A Synthetic Biohazard Non-proliferation Proposal
George M. Church http://arep.med.harvard.edu
Harvard Medical School, 77 Ave Louis Pasteur, Boston, MA 02115
(18-June-2004, updated 6-Aug-2004)
Introduction. In light of the significant past and future contributions of
oligodeoxyribonucleotide (hereafter "oligo") synthesis and the widespread knowledge of
DNA, RNA and protein sequences, halting information flow or DNA synthetic flow is
undesirable and unlikely. While the likelihood of misuse of oligos to gain access to
nearly extinct human viruses (e.g. polio) or novel pathogens (like IL4-poxvirus) is small,
the consequences loom larger than chemical and nuclear weapons, since biohazards are
inexpensive, can spread rapidly world-wide and evolve on their own. No protective
system will be risk-free. Below is a very specific proposal which could decrease risks
while minimizing impact on legitimate research. It focuses on a few potentially effective
bottlenecks which currently exist and which could be quickly strengthened. The goal of
this document is to solicit comments and participation from potential non-profit funding
and oversight agencies to expedite the establishment and testing of a system with benefits
to oligo suppliers for monitoring a full list of select agents.
Proposal
I. Instrument and Reagent Licensing: Sales and maintenance of oligo synthesis
machines and supplies would be restricted to licensed non-profit, government and forprofit entities. All use of reagents and oligos would be automatically tracked and
accountable (as is done for nuclear regulations). This licensing would initially be
voluntary, then expanded by economic incentives (e.g. via government grant restrictions,
and awarding "seals-of-approval"), then international agreements. The option of doing
this secretly has been discussed at least as early as 1998, but it appears that the risk of
continued slow adoption may be too high. An "open" strategy would be designed to
minimally impede (possibly stimulate) additional secret strategies. Manufacturers would
tag each new machine with an ‘IIN’ (Instrument Identification Number like VIN numbers
used in cars). A non-profit or government DNA Instrument & Reagent Registry
(DIRR) database and web site would allow manufacturers and customers to register their
instruments. Reagent vendors would be required to check that a customer had registered
prior to shipping phosphoramidites to them. The registry would also supply IINs for
existing machines, the resale of used machines and confirmed destruction of machines.
Additional "discovery" would be achieved by methods analogous to current Drug
Enforcement Agency (DEA) chemical monitoring.
II. Screening for Select Agents: Under the above licenses, all synthetic oligos would be
screened for "similarity" to known "select agents". This is already practiced by some
commercial suppliers. The information should be pooled to counteract the potential
strategy of spreading a set of oligo purchases among different suppliers. Any alarming
results would be reported rapidly (along with a statement of potential false positives) to
appropriate government security agencies, e.g. Dept. Homeland Safety (DHS), Center for
Disease Control (CDC), and FBI. The agencies currently responsible for maintaining the
select agent lists would consider proposals for adding new sequences from a broad
spectrum of biomedical researchers on a timely basis. Changes in the sequences
compatible with gene function would be expanded as they are anticipated (e.g. nearly
neutral codon or aminoacid changes, degenerate oligos, etc.) Analyzing customer oligo
sequences for anything other than biohazards would be strongly discouraged (as it is
already).
One reading of the current CDC regulations on select agents is that oligos are not subject
to those regulations and the Dept. of Commerce doesn’t care about anything less than 200
bases in length. These regulations could be tightened and clarified publicly, while
ensuring that oligo companies see a net benefit, not cost and liabilities. One scenario
would be to set up a non-profit DNA Agent Clearinghouse (DAC) with NIH, USDA,
and/or DARPA funding and CDC, DHS, and FBI oversight. The software (e.g.
BlackWatch), installation and testing would be made freely available to the oligo vendors.
Any positive matches that are found on site would be sent to the Clearinghouse, with a
copy to the site managers. Staff at the Clearinghouse (with security clearance) would
evaluate the sequences and make an immediate preliminary assessment. They would also
add those sequences to a second system that would look for patterns of activity like
related oligos being ordered from multiple vendors. If something significant turns up, the
Clearinghouse can contact the vendor and go directly to someone at the FBI, DHS, etc
who can initiate appropriate follow up action. The instrument registry above could be
coordinated with the Clearinghouse. This approach frees the vendors from a lot of effort
and responsibility. Customer data remain confidential unless a match is found. It puts the
task of assessing sequence matches and false positives in the hands of experts. The FBI
only has to deal with incidents being reported from a single source. They can be
confident that any report they get is significant and worthy of their attention. The
Clearinghouse could be affiliated with other bioinformatics activities to attract top
intellectual input.
