Research highlight by Kazuharu Arakawa and Masaru Tomita, Institute for Advanced Biosciences, Keio University, Japan

Gene duplications and mutations are central driving forces in the evolution of genomes. Genomes must be robust to such changes in order to be evolvable, and many studies have probed genome robustness using systematic gene knockouts or overexpression experiments. In a recent paper, Isalan et al. (2008) took a new approach to test the robustness of Escherichia coli gene circuitry by reconstructing gene duplication events by shuffling the promoter-ORF pairs for about 300 transcription factors and introducing 598 recombined pairs one-by-one into E. coli to rewire its transcriptional network. Surprisingly, ~95% of such additions are robustly tolerated, and some networks even exhibit greater fitness under various selection pressures. Moreover, the study shows that, in contrast to naive expectations, the introduction of positive or negative feedback loops has little effect on the protein expression levels of regulated ORFs.

Since radical rewiring of the gene circuitry appears to have only a limited impact on expression levels, this work suggests that gene regulatory networks are highly dynamic and underscores the potential importance of post-transcriptional mechanisms for the robustness of transcriptional regulation. Moreover, this work illustrates the fundamental robustness and evolvability of gene regulatory networks, which is reassuring news for synthetic biology.

Isalan M, Lemerle C, Michalodimitrakis K, Horn C, Beltrao P, Raineri E, Garriga-Canut M, Serrano L (2008) Evolvability and hierarchy in rewired bacterial gene networks. Nature 452:840

Two rather contrasting videos on synthetic biology this month. In the first videocast, released by TED, Craig Venter exposes his grand vision of synthetic genomics. He insists on the notion of 'combinatorial genomics', that will combine the power of large scale DNA synthesis ('robots that can make a million chromosomes a day') with a database of 20 million genes, 'the design components of the future'. This approach, a pragmatic mixture of rational function-oriented design and empirical large-scale selection, is envisioned to prepare a modern 'Cambrian explosion' of new synthetic species. It is good to see Craig Venter laughing when announcing casually the 'modest goal of replacing the entire petro-chemical industry'. In any case, Craig Venter appears to be more concerned that the technology may not develop sufficiently rapidly to match the urgency and scale of the major ecological and medical challenges faced by our planet than by potential threats represented by harmful biohacking and bioterror.

The second video, admittedly less entertaining, is a recording of the recent deliberations of the National Science Advisory Board for Biosecurity (NSABB). In his presentation entitled 'Assessing Biosecurity Concerns Related to Synthetic Biology', David Relman presents some preliminary findings and recommendations of the Working Group on Synthetic Genomics (jump to 1hr:34min:37sec). It is interesting to see that no consensus definition of synthetic biology exists among the various practitioners of the field, who all use different blends of the typical bottom-up engineering approach assembling circuits from standard components and top-down strategy, based on the modifications of existing genomes. Beyond the lack of definition, the current ability to predict biological functions from sequence (eg virulence) remains very limited complicating the possibility of realistic risk assessment. Finally, the development of synthetic biology can be seen as an extension of the success of 'kit-based' molecular biology, which facilitates access of these technologies to groups outside the traditional Life Sciences communities and institutions, making the mission of oversight, outreach and eduction more challenging. David Relman also clearly emphasizes the importance of not discouraging the enthusiasm directed towards potentially beneficial research and applications by overzealous oversight and regulations.

The intersection between the two talks above was perhaps made when the question of virulence was raised (jump to 1hr:59min:35sec). The fraction of pathogenic agents is very small compared to the number of existing species, a point also made by Craig Venter, and the rate of appearance of new pathogens is low. The idea was then raised as whether it would be possible to roughly estimate the risk of creating synthetic pathogens by calculating the likelihood that the amount of natural recombination responsible for the emergence of new pathogens 'in the wild' could be matched by an equivalent amount of experimental recombination in the laboratory. In other words, is there any way to estimate the probability that new forms of virulence could emerge from the announced synthetic 'Cambrian explosion'?

