Sara Boettiger and Brian D. Wright

Sara Boettiger and Brian D. Wright

Open Source in Biotechnology:
Open Questions

Innovations Case Discussion: CAMBIA-BiOS

The case narrative by Richard Jefferson in this issue of Innovations shows how the
rate and direction of progress in biology is constrained by available tools; a novel
tool can set the field on a new and more productive course, but only if creative sci-
entists are free to use it. The history of ¼-glucuronidase (GUS) reporter genes illus-
trates the great impact a technology can have when it is novel, useful, and globally
available on reasonable terms. Now Jefferson’s energy is directed at restoring
biotechnologists’ global freedom to innovate, by “inventing around” essential, Ma
proprietarily owned, research tools, and trying to ensure that the new alternatives
remain freely available for use and improvement.

A key part of his program is the development of BiOS, an institutional inno-
vation that applies aspects of the open source software model to biotechnology.
Although the jury is still out on the effectiveness and sustainability of BiOS,
Jefferson’s detailed account provides a good foundation for initial analysis. Forse
more important than his discussion of the BiOS model itself, Tuttavia, is Jefferson’s
articulation of the intellectual property problems faced by innovators in biotech-
nology who want to see their efforts make a difference to end-users globally. Che cosa
he has to say demands the attention of the many lawyers and economists who see
no problems with intellectual property protection in biotechnology.

Open source is currently one among several approaches designed to encourage
broad based participation in research in biotechnology in the face of the restric-
tions imposed by intellectual property rights on key enabling technologies.1 Open
source in biology is a work in progress, highly experimental and controversial. Questo
essay seeks to reach beyond the rhetoric of openness and transparency, to consid-
er some of the challenges that confront the BiOS project, and some of the oppor-

Sara Boettiger is Director of Strategic Planning and Development at PIPRA (IL
Public Intellectual Property Resource for Agriculture, ). Brian
Wright is professor of Agricultural and Resource Economics, University of California,
Berkeley, and member, Giannini Foundation.

© 2007 Sara Boettiger and Brian Wright
innovazioni / autunno 2006

45

Scaricato da http://direct.mit.edu/itgg/article-pdf/1/4/45/704113/itgg.2006.1.4.45.pdf by guest on 08 settembre 2023

Sara Boettiger and Brian D. Wright

tunities that might be created in biotechnology in general, and agricultural
biotechnology in particular, by open source innovation.

A DIFFERENT PATH

The GUS reporter gene and subsequent innovations, (and especially Jefferson’s
publication with almost 4,000 citations) are achievements for which many a pro-
fessor would contemplate homicide. This narrative has a familiar ring to those who
like to read the lives of the academic super-heroes. Outstanding student meets cre-
ative mentors on the cutting edge,
encounters the right research prob-
lem in the wrong field, and ports the
solution to the right application just
when it is needed. The accomplish-
ment is widely celebrated and the
just patent reward is claimed.

While most lawyers and
economists were still
debating whether access to
technology and freedom-to-
operate problems even
existed, Jefferson designed
CAMBIA and, in turn, BiOS
to tackle those problems.

But at this point Jefferson
begins to steer his career away from
the conventional, exhibiting reckless
disregard of academic disciplinary
boundaries and fiscal prudence.
Many of those GUS citations, some
might have noticed, were generated
by his own efforts to disseminate
the reporter gene technology far
and wide in useful kits which
enabled disenfranchised scientists in obscure corners of the world to do more
effective plant breeding. Eschewing the single-minded pursuit of further publica-
tions and attainment of tenure, Jefferson turned to champion an international
community of scientists, entrepreneurs, and farmers and their capacity to embrace
the emerging scientific opportunities offered by biotechnology. Against the back-
drop of such auspicious scientific potential, the constraints imposed by lack of
resources and encroaching patent claims caught his attention. Had he followed
Adam Smith’s recognition of the key role of specialization in innovation and the
social merits of selfish pursuing profit maximization, Richard Jefferson’s career
would have taken a very different, and less interesting, sentiero.

Because he has played on both sides of the patent game in a rapidly evolving
commercial field, he has had the opportunity to observe how patents can restrict,
or even kill, promising technologies, and stifle the formation of startup firms that
generate the flow of innovations to the end users. While most lawyers and econo-
mists were still debating whether access to technology and freedom-to-operate
problems even existed, Jefferson designed CAMBIA and, in turn, BiOS to tackle
those problems. His experience has earned him notable credibility in this debate.