III. Testing: The system can fail if it is not tested frequently, so a "testing group" can be
assigned to submit orders for oligos and/or reagents from each licensed entity at frequent
but random times. Both the testing and screening groups should have some form of
commercial and/or academic/social reward for improved performance (i.e. lower false
positives and negatives). While PCR primers will be harder to detect than whole gene
syntheses, they may be less common since their use requires access to the intact select
agent.
IV. Exemptions: Legitimate uses of oligos for select agents would be pre-approved by
relevant CDC permit, Biohazards Safety, grant peer review and other existing agencies.
A secure, computer-readable list of allowed target DNA/protein sequences would be
prepared for each such end user and transmitted to the screening groups. Testing group
submissions would not usually be accompanied by such pre-approval, but would be presubmitted to the government security agencies only to eliminate that type of false positive.
No doubt there will be government agencies which are exempt from the entire system.
V. Costs: The cost of setting up the non-profit DIRR and DAC could be kept low and
responsible administration high by embedding them in existing grant-funded non-profit
institutions. The cost of maintenance would be the salaries of the staff, which would
include at least a director, legal counsel, a database expert, a web programmer, and a
synthetic biologist. The director and biologist could be 50% FTE. For the initial phase
of installing the software in the laboratories of over 100 universities, biotech companies
and commercial oligo vendors will require training and travel for 4 technicians.
Hopefully much of the installation can be done via web downloads, but having the
technicians available by phone, email and personal travel should greatly increase
community acceptance. Possibly some financial incentives could be provided to the early
adopter beta-test sites to pay for lost staff-time while they work with DIRR & DAC staff.
Attendees nominated for "Risk committee" at the Synthetic Biology 1.0 meeting
(June 10-12, 2004, Cambridge MA):
Roger Brent, George Church, Drew Endy, Daniel Fletcher, Jay Keasling, Ken Nesmith,
Carl Pabo, George Poste, Paul Rabinow, Pam Silver.
Acknowledgments: Working on array DNA synthesizers with Steve Kieffer-Higgins, on
"genome engineering" with Andy Link, and on DARPA projects with Tom Knight were
crucial in formulating early thoughts on these subjects. Recently these gelled further in
writing a white paper for DOE (chaired by Ray Gesteland) and conversations with most
of the above "Risk" group. Especially significant are contributions from Tom Knight,
John Mulligan, Ron Davis, and Robert Jones.
References:
CIA, Academy of Sciences "Darker bioweapons future"
http://www.fas.org/irp/cia/product/bw1103.pdf
Cello J, Paul AV, Wimmer E. Chemical synthesis of poliovirus cDNA: generation of
infectious virus in the absence of natural template. Science. 2002;297(5583):1016-8.
Church, G.M. and Kieffer-Higgins, S. (1992) WO 92/21079A1 Parallel Sequential
Reactor.
Hoult M, & Mulligan,
J Public Comments on the New Regulation 03-SAR-049
http://www.cdc.gov/od/sap/comments/03sar049.htm
Jackson RJ, Ramsay AJ, Christensen CD, Beaton S, Hall DF, Ramshaw IA. Expression
of mouse interleukin-4 by a recombinant ectromelia virus suppresses cytolytic
lymphocyte responses and overcomes genetic resistance to mousepox. J Virol. 2001
Feb;75(3):1205-10.
Jones, J. Craic (2003) Computing LLC. Hazardous Biological Agent Sequence Detection
https://biotech.craic.com/blackwatch/index.html
http://www.cdc.gov/od/sap/comments/03sar110.pdf
Joy, W (2000) Why the future doesn't need us. Wired 8.04.
Morgan B and Youderian, P (1977) Doing DNA at home: A recipe for Botulism
http://arep.med.harvard.edu/gmc/Real77.pdf
Mulligan, J (2003) Bioterror threat from gene synth is immediate, not far off.
