I always appreciate the opportunity to learn from your comments, the more so when I have ventured outside my expertise. I was particularly excited to receive extended comments from Gary Hurd with references that were new to me and a pointer to one of my favorite biological free-thinkers, Carl Woese. This is a brief response to Gary, with some expanded thoughts and clarifications.
I view evolution as an accelerating process. At the same time that life has been evolving, evolution has been evolving. That is to say, the process of evolution has become more and more efficient over time, a phenomenon I once referred to as Evolution Squared. Recent stages of evolution have been remarkably efficient. Multi-cellular life is only half a billion years old (at least in the form we know it, with cells specialized to form organs, appendages, circulation, nerves etc). So the first 85% of life’s history was spent working on individual cells, their structure and metabolism. I expect that the earliest stages of evolution were remarkably slow.
The very first self-reproducing systems had to appear by chance, and had to evolve progressively before life had really learned how to evolve. My premise last week was surprise that the earliest cyanobacteria appeared so soon after the earth first cooled and solidified. Cyanobacteria have a lot of cell structure and some very fancy metabolic chemistry. This all had to be come together at a time when the evolutionary process was groping around in the dark.
I estimated that life had only 300 million years from loose alliances of molecules to cells that left a fossil record. Gary said this number should be 500-700 million. He’s probably right. 700 million years is the full time available from the first liquid water to the oldest fossils, and it is impossible to know whether it was late or early in that 700 million years that molecular life first self-organized.
Maybe LUCA was the vast blue sea
Gary pointed me to an article by Carl Woese with an alternate picture of the origin of life.
We’re accustomed to think that the Darwinian struggle for existence is a condition of life, and it has always been thus. We imagine that cooperation arose after competition, as alliances became a potent aid in the struggle.
We get our genes from our parents, and micro-organisms get their genes from progenitor cells. We read about bacteria that promiscuously share plasmids—“horizontal gene transfer”—and we think this is a bizarre, chimerical monstrosity.
We’re accustomed to thinking of life lived by individuals, each with its unique lineage extending back in time. We imagine that individuals evolved first, that they diverged and competed, and later organized into predator-prey communities and more complex ecosystems.
The technology we have for inferring genetic relationship and ancient lineage is statistical analysis of the similarity of DNA sequences in widely-varying life forms today, using genes that perform the same core functions. Presumably, these genes all derive from a common source, and the variations that they assume among present life forms tell the story of the path by which they were passed down.
Woese, who knew this evidence as well as anyone and thought about it in the broadest context, concludes that diverse branches of the tree of life cannot be traced to a single root.
The further back in evolutionary time we look, the more the notion of an “organismal lineage”—indeed, the very definition of “organism” itself—comes into question. It is time to release this notion of organismal lineages altogether and see where that leaves us.
The further we look back in time, the more horizontal gene transfer was the rule, and strict lineage the exception. The picture that Woese invokes is of a time before separate selves, when all partook of the chemical commons, and genes were free-floating templates belonging to no one and everyone.
The universal ancestor is not a discrete entity. It is, rather, a diverse community of cells that survives and evolves as a biological unit. This communal ancestor has a physical history but not a genealogical one. Over time, this ancestor refined into a smaller number of increasingly complex cell types with the ancestors of the three primary groupings of organisms [archaea, bacteria & eukaryotes] arising as a result.
One day, an oily film walled off one little portion of the sea, and the chemicals therein spoke the word “mine” for the first time in history.
I am fascinated by this picture. It comes from an eminent scientist, it seems plausible, and it aligns with diverse spiritual wisdom about the unity of life and with mythology of a time before our current Age of Separation. It will be awhile before I am able to assimilate its implications.
Is it hard to create a self-replicating network of molecules?
I claimed last week that lab scientists had tried and failed to come up with self-replicating molecular cycles. Gary pointed me to five reports in the literature of self-replicating chemistry. After looking at them, I think we’re both right. Indeed, there are examples of molecular systems that are able to copy themselves, but they work only when immersed in a soup of chemical feed that is already too complex to have arisen by chance. Not only are the simplest molecules capable of self-replication not simple enough that they might ever have arisen in a whole pre-biotic sea of random molecules; but even these are able to assemble copies of themselves only when provided with constituents that are also too complex to be plausibly present before biology.
