Can (non-biological) life be created de novo?

Publiceret Juli 2006

The “essence of life” has fascinated people through all times. What is life, how did it arise or where did it come from? Scientifically, much research and speculation has been devoted to the question of the origin of life on Earth, whereas less attention has been focussing on the prerequisites for life and the possibilities of creating “artificial life” in the laboratory.  In two ambitious, newly started, “intercontinental” projects, a European-US consortium is exploring routes that may eventually lead to non-biological artificial life.

Although we all believe that “we know what life is”, and are able to determine if an object is alive or not, a strict and rigorous scientific definition of “life” is far less straightforward. However, most scientists will agree that life must posses several key properties such as the ability to replicate, i.e. make copies of it self, and thereby multiply, as well as the ability to evolve, such that the progeny is not an exact (genetic) copy of the parent. In order to replicate and evolve an ordered structure and free energy are required. Therefore, to produce an ordered progeny from unordered components of the environment, free energy, and a flow of resource materials, are needed (in thermodynamic terms: as the entropy of the living entity is decreasing, the entropy of the environment increases and the free energy is “consumed” by the living entity).

Thus if we limit ourselves to chemical systems, as opposed to purely electronic systems (life within a computer [1]), life must consist of a genetic information bearing material and a metabolic system and these must be physically confined, e.g., in a container. Furthermore, the genetic system must be coupled to the metabolic system and/or the container, in order to establish what in contemporary life constitutes the genotype-phenotype coupling.

One of the most fundamental questions to ask is of course what are the minimum chemical and physical requirements for life. This would help us understand how life could have originated on Earth and also provide a much better basis for predicting the possibilities of discovering life elsewhere in the Universe. Furthermore, if it can be demonstrated that new life forms can be based on “non-biological” chemistry, i.e., life forms which are orthogonal to the biological life we know from Earth, this would profoundly change our view of Life biologically as well as philosophically.   

Chemists have within recent years attempted to develop self-replicating “living” systems based on liposome contained catalytic RNA [2], as a model for the origin of the RNA world. Such a system would not be conceptually or chemically different from contemporary life, and could be seen as a predecessor of this. However, self-replicating chemical systems has been demonstrated based on small oligonucleotides [3], or peptides [4] or even simple non-biological organic compounds [5]. Such systems could in principle lay the foundation for “non-biological life”.


Some years ago a very simple and primitive model for artificial life was proposed by Rasmussen and Chen [6,7]. It consists of the three central components, genetic material, a metabolic system, and a container. The key essential features involve a physical coupling of the genetic material to the container (in one version the container is a liposome in which the genetic component is embedded), and a coupling between the genetic material and the formation of container components via a photochemical metabolism, which is sequence dependently catalysed by the genetic material. The principles of the model are illustrated in Figure 1.

Minimal life and Los Alamos minimal protocell
Figure 1: We utilize an operational definition of minimal life and our protocell design principle is based on cooperation between a genetic, a metabolic, and a container structure. Our protocellular container is composed of fatty acid molecules, which are amphiphilic (having a polar water-soluble head group attached to a non-polar water-insoluble chain), and thus self-assemble in water. Also the gene molecules are amphiphilic (or even hydrophobic) and they are composed of modified peptide nucleic acid (PNA). The metabolic molecules, the light sensitizer, can either be free or linked to the gene templates (as shown here). Larger version.

In an attempt to physically materialize this model, an intercontinental research effort was initiated in 2004 by a European 6th framework sponsored project, Programmable Artificial Cell Evolution (PACE) together with a US Los Alamos National Laboratory sponsored project, Protocell Assembly (PAs). Together these two projects involve groups in Denmark, US, Germany, Sweden, Spain, Italy, Ireland and Lithuania. The ultimate aim, of course, is to demonstrate the feasibility of an artificial, replicating, and evolving cell (artificial life), but many challenges remain before that goal can be realized.  Multiple alternative ways are suggested to assemble a minimal life form as proposed for example by Luisi in Rome [8], Szostack at Harvard, and Deamer at Santa Cruz [9].

Our protocell design is quite different from a modern biological cell.  First of all, the notion of a cell container is not a traditional vesicle composed of a lipid double layer separating the cells contents from the outer world. One of the difficulties with a vesicle design is the barrier of transporting hydrophilic and maybe charged metabolites in and out of the container through the lipid bilayer. This problem, however, can be sidestepped by using a different and much simpler container concept based on micelles or (oil) droplets onto which (or in which) the cell components (“genes” and catalysts (“enzymes”)) are associated (dissolved), and thereby being freely exchangeable with the environment (Figure 1). The minimal container may be formed by simple fatty acids, which self-assemble into micelles (or vesicles) in water due to their amphiphilic character.