How to make biology easy to engineer and what are the consequences of success? Drew Endy exposes his views on these key issues in the field of synthetic biology in a video released in the last issue of EDGE.

As a teaser, here are a few quotes from this interview, summarizing in a nutshell his opinion on the current priorities of the field and its future development:

Engineers hate complexity. I hate emergent properties. I like simplicity. I don't want the plane I take tomorrow to have some emergent property while it's flying.

How do you manage the information going into a DNA synthesizer so that you can construct some useful object that'll help you do genetics? [...] I think George Church and Craig Venter have a lot to contribute to it, which will be terrific. It will be part of synthetic biology, but it will be synthetic biology impacting science, which is the worst case scenario for synthetic biology.

Five years from now, we may have just begun to make some good progress on reliable functional composition of standard biological parts. Nobody knows how expensive solving that problem will be, but because biology works there's plenty of existence proofs. [...] If I had to guess, I'd say we'll have a collection of tens of thousands of genetic objects that support reliable functional composition between ten and 15 years from now.

Drew Endy also mentions the need to develop an "ownership sharing and innovation framework" that will be appropriate to this pure engineering approach to synthetic biology. A related question might be to find the appropriate publishing instruments that would provide suitable incentives and (micro)attribution mechanisms for those who will embark in contributing, probably often incrementally, to the projected "tens of thousands of genetic objects". One idea could be here to adopt a two-layered system inspired from the one proposed for "Human Variome Microattribution Reviews". In such a system, a "Part Browser" would provide the list and number of all articles/database entries referring to a specific part while partner journals would commission high-level Part/Device Review articles to highlight a "family" of parts or device that might be of particular relevance to the community. Would this make sense (eventually)? How did the electronic engineering field deal with this problem in its early days?

A few weeks ago, Jason Kelly explained in his post how Itaya and colleagues (2007) assembled the complete 135 kb rice chloroplast circular genome starting from a collection of 5-6 kb fragments and using sequential in vivo homologous recombination in Bacillus subtilis. Now, Hamilton Smith, Craig Venter and colleagues have achieved the assembly of a complete 583 kb Mycoplasma genitalium genome ("JCVI-1.0", Gibson et al, 2008). The starting fragments were of similar length, 4-5 kb fragments with 80-360 kb overlaps, albeit synthesized chemically rather than by PCR. In contrast to Itaya et al, Ham Smith's team used in vitro recombination (using T4 pol digestion/annealing/Taq pol repair and ligation) in a 3 step hierarchical assembly process and completed the fourth step, the assembly of 4 quarter genomes, using in vivo homologous recombination in yeast (TAR cloning, Larionov et al, 1996). The use of yeast for the last step might be a little worrying, given the high recombination activity in yeast and the propensity for large constructs to rearrange (I used to work with YACs to construct mouse transgenes and I can still feel the pain... but I don't know about the stability of circular TAR clones). In any case, it worked! One final clone was sequenced (7X coverage) and, remarkably, was shown to match exactly the sequence designed!

This impressive technical feat may eventually have tremendous consequences when combined with the transformation procedure ("genome transplantation", ) Venter and colleagues reported last year (Lartigue et al, 2007). As Dawkins noted at the Digital Life Design meeting in Munich a few days ago (see video below for some excerpts of his discussion with Craig Venter and the transcript in Edge), "genetics has become a branch of information technology".

JCVI-1.0 has obviously not been assembled "from scratch". In fact, beside some "watermark" sequences inserted to distinguish the synthetic genome from the native one, the fact that its sequence is a remarkably accurate copy of M. genitalium genome is probably one of the major achievements of the study. The technology for the synthesis of very long DNA of arbitrary sequence (in principle...) is thus progressing at an impressive pace. But writing a genome is not (yet) equivalent to designing it. Exciting (and hard) work remains to be done to bridge this gap and to improve our understanding of how biological functions can be created by assembling genes into a synthetic genome and developing the tools that will make this process rational and efficient, a challenge the synthetic biology community is eager to tackle (see The BioBricks Foundation)...

