Arabidopsis Genome

This analysis was prepared by Dr. Brett Baillie and Dr. Marie Connett Porceddu using patent application and patent data provided by Neil Bacon and Doug Ashton from within a larger project funded by the OS4 Initiative (link) .  We are grateful for production assistance on many of the pages here by Dr. Kerry Fluhr, Dr. Wei Yang , Dr. Nick dos Remedios, and Nik Hatta.

Patent Landscape of the Arabidopsis genome and related plant species

This landscape deals with the Intellectual property and IP issues surrounding the genome of Arabidopsis thaliana. The landscape is intended to provide insights into the deliverability of innovations in the plant sciences for the public good. To this end the landscape investigates patenting activity directed towards the nucleotide and gene sequences from Arabidopsis - with a particular focus on patent applications involving large groups of DNA sequences. The results of the analysis will be of interest to those involved in public sector research and innovation who are unaware of the ways in which the fruits of public-funded research can be sequestered by private for-profit organisations.

Patent applications that claim Arabidopsis genes have the potential to also claim related genes from other organisms. Hence granted patents on genes in Arabidopsis may effectively grant exclusionary rights to similar genes from important industrial and subsistence crops such as canola, soya, potato, cassava, cotton, poplar and Eucalyptus.

Understanding the landscape of patents on Arabidopsis sequences also allows a better understanding of the degree to which these overlapping patent rights may influence future research and innovation. If findings about the function of Arabidopsis genes provide stronger enablement for composition of matter claims already filed, research done in the public sector may not actually be in the public domain or available for public good uses!

What this Landscape is about

This landscape will provide an overview of some important aspects of the intellectual property claims covering the Arabidopsis genome, which many in the biological sciences consider to be a public good, a model system that enables critical insights into many of the important pathways of crops, weeds and other plants.  In particular, we investigated "bulk sequence claims", patent applications that attempt to claim  100 or more Arabidopsis sequences, and found a large number of such patent applications by Monsanto, Syngenta, Dow, Dupont, Bayer, and BASF and their subsidiaries and researchers who work with these companies under exclusive contracts.

What this Landscape is NOT about

This landscape is not intended to provide a legal opinion about freedom to operate using Arabidopsis sequences.   The nature of patenting systems worldwide means that new patents and patent applications may appear at any time.  Similarly, patents may lapse or patent applications may expire or be replaced by new applications.  During the course of preparation of this landscape in early 2007, a series of new publications appeared, providing information for the first time on applications encumbering thousands of sequences, filed up to seven years earlier.  This provided a firm reminder that it is not possible to take for granted that if a topic is not readily found in a search of published patents and patent applications, there is freedom to operate!  So, although we have tried to give the best coverage of the intellectual property surrounding Arabidopsis sequences, this landscape should not be viewed as a comprehensive coverage of the subject.

What You Need to Know About Patents

Below are some important patent law concepts that are important for understanding the Rice Genome Landscape:

1. The claims define the scope of protection of a patent.

2. A patent application is not the same as a patent.

3. There is no such thing as an international patent.

4. The entity that has the patent rights may not be the owner listed on the patent document.

5. Individual patents have a limited lifetime.

What is Arabidopsis thaliana?

Arabidopsis thaliana is a model plant used in research to study many aspects of plant biology.  Originating in Europe and Central Asia, it is a dicot, as are many important and culturally significant staple crops, such as potato;  commercially important food crops, such as soybean; and fiber crops, such as cotton and hardwood trees.  Unlike many other members of the family Brassicaceae (the same family contains canola, broccoli, cabbage, mustard and other food and oilseed crop species), Arabidopsis thaliana is not grown for food or oil production, and is not agronomically important in itself.  Plants are quite small (see Figure 1 below).

Although in the genus Arabidopsis there are approximately 9 species and 8 sub-species (see the entry for Arabidopsis in Wikipedia), it is not uncommon for plant biology researchers to use "Arabidopsis" to mean Arabidopsis thaliana, the convention used in this landscape.

Figure 1: Arabidopsis thaliana seedlings growing in tissue culture (seedling ~10mm in width)

The popularity of  as a model for all plant processes has given rise to numerous methods and protocols for its use in research. Because germination through to senescence is only approximately 50 days (The Arabidopsis Information Resource (TAIR)), Arabidopsis offers a fast system in which to study processes that may take months or years in other flowering plants. The availability of large numbers of mutants and T-DNA insertion lines (e.g. The European Arabidopsis Stock Centre, NASC) has made it useful in understanding the role of many plant genes, particularly those involved in development, metabolism, and disease resistance.

Arabidopsis was chosen as the first plant species for a public whole genome sequencing effort because of its small genome size. The table below shows the genome sizes of some commonly-studied organisms, including Arabidopsis.

Organism	Common Name	Genome Size
Escherichia coli K12	E.coli	4.64 Mb
Saccharomyces cerevisiae YJM789	baker's yeast	16 Mb
Arabidopsis thaliana	thale cress	120 Mb
Oryza sativa	rice	430 Mb
Lycopersicon esculentum	tomato	950 Mb
Zea mays	maize	2365 Mb
Homo sapiens	man	3038 Mb
Triticum aestivum	wheat	17000 Mb

(Genome sizes were taken from the NCBI's Genomic Biology web pages.)

The future of Arabidopsis thaliana research?

The availability of sequence data for the Arabidopsis genome and EST projects has ensured that Arabidopsis is now more widely used than ever as a model for the study of flowering plants, and new or improved enabling technologies are being applied to it as they become available for research.  For example, with the availability of genomic sequence and EST data, gene silencing or RNAi technology is being used as a research method to determine the function of Arabidopsis genes.  The Arabidopsis Genomic RNAi Knock-out Line Analysis (AGRICOLA) project was started in 2002 with the goal of providing the scientific community with research tools (specific vectors and cloned Arabidopsis DNA) necessary to study gene function via gene silencing. 

Understanding the patent landscape surrounding Arabidopsis is important in planning research that aims for application in products, because of the continued importance of Arabidopsis as a research tool to many thousands of scientists. Furthermore, use of the genomic information of Arabidopsis in patents and patent applications has resulted in broad claims for patent coverage of genes in other flowering plants.

Khush GS (1997) Origin, dispersal, cultivation and variation of rice. Plant molecular biology 35 (1-2), 25-34 (Sep 1997)

The Genomic Structure of Arabidopsis thaliana

See background information about the structure of chromosomes. The genome of Arabidopsis thaliana consists of 5 chromosomes, one circular mitochondrial DNA, and a circular chloroplast DNA. Chromosomes are often drawn as in Figure 1 below: with a short and long arm and an intervening centromeric region. Figure 2 depicts the shapes and relative sizes of the A. thaliana chromosomes. Table 1 shows the size (in Mbp) and predicted number of genes for each chromosome. 

Figure 1	Figure 2

Table 1

Arabidopsis thaliana Genomic Sequencing

The Arabidopsis Genome Initiative (AGI) was set up in 1996. Consisting of 6 groups, it's mission was to sequence the genome of Arabidopsis thaliana by 2004 at high accuracy. Various aspects of the sequencing mission were divided between the 6 groups:

The sequencing project was completed in 2000, four years ahead of schedule.

The sequence derived from the AGI's efforts were the subject of a number of publications in nature, including:

• Sequence and analysis of chromosome 2 of the plant Arabidopsis thaliana (Nature. 1999 Dec 16;402(6763):761-8)
• Sequence and analysis of chromosome 4 of the plant Arabidopsis thaliana (Nature. 1999 Dec 16;402(6763):769-77)
• Sequence and analysis of chromosome 1 of the plant Arabidopsis thaliana (Nature. 2000 Dec 14;408(6814):816-20)
• Sequence and analysis of chromosome 3 of the plant Arabidopsis thaliana (Nature. 2000 Dec 14;408(6814):820-2)
• Sequence and analysis of chromosome 5 of the plant Arabidopsis thaliana (Nature. 2000 Dec 14;408(6814):823-6)
• Analysis of the genome sequence of the flowering plant Arabidopsis thaliana (Nature, 408:796-815).

Before, during, and subsequent to the AGI's efforts, groups, including some of those involved in the genomic sequencing, were involved in sequencing Expressed Sequence Tags (ESTs) from Arabidopsis. The following companies and institutes were involved in these efforts:

• Ceres Inc (~5000 ESTs deposited with TIGR)
• Kazusa DNA Research Institute, 1532-3 Yana, Kisarazu, Chiba 292-0812, Japan (12,028 non-redundant ESTs)
• TIGR
• Cold Spring Harbor Laboratory
• Washington University School of Medicine, Genome Sequencing Center
• Laboratoire de Biologie Cellulaire, INRA, Versailles Cedex, France/CNRS/ consortia (Plant J. 1993 Dec;4(6):1051-61; Plant J. 1996 Jan;9(1):101-24)
• Department of Energy Plant Research Laboratory, Michigan State University and Consortia (Plant Physiol. 2000 Dec;124(4):1582-94)

Today there are more than 622,000 Arabidopsis ESTs in the dbEST database at the National Center for Biotechnology Information (NCBI).

Patenting of genes and sequences from Arabidopsis

Analysis of the recent history of patent documents involving Arabidopsis provides some interesting information regarding the evolution of present day patenting strategies for organisms whose genomes are available. It also offers the opportunity to investigate Arabidopsis as a "model" for future patenting of other plant genomes.

The following simple search was performed using PatentLens to search the term "Arabidopsis" in full text patent documents (detailed search parameters here). Although it must be emphasized that this search is likely to overestimate the number of patent documents making claims to Arabidopsis "inventions" (it will also uncover documents making reference to Arabidopsis), its main purpose should be seen as a rough measure of the impact of Arabidopsis on patent publications.

One of the caveats with the analysis above is that it does not focus on patents making claim to sequences from Arabidopsis.  Such an analysis is difficult due to the fact that more than one sequence can be claimed in a single patent or patent application.  Hence a plot of the number of Arabidopsis sequences in claims against the year of application may look very different from a plot of patent documents containing claimed Arabidopsis sequences versus year of application.

Hybridization Language in Patent Claims

There are many ways to define the limits of a claim towards related genes, nucleotides, or protein sequences. One way is through a direct comparison of two aligned sequences in order to determine if they are "similar" enough to each other so that the claim language is satisfied. Typically this requires a quantifiable comparison, often in the form of "% identity":

US 2005/50583 (assigned to Genesis/Agrigenesis, according to press releases now subject to the rights of ArborGen) Claim 3

An isolated polynucleotide comprising a sequence selected from the group consisting of:

(a) sequences having at least 75% identity to a sequence of SEQ ID NO: 1-67, 131-481, 833-888, 946-952 and 960-974;

In this case if a compared sequence (even a sequence not in the possession of the patentee, such as one discovered in another species) falls within the 75% sequence identity range, it could potentially be subject to the patent applicant's claims.  Alternatively, subject to a number of detailed caveats (for the US see 35 USC 102), if a compared sequence more than 75% identical was made public or submitted for patent by someone else prior to the priority date of the patent application ("prior art"), the patent claim may be invalid. 

Such a comparison is technically straightforward, although for a long sequence listing with more than a handful of sequences (almost 500 are claimed here) it would require a large amount of time for the examiner to do a thorough analysis to determine whether or not the claim is valid, and for a researcher or a member of the public to determine whether any particular sequence might be infringing.

Even more difficult for the patent examiner and the public to compare, to clearly define the bounds of what is subject to patent claims, is the use of such claim language as:

US 2005/91708 (assigned to Dow Chemical) Claim 1

An isolated nucleic acid selected from the group consisting of SEQ ID NOs: 1-7554 and nucleic acid sequences that hybridize to any thereof under conditions of low stringency, wherein expression of said isolated nucleic acid in a plant results in an altered metabolic characteristic.

Hybridisation experiments would seem to be a definitive way to define similarity in patent documents. Unfortunately, the tools to perform fast and accurate comparisons of sequences based on such language, without the use of a research laboratory, are lacking in the patent office, and what is more, a skilled practitioner will realise that many (possibly only distantly related!)  nucleic acid sequences will hybridise together under low stringency conditions.   At some frequency, any DNA at all will hybridise to the claimed sequence under any stringency conditions. 

Hence hybridisation language, when used in patent documents, has the potential to create uncertainty in understanding the scope of the claim.  The probability that an applicant will achieve such claims to all 7554 sequences is small.  However, it is possible not only that some of these sequences will be successfully claimed in issued patents, but that most remain pending in future applications. The uncertainty with this claim is two-fold:

• Which sequences are going to be successfully claimed in future?
• Will the applicant also acheive claims to related sequences?  On the face of it, this claim would apply to sequences from any other species, not necessarily a plant species, and not necessarily sequences in the possession of the patentee, that hybridise to any of the 7554 claimed sequences and alter anything about a plant when expressed in the plant.

Below we explain, for those unfamiliar with the concept, how hybridisation works and how it is often presented in a patent application or patent.

The Basis for Hybridisation Language

DNA is a biological polymer consisting of "simple" repeated subunits, or bases.  It is within the sequence of these bases that genetic information is stored. Although DNA can exist as a single strand in solution, it is generally found within living systems together with its complementary strand (see the figure below).  If we consider a simple representational diagram of a DNA molecule, consisting of a sequence of bases: A, G, C, and T. We can see that some of the bases interact with one another in the figure below.