Concerning the genetic material, Nature’s chosen molecules, DNA and RNA are highly hydrophilic and charged macromolecules which do not freely dissolve in a lipid environment and will only associate with liposomes if these are positively charged.  However, the synthetic and non-natural DNA mimic peptide nucleic acid (PNA) [10] (Figure 2) is inherently charge neutral and can be modified to greatly increase its lipophilicity. Indeed we believe that it will be possible to create PNA molecules that are freely soluble in an organic (lipid) phase and have indirect evidence that stable PNA-PNA duplexes, which are crucial for a replication process, are stable under such conditions.

Chemical structure of PNA compared to DNA
Figure 2: Chemical structure of PNA compared to DNA

We envision that with such a container and lipophilic PNA genes a geno-phenotype coupled protocell may be created. This protocell would of course be far from self-sustained even if one could establish an effective PNA self replicating system as it would have to be fed with activated precursors for the PNA and lipid precursors for the container. In order to approach a more self sustained situation and also one being conceptually closer to the properties of biological life, a metabolism must be included.

We have chosen to focus on light as the source of energy as this is easy to dose and can be conveniently converted to chemical energy via relatively simple chemical photosensitizers. The wavelength of the light required and thus the free energy available to drive the metabolic reactions depends on the properties of the sensitizer, and the sensitizer has to be energetically compatible with the molecules it interacts with in the light-driven metabolism.

Metabolism, gene replication, and life cycle

In the protocell design a (trisbipyridyl)ruthenium complex has been chosen as the sensitizer because it absorbs in the visible region and effectively engage in specific electron transfer processes that may be exploited for “metabolism”. In the model system, the light-driven metabolic process produces fatty acid building blocks for the protocell from a lipid precursor. Furthermore, the (trisbipyridyl) ruthenium complex will be regenerated by electron transfer via the nucleobases of the genetic material (PNA). Theoretical calculations predict that the redox potential of the natural nucleobases is too large for such a process but model experiments have shown that 8-oxo-guanine will serve the purpose (Figure 3) [11]. This scenario ensures a crucial direct coupling between the genetic material and the growth of the container as the gene is indirectly catalyzing the metabolic production of new container material, and therefore a balanced growth of the protocell should be achieved [12]. In addition, it is predicted that the sequence of the genes will influence the catalytic potency, and thereby can be evolutionarily optimized. Finally, it is anticipated, as shown for analogous systems (e.g. growing micelles) that as the protocells grow in size they will at some stage spontaneously divide.

Schematic of the protocell's metabolic cycle. Click for a larger version.
Figure 3: Schematic of the protocell's metabolic cycle. The first experimental evidence is by DeClue, Bailey and Boncella at the Los Alamos National Laboratory, May 2006. This system supports the feasibility of the protocell design as it demonstrates how in principle an informational molecule may regulate the metabolic production of container components, establishing the desired catalytic connection between the genes, the metabolism, and the container. The process starts with the absorption of a photon that causes a charge separation within the Ru-bpy sensitizer molecule. As the gene molecule (in this initial case realized only by the 8-oxo-guanine base) donates an electron the sensitizer center is neutralized and the energy rich electron can be utilized for the cleavage of a picolinium ester (resource molecule) to form fatty acid and waste. Finally, the "ground state" gene has to be regenerated by a hydrogen source (perhaps water). (Picture courtesy Michael DeClue, James Bailey and James Boncella, LANL). Larger version.

Potential microfluidics life-support and programming

In order to experimentally realize the full life cycle of the protocells, and particularly their chemical evolution, one needs to rely on more controllable experimental methods compared to the ordinary closed-tube systems. We thus make use of micro/nanofluidic lab-on-a-chip systems, which can manipulate fluids and objects within networks of micro/nanochannels by means of viscous, surface and electromagnetic forces. Lab-on-chips have complex channel geometries (both 2D and 3D structures in silicon, glass or polymers) with specialized modules (for mixing, reactions etc.) and various integrated parts (electrodes, temperature- and pH-sensors, waveguides, etc.), [13]. The latest development in the field is programmable microfluidic technology that enables computer-controlled flows (by programmable electrodes and actuators), and real-time monitoring (by processing read-outs from within the system), [14]. Microflows and bio-chemical reactions can thus be controlled and programmed, e.g., in/out of a chemical microreactor, but also feedback adjusted based on, say, fluorescent intensity of interacting reactants. Programmable microfluidics is particularly suitable for realization of the protocells as it allows for separate controlled optimization of different parts of their life-cycle, i.e., the gene replication and the metabolism, which can then be synergistically combined in a cyclic manner (Figure 4). In this way we plan to evolve the protocells through repetitive selection of their replicating and metabolic properties, [15]. Eventually, we believe, the microfluidic platform could enable assembling and “life-support” of protocells with even more complex chemistries.