by Jason Kelly, MIT

Costs for de novo synthesis of DNA fragments (<10kb) are decreasing rapidly, and challenges now lie in the assembly of these fragments into ever-larger sequences. One of the main challenges is the fragility of long DNA sequences during the in vitro steps associated with traditional methods for assembling DNA. In a recent publication, Itaya et al (2007) describe a method for assembling 4-6kb DNA fragments in vivo via incorporation in the B. subtilis genome. They demonstrated this homologous recombination-based method by assembling the 134.5 kb rice chloroplast genome from 31 smaller fragments.

The process involves:

1. Cloning alternating, overlapping 4-6kb DNA fragments into one of two custom vectors with different selective markers.
2. Mixing these vectors sequentially with competent B. subtilis and taking advantage of native homologous recombination to add each fragment to a growing chain within the B. subtilis genome.
3. Each new fragment replaces the selective marker added by the previous fragment, allowing the chaining process to continue by switching the antibiotic selection at each step.
4. Removal of the fully assembled DNA construct from the genome and re-circularization via previously described methods (Tsuge and Itaya, 2001).

Due to it’s reliance on homologous recombination, this method faces challenges in assembling sequences with repeated regions. The rice chloroplast genome contains two such repeated regions (21kb each). The authors demonstrate a work-around for this problem by first using their method to assemble three blocks (72.9, 36.7, and 34.4 kb) of the rice chloroplast genome without internal repeating regions, then assembling these blocks as the final construction steps.

This work-around also demonstrates one method for parallelization of their sequential process. Parallelization provides the speedup necessary for construction of larger DNA segments or genomes. Each addition of a 6kb fragment takes a couple days, so building a synthetic E. coli genome (4.6Mb) through purely serial addition of small fragments would take over four years. A parallelized assembly process combined with Itaya’s previous work (Itaya et al, 2005) incorporating a 3.5Mb natural genome into B. subtilis brings synthetic E. coli-sized genomes closer to reality – will be exciting to watch where this goes.

Note from Thomas: welcome to Jason's new blog, Free Genes

In a very elegant study published last week in Science, Mark Mayford and colleagues use a synthetic bistable genetic switch to visualize the activity of neurons during associative learning in mice (Reijmers et al, 2007). The reporter system (called TetTag) has two components (see drawing, adapted from Reijmers et al, 2007): 1) the tTA transactivator (tetR-VP16) is placed under the control of the immediate-early gene fos; 2) a tetO-regulated bidirectional promoter drives the expression of both a tau-LacZ reporter and a mutated tetracycline-insensitive version of the tet transactivator, *tTA (tTA^H100Y).

Strong neuronal activity activates the fos promoter and stimulates production of tTA. Under permissive conditions, that is, in absence of doxycycline, tTA will activate the "toggle switch". Even after putting the mice back on doxycycline, the positive feedback maintains the switch active, apparently for up to five days after the initial stimulation. Temporary removal of doxycycline defines thus a time window during which active neurons can be permanently "tagged" (at least over a time scale of several days).

The authors use the TetTag system in transgenic mice to visualize neurons activated during the learning phase of fear conditioning, a behavioral task in which mice learn to associate a given stimulus (eg the "context" represented by the test cage) with a noxious shock. In the case of contextual fear conditioning, the memory of the learned fear response depends on the basolateral amygdala and can be assayed by exposing the mice to the stimulus several days after training and measuring their "freezing" behavior. By enabling TetTag activity during the learning phase only, neurons involved in the learning process are permanently tagged. Three days later, retrieval is tested by exposing the mice to the training cage and immunostaining of the immediate-early gene zif268 (egr1) is used as surrogate measure of neuronal activity during the retrieval phase.

Reijmers and colleagues observe that a significant fraction (12%) of the LacZ-tagged "learning" neurons are also "memory" neurons which are reactivated during retrieval. As the authors write, "reactivated neurons seem to be a likely component of a stable engram or memory trace for conditioned fear". With a clever application of an memory extinction protocol, they further show that the number of reactivated neurons correlates with the strength of the learned association.