46

innovazioni / autunno 2006

Scaricato da http://direct.mit.edu/itgg/article-pdf/1/4/45/704113/itgg.2006.1.4.45.pdf by guest on 08 settembre 2023

Open Source in Biotechnology: Open Questions

OPEN SOURCE: FROM SOFTWARE TO BIOLOGY

The merits of open source (OS) in software, though still debated, are widely
acknowledged. Some advocates of OS software, from its beginnings, have promot-
ed its development with mystical zeal. Ma, over time, its success as a production
model has garnered the respect of hard-headed lawyers and businessmen. OS has
proved to be an efficient, thus far sustainable, and competitive system for develop-
ment of some software applications, delivering high quality products, with faster
development time, at a fraction of the cost of firm-based production models.

In OS, self-selected volunteers develop ideas that might make their own lives a
little easier. For example, they remove bugs encountered in their idiosyncratic
work environments (Bessen 2005), some of which could only be detected by a cen-
tralized research authority with great difficulty and expense. This activity is known
as user innovation Von Hippel (2005). Often, they share their results with others,
and enjoy the resulting peer acknowledgement of their contributions. But none of
this started with software. Not by a long shot.

The first modern economist, Adam Smith, described the phenomenon in 1776.

A great part of the machines … in those manufactures in which labor is
most subdivided, were originally the inventions of common workmen,
who, … employed in some very simple operation, naturally turned their
thoughts towards finding out easier and readier methods of performing
Esso. Whoever has been much accustomed to visit such manufactures must
frequently have been shown very pretty machines, which were the inven-
tions of such workmen in order to facilitate and quicken their particular
part of the work.

Note the lack of any hint of monetary awards for the inventions, and the
assumed willingness of the employers to share them with all comers. Long before
Smith, farmers were solving biological problems without thought of monetary
award, and sharing their inventions with their peers. Open source agriculture is
more a restoration than a revolution.

To agricultural scientists, OS offers a promise of a return to the scientific envi-
ronment of decades past, where materials and ideas were exchanged with greater
fluidity, and today’s preoccupation with intellectual property rights that was
absent. But BiOS’ wet lab plant biotechnology constitutes a young field very differ-
ent from that of software production, or traditional plant breeding before the twin
revolutions in biotechnology and intellectual property rights. Jefferson’s initiative
accordingly provides an interesting lens through which to examine the prospects
for the open source model in novel terrain.2

The shift from copyright to patent law, and the complex and expensive regula-
tory regime, profoundly affect the prospects for open, distributed innovation and
the creation of protected commons of easily accessible technology in plant
biotechnology. The appropriate architecture for an OS model in biotechnology,
like the appropriate design of any innovation, is hard to predict ex ante. The fate of
BiOS, as a practical implementation of the model, will be highly instructive.

innovazioni / autunno 2006

47

Scaricato da http://direct.mit.edu/itgg/article-pdf/1/4/45/704113/itgg.2006.1.4.45.pdf by guest on 08 settembre 2023

Sara Boettiger and Brian D. Wright

Patent vs. Copyright Law

Free access to technologies in the OS model fundamentally depends on the protec-
tion of those technologies from encroaching IP claims. This is accomplished
through an open source license in which the right to use the technology is
exchanged for the promise not to privately appropriate it. In software, the strategy
designed to create a protected commons of accessible technologies involves the
dominant IP form, copyright, as the key legal instrument in the open source. In
biologia, the dominant form of intel-
lectual property protection is not
copyright but patents. Several char-
acteristics of patent law pose serious
challenges to the translation of the
OS software model to biotechnolo-
gy.3

The shift from copyright to
patent law, and the complex
and expensive regulatory
regime, profoundly affect
the prospects for open
[source]…in plant
biotechnology.

Whereas

copyright attaches
instantaneously and with zero cost to
new software code, obtaining patent
protection (“patent prosecution”)
for an innovation in biotechnology
costs tens of thousands of dollars,
and entails months if not years of
back and forth between the applicant
and the patent office. The traditional
OS model depends on the collaborative contributions of programmers who
engage in the project for any number of well-researched motives (reputation, fun,
improved skills, connection to community, eccetera.), but if their innovations were to
be protected by patents, and the cost of patenting were shared by all research col-
laborators, the community of contributors would likely collapse.