A potential microfluidic design that could co-evolve optimal conditions for sequential metabolic (lower channel) and genetic (upper channel) functionalities of protocells. Click for a larger version.
Figure 4: A potential microfluidic design that could co-evolve optimal conditions for sequential metabolic (lower channel) and genetic (upper channel) functionalities of protocells. Such systems could provide initial life-support for protocells and a controlled evolution. In the long term we believe that such systems can be utilized to "program" protocells for useful tasks such as to grow in particular chemical environments. Larger version.

Is it all science fiction?

Perhaps, but it involves new science and will despite its naïve simplicity guide us closer to answers of the essence of life and its possible creation. Also being able to make general self-replicating materials is a major challenge that stands between humans and our ability to develop a new so called “living technology”. Living technology is based on robustness, autonomy, local intelligence, self-repair, adaptation, and self-replication, which are all desirable properties that our current technology lacks. It is of course not possible to predict when chemically based artificial life will be produced and which form and properties it will have, but we are confident that artificial life can be created although it will most certainly (at least initially) not have the robustness, autonomy, and evolvability of biological life. However, once created, such artificial life forms will define exciting scientific as well as technological challenges to study, as they will be able to evolve according to Darwinian principle in analogy to biological life. Such evolutionary processes could give rise to living entities, which are better suited to certain environments of technological interest than the biological ones first created some 4 billion years ago.


[1] M. Bedau, J. McCaskill, N. Packard, S. Rasmussen, C. Adami, D. Green, T. Ikegami, K. Kaneko, and, T. Ray, Open questions in artificial life, Artificial Life 6 (2000) 363–376

[2] J.W. Szostak, D.P. Bartel, P.L. Luisi, Synthesizing life, Nature 409 (2001) 287-390

[3] A. Luther, R. Brandsch, G. von Kiedrowski, Surface-promoted replication and exponential amplification of DNA analogues Nature 396 (1998) 245-8

[4] A. Saghatelian, Y. Yokobayashi, K. Soltani, M.R. Ghadiri, A chiroselective peptide replicator, Nature 409 (2001) 797-801

[5] Q. Feng, T.K. Park, J. Rebek Jr., Crossover reactions between synthetic replicators yield active and inactive recombinants, Science 256 (1992) 1179-80

[6] S. Rasmussen, L. Chen, M. Nilsson, S. Abe, Bridging nonliving and living matter, Artificial Life 9 (2003) 269-316

[7] S. Rasmussen, L. Chen, D. Deamer, D. Krakauer, N. Packard, P. Stadler, M. Bedau, Transition from nonliving to living matter, Science 303 (2004) 963-965

[8] P.L. Luisi, P.S. Rasi, F. Mavelli, A possible route to prebiotic vesicle reproduction, Artificial Life 10 (3) (2004) 297-308.

[9] A. Pohorillle and D. Deamer, Artificial cells: Prospects for biotechnology, Trends in Biotechnology 20 (2002) 123–128

[10] P.E. Nielsen, M. Egholm, R.H. Berg, O. Buchardt, Sequence-selective recognition of DNA by strand displacement with a thymine-substituted polyamide, Science 254 (1991) 1497-1500

[11] Eds. S. Rasmussen, M.A. Bedau, L. Chen, D. Deamer, D.C. Krakauer, N.H. Packard, & P.F. Stadler, Protocells: From nonliving to living matter, MIT Press, in press (2006).

[12] T. Rochelau, S. Rasmussen, P.E. Nielsen, M. Nilsson, and H. Ziock, Emergence of protocellular growth laws, Phil. Trans. B in press (2006).

[13] Eds. O.Geschke, H.Klank, and P.Telleman, Microsystem Engineering of Lab-on-a-Chip Devices, Wiley-VCH (New York, 2003)

[14] U. Tangen, P. F. Wagler, S. Chemnitz, G. Goranovic, T. Maeke, J. S. McCaskill, An electronically controlled microfluidic approach towards artificial cells, Complexus 3 (2006) 48-57

[15] G. Goranovic, S. Rasmussen, P. E. Nielsen, Artificial life forms in microfluidic computers, to appear in Proc. MicroTAS 2006, Tokyo, Japan, Nov 2006