The bases A,T,G, and C form hydrogen bonds with their respective bases T, A, C and G, in the complementary strand.  These bonds provide a stability to the double strand DNA (dsDNA), under normal circumstances, preventing it from dissociating into single strand DNA (ssDNA).  Significantly, dsDNA can be "melted" to produce ssDNA under certain physical circumstances and can be influenced by many factors such as temperature, salt concentration, inorganic solvents, sequence composition, and (importantly) the degree of sequence similarity or missmatch between ssDNA molecules.

The Concept of Melting Temperature (Tm)

Mathematical relationships between the various physical parameters listed above and the melting point of DNA have been established.  Historically, the point at which 50% of the DNA molecule is ssDNA is referred to as the melting temperature or Tm.

The Tm for a DNA sequence can be estimated via the commonly-used calculation:

Where     Tm         = melting temperature in ^oC
               [Na⁺]     = Molar concentration of sodium ions in
            %[G+C]    = percent of G+C bases in DNA sequence
                 n         = length of DNA sequence in bases
                 P         = temperature correction for % mismatched base pairs (~1^oC per 1% mismatch)
                 F         = correction for formamide concentration (= 0.63^oC per 1% [formamide])

Note:  A particular DNA sequence will have different Tm values under different conditions.  This calculation is sometimes referred to as the effective Tm.

The fact that ssDNA can re-anneal to form dsDNA is used in the process called hybridisation. 

NOTE that the Tm refers to 50% hybridisation, not complete hybridisation!

A typical hybridisation experiment consists of a number of steps:

1. Binding of target ssDNA to a solid support (often a thin nitrocellulose, nylon, or polymer membrane), referred to here as a "blot".
2. Blocking any remaining free binding sites for DNA on the blot.
3. Labeling of the "query" ssDNA or probe.
4. Creation of a hybridisation solution containing the probe at a known concentration of salt, formamide,...
5. Equilibrating the "blot" with the hybridisation solution at a defined temperature (for a specified length of time).
6. Washing the unhybridised probe off the blot using a specific wash protocol;  the wash conditions are actually more important than the hybridisation conditions for determining what will remain hybridised, but are less often specified in patent applications!
7. Detecting the bound probe by some means.

If the probe (or query DNA) is similar enough to the target ssDNA (on the blot) then the probe will remain bound to it after the wash step.

"Stringency" refers to the concentrations and temperatures in steps 4-6 above, and it is most often "strigency" that is referred to in order to determine the scope of claims.  Strictly speaking, this should refer to the "stringency" of the wash step required to leave the probe bound to target sequences covered by the claims:  the "query" is often a sequence mentioned in the claims section (as in the example given above).

Although the table below is not legally binding and indeed patent applications may make their own definitions of stringency levels, the typical usage is roughly:

Stringency	[NaCl] (Molar)	Range from Tm (oC)	Description
High	0.0165 - 0.0330	5 to 10oC below Tm	Only highly similar sequences will bind under these conditions (typically >95%)
Medium/Moderate	0.165 - 0.330	20 to 29oC below Tm	Those sequences above and less homologous ones (typically >80%)
Low	0.330 - 0.825	40 to 48oC below Tm	Sequences with lower homology will bind under these conditions (>50%?)
1. Note that the values here are for the purpose of discussion.  Actual values and conditions must be acquired from the patent document specifications. 2. The "Range from Tm" is the temperature of the wash solution required to achieve the level of stringency and is dependent on the factors described above. 3. Values in Column 2 are typical only! 4. Values in column 3 were obtained from US 2006/0057724 A1 5. Values in Column 4 are typical only and sourced from the web.  Refer to patent specifications for your details! 6. Definition of stringency levels may vary between patents (BEWARE). Stringency conditions are usually defined in the patent document!

Sometimes the patent examiner will require the applicant to stipulate the stringency conditions: high/medium/or low stringency, or specifying the [NaCl] and temperature of the wash conditions.  Few bother to give a reference with respect to Tm (a value that can be calculated).  It is possible that the conditions defined as stringent in one patent document may not be as stringent as those defined in another!

Summary

Arabidopsis is an extremely important model system used to study many aspects of plant biology. From the graphs and figures presented in this chapter, it is obvious that there are a great many references to Arabidopsis in the scientific literature and in patent documents. These observations are consistent with the importance of Arabidopsis as a model organism for the study of plant biology. The interesting aspect of the appearance of "Arabidopsis" in the patent literature is the fact that Arabidopsis is NOT a commercial crop. Compared to important crops such as canola, soybean, and potato there would appear to be little value in patents directed towards Arabidopsis.

While the term "Arabidopsis" appears frequently in patent documents, it is the contents of the "claims" section of these documents that is most important. Since the focus of this landscape is towards the Arabidopsis genome, it is of interest to understand to what extent these claims are directed towards sequences and genes from Arabidopsis. Also of interest is an applicant's reasons for claiming sequences from an organism with little commercial value. In the following chapters we will analyse example patent applications and granted patents more closely to answer some of these questions.

Since patents containing claims to Arabidopsis sequences are the focus of this landscape, much of the analysis in the following chapters will involve patent documents making claims to large groups of genes and sequences from Arabidopsis. These patent documents will be described here as "bulk sequence" applications. For the sake of convenience and clarity, we will define "Bulk Sequence" applications as those that claim protection for 100-or-more Arabidopsis sequences. It is these patent documents that form the core of the following patent analysis.

Bulk Sequence Applications

Within the last decade there has been a trend towards attempts at patenting large groups of nucleic acid sequences from sequencing of EST libraries and genomes. There are “bulk sequence” claims over large portions of many genomes, including that of Arabidopsis.  Since Arabidopsis was the first plant to have its genome sequenced, is a popular model organism, and there are extensive expressed sequence data (ESTs and cDNA sequences) available, examination of the patent landscape of Arabidopsis offers the opportunity to examine what may happen to other plant genomes. 

Bulk sequence applications are important because they potentially cover genes involved in a very wide variety of processes in development, metabolism, and disease resistance. Hence it is a concern to examine the success of such patent applications in granting exclusionary rights to early applicants who can control what can be done with later research in these far-reaching areas.

The effects are not limited to those working on Arabidopsis.   Arabidopsis is not a crop plant.  If the economic gain to be had from patents on Arabidopsis genes was limited to royalties from Arabidopsis growers, there would be few patent applications pending on the genome of this plant.  Most who work on it are using it as a model for important related species such as soybean and cotton.  The claims in many of the bulk sequence applications described below would allow the rights-holders of patents on them to exclude all others from the use of genes not only in and from soybean and cotton, but many other leguminous food crops, fiber crops including trees such as Populus, fruit crops and ornamentals, and subsistence crops such as cassava and cowpea that are staples in the diets of millions of people. 

Thus Arabidopsis offers a further dimension as a plant model: as an example of plant genome patenting.

Patenting Organisations

Historically there have been a number of companies interested in the genomic sequences and expressed sequence tags (ESTs) from Arabidopsis.  Many companies, including Paradigm Genetics Inc. (now Icoria Inc.), Ceres Inc., Mendel Biotechnology Inc., Syngenta, Dow Chemical, and BASF have applied for patents over Arabidopsis sequences.

The ownership of rights to Arabidopsis sequences is complicated by the pattern of mergers, acquisitions, and exclusive research licenses that give control of the intellectual property to entities different from those named as the patent applicants.  In some cases when the ownership of a patent or patent applications changes hands, an assignment of rights is filed with the patent office, but in the U.S. there is no requirement that rights controlled through exclusive licenses be publicly recorded.

Monsanto has equity stakes in or controls the rights over the intellectual property resulting from exclusive research contracts with Mendel, Ceres and Paradigm/Icoria, as noted in press releases.  Monsanto has actually acquired a portion of Paradigm/Icoria (Agrinomics) but in this acquisition and in the exclusive research agreements, information about the extent of the control over the intellectual property has not been made public.  Monsanto has also had extensive historical ties to the Arabidopsis genome sequencing project, and has extensive bulk sequence claims over genes from other species such as cotton and soybean.  Monsanto's efforts to patent ESTs from maize were rebuffed in a court case suggesting a higher standard for utility should be applied for sequence patents (re Dane Fisher case; See also this commentary link), but hundreds of thousands of sequences remain in patent applications over Arabidopsis as well as soybean, maize, rice, and other crops.

Syngenta also controls intellectual property claims through granted and pending patents nominally assigned to a number of companies that have been partially or completely acquired by Syngenta or its parent companies, and through research alliances.   Nominal assignees of Arabidopsis intellectual property probably controlled by Syngenta include Torrey Mesa, Zeneca, Ciba Geigy, Novartis, Advanta, Garst, Agripro and Danisco. 

See bar graphs showing the portions of the Arabidopsis genome that are covered by patent applications and granted patents that are assigned to the following companies and others with whom they have equity stakes and exclusive research alliances:  DuPont, Monsanto, Paradigm Genetics, and Syngenta.

Legal Aspects of Bulk Sequence Claims

Much of the focus of patenting attempts and granted patents for Arabidopsis genes (and those of many other organisms) have been towards patent protection in the US. Other countries, such as Australia, often use the US as a model for guiding their own patent office procedures. Hence, when considering the legal aspects of bulk sequence claims, we will be referring to the current state of affairs in the US and in particular the actions of the United States Patent and Trademark Office (USPTO) towards bulk sequence applications. To understand the basis for patent protection in the US, one must understand that:

The default, with respect to applications, is that the applicant is entitled to the patent, and the applicant is given the opportunity throughout the examination process to edit the claims of the application to maximise the chance that a patent is issued. Hence the onus is on the patent office to prove that a patent application fails to meet the criteria for patentability. Patent examiners have a limited amount of time to spend examining an application.

It can be argued that such bulk sequence applications represent attempts to patent "discoveries" (which are not patentable) rather than "inventions". Specifically, the applicant has added to the amount of human knowledge by disclosing something that occurs in nature, rather than adding to human knowledge through the creation of something new (a useful invention). Therefore, it could be argued that bulk sequence applications lack an inventive step.

For an application that claims many sequences, applicants have historically argued that novelty, inventiveness, and utility are met by such large sequence applications. They have done so by pointing out that such sequences can be used (inventiveness and utility) for particular purposes, including:

• as a transgene to modify the phenotypes of other organisms
• as probes to help locate homologous sequences in other organisms
• to design primers for use in amplifying the gene, in diagnostics,...

These proposed uses bring to light another problem with such groups of sequences: Enablement.

The disclosure of information made in "payment" for the grant of a patent must enable a person "skilled in the art" to use the invention. A bulk sequence application may fail to adequately address this issue, since it would be extremely difficult for any applicant to give specific instructions on how to use all the sequences for the purposes listed above.

The USPTO has responded to attempts to patent large groups of sequences by strictly adhering to the requirement for enablement and substantial utility for such applications.

Recently, the USPTO's commitment to these standards were challenged by Monsanto, who had previously applied for patent protection for 32,236 maize ESTs (US patent application 09/619,643), and were refused a patent. Monsanto appealed the USPTO's decision in the Court of Appeals for the Federal Circuit. The appeal was denied by a majority of 2 out of 3 judges, based on a lack of utility and enablement (re Dane Fisher case; See also this commentary link).

Since the re Fisher case was not a unanimous decision, a question still remains as to whether-or-not future large sequence claims will be successful.

Another important point to note about granted patents:

This fact, in addition to the high cost of contesting a granted patent (generally around one million US dollars), provides an effective barrier to those interested in challenging a patent's validity.

Summary

Bulk sequence applications have (and are still) used to make claims to large numbers of sequences from an organism (including many animals, bacteria, and plants). There is no upper limit to the number of sequences that can be claimed within a single patent application within the US system. Some plant applications may contain more than 100,000 sequences. Hence the success of even a single bulk sequence application would mean that a large portion of a plant genome would be protected by patent claims. Another interesting point is that it is common for companies to license technologies that have a "patent-pending" status (i.e. that have not yet received a granted patent). Hence the protection that might be derived from a granted patent is incentive enough to provide protection in itself. Therefore patent applications (including bulk sequence applications) are also important when considering issues of deliverability, innovation and freedom-to-operate.

Whilst Monsanto's efforts to patent large numbers of maize ESTs were recently unsuccessful, examples of granted bulk sequence patents exist, and applicants continue to pursue bulk sequence applications at the USPTO.

In the following chapters we will investigate bulk sequence applications for Arabidopsis genes.

Mendel Biotechnology Inc.

Mendel is a US company located in Hayward, California. Deriving its name from the "father of genetics", Gregor Mendel, the company was founded in early 1997 with the idea that controlling gene expression would lead to new ways of manipulating plant growth and development.