Given that the cost of the patent system discourages the patenting of every iter-
ative improvement to open sourced biotechnologies, it would be necessary to make
informed bets as to what ought to be patented in order to achieve a cost-effective
degree of protection for the growing commons of the project. Remaining tech-
nologies might be defensively published. The Single Nucleotide Polymorphism
(SNP) consortium4 provides an example of effectively combining defensive pub-
lishing and defensive patenting to reach a similar goal of sustained open access, Ma
without the complication of maintaining access to “improvements” of key
enabling technologies.

A priori decisions must also be made regarding where to patent. One strength
of the OS model in software is its ability to cross national boundaries, gaining from
the talents of a truly international set of developers.5 While copyright lends itself
to virtually costless international coverage,6 patents are national in scope. Applying
for patent protection worldwide can be prohibitively expensive; even filing in a
handful of wealthy countries can cost hundreds of thousands of dollars in fees and

48

innovazioni / autunno 2006

Scaricato da http://direct.mit.edu/itgg/article-pdf/1/4/45/704113/itgg.2006.1.4.45.pdf by guest on 08 settembre 2023

Open Source in Biotechnology: Open Questions

associated expenses.

In patents, as in copyright, the utility of the protection gained from intellectu-
al property rights depends on the ability to enforce. All the expenditures and effort
involved in patent prosecution are in vain unless the OS commons has the credi-
ble financial capacity to sue infringers and finance the necessary litigation through
to a decision, if necessary. In patents this capacity does not come cheap; each law-
suit in the United States costs millions of dollars. It is not clear whether the issue
of enforcement is less serious in OS software, which is itself a pioneering commons
institution, E, as such, still a work in progress.

One practical distinction between software and plant biotechnology in this
regard is that infringers may have less incentive to fight to the end if they can, at
low cost to themselves, cease infringing by substituting lines of new code in a rel-
atively short time period. In plant biotechnology, Tuttavia, an accused infringer is
likely to have less attractive alternatives to legal warfare; switching to a non-
infringing technology may forfeit an investment of years of development, back-
crossing and regulatory testing because patented technology is often locked into
the genome of a novel plant variety.

In light of these constraints, BiOS, for effective management to achieve unfet-
tered access to crucial technologies, needs to be able to make centralized decisions
about patenting and publishing, and to have the financial capacity to enforce its
rights. Centralized decisions are not foreign to the traditional open source model;
despite claims of democratic innovation by OS protagonists, the system most often
depends on a hierarchy of reviewers ensuring quality control and assigning cred-
it.7 But even with this hierarchy, the OS quality control process lends itself to, E
indeed finds strength in, its openness and immediacy. To our knowledge, BiOS and
other OS biology initiatives have not addressed the issues of confidentiality, delays
and capital requirements associated with extension of the OS model to patentable
biotechnologies.

Open Access to End-Products

Beyond the challenges posed by the shift from copyright to patent law, further con-
straints to the translation of the open source model into applied biotechnology
arise from fundamental differences in the characteristics of product commercial-
ization paths. In the life sciences a significant amount of capital is often necessary
to move inventions through development, field testing, manufacturing, and distri-
bution. OS software, on the other hand, has no expensive regulatory hurdles to tra-
verse and can be replicated and distributed at zero marginal cost.

If the goal is open access to an end-product, not to a research tool (as in BiOS),
then widespread delivery of the product may depend on engaging capital to get it
from the lab out into the hands of consumers. The ability to leverage patent rights
can, in some cases, play a critical role. If the product has both commercial and
humanitarian markets (consider, Per esempio, an AIDS vaccine), the patent owner
may license the patent rights to a company for use in the lucrative developed coun-

innovazioni / autunno 2006

49

Scaricato da http://direct.mit.edu/itgg/article-pdf/1/4/45/704113/itgg.2006.1.4.45.pdf by guest on 08 settembre 2023

Sara Boettiger and Brian D. Wright

try market in exchange for the company’s promise to manufacture and deliver the
product into developing country markets at a reasonable price. This logic is not
new, Ovviamente. Product development public private partnerships (PDP’s), among
others, have demonstrated how to leverage intellectual property rights, segmenting
the market in their licensing agreements in order to achieve the ultimate goals of
delivering biomedical innovations to poor and underserved populations where
there are very limited commercial markets. It is in cases like these that open source
licenses may hinder the product’s commercialization by precluding the engage-
ment of private capital. An understanding of this dynamic is in part what drove the
BiOS model to focus on enabling technologies, preserving the potential for patent
rights on application-level technologies.

