotechnology patent examinations that is aimed to improve patent quality. The five-point plan is directed at the way the Office (mis)handles multiple inventions.

For those of you unfamiliar with what is called “restriction requirement”, the Patent Office limits the number of inventions in a patent to one. When there are more than one invention, the patent applicant chooses one to be examined and the others are “restricted out.” But they are not lost. The other inventions can be pursued in patent applications, which are called “divisional applications”. While this sounds reasonable, the implementation of the requirement has been anything but reasonable in the opinions of many practicioners. For example, a nucleic acid sequence encoding a protein and the protein product itself are considered to be two different inventions. The Patent Office’s justification for this division is that there is too high a burden on the Examiner to search the prior art.

Yet, the burden doesn’t seem to be excessive in Europe, where the example cited would be considered a single invention. The European Patent Office has the same requirement — only one invention per patent — but uses a different standard, the “unity of invention” standard. Essentially, all the claims have to be tied together under a common concept. For most applications, this is not a difficult standard to meet.

The much-touted “action plan” of the USPTO is a step toward the unity of invention standard. The problem with the action plan is that the action plan doesn’t seem to be public. I followed links on the USPTO web site from page to page and back again without ever finding out any more information than the list of action items. If you want to try, go here.

U.S. Patent number 4,569,794, claims fusion protein purification technique known commonly as “His Tag”. This patent and its foreign counterparts was recently licensed to Inproteo, formerly known as the Indiana Proteomics Consortium, from Eli Lilly and Company. Inproteo has the exclusive right to issue sub-licenses. It is rumored to be in negotiations with several biotech companies. So far two licenses have been announced, one to Biogen (7 Oct 2003) and one to Genentech (21 Jan 2004). Terms are not publicized however.

U.S. Patent 4,569,794 issued to Eli Lilly on 11 Feb 1986. Its patent term expires 17 years from date of issuance — 11 Feb 2003. Thus, it appears that this patent is no longer in force. How is it possible then that Inproteo is licensing it? Court decisions in the United States have unambiguously declared that royalties cannot be collected on an expired patent. One answer may be that licenses are being granted for non-US patents. The expiration date of those patents would be 21 November 2005.

The first claim in the European patent EP 184355 B1 is:

A compound comprising a biologically active polypeptide or protein covalently linked to a peptide that is able to chelate an immobilized divalent metal ion and that has two to five amino acid residues, at least one of which is selected from the group of histidine and cysteine.

On the face of it, this claim appears to be rather narrow, requiring that the tag is from 2-5 amino acids. The corresponding claim in the U.S. lacked a limitation for length of peptide tag as well as lacked a limitation for the composition of the tag. Most researchers use a His6 tag, which falls outside the literal wording of the European claim, but wouldn’t have fallen outside the scope of the U.S. claim.

Another main claim in the European patent encompasses the method.

A process for separating a biologically active polypeptide or protein in the form of its precursor from a mixture containing said precursor and impurities, which comprises contacting said precursor with a resin containing immobilized metal ions, said precursor comprising the biologically active polypeptide or protein covalently linked to a peptide that is able to chelate an immobilized divalent metal ion, binding said precursor to said resin, and selectively eluting said precursor from said resin, and wherein
said peptide comprises at least one amino acid selected from the group consisting of histidine and
cysteine.

This claim does not have a limitation on the length of the tag sequence, only that at least one of the amino acids in the tag be either histidine or cysteine. The corresponding U.S. claim was similarly worded but again did not contain a limitation for the composition of the peptide tag.

It’s interesting that this very broad patent on what many may feel is a basic research tool has escaped the attention of either the press or the “anti-patent” groups.

On Jan. 20 the Supreme Court of Canada will decide if Monsanto’s patent on a process that produces genetically engineered (GE) canola (oilseed rape) is valid. Unlike in the United States, the Canadian court has previously ruled that plants and animals cannot be patented. Many other countries also do not allow patenting of plants or animals.

Briston-Myers Squibb, a US pharmaceutical firm, has given up its patent in Thailand protecting the anti-retroviral drug didanosine (ddi) in tablet form, marketed under the tradename Videx. The patent was granted in 1998. Activists in Thailand had pursued legal action for two years to get the patent revoked. About one million Thais have been infected with HIV over the last 20 years.

The drug has been produced in powdered form by the Government Pharmaceutical Organization, but the tablet form has advantages of convenience and consistant dosage. Now, any factory in Thailand will be able to produce the drug in tablet form. The price is expected to drop from 3,000 baht (about USD 77) to less than 900 baht per month. The drug is used in combination with other drugs to treat AIDS.

In USA Today on 18 January 2004, there was yet another article exposing “bad” patents, especially in the fields of software and business methods. The press seems to thrive on splashing across headlines every instance of small players being asked to cease and desist because of a patent. While the press is probably right about poor standards being applied in examining patents in these fields, the solution or conclusion they seem to favor reporting is unreasonable in my opinion. That conclusion is that patents are “evil” and stifling innovation.

The issue has become serious enough to discourage some small software and biotech firms from innovating for fear of being sued. No new companies are entering mainstream agricultural biotechnology because of overly broad patents on genes and research methods that limit their lab work, says Brian Wright, economics professor at the University of California-Berkeley

As highligted in the quote, biotechnology (and especially agricultural biotechnology) suffered similar problems when filing patent applications in that field started to skyrocket. Currently, patent claims in biotechnology tend to be quite narrow. It would be prudent for the US Patent Office to learn from that and slow down the granting of patents in software and business methods until the examiners are better trained and patenting standards increased.