It will be interesting in the future to know more on the quantitative and dynamical characteristics of the TetTag system. But is likely that it will be a useful reporter system in the brain and possibly in other systems as well or in developmental studies. I find it also very nice to see that such a simple switch coupled to an immediate-early gene is already a sufficient device to keep a long-term trace of past neuronal activity. Would it not be nice to identify endogenous neuronal multistable genetic circuits coupled to electrical activity that could explain long-term changes of neuronal properties after given stimuli (as is drug addiction)?

The Royal Society seeks your views on the emerging area of synthetic biology. This is your opportunity to shape the focus of the Royal Society's policy future work in this important area. We welcome views from individuals or organisations by 27 August 2007.

• Potential developments and applications
• Current research capacity and geographical distribution
• Societal implications
• Ethical concerns
• Biosecurity risks
• Implications for the environment
• Research support and funding
• Implications for human health
• Legal issues and implications for regulation (national and international)
• Ownership, sharing and innovation frameworks (including intellectual property)
• Biosafety concerns
• Education and training
• Governance and oversight of research
• Economic considerations for developed and developing countries

The Synthetic Biology 3.0 meeting ended yesterday with a closing keynote lecture by Tom Knight. In his lecture, Tom Knight explained how engineering principles can or could be applied in synthetic biology to facilitate assembly of "new systems that never existed in nature in a productive and safe way". Key notions are the principles of hierarchical abstraction, modularity, standardization and flexibility. Defining appropriate levels of abstraction in the description and design of biological objects is essential to cope with the otherwise daunting complexity, says Tom Knight. The power of these concepts was illustrated by showing the many levels of abstraction used in electronics and computer science: at the top of the hierarchy are concepts as used at the level of application software, operating system and programming languages while the lowest levels of abstraction go progressively down to the levels of processing unit, logic gates, semiconductor physics and ultimately quantum mechanics at the very bottom of the hiearchy.

Listening to the talks given at this conference, it seems that only a modest fraction of the presented research relied on engineering concepts as presented by Tom Knight. The synthetic approach applied to the understanding of fundamental aspects of living organisms is conceptually extremely powerful and from the scope of this meeting it is apparent that synthetic biology is not limited to its engineering side. But it is also clear that applying principles of engineering to biology needs very significant- in fact enormous – efforts and more time is probably needed for synthetic biological engineering to come of age. One of Tom Knight's slides depicting a VLSI circuit made me wonder whether the type of work needed to found synthetic biology as a true engineering discipline- systematic standardization, defining the interfaces between abstraction levels, etc- is well rewarded within the academic and publishing context in which conventional biological research is conducted. How was the situation in the early days of electronics? What was the part of academic research and what was the influence of companies in pushing standardization? What are the lessons to be learned by synthetic biology on how to develop the appropriate strategies and infrastructure to promote the foundational work (Endy, 2005) required for rigorous engineering practice in synthetic biology?

It will be interesting to re-examine these questions next year, at the SB 4.0 meeting (apparently to be held somewhere in Asia), to see how the many facets of this field are evolving.

Ron Weiss just gave his talk at SB 3.0 on the application of synthetic circuits to control stem cell differentiation. Ron Weiss has already elegantly shown how synthetic circuits coupled to cell-cell communication can generate pattern forming multicellular systems (Basu et al, 2005). This previous work was performed in yeast and Ron now plans to port it to mamamlian stem cells. Pattern formation is crucial in developmental processes and the hope is that the synthetic approach will not only help to control and use stem cell differentiation and patterning but will also provide insight into the endogenous specification/differentiation mechanisms. One of the application envisioned is to develop a system regulating differentiation of pancreatic beta cells in function of cell density to enable a kind of "synthetic" homeostasis able to compensate autoimmune attacks in type I diabetes.