The polio vaccine provides a historical example that seems to contradict the
cautions above. It is often cited as a case where the choice not to patent resulted in
a major public health success. Jonas Salk famously stated: “Who owns my polio
vaccine? The people! Could you patent the sun?” It’s true that Salk did not patent
his work and open access was achieved, by almost anyone’s standard, as the polio
vaccine represents one of history’s great public health success stories.

The polio vaccine was delivered through an extraordinary collaboration
between individual volunteers and a public charity, the National Foundation for
Infantile Paralysis (now known as the March of Dimes), founded by Franklin
Delano Roosevelt. Salk’s work was funded by the National Foundation. The field
trials were the biggest peace-time mobilization of volunteers in U.S. history. Nearly
two million school children, called the “Polio Pioneers,” and thousands of health-
care workers and lay people volunteered to take part in or assist with the vaccine
field trials. The results of the trials were analyzed at the University of Michigan.
Millions of Americans participated by raising funds in their communities. IL
National Foundation for Infantile Paralysis even funded the manufacture of the
vaccines by subsidizing the production of nine million dollars worth of vaccines.
The story of the polio vaccine is, Infatti, an inspirational illustration of a
nation mobilizing its resources to address a public health crisis. But it was devel-
oped with ample funding and without a thicket of potentially blocking patents.
Remember, pure, that vaccines are currently under-supplied globally. The Salk
model has not been sustained. A major source of vaccines for tropical diseases is
the U.S. government, which funds the necessary research to protect soldiers who
might one day fight in tropical lands; any gains that accrue to locals in such coun-
tries are more or less incidental. Where there is still some doubt as to whether pri-
vate sector resources may need to be engaged, the option to use IP rights as a tool
to achieve the goals of open access may be valuable.

The larger point is that different IP management tools fit different circum-
stances. There are many instances where publishing and not patenting is the path
to ensuring open access. Yet another important strategy, widely praised as judi-
cious, is exemplified by the broad and non-exclusive licensing strategy implement-
ed for the key Cohen-Boyer patents. Effective IP management plans require flexi-
bility and knowledgeable professionals. They should be designed to support par-

50

innovazioni / autunno 2006

Scaricato da http://direct.mit.edu/itgg/article-pdf/1/4/45/704113/itgg.2006.1.4.45.pdf by guest on 08 settembre 2023

Open Source in Biotechnology: Open Questions

ticular goals, and depend on the characteristics of a technology and surrounding
circumstances as they unfold. Open source mechanisms, Anche se, are not flexible;
in terms of IP, the fate of a new invention is mandated ahead of time. This can
mean missed opportunities.

Inter-operability and Parallels to Linux

The burgeoning of the OS model in software and its ability to generate serious
rivals to commercial products in some market segments was dependent, in parte, SU
two critical elements. Primo, the contribution of a kernel by Linus Torvalds in 1991
enabled Linux to become a complete, functional alternative to proprietary operat-
ing systems, and subsequently the flagship for OS success. Secondo, the creation of
a set of OS licenses with different degrees of virality, allowed OS code to be used in
combination with proprietary software, thereby broadening the range of business
applications that could integrate OS code. The original OS license, the GNU
General Public License (GPL), has a viral quality which mandates that products
incorporating the original code also become additions to the commons and must
be licensed under the same GPL terms. In response to needs for an OS license
where interoperability brought fewer restrictions, other licenses were developed8
which allowed OS code to be incorporated into proprietary commercial products.
The range of degrees of virality among licenses reflects a trade-off. More viral
licenses promote greater growth in the protected commons of code, but at a cost
of reducing the range of applications for the code. Less viral licenses still can work
to preserve the commons of code, but lean more toward the direction of a static
commons which does not grow as quickly.

It is natural to look to OS in software to find a model for the protected com-
mons of technology that BiOS seeks to create. Jefferson rightly identifies the need
for a complete platform of enabling technologies, tools for plant genetic transfor-
mazione, as an important element of OS application for agricultural biotechnology.
Along with Jefferson, the press has highlighted parallels between his Transbacter™
technology, designed to work around existing, proprietarily-owned, plant trans-
formation methods that form a crucial bottleneck in agricultural biotechnology,
and the kernel of what we now know as Linux. The analogy, Anche se, is premature,
for two reasons. Primo, Torvald’s kernel was the lynchpin to the system—with it a
truly self-sufficient operating system was born. Transbacter™, Anche se, removes
only one of two current bottlenecks. There is yet another technology that remains
a critical impediment to operability.9 Second, Transbacter™ is a young technology.
Its utility for plant breeders is not yet established.10