Mendel's main interest in Arabidopsis is the 1,800-or-so transcription factors that control the expression of ~27,000 Arabidopsis genes. Accordingly, Mendel has applied for patents over a large number of Arabidopsis transcription factors. Presumably this strategy relies on the fact that control of expression of genes via transcription factors is one way in which useful new phenotypes may be engineered into flowering plants (including many agronomically important crops). Since transcription factors are expected to play important roles in plant development, it is expected that Arabidopsis transcription factors may be functionally and structurally similar to, and highly homologous with transcription factors of other flowering plants.  Thus a claim on Arabidopsis transcription factors may also effectively be a claim over the same gene in many other flowering plants.

Transcription Factors

RNA polymerases (RNAPs) are enzymes involved in the process of transcription (transcribing an RNA molecule using DNA as a template).  While prokaryotic RNAP can recognise gene promoters and initiate transcription from the proper start point, the efficiency of the process is further controlled by trans-acting proteins called Transcription Factors (TFs). In both eukaryotes and prokaryotes, the process of transcription is tightly regulated for many genes and one way in which the cell is able to control transcription is through the use of TFs. In eukaryotes RNAP II requires the presence of a group of auxiliary proteins (general Transcription factors) to be able to recognise gene promoters. Hence production of mRNA in eukaryotes requires the presence of general TFs, and may also be further controlled by the presence of gene-specific TFs.

TFs are proteins that bind to specific DNA sequences within the promoter or enhancer regions of a gene and in so doing are able to control the transcription of the gene. They may either decrease (repressors) or increase (activators) the level of transcription. They are often involved in the final step in a signal transduction pathway and can be activated or deactivated by other proteins higher in the pathway. TFs can be recognised from EST and genomic DNA sequencing projects based on the fact that they often contain sequence motifs such as: zinc fingers, helix-turn-helixs, and basic-helix-loop-helix (for DNA binding) and leucine zippers (for interaction between TFs).

Patent Documents Involving Transcription Factors

We have chosen one patent application by Mendel Biotechnology as an example of its patenting strategy towards transcription factors.  The patent application US 2004/0019927 A1 attempts to claim a large group of Arabidopsis transcription factors and related sequences in plants.  This is the largest of the Mendel patent applications for bulk sequences that we were able to find (although there may be more), others also exist that claim either individual or small groups of TFs. Analysis of granted TF patents are discussed in Chapter 7.

Mendel Biotechnology Announces Agreement With Monsanto

HAYWARD, Calif., July 10 /PRNewswire/ -- Mendel Biotechnology, Inc. ("Mendel") today announced a new research and commercialization partnership with
Monsanto Company ("Monsanto"). The new program represents a significant expansion of the ongoing collaboration between the companies and extends
the agreement through 2011. Mendel and Monsanto have agreed not to disclose financial details of the agreement.

Monsanto and Mendel will continue to collaborate on the commercialization by Monsanto of technologies developed by Mendel. In addition, the
companies will initiate a systems biology program to develop an integrated framework for predictive control of plant gene expression that is anticipated
to streamline discovery and product development activities...

This new partnership extends the relationship initiated in 1997, under which Mendel executed a functional genomics program in the model plant
Arabidopsis focused on understanding the factors that control plant gene expression. That program identified a number of genes that have been shown to
enhance the yield of major crops in extensive field trials and that are in early commercial development by Monsanto.

Press Release    Source: Mendel Biotechnology, Inc.    July 10, 2006

"Four years ago, Chris Somerville, head of plant biology at the Carnegie Institution of Washington, slipped a gene for making plastic into Arabidopsis, a type of mustard plant. The gene turned the plant into a biological plastics factory.  Now, Monsanto Co. scientists are turning the concept into commercial reality....Just as exciting is a recent discovery by Calgene [now owned by Monsanto] scientists of the gene for the enzyme controlling the formation of cellulose in plants. After 30 years of fruitless biochemical search for the enzyme, ''this is our first break in understanding how to control biomass,'' Somerville explains. Genetically boosting the enzyme could make it possible to create trees with much higher proportions of cellulose--the plant kingdom's structural fiber--and less than the normal amounts of other cell wall components. Because these secondary components are what make the pulp- and papermaking process polluting and inefficient, scientists say, the engineered trees could help clean up a major industry. Beyond that, ''there is a mad scramble inplant biology to find the most useful genetic sequences,'' says Somerville. ''The world hasn't even seen the tip of the iceberg.'’ ‘“THE BIOTECH CENTURY”, Business Week, 1997 Chris Somerville is still at the Carnegie Institution, but is also the CEO of Mendel.

This diagram identifies a locus in the Arabidopsis genome with many patent applications covering genes in a pathway important for biofuels.   Mendel's patenting (and that of other large companies such as DuPont) is not only over crops critical to worldwide food security, but also industrial uses that could add to the prosperity of developing countries if they were available to exploit.  Who gets to exploit them?

www.mendelbio.com/news/

Moderna, Monsanto link up in new venture

FOCUS - Moderna, Monsanto link up in new venture Reuters, Monday, November 24, 1997 at 20:09 By David Luhnow

MEXICO CITY, Nov 24 (Reuters) - Mexican conglomerate Empresas La Moderna (MEX:MDA.A) and U.S. agricultural giant Monsanto Co (NYSE:MTC) on Monday said they had bought a stake in newly formed agricultural genetic research company Mendel Biotechnology Inc. Each company paid $15 million to fund a five-year research and development project at Mendel, a start-up firm based in San Francisco and headed by some of the world's leading plant genome experts, according to La Moderna officials in Mexico. In return, Moderna and Monsanto each get a 20 percent equity stake in Mendel and rights to develop and commercialize Mendel's technical capabilities in the next generation of value-added agricultural products.

Both companies also have an option to buy an additional 10 percent stake in Mendel at an undisclosed future date, La Moderna company spokesman Dieter Holtz told Reuters. Each firm will have rights to Mendel's research in the areas they are interested in and dominate, officials said. Monsanto will focus on Mendel's advances in agronomics, or the capabilities in plant genetics and genomics for many crops, including corn, soybeans, fruits and vegetables. "This collaboration gives us the abilty to better understand the function of specific genes in plants, thereby allowing us to more quickly introduce crops with improved agricultural traits," said Ganesh Kishore, Monsanto's assistant chief scientist and chief biotechnologist. Kishore said the venture will greatly reduce research and development time for products that increase yield or otherwise enhance the growing, processing or nutritional characteristics of food.

La Moderna, the world's largest vegetable seed maker, will have access to Mendel's research in vegetable seeds as well as fresh and processed fruits and vegetables. "We will each focus on our core businesses: Monsanto in agronomics and (its herbicide) Roundup and Moderna in vegetable seeds and fruits and vegetables," Holtz said in a telephone interview.

Earlier this year, Moderna sold its Asgrow agronomics division to Monsanto for about $240 million in order to focus more on the Mexican company's strength in vegetable seeds. Mendel's president and CEO is Michael Fromm, a former research director at Monsanto, and its chairman of the board is Christopher Sommerville, director of the Department of Plant Biology at the Carnegie Institution of Washington. "These are the best of the best in the field," Holtz said. The top researchers at Mendel were responsible advances in areas such as disease resistance, the molecular genetics of nitrogen fixation in plants and the genetics of photosynthesis, he added.

"We didn't want to have majority control of the company because we didn't want to cramp their entrepreneurial spirit. We thought it was better to see what they came up with on their own," Holtz said.

Copyright 1997, Reuters News Service

Monsanto, Empresas in Joint Venture with Mendel

MONSANTO, EMPRESAS IN JOINT VENTURE WITH MENDEL Reuters, Monday, November 24, 1997 at 16:22 ST. LOUIS, Mo., Nov 24 (Reuters) - Monsanto Co (NYSE:MTC) and Empresas La Moderna S.A. (MEX:MDA.A) said Monday they signed a broad technology collaboration agreement with Mendel Biotechnology Inc in the field of agricultural functional genomics.

The agreement gives Monsanto and Empresas exclusive access to Mendel's technical capabilities in plant genetics and genomics for many crops, including corn, soybeans, fruits and vegetables.

As part of the agreement, Monsanto and Empresas will each acquire a substantial minority equity interest in Mendel and fund a research and development program over five years. Financial terms of the agreement were not disclosed. "This collaboration gives us the abilty to better understand the function of specific genes in plants, thereby allowing us to more quickly introduce crops with improved agricultural traits," said Ganesh Kishore, Monsanto's assistant chief scientist and chief biotechnologist.

Kishore said the venture will greatly reduce research and development time for products that increase yield, or otherwise enhance the growing, processing or nutritional characteristics of food.

Copyright 1997

Monsanto and ELM Enter Into Joint Technology Venture With Mendel Biotechnology

PR Newswire, Monday, November 24, 1997 at 19:46 ST. LOUIS, Nov. 24 /PRNewswire/ -- Monsanto Company and Empresas La Moderna, S.A., (ELM) jointly have signed a broad technology collaboration agreement with Mendel Biotechnology Inc. in the field of agricultural functional genomics. The agreement gives Monsanto and ELM exclusive access to Mendel's technical capabilities in plant genetics and genomics for many crops, including corn and soybeans, and fruits and vegetables. Leaders in the plant genetics and genomics fields to identify the function of genes founded Mendel Biotechnology and patent the corresponding DNA sequences that will produce the intellectual property basis for the next generation of agricultural products created through biotechnology.

As part of the agreement, Monsanto and ELM will each acquire a substantial minority equity interest in Mendel and fund a research and development program over a five-year period. Financial terms of the agreement were not disclosed.

"This collaboration gives us the ability to better understand the function of specific genes in plants, thereby allowing us to more quickly introduce crops with improved agricultural traits," said Ganesh Kishore, assistant chief scientist and chief biotechnologist of Monsanto. Kishore said the venture will greatly reduce research and development time for products that increase yield, or otherwise enhance the growing, processing or nutritional characteristics of food.

"This is our first venture in genomics, and it builds nicely on our original technology agreement with Monsanto as a preferred provider of agronomic and quality traits we're already using in our fruit and vegetable seed and produce businesses," said Alfonso Romo Garza, chairman and chief executive officer of ELM. Through this partnership, ELM will have exclusive rights to Mendel's technology for the development of proprietary, transgenic fruits and vegetable products that create value for growers, processors and consumers.

Mendel's founding partners include several of the world's most renowned plant geneticists. Their scientific contributions include the isolation of genes responsible for disease resistance, the molecular genetics of nitrogen fixation in plants, the genetics of photosynthesis and plant oil biosynthesis, the development of synthetic plastic in plants, and the gene transfer technology for corn.

"Applying cutting-edge genomics capabilities to the agricultural biotechnology strengths these two leaders have developed is a natural next step in discovering and commercializing improved-trait crops," said Christopher R. Somerville, chairman of Mendel. Somerville also is the director of the Department of Plant Biology at the Carnegie Institution of Washington, and professor of the Department of Biological Sciences at Stanford University. The president and chief executive officer of Mendel Biotechnology is Michael Fromm, a former research director of enabling technologies and plant genomics for Monsanto.

As a life sciences company, Monsanto is committed to finding solutions to the growing global needs for food and health by sharing common forms of science and technology among agriculture, nutrition and health. ELM is a leading agribusiness biotechnology company focused on developing and marketing premium branded vegetable seeds, as well as fresh and processed fruits and vegetables. Mendel, a newly formed biotechnology company, specializes in applying functional genomics techniques to create valuable new traits in plants.

SOURCE: Monsanto Company 11/24/97

CONTACT: Lori J. Fisher, 314-694-8535
Lori.J.Fisher@monsanto.com

Search Strategy

Since Mendel Biotechnology Inc. is known to have interests in Arabidopsis, the company name was used as a search term to uncover new patent documents.

The Mendel Biotechnology Inc. patents discussed in this section were identified using the following search strategy.

As a precaution, a further search was performed:  this time no assignee/applicant/correspondent names were used.  Instead the strategy conducted was:

Patent applications for Arabidopsis bulk sequences made by Mendel Biotechnology Inc.

The following is an example of one of the 11 patent documents identified as belonging to Mendel Biotechnology Inc.  The '927 application is discussed here for the following reasons:

• It involves a large group of Arabidopsis sequences (i.e. a bulk sequence claim),
• deals with a specific class of target sequences that are likely to have numerous applications across many flowering plant species (of interest to many), and
• demonstrates Mendel's stated strategy of targeting transcription factors (TFs) from Arabidopsis (strategic IP?)

(Note that this patent application contains by far the most sequence claims compared to other applications.  Note also that there are ~1800 TFs in the Arabidopsis genome and there are >1900 SEQ IDs listed in this application alone! Hence this application is likely to be a significant patent application with respect to TFs for many plant species.)

US 2007/22495

This Mendel patent application, one of a family of, to date, over 180 patent applications filed in Australia, Brazil, Canada, Japan, Mexico, the PCT, and possibly other jurisdictions that do not report to INPADOC, relates to polynucleotides and polypeptides associated with transcription factors from A. thaliana.   Note how non-specific the utility is:  particularly, the transcription factors are those responsible for modifying phenotypes of plants.  Very broad language is used for the phenotypes in claim 1b below;  a patent examiner could rightly question whether applicant has described how any one of the sequences of part (a) of the claim relate to any one of the phenotypes in part (b) of the claim, because on its face the claim is to any plant having any characteristic listed in claim 1b and containing any sequence, regardless of source, that is somewhat like any of the sequences in claim 1a.

