Many people, perhaps most, hate the idea that life might depend on chance processes. It is a human tendency to search for meaning, and what could be more meaningful than the belief that our lives have a greater purpose, that all life in fact is guided by a supreme intelligence which manifests itself even at the level of individual molecules?
Proponents of intelligent design believe that the components of life are so complex that they could not possibly have been produced by an evolutionary process. To bolster their argument, they calculate the odds that a specific protein might assemble by chance in the prebiotic environment. The odds against such a chance assembly are so astronomically immense that a protein required for life to begin could not possibly have assembled by chance on the early Earth. Therefore, the argument goes, life must have been designed.
It is not my purpose to argue against this belief, but the intelligent design argument uses a statistical tool of science -- a probability calculation -- to make a point, so I will use another tool of science, which is to propose an alternative hypothesis and test it. In living cells, most catalysts are protein enzymes, composed of amino acids, but in the 1980s another kind of catalyst was discovered. These are RNA molecules composed of nucleotides that are now called ribozymes. Because a ribozyme can act both as a catalyst and as a carrier of genetic information in its nucleotide sequence, it has been proposed that life passed through an RNA World phase that did not require DNA and proteins.
For the purposes of today’s column I will go through the probability calculation that a specific ribozyme might assemble by chance. Assume that the ribozyme is 300 nucleotides long, and that at each position there could be any of four nucleotides present. The chances of that ribozyme assembling are then 4^300, a number so large that it could not possibly happen by chance even once in 13 billion years, the age of the universe.
But life DID begin! Could we be missing something?
The answer, of course, is yes, we are. The calculation assumes that a single specific ribozyme must be synthesized for life to begin, but that’s not how it works. Instead, let’s make the plausible assumption that an enormous number of random polymers are synthesized, which are then subject to selection and evolution. This is the alternative hypothesis, and we can test it.
Now I will recall a classic experiment by David Bartel and Jack Szostak, published in Science in 1993. Their goal was to see if a completely random system of molecules could undergo selection in such a way that defined species of molecules emerged with specific properties. They began by synthesizing many trillions of different RNA molecules about 300 nucleotides long, but the nucleotides were all random nucleotide sequences. Nucleotides, by the way, are monomers of the nucleic acids DNA and RNA, just as amino acids are the monomers, or subunits, of proteins, and making random sequences is easy to do with modern methods of molecular biology.
They reasoned that buried in those trillions were a few catalytic RNA molecules called ribozymes that happened to catalyze a ligation reaction, in which one strand of RNA is linked to a second strand. The RNA strands to be ligated were attached to small beads on a column, then were exposed to the trillions of random sequences simply by flushing them through the column. This process could fish out any RNA molecules that happened to have even a weak ability to catalyze the reaction. They then amplified those molecules and put them back in for a second round, repeating the process for 10 rounds. By the way, this is the same basic logic that breeders use when they select for a property such as coat color in dogs.
The results were amazing. After only 4 rounds of selection and amplification they began to see an increase in catalytic activity, and after 10 rounds the rate was 7 million times faster than the uncatalyzed rate. It was even possible to watch the RNA evolve. Nucleic acids can be separated and visualized by a technique called gel electrophoresis. The mixture is put in at the top of a gel held between two glass plates and a voltage is applied. Small molecules travel fastest through the gel, and larger molecules move more slowly, so they are separated. In this case, RNA molecules having a specific length produce a visible band in a gel. At the start of the reaction, nothing could be seen, because all the molecules are different. But with each cycle new bands appeared. Some came to dominate the reaction, while others went extinct.
Bartel and Szostak’s results have been repeated and extended by other researchers, and they demonstrate a fundamental principle of evolution at the molecular level. At the start of the experiment, every molecule of RNA was different from all the rest because they were assembled by a chance process. There were no species, just a mixture of trillions of different molecules. But then a selective hurdle was imposed, a ligation reaction that allowed only certain molecules to survive and reproduce enzymatically.
In a few generations groups of molecules began to emerge that displayed ever-increasing catalytic function. In other words, species of molecules appeared out of this random mixture in an evolutionary process that closely reflects the natural selection that Darwin outlined for populations of higher animals. These RNA molecules were defined by the sequence of bases in their structures, which caused them to fold into specific conformations that had catalytic properties. The sequences were in essence analogous to genes, because the information they contained was passed between generations during the amplification process.
The Bartel and Szostak experiment directly refutes the argument that the odds are stacked against an origin of life by natural processes. The inescapable conclusion is that genetic information can in fact emerge from random mixtures of polymers, as long as the populations contain large numbers of polymeric molecules with variable monomer sequences, and a way to select and amplify a specific property.
I will close with a quote from Freeman Dyson, a theoretical physicist at Princeton University who also enjoys thinking about the origin of life:
“You had what I call the garbage bag model. The early cells were just little bags of some kind of cell membrane, which might have been oily or it might have been a metal oxide. And inside you had a more or less random collection of organic molecules, with the characteristic that small molecules could diffuse in through the membrane, but big molecules could not diffuse out. By converting small molecules into big molecules, you could concentrate the organic contents on the inside, so the cells would become more concentrated and the chemistry would gradually become more efficient. So these things could evolve without any kind of replication. It's a simple statistical inheritance. When a cell became so big that it got cut in half, or shaken in half, by some rainstorm or environmental disturbance, it would then produce two cells which would be its daughters, which would inherit, more or less, but only statistically, the chemical machinery inside. Evolution could work under those conditions.”