In any case, because of the territorial nature of patents, these bottlenecks of
proprietarily owned enabling technologies exist for the most part in only a few
countries (including the United States). The key patents creating the bottlenecks
in the U.S. were either never issued, or have expired in many other countries.
While it is true that products exported back into territories where these patent bot-
tlenecks exist will have problems, there are many countries without these patents

innovazioni / autunno 2006

51

Scaricato da http://direct.mit.edu/itgg/article-pdf/1/4/45/704113/itgg.2006.1.4.45.pdf by guest on 08 settembre 2023

Sara Boettiger and Brian D. Wright

in which researchers can use a full set of technologies in the public domain with
impunity, and with no need to consider BiOS license terms. There are other, non-
IP, reasons why this is not done; access to materials, biosafety issues, liability and
stewardship issues, and a weakness in scientific capacity can be more serious
impediments than foreign patents in hindering progress in plant biotechnology in
developing countries.

Tuttavia, the path pioneered by BiOS could become a route to freedom to
operate for poor countries in the future. The full effects of the global spread of
patenting fostered by the TRIPS Agreement of the World Trade Organization, E
even more onerous bilateral agreements, are now coming to bear on agricultural
researchers in developing countries. As their scientific capacities develop, the full
force of patent claims might well become a serious obstacle.

Because BiOS does not currently provide a complete and viable alternative
interoperability concerns are not just important, but essential.
platform,
Researchers have no alternative to using technologies licensed under the BiOS
terms in conjunction with patented technologies owned by others. Unfortunately
the BiOS license mandates encumbrances that “infect” key complementary
enabling technologies. Owners of patents on such technologies might well find
these encumbrances unacceptable.

Suppose, for instance, a scientist creates a plant transformation vector (IL
research tool that enables a researcher to insert a gene into a plant’s DNA), con un
BiOS enabling technology as one of its many component technologies. Under the
terms of the license, the entire vector system must be granted back to BiOS. IL
BiOS license, in its reservation of rights for the licensee to own application tech-
nologies (i.e. not enabling technologies), falls short of what is known in open
source software licensing as “viral.”11 However, the BiOS license does have a viral
quality to it that affects enabling technologies.

The terms of the BiOS license could mean that researchers become limited in
the set of tools from which they choose. In a sense, the BiOS license could, count-
er-intuitively discourage, rather than encourage collaboration. To see this effect,
imagine again the vector system referred to above where one component is a BiOS
technology. The researcher would like to use another component that happens to
be patented by a commercial firm. The commercial firm will not agree to the use
of their technology knowing that the vector system, incorporating their technolo-
gy will be available for free under the BiOS license. Therefore the researcher’s
choice of tools is effectively diminished by having chosen to use the BiOS technol-
ogy; he has relinquished the ability to use certain tools because he has brought a
technology into his lab under the terms of the BiOS license.

It is true that the BiOS license allows the licensee to refrain from granting back
improvements if they are kept as trade secrets. (Not only can the licensees benefit
from access to enabling technologies, and the improvements of others, but they
can use trade secrecy, where feasible, to avoid making their own improvements
available to other licensees.) The trade secrecy option is, Tuttavia, unlikely to be a
useful concession for universities, where disclosure is an important part of the cul-

52

innovazioni / autunno 2006

Scaricato da http://direct.mit.edu/itgg/article-pdf/1/4/45/704113/itgg.2006.1.4.45.pdf by guest on 08 settembre 2023

Open Source in Biotechnology: Open Questions

ture and materials are regularly shared among researchers in a lab (or among labs)
informally.12

In sum, creation of a transformation platform as a flagship application with
freedom to operate has been the subject of much effort and creativity, but it is still
a work in progress.

Is the BiOS Model Sustainable?

In order for BiOS to be a viable and replicable model, sustainability is essential.
Richard Jefferson has generously seeded the model by placing his own patents
under the BiOS license; the Rockefeller Foundation and IBM, among others, Avere
provided financial or material support. The initiative aims to increase the com-
mons with improvements to the existing technologies. But what about new tech-
nologies? The incentives for participating by signing the BiOS license are separate
from the incentives needed to get people to donate new technology. Will the latter
be forthcoming to BiOS as contributions from the private sector or public sector?
Jefferson’s own experience with BiOS is not encouraging on this point. Perhaps the
BiOS approach will be sustained by replication, with each new collaboration initi-
ated by a creative leader who sets the broad agenda. The key enabling technology
for widespread adoption of the BiOS model might be a perfected BiOS license.