US 6946586

US 7193129

US 7196245

US 7223904

Patent claiming Transcription Factors from Arabidopsis thaliana

The following is an example of one of the granted patents for transcription factors from Mendel Biotechnology Inc.  As previously noted, the patent is part of a much larger family of documents containing applications claiming many Arabidopsis sequences, many of which are still pending in many countries:

This patent application is related to nucleotide sequences encoding plant transcription factors from soybean, rice, maize, barley, wheat and other plants as well as Arabidopsis thaliana.  Like the transcription factor patent application shown before, it is part of a family of over 180 patent applications filed in Australia, Brazil, Canada, Europe, Japan, the USA and the PCT as well as possibly in other jurisdictions that do not report to INPADOC.

The following are US patents that have issued from the above-described Mendel patent applications on "Polynucleotides and Polypeptides from Plants". Mendel, and probably its research ally Monsanto, has the right to exclude anyone from using technology covered by the claims below in the US or in products to be imported into the US. Note that while the claims are narrower than those in the corresponding patent application, this is no cause for feeling relieved about how many research avenues they leave open for development into deliverable products.

For example, in the patent 6,717,034 shown below, the scope of the claims is towards transgenic plants having increased biomass (compared to a control), and the methods involved in producing them.  The scope is further limited with respect to the sequence used to obtain this phenotype.  Specifically the DNA sequence in SEQ ID 1 is used, along with the peptide sequence in SEQ ID 2, to define the scope in plants. Claims include DNA sequences (and complementary sequences) that code for SEQ ID NO: 2 (including SEQ ID NO 1), and sequences that "bind under stringent conditions" to SEQ ID 1 (and complementary sequences). Such claim language opens the possibility of using other transcription factors that might result in similar useful phenotypes (e.g. would the equivalent transcription factor from another dicot, that codes for a peptide sequence other than SEQ ID NO 2, and is sufficiently different to SEQ ID NO 1, be a possible work-around for this patent?)  However, before making such a conclusion it would be important to examine the prosecution history, i.e. the interactions between the patent office and the applicant, about how the doctrine of equivalents would need to be taken into account).  

Also, this patent is one of a large INPADOC family, and broader claims to the same sequences, as well as to all the other sequences in the application, could still be pending.  This means that at any time patents with broader claims could still issue from any of the over 100 patent applications that are still pending.

Summary

Mendel's interest in transcription factors of Arabidopsis are clearly demonstrated in their patenting strategy.  Like Paradigm Genetics, Mendel's early patent applications contain claims to large groups of sequences.  However, unlike Paradigm, Mendel's patenting activities have extended beyond the initial bulk sequence claims.  A large and complex patent family has arrisen from a small number of initial patent applications.  Unlike Paradigm's applications, the '927 application discussed above is the result of a series of continuations of previous applications.  The '927 application has itself led to still further continuation applications.  Many of the Mendel applications are still active, and even those that have abandoned or expired have led to further applications. Unlike Paradigm, Mendel has granted patents for some transcription factors and methods relating to transcription factors (e.g.US 6664446 , US 6835540, and US 6717034).

Mendel has invested significant resources in developing this patent document family. It is likely that Mendel will continue this strategy to eventually obtain patents on more transcription factors from Arabidopsis.  The language of the granted claims is crucial in determining whether claims over other dicot species will be possible.  The recent granted patent, US 6717034 (towards a transcription factor involved in modifying plant biomass), has the following claim language:

Such language is more limiting in scope than that of the '927 patent.  However, it is not unreasonable to expect that such hybridization conditions may be met by at least homologs from other species within the mustard family. Whether-or-not homologs from other important crop species meet the same criteria cannot easily be deterimined, and would require experimental analysis. In part, this uncertainty evolves from the use of hybridization language in the claims.  Such claims, based on hybridization conditions, are difficult to quantify and the huge cost of possible litigation may act as a barrier to those interested in developing IP around transcription factors in other dicots.

To demonstrate the complexity of the patent document family that Mendel has created, consider the following diagram (the '927 application family):

(Hexagons: represent pending applications; Diamonds: represent adandoned or expired applications; F = filing date; CIP = continuation-in-part application)

Broad claims in Applications:
Arguebly, it is the role of a good patent attorney to provide the best possible outcome for a client.  It is probably for this reason that the scope of initial claims in a patent application is often as wide as possible.  Such broad intent is demonstrated in the text of the '927 application, where the following is stated:

The importance of using hybridisation and sequence similarity language to broaden the scope of claims can be inferred from the following two figures from the '927 application.  These figures relate the evolutionary position of Arabidopsis to other important crops.  One may suspect that those crops most closely related to Arabidopsis would be most likely to have genes with highly homologous sequences (and thus satisfy the hybridisation and/or similarity ranges stated in the claims):

Paradigm Genetics Inc. (Icoria Inc.  and Monsanto Co.)

Paradigm Genetics Inc. (Research Triangle Park, N.C., US) was founded in 1997 and went public in 2000. In its early years Paradigm's business statements emphasised making use of genomics information obtained from Arabidopsis, for extensions into crop plants via agricultural biotechnology.  During this time it sequenced a large number of Arabidopsis ESTs.

In 2004 Paradigm changed its name to Icoria Inc.  Icoria moved into medical biotechnology and sold assets relating to its agricultural genomics business to Monsanto Co.  In the period 1997-2004 Monsanto had extensive ties to the Arabidopsis genome sequencing project, and extensive bulk sequence patent filings in crop plants (e.g. rice, maize and cotton), although because of non-publication requests by Monsanto and other procedural matters, many of these were not known to the public until 2004 or later (see for example US 2004/31072, which claims 285,684 sequences mostly from soybean, with applications in many other related crop plants).

We have been able to find 9 patent applications with Paradigm Genetics Inc. as the applicant or correspondent.  Since the 2004 acquisition these may now be assigned to Monsanto, although USPTO assignment information provided by the applicants is not always kept up to date.  

Since all 9 patents are very similar, and each claim ~1000 ESTs from Arabidopsis, we have chosen one, US2001/44940 A1 (the '940 application) to discuss below.

See bar graphs showing the portions of the Arabidopsis genome that are covered by patent applications and granted patents that are assigned to Paradigm Genetics Inc.

Search Strategy

Since the major players in patenting of Arabidopsis sequences are known (from press releases and biotechnology publications), company names can be used in searches.  In this case Paradigm Genetics was used as part of a search strategy.

The Paradigm Genetics Inc. patents discussed  in this section were identified using the following search strategy.

As a precaution, a further search was performed:  this time no assignee/applicant/correspondent names were used.  Instead the strategy conducted was:

Patent applications for Arabidopsis bulk sequences  made by Paradigm Genetics Inc. (Icoria Inc. - Monsanto Co.)

The following is an example of one of the 9 patent applications by Paradigm Genetics Inc (now Icoria Inc.). The '940 A1 application has the earliest filing date of the 9 applications. All 9 applications are listed as "abandoned" on the USPTO website database.  However, one must remember that "abandoned" does not necessarily mean that Icoria or now Monsanto Co. have actually given up their attempts at patenting all or part of the claims. It may still be possible that future continuations, continuations-in-part, or divisional applications might appear as patent document publications.

The 8 remaining applications are:

Summary

Paradigm Genetics Inc. was essentially a start-up company, created to make use of (and patent) DNA and EST sequences generated from Arabidopsis. By making use of the new technology of EST sequencing it could potentially gain patents over a large number of Arabidopsis genes.  Due to the language of the patent claims:

"...hybridizing under stringent conditions to a sequence set forth in..."

it is entirely likely that successful claims would "read on" large numbers of related genes from other dicot species.  In particular those that have agricultural importance: potato, soya, canola,...

All 9 patent applications of Paradigm are classified as "abandoned" in the USPTO's database PAIR.  The assignee of these applications at the USPTO is still listed as Paradigm Genetics. Note also the fact that these are US applications, and as far as we are able to tell, were not the focus of applications elsewhere in the world.

Unfortunately, although classified as "abandoned" it is still too early to say for certain that these applications are no longer "active" for the following reasons:

1. Continuations, continuations-in-part (CIP) and divisional applications may be filed prior to abandonment (and may still be in the process of publishing)
2. Alternatively payments can be made to the USPTO so that some "abandoned" applications can be recovered
3. There is some scope for US applicants to request that their applications not be published immediately

It is unlikely that, in their present state, patents would be granted on these applications, since the claims appear to be very broad. However, Paradigm's applications may demonstrate one early step in a bulk sequence patenting strategy: 

This strategy may have the following benefits:

• Setting an early priority date negates the possibility of confounding prior art for future continuations, CIPs, and divisional applications
• An early bulk sequence application might have a better chance of granting while such applications are still relatively "novel" to examiners
• Work done on the genes claimed in the applications after the priority date can potentially be used to add value to future applications (and won't be considered prior art)
• Publication of such an application immediately increases the value of the company based on a perception of future income created by the IP
• Such potential IP might be saleable to larger companies (like Monsanto)
• Licenses may be negotiated based on technology claimed in the applications.

By creating such important potential IP, a small company such as Paradigm became potentially valuable to larger companies such as Monsanto, who have large agricultural biotechnology interests.  The decision by Paradigm to move into human health biotechnology, was almost certainly facilitated by funds generated in the sale of agricultural assets to Monsanto. In turn, Monsanto acquired potentially useful IP and the loss of a competitor in Paradigm/Icoria.

Also, Monsanto is probably in a far better position to make use of such patented technologies. In fact Monsanto has its own bulk sequence claims for maize and cotton.

As we will see in the following chapters, others have refined this patenting strategy in Arabidopsis.

Ceres Inc.

Ceres is a privately-owned, US-based, plant biotechnology company (www.ceres.net) with headquaters in Thousand Oaks, California. It makes use of genomic technologies towards developments in agriculture and human health.  Ceres specialises in high throughput screening, full length cDNA sequencing, plant breeding, and targeted gene activation. Interestingly, as with Paradigm and Mendel, Ceres has ties to Monsanto, with whom it has extensive license-based agreements.

The company states clearly on its website that the development and protection of IP is an important goal. On the same site it claims that Ceres has filed applications on over 50,000 full length genes and 10,000 promoter sequences.  Ceres' goal of sequencing full-length cDNA sequences compared to ESTs or genomic sequences appears to arise from the need to gain composition of matter claims in US patents (and it seeks at least one substantial utility per gene sequence for patenting purposes). Ceres' genomics facilities apparently allow it to be able to investigate gene function such that descriptions of multiple utilities for each gene of interest can be obtained.

We chose to analyse a some recent applications by Ceres covering both large numbers of functional Arabidopsis sequences and also promoter sequences. Ceres' patent applications provide a more recent example of a bulk sequence application for analysis, and for comparison to the Paradigm and Mendel applications.  Analysis of one example promoter application is also useful since Ceres has interests in this technology.

Ceres' IP interests in Arabidopsis are focused towards claims on promoter sequences and full length genes with agricultural utility.

To date, Ceres has filed patent applications on over 50,000 full-length functionally annotated genes and over 10,000 promoter sequences

Intellectual Property

Ceres places a large emphasis on the generation and protection of intellectual property. Ceres' intellectual property strategy leverages three major competitive advantages: (1) the ability to discover and sequence full-length genes in a high-throughput manner, (2) an integrated suite of genomic technologies which enables Ceres to ascribe specific utilities to a very high proportion of the genes that it discovers, and (3) information management systems which enable the rapid and accurate filing and follow-up of thousands of patent applications.

Ceres practices a proprietary technology for capturing and sequencing full-length cDNAs. Unlike ESTs, these sequences cover the complete and accurate open reading frame of a gene giving the information necessary for a cell to manufacture the encoded protein. Genomic sequence can sometimes give equivalent information, but accurate predictions of gene structure from genomic sequence are only accurate in approximately 50% of the cases tested. The US Patent and Trademark Office has indicated that accurate full-length genes are necessary for the issuance of composition of matter claims that cover the production of the encoded protein. In addition, at least one substantial and practical utility must be described for a particular gene sequence. Ceres' integrated genomics engine enables the description of multiple utilities for any given gene. The combination of this utility information and the full-length gene sequences should enable Ceres to gain intellectual property protection for a large portfolio of important plant genes.

The generation and integration of such a large amount of data into accurate patent application submissions requires a sophisticated information management effort. Ceres has designed and incorporated information management technologies into its technology platforms to enable the timely and accurate generation of thousands of patent applications. To date, Ceres has filed patent applications on over 50,000 full-length functionally annotated genes and over 10,000 promoter sequences.

http://www.ceres-inc.com/about/intell.html

April 3, 2002

Monsanto and Ceres Announce Important Product Discovery and Development Collaboration

St. Louis, Missouri and Los Angeles, California -- Monsanto Company and Ceres, Inc. today announced an important product discovery and development collaboration focused on applying genomics technologies to provide improvements in certain agricultural crops.

"Our new alliance with Ceres provides us access to a proprietary knowledge base that will accelerate Monsanto's product pipeline. By marrying the complementary genomics capabilities of Ceres with the product development capabilities of Monsanto, we will be in a position to deliver additional value to farmers worldwide," said Hendrik Verfaillie, chief executive officer of Monsanto.