To achieve this type of system, cell-cell communication and control circuits regulating the necessary differentiation regulators have to be developed and optimized. Ron presented the adaptation of the AHL LuxI LuxR system (apparently with some effort...) to the mammalian system and the generation of a toggle switch with large dynamic range, which is essential to achieve the high expression levels required to trigger cellular differentiation. Finally, Ron showed the plan for a complex 22 component system (BTW what is the most complex synthetic circuit assembled so far?) that would integrate the various features necessary to achieve a homeostatic differentiation system. Ron made no mystery that setting up even simple part of thes types of circuits still require a fair amount of "tweaking" and also pointed out the importance for this project that the behavior of the synthesized circuit stays reliable within a changing cellular context of a differentiating cell. Finally, the large 22 component circuit extensively interacts with the endogenous differentiation regulatory network, adding an additional level of complexity in the design.

How to develop organisms with modified and extended genetic codes? Can bacteria be evolved to replace given metabolic pathways by "exotic" alternative ones or eliminate altogether given amino acids from the entire proteome? These were the topics of a few talks presented at the third Synthetic Biology conference, Zurich, Switzerland.

I already wrote in my first SB3.0 post on the technology being developed in George Church' laboratory to enable automated editing of the bacterial genome, for example to eliminate or mutate a given codon. Yesterday, Jason Chin presented his work on the engineering of an "orthogonal" translation system that works independently of the endogenous translation machinery. As a first step, pairs of tRNA / aminoacyl tRNA synthetase (aaRS) with altered specificity were developed, by adopting a heterologous aaRS from Methanococcus jannaschii and evolving it to shift its activity towards loading of an unnatural amino acid (Chin et al, 2002). The second step was to develop orthogonal ribosome / mRNA pairs by selecting new 16SRNAs recognizing mutated mRNA header sequences (Rackham and Chin, 2005). Finally, to enable efficient translation of amber codon, the resulting orthogonal ribosomes were evolved to escape competition with release factor RF1. Put together, these components (unnatural amino acid/aminoacyl tRNA synthetase / mRNA / ribosome) reconstitute a complete orthogonal translation apparatus that works with high fidelity in parallel with endogenous translation. Such orthogonal systems enable "programmatic" incorporation of fluorescent groups, stabilizing groups or photo-crosslinking agents into proteins and may even be substrate for novel cellular logic (Rackham and Chin, 2005).

Volker Döring was arguing that reassignment of portions of the genetic code might even be a mean to increase safety of engineered organisms. I am not sure I fully followed this. Such organisms are probably resistant to lateral gene transfer and would, for the same reason, avoid "genetic pollution". But as Sven Panke pointed out in a question, these bacteria would also loose some "enemies", eg bacteriophages?, given that they are likely to be resistant to infection by virus, which expect a given genetic code. Not sure it really increases the safety of these organisms, but I may have missed something...

More fundamentally, Philippe Marliere who works on re-designing entire portions of the metabolism by evolving alternative pathways replacing the natural ones, suggests that applying sophisticated and imaginative selection pressure schemes can force organisms to evolve in regions of the metabolic/functional space they would otherwise never explore: "the task of synthetic biology (in metabolism) could be to widen the search performed by natural selection, to diversify the set of reactions and metabolites useful to the cell: we are their only chance"

"Life, like a machine, cannot be understood simply by studying it and its parts; it must also be put together from its parts. Along the way to synthesizing a cell, we might discover new biochemical functions essential for replication, unsuspected macromolecular modifications or previously unrecognized patterns of coordinated expression." (Forster and Church, 2006) .

The quest for generating a minimal cell may follow different paths, as illustrated by several talks presented at the Synthetic Biology 3.0 conference. Yesterday we heard from Giovanni Murtas how a purified transcription/translation system (36 recombinant enzymes + purified ribosomes) could be introduced into vesicles to successfully produce GFP. By extension incorporation of lipid synthesis enzymes, DNA replication enzymes and tRNA genes may enable the assembly of a minimal cell ultimately able to replicate both its core machinery and its envelope.