The culture of hackers that continues to fuel the advances of OS in software
may not be replicable in the field of biology. But to the extent that it is, researchers
in the public sector are likely to be crucial participants, especially in less developed
economies where almost all agriculture-related research is public. In its present
form, the BiOS license remains a poor contractual fit for universities (particularly
those in the U.S.13). Some would argue that the goal of the BiOS model to provide
for dissemination and access is already part of the university process. Universities
publish, they collaborate, they share and exchange. Having learned from painful
experiences akin to Jefferson’s loss of access to a positive selection strategy, Quando
universities license their technology they typically retain the right to publish, con-
duct research, and allow other universities/non-profits to do the same for research
and educational purposes.

The reach of the BiOS grantback14 goes beyond what is often found in licenses
from the nonprofit sector, and could impose obligations that public sector scien-
tists are not free to satisfy, since the rights to their inventions are often mediated by
their employer institutions. This is a problem that arises when the OS approach is
ported from the world of copyright to the world of patents. Universities have not
wrested control of copyright on texts from the grasp of their academic authors,
even if the work is produced on campus. By contrast, patents on inventions origi-
nating in U.S. university labs are assigned to the institution. It may be that a sus-
tainable OS model in patentable biotechnologies will need to utilize a legal mech-
anism that is better-suited to the peculiarities of academic institutions.

innovazioni / autunno 2006

53

Scaricato da http://direct.mit.edu/itgg/article-pdf/1/4/45/704113/itgg.2006.1.4.45.pdf by guest on 08 settembre 2023

Sara Boettiger and Brian D. Wright

Some Perspective

At this point, it is appropriate to put some perspective on the relevance of open
source initiatives for global agriculture. To date, it is unlikely that massive numbers
have died of hunger due to the current state of agricultural intellectual property
rights. Subsidized by rich countries’ agricultural policies, the world markets have
offered basic foods at prices lower than ever recorded.

Inoltre, almost all currently useful agricultural biotechnology has been
available, without patent protection, to most developing countries, for all purpos-
es except incorporation in exports to countries with relevant patents in force. It has
not been widely adopted for basic food production, for two reasons. Primo, wide-
spread opposition to genetic modifica-
zione, for reasons of biosafety and pub-
lic acceptance, has discouraged this
type on innovation. (For example, COME
Di 2006, South Africa was the only
country in Africa where genetically
engineered crops are grown commer-
cially).15

At present, open source is a
promising, but problematic,
way to preserve some
freedom to innovate in a
world of patent thickets.

Secondo, the less developed coun-
tries, with a handful of notable excep-
tions including China, India, Brasile
and Argentina, lack any real capacity to
exploit the new technologies, because the substantial, sustained investments in
formazione scolastica, research and facilities necessary to get the process under way have not
been made.

Nevertheless this is the right time to be addressing patent problems in devel-
oping countries. The long downtrend in food prices has been interrupted, UN
reminder that the food yield increases behind recent declining price trends did not
come automatically, but reflect sustained, large, largely public, investments in
research. In the past year, the world has awakened from complacency about atmos-
pheric carbon, on the one hand, and reliance on imported fuels, on the other, A
support massive increases in ethanol production from crops. If these increases
continue, large yield improvements in the productivity of crops will be needed to
ensure that competition from gasoline consumers does not cause an increase in the
numbers of the world’s poor and hungry.

Considerable investments have already been made into researching the genet-
ic modification of developing-country crops (for instance, biofortification, disease
and pest resistance, and drought tolerance). These projects must consider con-
straints and opportunities associated with intellectual property rights in order to
ensure the intended delivery of the products of their research into the hands of
farmers. In its short history, there is already an accumulation of anecdotal evidence
of agricultural biotechnology research projects being delayed, re-directed, or halt-
ed all together because of intellectual property rights problems (Wright and

54

innovazioni / autunno 2006

Scaricato da http://direct.mit.edu/itgg/article-pdf/1/4/45/704113/itgg.2006.1.4.45.pdf by guest on 08 settembre 2023

Open Source in Biotechnology: Open Questions

Pardey 2006a, 2006B). A recent survey of agricultural biologists at U.S. Land Grant
Universities reveals that they, as a group unusually familiar with patenting and the
exchange of tools, believe that intellectual property rights, through their effects on
transactions with their peers, are on balance hindering progress in their research
areas (Lei et al., 2007).