Walter De Logi, chief executive officer of Ceres, said, "Monsanto is an ideal partner for Ceres because it is a leader in providing biotechnology-based agricultural products to the world's farmers. Monsanto has the resources to give Ceres' technology the widest possible distribution. Furthermore, we share a vision of providing abundant food and a healthy environment around the world."

Under the collaboration, Monsanto will acquire rights to Ceres technologies in certain crops and applications in exchange for payments over several years. Ceres will receive additional payments subject to meeting specified objectives for developing additional related technology, as part of its continuing commitment to genomics-based product discovery. Monsanto will also fund a jointly implemented research program and has made a minority equity investment in Ceres. Expected payments to Ceres under the collaboration are $137 million over several years plus potential royalties.

As part of the collaboration, Monsanto and Ceres have agreed to make technologies accessible to farmers in developing countries, including non-profit humanitarian applications not served by commercial markets or research and development investments. This innovative feature of the collaboration reflects the commitment of both companies to bring the benefits of science and technology to society through public-private partnerships to improve subsistence crops, in addition to the commercial development of products around the world.

Ceres, Inc. is a privately held biotechnology company utilizing multiple integrated plant genomics technologies to develop innovative products. In addition to pursuing opportunities in the food, feed and fiber industries, Ceres' business strategy consists of leveraging its technology platform in the agrochemical, chemical and pharmaceutical industries.

Monsanto Company (NYSE: MON) is a leading provider of agricultural solutions to growers worldwide. Monsanto's employees provide top-quality, cost-effective and integrated approaches to help farmers improve their productivity and produce better quality foods. For more information on Monsanto, see: www.monsanto.com.

http://www.ceres-inc.com/PressRelease_News/04_03_02.html

Ceres Announces Genomics Milestone in Energy Crop Enhancement Program (info) 
12,000 full-length switchgrass genes sequenced and genetic variation characterized   

THOUSAND OAKS, CA – July 10, 2006
Contact:   Ceres, Inc.  Shirley Bell  sbell@ceres-inc.com
www.oxbio.com/pressreleases/071006_Ceres.pdf 

MONSANTO COLLABORATES WITH CERES 

Monsanto Company and the privately owned biotechnology company, Ceres Inc, Los Angeles, have announced collaboration to apply genomics technologies to provide improvements in agricultural crops. "Our new alliance with Ceres provides us access to a proprietary knowledge base that will accelerate Monsanto's product pipeline.  By marrying the complementary genomics capabilities of Ceres with the product development capabilities of Monsanto, we will be in a position to deliver additional value to farmers worldwide," said Hendrik Verfaillie, CEO of Monsanto.

Under the collaboration, Monsanto will acquire rights to technologies in certain crops and applications in exchange. Expected payments to Ceres through the deal will be $137 million over several years as well as possible royalties. Monsanto and Ceres have also agreed to make some technologies accessible to farmers in developing countries, including non-profit humanitarian applications.

Crop Protection Monthly - 30 April 2002, Issue No. 149
Published by: Market Scope Europe Ltd     
Website: http://www.crop-protection-monthly.co.uk
Editor: Brian R. Hicks
E-mail: brianralphhicks@aol.com

Search Strategy

Since the major players in patenting of Arabidopsis sequences are known, the company names could be used in searches.  In this case "Ceres" was used as part of a search strategy.

The Ceres Inc. patents discussed  in this section were identified using the following search strategy.

Patent applications for Arabidopsis bulk sequences made by Ceres Inc.

The following is an example of a recent patent application by Ceres towards a large group of Arabidopsis sequences. Whilst there are many-more-than 100 sequences in this application (2523 sequence IDs listed), approximately half of these are for the corresponding protein sequence, and some sequences are from corn, soya bean, wheat, brassica, and others. This patent document also provides an example of a more recent bulk sequence application, than those of Paradigm or Mendel. The focus of this application is towards Arabidopsis polypeptides and the sequences encoding them,...

Patents claiming Promoter Sequences from Arabidopsis thaliana

The following is an example of granted patents for promoter sequences from Ceres Inc. Although the application shown here contains less than 100 sequences, they are included to demonstrate Ceres' interest in Arabidopsis promoters and the relationship between the claims in granted patents and patent applications. Additionally, Ceres has many applications and patents for promoters, which together might constitute claims equivalent to a bulk sequence application.

The following 816' application does not list Ceres as the applicant or correspondent!  By performing google searches using the inventors names (shing kwok yu-ping) it was possible to infer that these inventors were employed by Ceres around the same tme that the patent application was made (the publication number 20060008816 is not present in the USPTO assignee database at the time of this analysis).  Hence the inclusion of the application below:

Search Strategy Used

The patents for promoter sequences discussed in this section were identified using the following search strategy.

US 2006/150283

This patent application is directed to nucleotide sequences and their encoded polypeptides from soybean, wheat , maize and canola, as well as Arabidopsis thaliana. According to sequence information published in the USPTO's PSIPS database, 1032 sequences are included in this application.

US Application with missing sequence identifiers

Summary

Cere's stated interests in patenting of whole cDNA sequences and promoters is supported by the patent documents we found for Arabidopsis. 

Our analysis of bulk sequence applications focused on their 724' A1 application. This document was chosen becuase it represents a more recent example of a bulk sequence application than the Paradigm or Mendel applications.  Interestingly, there are a few differences between such applications in the past 6-or-so years.  Differences between the Cere's and Paradigm applications include the change from EST sequence claims (whch are effectively partial coding sequences only), towards complete cDNA sequences. One significant difference between Ceres and Paradigm's applications is the fact that Ceres has analysed missexpression of the claimed peptides in plants and made claims based on sequences that are more likely to have utility.

Mendel, in the patent application describes previously, also investigated the function of claimed transcription factors by over expression or knock-down experiments.  However, unlike Mendel, the Ceres' bulk sequence application appears to be at a much earlier stage of the patenting process.  The patent family surrounding the Mendel application is far more extensive and complex than that of the present Ceres application.  It is not unreasonable to suspect that Ceres patenting strategy for Arabidopsis has been guided by knowledge and experience of previous bulk-sequence claims, and perhaps by the USPTO's requirement for significant utility in sequence applications.

Some interesting points, underlined by the analysis of the promoter patent applications, were:

• Ceres Inc. does not appear as either the applicant or the correspondent in the patent document,
• It is not always straight forward to identify assignee information using the USPTOs patent assignment databases
• Inventors can assign their patents to companies such as Ceres, after the applications have been made
• The USPTO assigment databases may not reflect the present assigment details for patent documents

Considering the points above, it is a difficult and time-consuming task to actually identify the present assignee of a particular patent.  This is particularly true if there is little benefit for the assignee in disclosing this information in a  clear and public fashion.  It is entirely possible that many Ceres patent documents may not be identified from the analysis conducted here, which used a combination of both a standard patent document search (using "Ceres" in FULLTEXT) and a USPTO assignments database search (using "CERES" in ASSIGNEE).

Dow Chemical Company.

Dow Chemical Company is a multinational, US-based, corporation with headquarters in Michigan.  Followed by DuPont, it is the largest chemical company in the world (based on market capitalisation). It has interests not only in chemicals, but also in  agricultural products, plastics and specialised products and services.

Its wholly-owned subsidiary, Dow AgroSciences LLC, is based in Indianapolis, Indiana, USA. Dow AgroSciences has interests in crop protection and crop biotechnology.  Both Dow and Dow AgroSciences are applicants on a number of applications involving Arabidopsis.  Two of these applications that appear to deal with bulk sequence claims are discussed below.

Search Strategy

Since the major players in patenting of Arabidopsis sequences are known, the company names could be used in searches.  In this case "Dow" was used as part of a search strategy. Unfortunately using arabidopsis AND (Dow in applicant) AND ((SEQ AND ID) in claims) did not uncover any patent documents of interest for bulk sequnence claims.  Hence the general search strategy was broadened to the more general search below:

The Dow Chemicals Co. patents discussed  in this section were identified using the following search strategy.

Patent applications for Arabidopsis bulk sequences made by Dow Chemical Co.

The following are patent applications by Dow Chemical Co. These applications cover a large number of claimed sequences (7,554 SEQ ID NOs in the first below). However, these applications contain sequences from a number of different organisms. Thus, not all the sequences are specifically from Arabidopsis. Claims are targeted towards sequences that result in altered metabolic, visual, and disease tolerance characteristics (i.e. those that have possible utility). There are 177 Arabidopsis sequences claimed in the first application below, and the other plant sequences claimed (e.g. from Nicotiana benthamiana) may overlap with other dicot species (including some from Arabidopsis) - depending on the scope of claim language used.

Dow Agrosciences has a similar application for a group of 2,000 sequences:

Summary

Dow and its subsidiary DowAgroscience have interests in biotechnology R&D focused towards:

• Animal health and foods, including animal vaccines produced in plant cells
• Plants with altered oil characteristics for food production
• Modifying plants with traits such as improved disease resistance

The bulk sequence documents analysed above claim a mixture of sequences from a number  of different plant species, both monocots and dicots. For Dow, sequences from Arabidopsis appear to be less important than claims on sequences from organisms that have more direct economic value such as maize, cotton, sorghum, alfalfa and soybeans, and some species that are both model systems and crops, such as tobacco or rice.  Additionally the Dow applications studied also contain a large number of claims to fungal sequences.  For example, Trichoderma harzianum (sequences claimed in the patent document discussed above) is a fungus that is used as a biocontrol measure to protect crop plants against fungal pathogens such as Fusarium sp. (a fungal pathogen of cotton, amongst others).

Dow and DowAgroscience's diverse interests are reflected in their bulk sequence applications.  So although these applications contain some Arabidopsis sequences, Arabidopsis is less a focus of patenting activity for Dow than Nicotiana benthamiana, another popular model plant system. The interest in N. benthamiana may be due to a number of reasons, including:

• "Saturation" applications for Arabidopsis (applications known to have been filed by such companies as Mendel and Paradigm may have influenced Dow's patenting strategy)
• N. benthamiana is in family Solanaceae, and more closely related to important crops such as potato, tomato, and capsicum than Arabidopsis
• Gene silencing in N. benthamiana can be far more easily used to determine gene function (and utility) than in Arabidopsis

Applications claiming sequences from dicot crops

Monsanto Company has patent applications claiming bulk sequences from dicot crops such as soybean and cotton. We include these patent applications here because many of the claimed sequences from these crops are orthologous to sequences from A. thaliana.

US 2007/83945

This patent application covers soybean sequences . It remained unpublished for 83 months instead of the normal 18 months.

US 2007/61916

This patent application covers cotton (Gossypium hirsutum) sequences.

US 2007/236419

This Monsanto patent application claims sequences from soybean, maize and rice, and it is related to a family of patent applications US 2004/31072, US 2004/34888, US 2004/214272, US 2004/216190 and US 2007/11783.

.

Patent application by CropDesign

Granted patents on genes from Arabidopsis

One of the characteristics of bulk sequence claims in plants is that they are almost always in the form of patent applications.  The applicants strategy in claiming bulk sequences is often to establish an early priority date and to later obtain granted patents on individual genes or small groups of genes derived from the initial applications.  Hence , to this point, we have considered only patent applicatons.  Whilst these applications have the potential to influence those working in dicot biotechnology, they are not granted patents.

Transcription Factors

In this section we will look more closely at examples of patents granted on cliams to Arabidopsis Transcription Factors (TFs). 

This group was chosen because it includes patents granted to Mendel Biotechnology Inc., who have made application claims for bulk sequences from Arabidopsis. Since the applications are to transcription factors from Arabidopsis, they may also cover homologous sequences from other dicots.  Thus analysis of granted patents for TFs offers the opportunity to analyse an important group of genes and also the ultimate patent claims that may evolve from bulk sequence claims.

Search Strategy

The patents for Arabidopsis Transcription Factors discussed in this section were identified using the following search strategy.

 A new search was performed to attempt to limit the number of useful identified patent documents:

 A new search was performed designed to identify granted patents from Mendel only:

Patents claiming other sequences from Arabidopsis thaliana

The following are some other examples of granted patents for Arabidopsis sequences. Note that there are many more granted patents for Arabidopsis,  far more than we can analyse at present.  We invite interested readers to suggest patented genes and sequences (or a particular area of interest) for further analysis.

The following is an example of a granted patent for an Arabidopsis transcription factor from Korea Kumho Petrochemical Co. This was chosen as a comparison to those granted transcription factor patents from Mendel Biotechnology:

Granted Claims

A comparison of the granted claims from the 466' patent and the claims of the parent application demonstrates one important difference between applications and granted claims
The following is claim #1 from the provisional patent application US 60/125,814:

Note that the term "...substantially the same DNA binding specificity,..." is difficult to quantify and that the claims are towards sequences from conserved DNA binding domains of transcription factors. Thus these claims are potentially very broad.
Compare the text above with the equivalent claim from the granted US patent US6664446 that evolved from the 814' application:

Whereas, the application claims refers to a broad group of conserved DNA-binding domains from transcription factors, the granted patent refers specifically to polynucleotides that encode a single protein - SEQ ID NO.18. This protein is a full-length 268-amino-acid transcription factor from Arabidopsis. Hence this claim is very-much limited in scope compared to the application. 

