Published 29 June 2010 by Martin Fenner

On artificial and synthetic cells

Monday morning Jack Szostak talked about his ongoing work on creating artificial cells, where he is trying to create simple protocells from chemically synthesized material that in their simplest form only contain a membrane and genetic material.

Later in the afternoon Hamilton Smith gave a detailed account of the work by the J. Craig Venter Institute cumulating in the synthetic cell paper published in Science on May 20 (doi:10.1126/science.1190719, fulltext freely available).

Jack Szostak is interested to understand more about the origins of life, and for this is studying the simplest possible forms of living organisms, only containing a membrane and genetic material. A main motivation for his work has been the discovery in the 1980s that RNA is not only encoding genetic information, but can also catalyze chemical reactions inside cells – indicating that RNA might have existed before DNA and proteins. For this work, Jack Szostak and his laboratory have spent a large amount of work studying primitive membranes (e.g. Nature 2008, doi:10.1038/nature07018). Efforts to replicate genetic material using nucleic acids without the help of biologic material so far have not been successful. The work on artificial cells by Jack Szostak and others is fascinating, but extremely complicated (his lab has been working on this since the early 1990s). But we will probably learn a lot about basic biologic processes along the way.
Artificial cells are started from scratch and are completely human-made. In contrast, a synthetic cell is a cell controlled by a chemically synthesized genome and the human-made part is that genome. Hamilton Smith used the analogy from computers to explain the concept of synthetic cells: the genome is the operating system and the cytoplasm is the hardware.
The second genome ever sequenced (in 1995) is from Mycoplasma genitalium, the smallest genome (580 kB) of an organism capable of independent growth in a laboratory. In addition to 485 known coding sequences, M. genitalium contains 100 genes without known function. The synthetic biology group at the J. Craig Venter Institute started synthesizing oligos (later outsourced to several companies) that were then assembled into larger pieces by homologous recombination. It was easy to get to the quarter genome (144 kB), but assembling those last four pieces together proved difficult. The group moved to yeast in which they could assemble the full genome. The next step of transferring the synthetic genome into a receptive bacterial cytoplasm turned out to be very difficult.
After unsuccessfully working on this problem for a few months, Hamilton Smith and his group decided to use faster growing bacteria (M. genitalium has a fairly long doubling time of 16 hours). They switched to the Mycoplasma mycoides genome (1.1 MB) for the donor genome and Mycoplasma capricolum as a recipient. The process used to assemble the genome was the same (oligo synthesis, combination of fragments using homologous recombination, and final synthesis of the full genome in yeast). Hamilton Smith and colleagues then finally succeeded to transfer the synthetic genome into recipient M. capricolum. Along the way they lost three months during the assembly phase because of a single base pair deletion in an essential gene (the chromosomal replication initiator DnaA).
The resulting M. genitalium cells were almost indistinguishable from wild-type M. genitalium. Almost indistinguishable because the synthetic cells not only contain a tetracycline resistance to allow for selection, but also have undergone a few genetic changes in the process: 8 single nucleotide changes due to mutations, a 85 bp insertion, and a 777 bp insertion of an E. coli IS1 sequence (E. coli was used during the cloning process). 
In order to distinguish the artificial genome (and to have some fun), the synthetic M. mycoides also contains watermarks, human-readable information using a new DNA code that is biologically neutral. The scientists at the J. Craig Venter Institute encoded 46 names of people involved in the project, an email address (to contact when you cracked the code) and quotes from James Joyce, Richard Feynman and the R. Oppenheimer biography into the genome. As expected, it took less than three weeks for the code to be cracked and the names of researchers and quotes to be revealed (Using Arc to decode Venter’s secret DNA watermark). 

Martin Fenner