There is, Perciò, a sound argument that we cannot wait to find out how the
global implementation of the TRIPS agreement, and subsequent bilateral negotia-
tions on intellectual property rights, affect global innovation over the next quarter
century. IPRs, among the many challenges in life sciences, require forethought.
Decisions today about the ownership of and access to technologies (through
patents and licenses) will affect the paths of research and development for decades
ahead.

CONCLUSION

To develop BiOS, Jefferson has had to dedicate years of effort and ingenuity, call-
ing on all his talents as scientist, entrepreneur, innovator, fund-raiser and cheer-
leader. To create a flagship application for BiOS, he and his colleagues have invent-
ed a novel technology for genetic transformation of plants, designed to be unen-
cumbered by prior patent claims. To ensure its development follows the open
source model, he has fashioned the BiOS license, porting the open source licensing
concept from copyright to a more complex world of patent protection and biosafe-
ty regulation.

Given the magnitude of the task, it is no surprise that the development of BiOS
as a sustainable institutional innovation is still a work in progress. But at this stage,
the story merits a close reading. Jefferson has indisputable credibility as a witness
to the multidimensional challenges of acquiring freedom to operate in agricultur-
al biotechnology.

From a policy perspective, the major lesson is implicit. Almost the entire effort
in creating BiOS constitutes expenditure of valuable, if not unique, resources that
would be unnecessary, absent a patent system, or a system of efficient license agree-
menti. This effort, Poi, constitutes a concrete example of the “excess burden” of
the patent system, as it exists in developed countries, questo è, its cost to innovators
that does not get transferred to others as benefits, but is lost as economic waste.

The availability of global communication at virtually zero cost offers unprece-
dented opportunities for exploiting specialization and the division of labor in
biotechnology research. Unfortunately, the recent revolution in patent protection,
and constraints imposed by biosafety regulations, have had the opposite effect,
forcing “in-house” aggregation of essential agricultural biotechnology innovation
capabilities within a few vertically-integrated firms. As this has happened, IL
innovation race has slowed to a crawl.

Thus far, the prudent caution regarding biosafety, and the slowdown in
biotechnology innovation, have had no serious effects on food consumption; past
research investments, and rich-country food subsidies, have kept prices low and

innovazioni / autunno 2006

55

Scaricato da http://direct.mit.edu/itgg/article-pdf/1/4/45/704113/itgg.2006.1.4.45.pdf by guest on 08 settembre 2023

Sara Boettiger and Brian D. Wright

supplies high. Given the current surge in biofuels demand, and the continuing
increase in world population, it would be foolhardy to assume that this situation
will continue. At present, open source is a promising, but problematic, way to pre-
serve some freedom to innovate in a world of patent thickets. Achievement of a less
restrictive patent regime would allow the full creative potential of open source col-
laboration to be realized in ensuring an adequate supply of food for the years
ahead.

We invite reader comments. E-mail .

1. Other models include patent pooling, clearinghouse mechanisms, and humanitarian licensing.
2. It is not clear that biotechnology per se is less amenable to specialization and open source collab-
oration, absent biosafety and intellectual property constraints. The potential efficiencies of spe-
cialization and collaboration in synthetic biology are illustrated by the BioBricks initiative
. See Endy (2005).

3. In the interest of brevity, we discuss only highlights of several differences between patent and
copyright law and their significance for the translation of the OS model are provided. Infatti, dif-
ferences in the legal systems have wide-ranging implications for OS that deserve more in-depth
analysis.

4. Robert Cook-Deegan (2003) describes how a group of academic institutions and thirteen private
firms formed a consortium to ensure the SNPs remained broadly accessible and were not private-
ly appropriated. He writes: “The SNP Consortium did not just dump the data. They filed patent
applications and then characterized the SNP markers enough so that they could be sure that
nobody else could patent them. At that point, they would abandon the patent. It is a very sophis-
ticated intellectual property strategy that in the end was intended to bolster the public domain. It
requires coordination, lots of paperwork, and it costs money to file and process applications, Ma
it appears to be an effective defensive patenting strategy.”

5. Lancashire (2001) reports 33 different nationalities among Linux contributors.
6. There are two principal international copyright conventions: the Universal Copyright Convention
(or UCC) and the Berne Convention. To protect copyright internationally the name of the author
is required and (for the UCC) the year of publication and a © symbol.