Even-so, it seems likely that this granted claim covers more than a single polnucleotide sequence.

This is due to the following:

• Degeneracy in the polynucleotides coding for the same transcription factor (many polynucleotide sequences can code the same protein sequence)
• Variations at the level of the nucleotide sequence can also be covered based on the fact that the same amino acid can be coded by 1 to 6 different trinucleotide sequences.

• Doctrine of Equivalents: might cover sequences that perform essentially the same function and that have similar sequences (compared to SEQ ID NO 18)
• Hence, even under the Doctrine of Equivalents, it is possible that transcription factors that are >95% similar to SEQ ID NO 18 may also be considered (at least by  the applicants) to be protected by the patent.   Although the decision of a court may be required to determine this.

Notice also that the 446' patent was filed in 2000, and granted in 2003 and claims benefit of the U.S. Provisional Application Serial No. 60/125,814 filed Mar. 23, 1999. There are at least 17-other patent applications that claim priority  from the same provisional application.  It is possible that other granted patents will arise from this family in the near future.

Summary

An important point demonstrated in this chapter, by comparison with previous chapters, is the difference between bulk sequence applications and the patents that are likely to evolve from them. Granted patents that provide exclusionary rights based on sequences from Arabidopsis are often very-much restricted in claims scope than their parent applications. So, while the claims made in bulk sequence applications by companies such as Paradigm and Mendel may be broad, descendent patents appear to be reasonably limited in scope.

In one sense this is an entirely obvious, and not unreasonable observation.  Patent applicants often attempt in the first instance to claim the maximum range possible, and these claims end up being moderated and limited by rounds of revisions required by patent examiners. This is a common feature of patent applications and is not limited to applications containing sequences.  In fact such a strategy is almost enshrined, if not promoted, by many patenting systems.

As stated in a previous chapter, granted patents are not necessarily the problem (although some may argue with this!).  Instead, it is the uncertainty generated by the ability of large companies to make such bulk sequence applications that may cause problems for those working with Arabidopsis (and flowering plants). The process of setting early priority dates with bulk applications and then keeping these applications pending through chains of continuations, continuations-in-part, and divisional applications, results in the following:

• Long latency periods (uncertainty in patenting intentions for large period sof time, often years)
• Uncertainty in which sequences, if any, will be granted patents (and when this will happen)
• Uncertainty as to the scope of eventual granted claims

Such uncertainty may not be at the forefront of much not-for-profit research, it will however be pivotal in decisions made by private R&D companies, where FTO is critical.  With the growing need for collaboration between not-for-profits, Universities, and private industry, such uncertainty is a growing concern for university researchers and not-for-profits.

This uncertainty has the potential to prevent innovation - a result which would be in conflict with the intent of the patent system.

Bulk Sequence Claims from other Dicots

As mentioned in previous chapters, one of the problems with bulk sequence claims (not just for researchers but also for patent examiners) is the fact that these applications often try to claim homologous sequences in many other organisms.  This is done typically by the use of such claim language as:

(Note that this is a claim from the Dow Chemical Co. application we discussed in Chapter 6, which contains many Nicotiana benthamiana and "Artificial" sequences )

Although this is not the only way of claiming many related sequences.  It also has the added "advantage" (if you are the applicant!) of being difficult to easily quantify, and hence offers the opportunity of obtaining the broadest possible claim scope.  Those "skilled in the art" will realise that many (possibly only distantly related!)  nucleic acid sequences will hybridise together under low stringency conditions. Further, the term ".., wherein expression of said isolated nucleic acid in a plant results in an altered metabolic characteristic." provides a way of claiming such nucleic acid sequences with utility.

The probability that an applicant will achieve such claims to all 7554 sequences is small.  However, it is likely (and no doubt the applicant expects) that some of these sequences will be claimed in successful future applications. The uncertainty with this claim is two-fold:

• Which sequences are going to be successfully claimed in future?, and
• Will the applicant also acheive claims to related sequences?

At this point we see the connection between such broad bulk-sequence claims in other dicoots (Nicotiana benthamiana claim above) and Arabidopsis.  It is very likely that the bulk of the genes/DNA sequences in claim 1 above will hybridise under low stringency conditions with the corresponding Arabidopsis DNA sequences.  Hence bulk sequence claims for other dicots influence Arabidopsis in the same way that the bulk sequence claims of Paradigm Genetics Inc. would read on sequences from Nicotiana benthamiana!

Search Strategy

The search strategy we used in this case was different from those used previously, as it made use of the term "seqdata".  Sequence listings of considerable size (greater than 300 pages, i.e. those likely to represent larger bulk-sequence applications) are now available for separate download from the USPTOs website (PSIPS).  Most recent, but perhaps not all, applications and patents with large sequence listings can thus be identified with the serach term "seqdata". A term that forms part of the referring URL found at the end of such patent documents.  Hence this term offers a way of limiting searches to recent bulk-sequence documents.  Another option available is to go to the USPTO's PSIP site and search the listings via publication date.  This latter option is difficult since there are >1000 patent documents to browse, which contain large sequence listings.   

One of the patents discussed in this chapter were identified using the following search strategy.

Patent applications for bulk sequences from other dicots

The following patent documents demonstrate attempts to patent large groups of sequences from other dicots.  These applications use claim language such that a successful claim would almost certainly cover large numbers of sequences from other dicots, and perhaps also some from Arabidopsis.  It is therefore useful to investigate briefly the claim contents of such applications - particularly since many of these applications are more recent than those of Paradigm Genetics, and, unlike the applications of Paradigm, appear to be in the process of active patenting efforts by the respective companies.

The claims of the Agrigenesis patent below refer to sequences derived from Eucalyptus grandis and Pinus radiata:

Summary

[Summary for overlapping claims from one organism to another.  Leading into and connection with genome mapping of patents in Chapter 10]

The Extent of Sequence Patenting in Arabidopsis

In Chapter 8 we expanded upon the idea that claims originating on sequences from one organism, for example Arabidopsis, may also claim similar sequences from other organisms. In particular claims to sequences from a model organism (that may have little economic importance) may include hybridisation and/or similarity language that may effectively claim sequences from economically important crops. So companies that seek to claim large numbers of genes from Arabidopsis are effectively also attempting to claim similar sequences from important crop plants.

The degree to which such claims may read on other organisms is very difficult to calculate due to:

• Difficulty in obtaining the sequences claimed in patent applications: Only a subset of patented sequences are present in popular public databases such as GenBank. Sequences from published patent applications, although in the public domain, need to be obtained from the various patent authorities.

• Mapped genomes: Whilst we can identify and map claimed sequences from organisms such as Arabidopsis, where we have extensive genome information, the genomes from many other organisms are not as well defined. Hence it is not possible to map the claims from Arabidopsis onto the genome map of soya, since the latter effectively does not exist.

• Data formats and availability are variable:  WIPO sequences for the most part are available as TIFF images only. The USPTO supplies recent sequence data within application documents, available here or at its website.  Bulk sequence data must be downloaded separately, and has a data structure different from that available at GenBank. Much of the older sequence data is not publicly available in a useful format (i.e. available as scanned pages or as OCRed data files).

• Software issues: Since data formats vary between published sequences, only that present within Genbank is easily analysed by the average user. Software is not yet readily available to allow the average user to perform in depth analysis of patented sequences and to map such sequences to the claimed organism, or to find matches in the genomes of related organisms. Issues of claim language (see below) make quantitative analysis of large groups of patent documents difficult.

• Claim language: The claim language used in patent documents is critical! While we can recognise groups of claims (based on hybridisation conditions and sequence similarity measures), each claim can be subtly different. This results in the need for complex and time consuming analysis - often involving a great deal of time and effort. Since such language is complex and very variable, it does not lend itself well to automated analysis via software solutions. Although, a software solution is needed to analyse bulk sequence claims and for finding the extent of inter-genome claims.

For each of the points above, we have endevoured to develop a solution, an improvement, or an approximation, that gives a better idea of the state of inter-genome claims.

US Applications that Claim Arabidopsis thaliana Sequences

The plot below shows the total number of Arabidopsis thaliana patent sequence filings (blue bars) as well as the number of applications claiming sequences that match a segment of the the Arabidopsis thaliana genome at least 150bp long (green bars) from 1994 to 2006. Because published US application data was not available prior to  2001, the low number of claimed sequences and sequence filings from 1994 to 2001 may be largely attributed to a lack of data from the USPTO rather than a lack of filings.

This plot shows the average number of sequences referenced in claims since 2001.

US Arabidopsis thaliana Patent Applications: Number of Sequences Referenced in Claims per Genome Segment

For the plots below, the Arabidopsis genome was divided into 300 kbp segments.  For patent applications that claim Arabidopsis gene sequences, the plots below show the number of sequences that match at least a 150 bp fragment of each 300 kbp genome segment.

In addition to initial patent filings, continuation and divisional applications are included in the count, so the same sequence may be counted multiple times. Each plot shows data for a single Arabidopsis chromosome.

See the corresponding plots for the rice genome landscape.

The Arabidopsis genomic data contains some vector contamination for chomosomes 2 and 3 resulting in some spikes that go off the scale of the chart.

US Arabidopsis thaliana Patent Applications: Number of Patent Applications that Claim Arabidopsis Sequences per Genome Segment

For the plots below, the Arabidopsis genome was divided into 300 kbp segments. The plots below show the number of US patent applications that claim sequences that have at least a 150 bp match to a 300 kbp genome segment.

In addition to initial patent filings, continuation and divisional applications are included in the count, so the same sequence may be counted multiple times. Each plot shows data for a single Arabidopsis chromosome.

See the corresponding plots for the rice genome landscape.

US Arabidopsis thaliana Patent Applications: Percent Coverage by Patent Claims per Genome Segment

For the plots below, the Arabidopsis genome was divided into 300 kbp segments. For each 300 kbp segment, the percent coverage by claims in US patent applications was calculated. For example, if an entire 300 kbp segment was claimed by US patent applications, the plot would read 100% for that segment. The Y axes show the percent coverage, and the X-axes show the region (locus) of the Arabidopsis genome. Each plot shows data for a single Arabidopsis chromosome.

See the corresponding plots for the rice genome landscape.

Granted US Arabidopsis thaliana Patents: Number of Sequences Referenced in Claims per Genome Segment

For the plots below, the Arabidopsis genome was divided into 300 kbp segments.  For granted patents that claim Arabidopsis gene sequences, the plots below show the number of sequences that match at least a 150 bp fragment of each 300 kbp genome segment.

In addition to initial patent filings, continuation and divisional applications are included in the count, so in some cases, the same sequence is counted multiple times. Each plot shows data for a single Arabidopsis chromosome.

See the corresponding plots for the rice genome landscape.

Granted US Arabidopsis thaliana Patents: Percent Coverage by Patent Claims per Genome Segment

For the plots below, the Arabidopsis genome was divided into 300 kbp segments. For each 300 kbp segment, the percent coverage by claims in granted US patents was calculated. For example, if an entire 300 kbp segment was claimed by US patents, the plot would read 100% for that segment. The Y axes show the percent coverage, and the X-axes show the region (locus) of the Arabidopsis genome. Each plot shows data for a single Arabidopsis chromosome.

See the corresponding plots for the rice genome landscape.

Granted US Arabidopsis thaliana Patents: Number of Patent Applications that Claim Arabidopsis Sequences per Genome Segment

For the plots below, the Arabidopsis genome was divided into 300 kbp segments. The plots below show the number of granted US patents that claim sequences that have at least a 150 bp match to a 300 kbp genome segment.

In addition to initial patent filings, continuation and divisional applications are included in the count, so the same sequence may be counted multiple times. Each plot shows data for a single Arabidopsis chromosome.

See the corresponding plots for the rice genome landscape.

Genome Coverage in Applications and Patents by For-Profit Companies

A number of companies have claimed Arabidopsis genes sequences in U.S. patent applications and issued patents. Here we present graphs showing the percentage of each chromosome that is claimed in patent applications and patents assigned to seven different companies.

For the plots below, the percent coverage for each rice chromosome is shown. For example, if the entire sequence of a particular Arabidopsis chromosome were claimed in patents or applications, the plot would read 100% for that chromosome. The Y axes show the percent coverage, and the X-axes show the chromosome of the Arabidopsis genome. Note: the scales of the Y axes have been adjusted to reflect the data for each company.