Genome Technology.
https://biotech.craic.com/blackwatch/BlackWatch_GenomeTechnologyLetter.pdf
Tian J, Gong H, Sheng N, Zhou X, Gulari E, Gao X, & Church GM (2004) Accurate
Multiplex Gene Synthesis from Programmable DNA Chips.
http://arep.med.harvard.edu/SBP/ Nature in revision
Select agents HHS, USDA, CDC
(including 42 C.F.R. Part 73, Federal Register, Vol. 240, No. 67 Dec 13, 2002)
http://www.cdc.gov/od/sap/docs/salist.pdf
http://www.cdc.gov/od/sap/
NIH Recombinant DNA and Gene Transfer Guidelines:
http://www4.od.nih.gov/oba/rac/guidelines/guidelines.html
Suppliers of 2-cyano-ethyl N,N-diisopropylchlorophosphoramidite (key precursor):
http://www.cmschemicals.com/CMSChemicals/CMSProd.nsf/0/d0e383d8b3f1b24580256c41003bb2
27?OpenDocument
http://www.chembuyersguide.com/partners/dalton.html
Suppliers of phosphoramidites, CPG, anhydrous solvents (key to oligo synthesis):
http://www.glenres.com/
http://egt.horus.be/code/en/page_08.asp?Page=276&smenu=5
Synthesis machines sources:
Beckman 1000M $11K
grizzlyanalytical.com
Biolytic BLP 192 $125K
biolytic.com
Biosearch 8700, 8750; CyClone $8K-11K
Cruachem PS250 $3K
MerMade 192 $90K
bioautomation.com
PE-ABI 3900 $70K; 394,392,391,390Z,3948, Expedite8909 $ 11-25K
Polygen 10 $90K
www.polygen.de
Polyplex Gene Machine $89K
genomicsolutions.com
TAGC96 $95K
tagc.com
Commercial oligo suppliers:
Illumina
Qiagen-Operon
OligosEtc
Invitrogen
IDT
MWG
illumina.com
oligos.qiagen.com
oligosetc.com
invitrogen.com
idtdna.com
mwg-biotech.com
Examples of Academic oligos suppliers:
www.albany.edu/genomics/oligos.html
ww2.mcgill.ca/sheldon/oligonucsynthesis.htm
msf.ucdavis.edu/oligo.html
hms
Custom oligo array suppliers:
Agilent
Febit
Metrigen
Nimblegen
Oxamer
Xeotron/Atactic/Invitrogen
Combimatrix
chem.agilent.com
febit.com/geniom
metrigen.com
nimblegen.com
oxamer.com
xeotron.com
combimatrix.com
US Synthetic gene suppliers:
Abetal LLC
Aldevron
AnaGen Technologies
Aptagen
Bio Applied Technologies Joint
BioNexus
BioTech Core
Blue Heron Biotech
Certigen
Commonwealth Biotechnologies
DNA 2.0
Egea Biosciences
GeneMed Synthesis
GenScript
IDT
MCLAB
Midland
Modular Genetics
Molecular Diagnostic Services
Picoscript
Qiagen
SeqWright
abetal.com
aldevron.com
ana-gen.com
aptagen.com
batj.net
genesynthesis.net
biotechcore.com
blueheronbio.com
certigen.com
cbi-biotech.com
dnatwopointo.com
egeabiosciences.com
genemedsyn.com
genscript.com
idtdna.com
mclab.com
oligos.com
modulargenetics.com
mds-usa.com
picoscript.com
qiagen.com
seqwright.com
Sigma Genosys
Retrogen
sigma-genosys.com
retrogen.com
Non US Gene Synthesis
Scandinavian Gene Synthesis
GeneArt
Entelechon
Evrogen (Russia)
BaseClear (Netherlands)
Bio S&T Inc. (Canada)
OZEX (Australia)
Metabion
Medigenomix GMBH
BioSpring
TWCBiosearch
MedProbe AS
Genosphere Biotechnologies
TechDragon Limited
IBA
Chemos CZ
Microsynth
GeneXpress
sgsdna.com
geneart.de
entelechon.de
evrogen.com
baseclear.com
biost.com
ozex.biz
metabion.com
medigenomix.de
biospring.de
twcbiosearch.com
medprobe.com
genosphere-biotech.com
techdragon.com.hk
iba-go.com
chemos.cz
microsynth.ch
genexpress.at