7. Jill Coffin (2006) notes: “For [an open source project] to function…an organizational and polit-
ical structure must support it. Hybrid, flexible political systems based upon meritocracy moti-
vates participants, provide rewards in the absence of capital, and encourage a community-wide
sense of project ownership. In addition to the bottom–up, peer–administered hierarchy described
in the analysis of Wikipedia, the benevolent dictator and consistently active personnel keep the
project alive and dialog open from above, so to speak…A transparent meritocratic structure also
allows for smooth succession in administrative and leadership positions.”

8. For instance the Berkeley Software Distribution (BSD) style licenses.
9. Generating genetically modified crops requires several indispensable technologies including those
necessary to transfer foreign DNA into a plant cell, selection gene markers to distinguish geneti-
cally modified cells from untransformed cells, and marker-excision technologies to remove super-
fluous DNA after successful integration of the trait gene into the plant genome. This packet of
core technologies is complemented with other research-specific technologies, which may also be
protected by IP. Of the transgenic crop technologies, transformation and selectable markers may
be considered “bottleneck” areas where the restricted access to the technologies can impede inno-
vation.

56

innovazioni / autunno 2006

Scaricato da http://direct.mit.edu/itgg/article-pdf/1/4/45/704113/itgg.2006.1.4.45.pdf by guest on 08 settembre 2023

Open Source in Biotechnology: Open Questions

10. Broothaerts et al. found that the transformation efficiency of non-Agrobacterium bacterial
species ranged from less than 1% to almost 40% of that of Agrobacterium-mediated transforma-
zione, depending on the transformation assays and species used. (Gene Transfer to Plants by
Diverse Species of Bacteria, Nature 433, 629-633 (10 Febbraio 2005)

11. The GPL (http://www.gnu.org/copyleft/gpl.html), requires in clause 2(B) that any works derived

from the licensed software must also to be distributed under the GPL.

12. Another anomaly differentiating the commercial use of the BiOS license from academic has to
do with federal funding and obligations that are common in university policy as a consequence
of the adoption of the Bayh Dole Act.

13. PIPRA, The Public Intellectual Property Resource for Agriculture, www.pipra.org, has complet-
ed an in- depth analysis of the BiOS license from the perspective of a U.S. university as a poten-
tial licensee.

14. For a legal discussion of the BiOS license see Boettiger and Burk (2004).
15. See Eicher (2006).

Riferimenti

Boettiger, S. and D. L.Burk (2004). “Open Source Patenting”. Journal of International Biotechnology

Legge, Vol. 1, pag. 221-231. Available at SSRN:

Coffin, J. (2006). “Analysis of Open Source Principles in Diverse Collaborative Communities,” First
Monday, 11:6, (avuto accesso 3
Marzo 2007).

Cook-Deegan, R. (2003). “The Urge to Commercialize: Interactions Between Public and Private
Research Development in The Role of Scientific and Technical Data and Information” in The
Public Domain: Proceedings of a Symposium, Julie M. Esanu and Paul F. Uhlir, Eds., National
Academies Press, Washington, DC.

Eicher, C., K. Maredia, IO. Sithole-Niang (2006). “Crop biotechnology and the African Farmer,” Food

Policy, 31 pag. 504-527.

Endy D.

(2005).
DOI:10.1038/nature04342

“Foundations

for engineering biology,” Nature,

24 novembre.

Lancashire, D. (2001). The Fading Altruism of Open Source Development, First Monday, 6:12
(Dicembre), (avuto accesso 3
Marzo 2007).

Lei, Z., R. Juneja and B. D. Wright (2007). “Implications of Intellectual Property Protection for

Academic Agricultural Biologists.” Mimeo, University of California, Berkeley.

Smith, Adam (1776). An Inquiry into the Nature and Causes of the Wealth of Nations, Book 1, Chapter

1. < http://geolib.com/smith.adam/won1-01.html>, last accessed March 4, 2007.
Von Hippel, E. (2005). Democratizing Innovation. Cambridge, MA: CON Premere, 2005.
Wright, B.D. and P.G. Pardey (2006). “Changing Intellectual Property Regimes: Implications for
Developing Country Agriculture.” International Journal of Technology and Globalization 2, nos.
1/2: 93-114.

innovazioni / autunno 2006

57

Scaricato da http://direct.mit.edu/itgg/article-pdf/1/4/45/704113/itgg.2006.1.4.45.pdf by guest on 08 settembre 2023
Scarica il pdf