BASF (includes American Cyanamid)

Applications	Granted Patents

Bayer (includes Aventis, Rhone-Poulenc, Plant Genetic Systems, Hoechst, Proagro, Agrevo, Biogentic, Planttec, Sementes, Nunza, Agrinomics, and Rorer)

Applications	Granted Patents

Dow (includes Mycogen and Dowelanco)

Applications	Applications

DuPont (includes Pioneer)

Applications	Granted Patents

HySeq (includes Nuvelo and Variagenics)

Applications

Monsanto (includes Calgene, Delta Pine, Seminis, Agracetus, Dekalb, Emergent, Produsem, Mahendra, Stoneville, and Pharmacia)

Applications	Granted Patents

Paradigm (includes Icoria)

Applications	Granted Patents

Sembiosys

Applications	Granted Patents

Syngenta (includes Torrey Mesa, Zeneca, Ciba Geigy, Novartis, Advanta, Garst, Agripro, and Danisco)

Applications	Granted Patents

Mapping Arbidopsis Patents to the Arabidopsis Genome

The first steps in understanding the extent to which claims from one organism overlap into the genome of another is to develop the databases, software, and methodology to answer the question.  The next step we chose was to perform the most straightforward analysis, both to test the methodology and to obtain a "baseline" for comparisons of future analyses.  We set out to:

• Developed reliable databases containing the defined sequences of interest. In the following analyses we have used data from a number of sources to develop a database containing sequences in the claims section of US granted patents and patent applications. To the extent possible, we have filled the known gaps in the existing GenBank patent sequence data using data from the USPTO.

• Defined a standard methodology. Our in-house bioinformatics team developed the methodology and parameters used during analysis (initially based on the work of Jensen & Murray, 2005). 

• Developed software tools. Our in-house bioinformatics team developed and identified software solutions allowing us to build the database above (the patent query databases), the target database (the Arabidopsis genome sequence), to perform the megaBLAST analysis, and to map the output onto Arabidopsis chromosomes.

These initial steps were used to attempt to reproduce aspects of the analysis performed by Jensen and Murray (Science. 2005 Oct 14;310(5746):239-40.) on the human genome map for Arabidopsis.  Essentially, claimed sequences from US-granted patents were megaBLASTed against the Arabidopsis genome sequences, and sequences of interest (longer than 28bp, with expect values of <1 x 10e-200 and highest bit scores) were identified.  The genome positions of each of these sequences of interest were identified and grouped together into regions.  Each region spans a 300kb region of an Arabidopsis chromosome. The number of patents in each region was then mapped to the region position on the respective chromosome. 

The end result of this analysis was a dataset of patents containing Arabidopsis sequences in their claims, linked to the specific map location of those sequences on the Arabidopsis genome.  The sequences and matched genomic regions are made using relatively highly specific search criteria. Resulting in "exact" matching of claimed sequences to their corresponding genomic sequence. No attempt was made to identify homologous sequences within the genome, which may also be claimed by the patent document.  Hence the maps obtained are likely to under-represent the actual sequence claims in Arabidopsis, from patents claiming Arabidopsis sequences.

Sequences from US Patents Mapped onto the Chromosomes of Arabidopsis

	By performing a modification of the analysis of Jensen and Murray (Science. 2005 Oct 14;310(5746):239-40.), a member of our group was able to map patented sequences  from US granted patents onto the Arabidopsis chromosomes.  The following 5 maps demonstrate the distribution of patents across Arabidopsis chromosomes. Note carefully the limitations of this analysis: Only granted US patents were analysed Only patent sequence details for DNA sequences were analysed not claims on protein sequences Only sequences recorded in NCBI's Patent database were included (June  2006 version) Only sequences appearing in the claims section were analysed No regard was made of the claim language, or of claim scope (no attempt was made to identify claims to other homologous or related sequences) No regard was given to claims originating from overlap of claims from other species (i.e. we identified only "identical" Arabidopsis sequences from patents). (see below for more details of these analyses) (Chromosome 1 = 30Mb)
	In all chromsome maps the Y-axis is the number of patents found within 300Kb sections of the chromosome map. Chromosome 2
	Chromosome 3: Note: The Y-axis range is 0 to 10 patents The locus map location 13,500,000 is a patent "hot spot" and contains 47 patents (with references to 78 individual GenBank sequences) For more details of this "hot spot" see the text below.
	Chromosome 4 Note: Sequences contained in US applications were not included in any of these chromosome maps. All of the bulk sequence applications discussed in the text below are not included in this map.
	Chromosome 5 The same Y-axis scale has been used throughout, to allow the user to compare the relative "intensities" of patenting activity on each of the chromosomes. Different X-axis scales have been used, to allow optimal resolution of patent activity.  The relative sizes of the different chromosomes are represented in the chromosome graphic below the X-axis.
Note:  Because of the nature of the analysis above (see the notes on the right of the maps above), the mapped sequences are likely to greatly under-represent the true degree of patent activity on the Arabidopsis genome!

Analysis of Patenting "Hot Spots" on the Arabidopsis genome map

During the mapping of patent documents to the corresponding regions of the Arabidopsis map it was noticed that there are a number of places on the map that have a higher-than-average patenting activity.

These regions included:

From this table further analyses revealed the following:

Arabidopsis Chromosome 3 and the 47-patent hot spot at Map position 13.5Mb:

Analysis of the sequences from these 47 patents (by BLAST at NCBI) demonstrated that a ~450bp fragment from a cloning vector contaminates the chromosome sequence at this position. The large number of patent documents matching at this site are a result of matches with this region from patent sequences unrelated to Arabidopsis sequences. Our method is sensitive to such "contamination" since it relies on almost identical matches with the chromosomal sequences of Arabidopsis to identify patented sequences (and hence patents of interest from Arabidopsis). It is interesting that this sequence contamination has persisted in the Arabidopsis genome sequence and suggests that the methods used to screen for vector contamination are not entirely reliable.

Arabidopsis Chromosome 3: The 10-patent-hot-spot at Map position 3.6Mb:

US patents appearing at this hotspot include those listed in the following table:

Essentially this hotspot refers to (amongst others) claims against fatty acid desaturase (FAD) genes from Arabidopsis and related organisms such as Brassica spp. Some of the sequences mapping to chromosome 3 of Arabidopsis are in fact the FAD gene from rape (Brassica napus). The importance of FAD claims is underlined by the presence of large private companies involved in agritech research and development: Monsanto, Du Pont, and Cargill. Manipulation of FAD levels in plants is one way in which the ratio of saturated to unsaturated oils might be controlled. Such manipulation is of interest due to the human health implications and industrial applications (including biodiesel) possible through the generation of crop plants with modified levels of unsaturated oils. The patents listed above are an example of a small patent thicket that has arisen around an industrially-important gene family (in this case FAD from Arabidopsis and Brassica sp.). Although beyond the scope of the present landscape: It is interesting to speculate that the FAD genes from other species (e.g. Gossypium hirsutum) may also be the focus of patenting activity.

Arabidopsis Chromosome 2 and the 6-patent hot spot at Map position 3.6Mb:

US patents appearing at this hotspot include those listed in the following table:

Again, most of the patenst above do not deal with direct claims to sequences from Arabidopsis or other plants. Instead there appears to be some vector contamination in the chr2 sequence from Arabidopsis, and this appears to account for much of the "hotspot" activity here.

Arabidopsis Chromosome 4 and the 8-patent hot spot at Map position 12.9Mb:

US patents appearing at this hotspot include those listed in the following table:

(Note that the Applicant in the above tables was that listed in the PatentLens database entry for the given patent.  Where no company name appeared (only inventors names) a further USPTO search was conducted to identify assignee information. Note: These assignments may not be accurate after the date of issue.)

US Applications - Sequence Filings by Year

The plot below shows the the total number of U.S. patent applications containing sequence listings from 2001 through the second quarter of 2005. The plot was constructed with data from applications published since 15 March 2001, when applications were first made available in the XML format. The data do not extend to the current date due to the 18 month lag between when an application is filed and when the application generally publishes.

The graph below shows the average number of sequences referenced in claims over the same time period.

Mapping Claimed Sequences to the Arabidopsis Genome

As we saw in the previous analysis of claimed sequences appearing in granted US patents (US-B patent documents), it is possible to obtain an overview of patent activity in Arabidopsis by:

1. Generating a query database consisting of the sequences appearing in the claims section of granted US patents,
2. Generating a target database consisting of the entire sequences of the chromsomes from Arabidopsis,
3. Performing a megaBLAST analysis with this data (With Expect <= 1 x 10^-200 to filter results),
4. Filtering the megaBLAST results to obtain matches of >=150bp in length (and taking only the match with the highest bit score), and
5. Marking the approximate positions of these results  onto a map of the Arabidopsis chromosomes.

From this we were able to detect patenting "hot-spots" and to identify various Arabidopsis genes of interest to those involved in exploiting related genes.

One limitation of this analysis is the fact that much of the Arabidopsis sequences appear in patent applications, not in granted patents.  To overcome this limitation a further query database was constructed, consisting of sequences appearing in the claims section of US patent applications (US-A patent documents).  Included in this dataset were sequences from bulk sequence applications for Arabidopsis.

(Note:  The construction of this database required considerable effort.  The database itself can be found here.)

We repeated the analysis performed above on this new dataset, to determine the extent of patent pending claims on the Arabidopsis genome.

Arabidopsis Genome Maps of Sequences appearing in US-patent applications

	By performing a modification of the analysis of Jensen and Murray (Science. 2005 Oct 14;310(5746):239-40.), a member of our group was able to map sequences appearing in the claims section of US patent applications onto the Arabidopsis chromosomes.  The following 5 maps demonstrate the distribution of sequences across Arabidopsis chromosomes. Note carefully the limitations of this analysis: Only sequences from US applications were analysed Only patent sequence details for DNA sequences were analysed not claims on protein sequences By our estimates >90% of relevant US-A patent documents have been included in this analysis Only sequences appearing in the claims section were analysed No regard was made of the claim language, or of claim scope (no attempt was made to identify claims to other homologous or related sequences) No regard was given to claims originating from overlap of claims from other species (i.e. we identified only "identical" Arabidopsis sequences from patents). (see below for more details of these analyses) (Chromosome 1 = 30Mb)
	In all chromosome maps the Y-axis is the number of patent applications found within 300Kb sections of the chromosome map. Chromosome 2 The hot-spot at 3.6MB has 1077 sequences from patent applications mapped to it (Y scale = 200 patents)
	Chromosome 3: Note: The Y-axis range is 0 to 200 patents The locus map location 13,500,000 is a patenting "hot spot" and 1336 sequences from applications are mapped here.  For more details of this "hot spot" see the comments in the previous section.
	Chromosome 4 Note: Sequences contained in US granted patents were not included in these chromosome maps. Although it is unlikely that the inclusion of such data would change the map significantly. As far as possible, bulk sequence applications have been included in this analysis.
	Chromosome 5 Different X-, and Y-axis scales has been used throughout, to allow the best possible view of all data.
Note:  Because of the nature of the analysis above (see the notes on the left of the maps above), the mapped sequences are likely to greatly under-represent the true degree of patent activity on the Arabidopsis genome!

Methods for Genome Mapping

In this genome landscape, we have determined the extent to which the Arabidopsis genome sequences are claimed in both issued U.S. patents and U.S. patent applications. Our initial approach was to reproduce aspects of the analysis performed by Jensen and Murray (Science 310:239-40, 2005) in their analysis of the human genome. Our process entailed a number of informatics steps that are outlined below.

In summary, we compiled a database of patent sequences that are claimed in patent applications, and compared these sequences to the published Arabidopsis genome using a BLAST-type interface. We then determined which portions of the claimed sequences had significant homology to sequences in the Arabidopsis genome, and mapped these sequences to their location on the Arabidopsis chromosome. The results of our analysis are shown in the subsequent pages of this landscape.

1. Compilation of a searchable Arabidopsis genome database

We used the most recent Arabidopsis genome sequences from the Arabidopsis Genome Project at NCBI. We then used the formatdb program from NCBI to convert the data to a searchable BLAST database.

2. Compilation of sequence databases for granted patents and patent applications

Applications

For patent applications, we acquired the sequences of the bulk sequence applications from the Publication Site for Issued and Published Sequences (PSIPS) web site. This web site provides sequence listings for U.S. patents and applications that are longer than 300 pages. We also acquired the sequence listings for the non-bulk sequence listings (fewer than 300 pages in length) that are published by the USPTO as an XML document. For each of the listing types (bulk sequence and non-bulk sequence), there was a separate file for nucleotides and amino acids. Data for U.S. applications have only been available since 2001.

The bulk and non-bulk sequence listings were then converted to a common data format (FASTA) and combined to create one database for nucleotide sequences, and one database for amino acid sequences. Additionally, each of these combined databases was converted to a searchable BLAST database for use with CAMBIA's patent sequence search tool.

Granted (Issued) Patents

For granted U.S. Patents, we had a data source that wasn't available for the applications; GenBank at NCBI has a searchable patent database of sequences disclosed in granted patents. To create our granted patents sequence database, we started by acquiring the U.S. patent sequences from GenBank. This required removing all sequences that originated from non-U.S. patents.

We then acquired the sequence listings from the bulk and non-bulk patents in the manner described above in the Applications section. The data from all three sources (GenBank, bulk, and non-bulk) were converted to a common format. We then carried out a filtering step that removed any duplicate sequences in the data provided by GenBank, and the sequences provided by the USPTO (bulk and non-bulk).

The identical process was carried out for nucleotide sequences and amino acid sequences. As with the applications, each of these combined databases was converted to a searchable BLAST database for use with CAMBIA's patent sequence search tool.

3. Identification of sequences that are claimed in granted patents and patent applications

A key feature of our analysis is that we determined which sequences were actually claimed in patents and applications, rather than just disclosed in the specification. To this end, we created  four databases that contain only the sequences that are claimed in patent applications. The four databases created correspond to nucleotide sequences in applications, amino acid sequences in applications, nucleotide sequences in granted patents, and amino acid sequences in granted patents.