Converse to this assembly strategy, Hamilton Smith presented a top-down approach in which a genome is reduced to its minimum length by elimination of all non-essential elements, synthesized de novo and finally transfered into a host cytoplasm to boot a new cell. Mycoplasma genitalium appears to be a suitable organism for this type of experiments due to its small size (485 protein coding genes 43 RNA genes). Incidentally, M. genitalium also uses UGA (stop in E. coli) to code for Trp, which is a useful property to prevent expression of toxic products when fragments are assembled together via homologous recombination in E.coli. However, this requires in turn to use a Mycoplasma cell as acceptor cell. To set up the "transplantation" of naked DNA into host cell, the team tested the transformation of purified M. mycoides LC genomic DNA (via beta agarase digested agarose plug prep) into M. capricolum host (PEG-mediated transformation protocol). Nicely, the two genomes appear to segregate (how? incompatibility of origins of replication?) and various lines of evidence indicate that selected cells (tet resistance on donor DNA) contains DNA from the transplanted M mycoides genome only. Along similar lines, Rene Warren plans to rebuild H. influenzae using E. coli as a host. Without attempting to define a minimal genome, the idea is to assemble a series of 61 overlaping fosmids using the lambda red recombination system (Yu et al., 2000).

Assembly of 100 fragments via 4 rounds of homologous recombination, as described by Ham Smith, may appear as somewhat tedious and tricky. In a beautiful talk, Miroslav Radman showed how Deinococcus radiodurans manages to do this just by itself. It is almost spooky: after strong irradiation the genome of D. radiodiurans is chopped--literally pulverized--into small 20-30kb fragments. From this DNA soup, D. radiodiurans reassembles then a complete genome with high fidelity, within a couple of hours.... Apparently, small fragments are assembled via a synthesis-dependent strand annealing mechanism followed by larger fragment assembly via RecA-dependent homologous recombination (Zahradka et al, 2006). This process, a "molecular basis for resurrection", was suggested by Miroslav to inspire the most imaginative projects, from "global sex" to "directed panspermia" and ressucitation of dead neurons…

see also: Brendan1 Brendan2 Brendan3 ETC blog

Today was the first day of the Synthetic Biology 3.0 conference, held this year in Zurich and organized by Sven Panke, Matthias Heinemann, Jörg Stelling and Martin Fussenegger. For those who would like to have a definition of what Synthetic Biology is, there are very instructive explanations here by Drew Endy.

Here are some subjective snapshots of some of the talks that were presented today.

The inaugural keynote lecture, "Reading, Writing and Evolving Genomes" was given by George Church (http://arep.med.harvard.edu, also Senior Editor of Molecular Systems Biology). Even if he did not make it too easy for the poor blogger to summarize this speedy talk :-), George provided a very impressive overview of the fascinating opportunities offered by advances in synthesis, automated recombination, evolution and next-generation sequencing technologies. Not satisfied with creating new life, the lab is now planning to synthesize a mirror form of life! As a starter, the complete in vitro synthesis of a (chirally) mirror version of Sulfolobus DNA polymerase IV has been undertaken. On the theme of massive DNA engineering, he presented the application of a highly efficient single-stranded DNA recombination method (see Costantino and Court, 2003) to oligonucleotide pools. Using automated cycles of electroporation with pools of oligos, this method will ultimately allow whole genome redesign in E. coli, for example to mutate all amber codons into UAA, thus "freeing" the UAG codon for orthogonal systems based on a redefined genetic code. Finally, after evolving an artificial symbiotic association between a strain deficient in tyrosine biosynthesis and another deficient in tryptophane biosynthesis, polony sequencing reveals that only a very few classes of mutations are selected (some data are presented in Shendure et al 2005), a result very similar to what Palsson and colleagues have observed (Herring et al, 2006).

The topic of evolution appeared also in the talk presented by Luis Serrano (yes, Luis is also on our board…), who introduced new links into the transcriptional network of E. coli by expressing rearranged promoter/transcription factor cassettes. From hundreds of strains screened, it appears that additional links in the network do not cause major growth phenotypes. However, some of the strains behave like "superstrains" that can overgrow the wildtype strain and thus can be selected under selective pressure, illustrating the potential for evolvability of the network by addition of new links.