All present life requires three kinds of molecules: Proteins, DNA and RNA. Proteins do the cell’s work, including the work of replicating DNA; but DNA holds the information that tells how to build a protein, so each is dependent on the other. DNA and RNA have the essential property that they can act as templates for their own replication. That’s the significance of the “double helix”—two strands of DNA or RNA can come unzipped, and each finds pieces to make a new mate for itself. But in biology, this only happens with the aid of protein molecules that do the zipping and unzipping, and finding the constituents to build the copy. To make a protein from a DNA template requires an RNA intermediary and a tiny molecular factory called a ribosome, which is itself a miracle of natural bioengineering. This 3-component system is so complex that it could not have been the basis of the first life on earth. So biochemists looking for a simple self-replicating system work either on proteins alone or with RNA alone. (DNA alone is not considered viable because it is a passive informational molecule and is not capable of doing anything on its own.)
The “protein world” hypothesis has the advantage that proteins are made of amino acids, which are relatively easy to make and are plausible constituents of the pre-biotic soup.
The “RNA world” hypothesis has the advantage that RNA is both a workhorse molecule and a template for replication. But the constituent pieces of RNA are nucleic acids which are, themselves, harder to synthesize and it is thought that the pre-biotic soup would contain only minuscule amounts of them compared to the amounts of amino acids, which were already rare (dilute) in an absolute sense.
The five papers Gary cites are interesting for what they accomplish, as well as for what they fail to accomplish:
- Lee et al, Nature 1997 This group at Scripps Inst in La Jolla works with a protein that is able to make more of the same protein (no RNA or DNA). A particular protein of length 32 can act as a template to make copies of itself by joining together two pieces of length 17 and 15 residues.
A protein (or “peptide”) is a chain of amino acid molecules. The individual pieces are called “residues” in this context. Individual residues qualify as simple enough to have appeared by chance in the pre-biotic soup. However, chaining them together can be done in many different ways.
For example, there are hundreds of known, simple residues. 20 of them are essential for today’s biology. The number of different chains of length 32 that can be made using just the 20 known residues is 20 raised to the power 32=1041.
In other words, this is a hugely improbable molecule to appear spontaneously, and even so, it can’t assemble copies of itself from individual residues, but only from two halves of itself.
My judgment is that this is tremendously exciting work, it’s on the right path, but still both too complex and not effective enough to be a candidate for the first self-replicator.
- Lincoln & Joyce, Science 2009. This is another group at Scripps that works with RNA alone (no proteins). The paper describes a pair of RNA molecules, each of which was effective in assembling a new copy of the other.
This is an example of a hypercycle, as I described last week—a set of molecules that are mutually auto-catalyzing. The fully-assembed molecules are more than 100 bases in length, and they can be assembled from smaller pieces of themselves. The smallest piece in the “feedstock” is 21 bases in length, and the largest is more than 60.
As with the protein work above, I would judge that this is tremendously exciting and promising work. I’m grateful to Gary for pointing it out to me. But we’re still a long way from having a candidate for the first pre-biotic chemical system. These molecules are too large (= too complex=too improbable) to have plausibly appeared by chance in all the world’s oceans in 500 million years. And even if one did appear by chance, it would require the other, and then the two would only be able to replicate if provided with smaller fragments.
- Turk et al, PNAS 2010. This group at UC Santa Cruz reports a crucial protein-building step performed by an RNA fragment that is only 5 bases long. In order to support biology that includes both RNA and proteins, you need ribosomes, which can make a protein to order from a from an RNA blueprint. But ribosomes are much too complex to be part of the first living things. This paper reports that one piece of the work done by a ribosome can be accomplished by this simple 5-base RNA fragment. 5 bases linked together is a small enough molecule that it could plausibly have appeared by chance, before biology.
The Bottom Line
In summary: I’m impressed with the progress that has been made in the search for the chemical basis of pre-biotic life, and I’m grateful to Gary for having pointed me to this literature. We have a long way to go before we can say we understand how life got started, but we’ve made some promising steps in the right direction.
Specialists in this area of research remain divided in their fundamental pictures of the origin of life. Some favor the protein world, some the RNA world, and some the view that I described last week, which is that life arrived on a meteor from an extra-terrestrial source. A common criticism is that the extra-terrestrial hypothesis is vacuous, or superfluous, because it just “kicks the can down the road”. You are still left having to explain how life got started on some other planet at some other time. But I’d argue that the extra-terrestrial hypothesis contributes three things:
- it helps explain why all life on earth is related, with a common chemical basis;
- it helps explain why life arose relatively quickly after the birth of Planet Earth;
- and it provides more time and space for the vastly improbable events that led to the first life.
The ET hypothesis makes a prediction that may someday be testable: If extraterrestrial life is discovered, or if we are visited, then the prediction is that these visitors will have metabolisms and genetics based on proteins and DNA, respectively, just like us.