We compiled a list of common phrases that are used to identify sequence listings in claims. This step was tricky, as there are many different phrases that patent applicants use to designate sequence listings in claims (see examples of phrases) . Using these phrases, we created a list of sequence ID numbers that are designated in patent claims. We then created four new databases that contain only the sequences that are claimed in applications and patents.

4. BLAST search of Arabidopsis genome database using the claimed sequences as input

After compiling a collection of sequences that are claimed in patents and applications, we then used those sequences to query the Arabidopsis genome database (see step 1) using mega BLAST to identify claimed sequences that have significant homology to sequences in the Arabidopsis genome. We performed this analysis only with the claimed nucleotide sequences in patents and applications.

5. Plotting the results of the analysis

The sequences that had significant homology to sequences in the Arabidopsis genome were plotted three different ways. The criteria for matches in the database was that they have a BLAST E value less than e-200.

1. Sequence Count. For these plots, the Arabidopsis genome was divided into 300 kbp segments.  For patent applications that claim Arabidopsis gene sequences, we  plotted the number of sequences that match at least a 150 bp fragment of each 300 kbp genome segment. For each sequence, only the highest-scoring genome match was counted.
   
2. Patent Count. As with the Sequence Count plots, the Arabidopsis genome was divided into 300 kbp segments. For patent applications that claim Arabidopsis gene sequences, we plotted the number patent applications that claim a 150 base pair or longer fragment of each 300 kbp genome segment. For each sequence, only the highest-scoring genome match was counted.
3. Percent Genome Coverage. For these plots, we plotted the percentage of each genome segment that was claimed in patent applications. Unlike the previous two plot types, there was no requirement that the fragments that matched the Arabidopsis genome sequences be a minimum length (e.g., 150 base pairs). However, the minimum size for a positive match using the BLAST interface is 26 base pairs, so matches shorter than 26 were not included in the analysis. In addition, if multiple sequences covered the same portion of the genome, that portion of the genome was counted as being covered only once.

The end result of our analysis was a dataset of patents containing Arabidopsis sequences in their claims, linked to the specific map location of those sequences on the Arabidopsis genome. The sequences and matched genomic regions are made using relatively highly specific search criteria. Resulting in "exact" matching of claimed sequences to their corresponding genomic sequence. No attempt was made to identify homologous sequences within the genome, which may also be claimed by the patent document. Hence the maps obtained are likely to under-represent the actual sequence claims in Arabidopsis, from patents claiming Arabidopsis sequences.

How Does Claim Language Affect the Genome Maps?

During the patent application process the applicant generally attempts to achieve the widest possible claims. Often this is facilitated by the use of similarity or hybridisation language such as:

[INSERT A FIGURE EXTRACTED FROM A PATENT APPLICATION - ONE CLAIMING ~60% IDENTICAL SEQUENCES]

The mapping analyses performed above only detect sequences that are ~100% identical to the claimed sequences, and miss altogether those that are less-than ~100% identical. This fact, and the existence of such claim language, implies that there are many-other sequence claims that do not appear on our maps. The maps above are therefore under-estimates of the true patenting activity against the Arabidopsis genome. 

Many of the sequences mapped to the chromosomes above will have homologous sequences elsewhere in the genome.  Similarly, sequences from other related-organisms (e.g. Brassica napus, canola) that are <100% identical to the corresponding Arabidopsis gene are probably not present on the map (although some that are almost 100% identical are there already).

The use of hybridisation language also contributes to these cross-genome claims. For example, the following claim would almost certainly claim similar sequences from many other genomes:

^{(Patent Application US20040249146: Nucleic acid compositions conferring altered visual phenotypes, DOW AGROSCIENCES)}

Due to the complexity and variation in the claim language it is not possible to accomplish an exhaustive claim-by-claim analysis for all patents.  For this reason we have chosen what we believe to be a reasonable computational approximation of the problem. The choices made with respect to this new method of analysis were:

Predicting the influence of cross genome claims from Arabidopsis gene patents on important dicot crops: Soybean versus Arabidopsis

Soya Bean (Soybean/Soya/Soy/Glycine max):

Soybean, or Glycine max, has a long history of domestication as a food source, and the seed has been harvested for food for many hundreds of years in Eastern Asia. This long history of domestication has resulted in the existence of a very large number of soybean variaties anbd cultivars. More recently, soybean has become an important global food crop, grown in Europe, Asia, Australia, Africa, and Northern and Southern America.  The plant itself is a legume, and will grow in many soil types, requireing a hot summer growing season. Being a legume, the plant establishes a symbiotic relationship with nitrogen-fixing bacteria and is able to grow in relatively nitrogen-poor soils.

Soybean is rich in both protein and oil.  Present day crops are grown for use mainly in processed foods, with only a relatively minor proportion being comsumed directly as food. Soybean and it's processed components finds its way into products such as breakfast cereals, oils, milk replacements, infant formulas, biodeisel, cosmetics, and animal feed.

Soybean and Biotechnology

Due to its economic importance and growing popularity in processed foods, G. max has been the focus of much research, including breeding studies and the production of GMOs (Genetically Modified Organisms).  As one important example of this:  Monsanto markets a Roundup-ready form of G. max which is resistant to glyphosate-based herbicides (e.g Monsanto's Roundup herbicides).  This modification offers a simple way for farmers to plant seed directly into unploughed fields.  Spraying with glyphosate-based herbicides eliminates competing weeds without the need for field ploughing prior to planting.

Patenting and genomics
The soybean genome is-as-yet in the early stages prior to genome sequencing.  The Purdue University consortium has started early work to develop the base for sequencing the genome of G. max.  Hence this genome is only in the earliest stages of sequencing.  However, there are presently 463,101 nucleotide and 2,615 protein sequences deposited in GenBank at the NCBI (as of 24th July 2006) for Glycine max.  Although this dataset is far less than that expected for the entire genome, it is possible to use the available sequence data to predict the amount of similarity between Glycine max and Arabidopsis thaliana genes and to determine the degree of cross genome claims in patents.

Glycine max has a genome size of 1100Mb, almost 10-times larger than Arabidopsis thaliana, with a diploid chromosome count of 40 (Genetics. 2005 Jul;170(3):1221-30.). It belongs to the pea family, Fabaceae, along with peas, beans, and other legumes.  We chose soybean as the Arbidopsis comparison for this analysis based on the following:

• Glycine max is an important crop dicot
• Large companies such as Monsanto are already interested in developing Glycine max
• The genome is in an early stage of sequencing, however there is still significant nucleotide sequence deposited with GenBank
• Glycine max is not as closely-related to Arabidopsis  as are rape, cabbage and lettuce, and is therefore a more realistic comparison for dicots in general.

Summary

Goal:  Demonstrate to users through bioinformatics analysis and graphical presentation of data:

• The current limited view of genome maps (due to limitations in data and tools available)
• A best-prediction genome map (based on our developed data and bioinformatics approaches)
• A demostration of the influence of claim language on claims to genes from other organisms (using a soya to Arabidopsis genome mapping approach)

Glossary

Expressed Sequence Tags (ESTs)[add a comment]

Summary

The availability of the Arabidopsis genome and gene sequences represent an opportunity to investigate the influence of such data on patenting, patenting opportunities, and patenting strategies of stakeholders. Such an investigation is useful, not only to those patenting related "inventions", but also to those who use the data, the model system, or who intend to use similar systems. Whilst other genomes, including those of bacteria (e.g. E. coli), invertebrates (e.g. C. elegans and Drosophila), and the human genome (Jensen and Murray, ref###) are also useful in this regard, the availability of an analysis for a popular plant genome offers many advantages for those in the plant sciences and/or agribusiness. Understanding what has happened in Arabidopsis with respect to patenting allows us to better understand:

• The evolution of patenting strategies of the stakeholders
• Freedom to operate issues relating to the data
• The extent to which the organisms genome is protected by patents
• The ability of the patenting system to deal with such "inventions"
• The influence of related patenting activities on public and not for profit research

Although most of the members of the AGI were not for profits or government institutions, funded through public money, companies such as Monsanto (throught Cereon Genomics Inc) and (###) Perkin Elmer Applied Biosystems (### CHECK THIS) did support their efforts.  Monsanto also sequenced the related Landsberg genome, and has supplied this data to the not-for-profit research community.  Hence the sequencing and sequence data that exists today for Arabidopsis was affectively acheived through collaboration between private and public organisations.  The end result of this collaboration, the genomic sequence, the ESTs, and the genetic markers are available to scientists throughout the world (either directly via public databases such as GenBank and dbEST, or under agreement with Monsanto for the Landsberg data).

However, a number of problematic issues exist with respect to Arabidopsis sequence data, particularly the data that was (and still might be) the focus of bulk sequence claims made by Paradigm Genetics Inc, and to a lesser extent those made by Mendel Biotechnology Inc. and Ceres Inc. Such companies have sought to protect their investment in Arabidopsis sequencing and research through patenting of large groups of genes.  The initial step in such a strategy has been to apply for patent protection over large groups of genes.  This "maximalist" strategy is not uncommon in patenting strategies focused towards other similar technologies, and arguably makes good business sense. Unfortuneately, once applications for large groups of sequences are made, this generates uncertainty for others in the field.

Patent Applications

Almost paradoxically, patent applications may have a similar or greater influence on downstream deliverability of products and services than patents themselves. Compared to granted patents, applications may have the following properties:

• Very large numbers of sequences in claims sections (sometimes many thousands)
• Extremely broad claims (possibly overlapping into the genomes of distantly related organisms)
• Applications can be re-written (and continuations made) to include additional details of utility and enablement
• Uncertainty

One drawback of the ability of applicants to make applications for many hundreds or thousands of genes and/or nucleic acid sequences is the uncertainty this generates in the minds of potential users of the technology. 

Uncertainty can take many forms, and this can result in researcher/strategists asking questions such as:

• How many, if any, of these claims will be granted?
• What influence will a granted claim on an Arabidopsis homolog have on the deliverability of products arising from my research?
• Who do I talk to about licensing gene "X" ?
• Are applications with "abandoned" status REALLY abandoned? Or are there continuations, CIPs, or divisionals just around the corner?

This uncertainty may form a barrier to use of the technology by others.  Additionally, as we have seen in previous chapters, the strategies used by some applicants seem to rely heavily on establishing an early priority date for bulk sequence applications and then supplying "inventiveness" and "enablement" at some much later date.  It is also possible for owners of the IPR to license technologies that are not-as-yet patented.

These properties of patent applications and the uncertainty generated by them produces a de facto protection for the technology under application.  Hence, although most of the claimed sequences present in bulk sequence applications will never produce a granted patent, they are in effect "protected" by the applicant and some aspects of the monopoly possible under a granted patent (de jure protection) is acquired by them.

Granted Patents

As we have seen in the case of bulk sequence applications, few sequences from applications have made their way into granted patents.  Possibly because applications themselves offer a form of protection that allows income to be derived from licensing without the need for a granted patent.  At this point granted patents appear to hold far less Arabidopsis sequences than patent applications.  This is almost certainly due to the economies involved in making and maintaining applications rather than generating, maintaining and defending granted patents.

It is unlikley that the originators of the US patent office envisaged that continuation and divisional applications would be used in this way,  since the original system was established essentially to allow monopoly protection of mechanical inventions and not genomic information. Arguably, granted patents offer a far higher degree of certainty for both applicant, licencees, and business than does a patent application.  However, the creation and maintenance  of uncertainty (either wittingly or unwittingly) provides in some cases a more economical business tool to monetise a technology.

Future for this Landscape

Charting an IP landscape surrounding any genome is a large undertaking.  To ensure that this landscape is useful to as many as possible and is functional as soon as possible, we have chosen to analyse briefly only a small part of the IP surrounding the Arabidopsis genome.  The choice to focus our analysis on only patent documents involving bulk sequence claims was made to ensure we could meet these initial goals.  As with any landscape, constant revision and additions are required to keep the cartography accurate and relevant.  With this in mind, we do not intend that this present version of the landscape is finished.  It is instead a first step towards a more detailed landscape. The intention has always been that this document should by dynamic and flexible, and that user comments would help shape future landscape efforts. [add a comment]

Towards this end, the landscape is marked throughout with user commentable regions.  We would be very interested to hear your comments about any of the regions of the landscape which you feel could be improved.  We are particularly interested to hear from people who have first hand experience in these matters (Arabidopsis research, patenting, licensing, and legal aspects).  User feedback will help us in directing future efforts to expand and update the landscape (and, if necessary, make corrections to the existing landscape). [add a comment]

Providing your input and help at an early stage of this project may very well create a landscape that provides yourself and users like you with a useful and valuable resource.[add a comment]

CAMBIA
29th June 2006[add a comment]

The information contained in this page was believed to be correct at the time it was collated. New patents and patent applications, altered status of patents, and case law may have resulted in changes in the landscape. CAMBIA makes no warranty that it is correct or up to date at this time and accepts no liability for any use that might be made of it. Corrections or updates to the information are welcome, please send an email to info@bios.net.