Nice presentations were given by two young teams of last years iGEM competition (Imperial and Ljubljana), illustrating the more applied side of synthetic biology, which makes full use of standardized reusable parts to quickly go through the typical engineering cycle of specification, design, modeling, implementation and testing.

Finally, Georg Seelig presented his recent work on DNA-based logical gates (Seelig et al 2006) that work exclusively on the basis of base-pairing and strand-displacement reactions, showing that synthetic biology can even extend to purely in vitro and cell-free systems (Simpson 2006).

Last April, the US National Science Advisory Board for Biosecurity issued a draft report on Biosecurity (see post below). One point of criticism expressed with regard to this report was that it did not address explicitly concerns specific to synthetic biology.

In a Commentary published in Nature Biotechnology (Bügl et al, 2007), a panel of scientists, executives from the DNA-synthesis industry and members of the US FBI present now their views and concrete recommendations on measures required for an efficient and practical oversight of DNA-synthesis activities.

Rapid progresses in DNA-synthesis technologies are increasingly challenging the current safety measures and oversight mechanisms that were tailored for recombinant DNA technologies (Berg et al, 1975). Two major sources of concerns are expressed by the authors with regard to the combination of facile DNA-synthesis, very short delivery time and internet-based communication: 1) the processes of design, assembly and use of engineered genetic material can be "decoupled" and performed in a fragmented way across different locations, rendering tractability of the overall process difficult; 2) DNA-synthesis may provide a workaround strategy to circumvent the existing physical barriers and containment strategies that currently regulate access to pathogens.

The goals of the proposed oversight framework are listed as follows in Bügl et al:

First, the framework should promote and later compel responsible behavior on the part of users of DNA-synthesis technology. Second, the framework should be sufficiently simple and robust be adopted as best practice throughout industry. Third, the framework should enable common improvement of needed technologies and promote sharing of operational wisdom throughout industry and government. Fourth, the framework should build on the existing practices that have enabled the safe development and application of recombinant DNA technology over the past three decades. Finally, the framework should foster and support international transparency and cooperation.

To achieve these objectives, the authors suggest an initial scheme on how users, industry and government may interact to implement the proposed framework. At the individual level, customers should identify themselves, provide relevant biosafety information and observe local accountability mechanisms. At the corporate level, companies would implement state-of-the-art screening methods and directly cooperate with governments to identify suspicious DNA orders.

In addition, a group of DNA-synthesis companies have formed an "International Consortium for Polynucleotide Synthesis" (ICPS), of which some authors are member, and is proposed to serve as an interface between government agencies and synthetic biology companies.

see also: NSABB Report "Addressing Biosecurity Concerns Related to the Synthesis of Select Agents" (pdf download)

In a News&Views just published in Molecular Systems Biology, Joachim Henkel & Stephen Maurer expose their views on the economics of synthetic biology (Henkel and Maurer, 2007):

Synthetic biology contains almost all of the same ingredients that make embedded Linux successful. First, synthetic biology's parts approach emphasizes strong modularity. This allows the work of creating a parts library to be spread over many companies. It also makes it possible for companies to earn profits by patenting some parts while making others openly available. Second, we expect companies to have fairly idiosyncratic parts needs. This means that they cannot simply 'free ride' by waiting for others to make what they need. It also suggests that companies can often share parts without losing their technological 'edge' to competitors. Third, different companies will have different expertise. This suggests that community-based libraries will often outperform company ones. Finally, the synthetic biology market will probably include large numbers of small, idiosyncratic customers. This makes patent licensing less lucrative and, by comparison, openness more attractive.

Synthetic biology is defined around the concept of standardized re-usable parts. A piece of C++ code is very very very well behaved and therefore highly suitable for a development model based on sharing parts. In synthetic biology, as Ron Weiss writes (Andrianantoandro et al, 2006),

the engineering strategies of standardization, decoupling, and abstraction can also be useful tools for dealing with the complexity of living systems...The above engineering strategies come from disciplines where components are well behaved, easy to isolate from each other, and can subsist in isolation. The strategies must be adapted to work well in the biological realm, where biological components cannot exist without being connected....Design and fabrication methods that take into account uncertainty and context dependence will likely lead to on-demand, just-in-time customization of biological devices and components, which need not behave perfectly. Building imperfect systems is acceptable, as long as they perform tasks adequately."

How close will synthetic biology come to something like object-oriented programming? The future will tell how far the analogy with the "embedded Linux" model can be extended and whether the economics of synthetic biology will be influenced by how "well behaved" and complex synthetic parts are.

How do proofreading and repair activities emerge and what is their impact on system-level properties such as robustness or evolvability?

Biological networks and the Internet may share common architectural principles common to "robust yet fragile" systems (Doyle and Csete, 2007). But the Internet does not (yet) repair itself while cells or tissues do. Repair may represent an example of "downward causation" (The Music of Life, Noble 2006 or a book review) in which a high-level functional property (eg repair activity) emerging from a multi-component system (eg the DNA repair machinery) acts on a component at a lower level of organization (eg one nucleotide). In other words, could repair be considered as an example of a "cross-scale" feedback motif? Can these types of motifs be generalized and how would they evolve? The topic could in fact be extended to the field of synthetic biology: how to assemble synthetic systems with self-repair capability?

Many questions... Who has answers?

see also: Fidelity and infidelity (2001) Radman, 2001

Observations suggest that current climatic models may underestimate how quickly the climate system is changing (in particular for sea level), according to a report in Science a few weeks ago (Rahmstorf et al, 2007). Another Science paper published last week shows that the capacity of the Southern Ocean CO₂ sink is weakening, which may result in increased atmospheric CO₂ levels in the long run (Le Quere et al, 2007).

I remember Hiroaki Kitano calling the systems biology community, in his talk at the ICSB meeting last October in Yokohama, for ideas on how system-level approaches could contribute to address the challenge of global warming. In response to the studies above, a similar call is now sent to the microbiology community by Jonathan Eisen on his blog. Research topics suggested in his post include:

Marine Microbiology

Carbon fixation processes

Hydrogen production

Carbon sequestration

Methane capture

Microbial fuel cells

A similar list of priorities related to energy challenges, environmental remediation and carbon cycling and sequestration can be found on the site of the Genomics:GTL research program from the US Department of Energy.

For all the topics listed above, systems biology and synthetic biology approaches are likely to be crucial not only to accumulate the necessary fundamental knowledge but also to find ways to translate it into technological applications. Proposals, insights and visionary suggestions are more than welcome...

some additional links:
Special issue on Energy and Sustainability
ASM Report on Microbial Energy Conversion
Microbial ecology meets electrochemistry: electricity-driven and driving communities. Rabaey et al, 2007, The ISME Journal 1:9

Literature reviews invariably get outdated. Could reviews benefit from being hosted within a wiki environment (be "wikified") and kept up-to-date by the community? This is the experiment the Openwetware community has just started on the OWWReviews page (see also post on Public Rambling).

One of the first articles selected for this interesting project is Ron Weiss' review on synthetic biology (Andrianantoandro et al, 2006, Mol Syst Biol 2:2006.0028) published in Molecular Systems Biology last year. It will be interesting to follow the evolution of this derivative work and see how new knowledge is incorporated but also how the community deals with the more personal opinions expressed in such a review.

The model chosen for this initial experiment is to wikify previously published reviews. It might certainly also be appropriate to explore the reverse process, that is, to write a Review in the wiki first and let it evolve. Should such a piece be "frozen" at some point as a peer-reviewed publication? Who would be the authors? Would the definition of a minimal template providing some unifying structure be helpful? What topics are best suited to stimulate "distributed writing" via multiple incremental contributions from the community?