Let us turn our attention to the origin of life on Earth. How did it come about in the first place? The writers of the Bible consistently claim that YHWH is the Creator of life. This sounds like an extraordinary claim, but I have already presented extraordinary evidence for it, showing how the story of Jacob’s life in Padan-Aram contains many analogies to molecular biological processes including transcription and translation, as well as concepts such as selection and mutation.
However, let’s consider the alternative, assumed by much of the scientific establishment, that life on Earth somehow arose by itself. First of all, what do I mean by “life”? It’s difficult to define precisely, and there is disagreement over what is living and what is not. Since I want to avoid falling into a word trap and debating the definition of the word “life” rather than the idea behind it, let’s put an exact meaning aside, and look at what life on Earth has in common.
One thing that seems common is the cell. But even the simplest of cells are actually very complex. A typical single-celled bacterium, for example, has a membrane and maybe also an outer wall, to keep itself separate from the outside world while still allowing certain materials in.
Another common feature of life on Earth is the DNA (deoxyribonucleic acid) molecule. Even the lowly bacterium has a circular DNA double strand which contains a blueprint for itself. Most cells also use the core biological processes of transcription and translation, where a DNA sequence is transcribed into a messenger RNA (mRNA) strand, and then the mRNA strand is converted by a ribosome into a protein.
A key requirement for life on Earth is the ability to reproduce or replicate. The bacterium can duplicate itself, in a process called “binary fission.” It also has a built-in system for producing energy, and an internal transportation system to pass materials around the cell and in or out of its environment.
These are all complex features. If life came about all by itself, obviously it can’t have started out being anywhere near as complex as even the humblest of single-celled organisms we see today. How did it happen? How did non-living chemicals become living organisms?
Scientists don’t know for sure, because no human was around at the time to observe it. However, various ideas have been suggested. One of the most popular ones is called “RNA World.” In this hypothesis, RNA (ribonucleic acid) molecules arose before DNA became prominent. There are several reasons why proponents of this idea think RNA helped life to get started.
First, some of the core cell machinery uses a lot of RNA, including the ribosome. Since RNA is quite similar to DNA, they argue that much of the initial machinery was made out of RNA strands, and then some of it was later replaced by DNA, which is more stable. Second, certain RNA strands can act as catalysts. This means they can dramatically increase the reaction rate of a chemical process, which is necessary for many of the processes critical to life. Third, short strands of RNA can act like miniature machines, much like proteins do.
In the hypothetical RNA World, pieces of RNA perhaps floated around in some liquid, or were sandwiched together in clay or ice or something else, and gradually formed into longer strands. At some point, these clumps of RNA acquired membranes, forming a simple “protocell,” and the protocell somehow acquired the ability to replicate, and eventually to store coded information and translate that information into useful proteins or some primitive equivalent.
The real question for this or any other hypothesis on the origin of life is, how could these things happen, so that we arrive at the equivalent of the first living cell? It’s one thing to tell a story of how it could have happened, but quite another to show that it’s actually feasible. There are enormous hurdles along the way.
Let’s conduct a thought experiment in the laboratory of our mind, so we can gain a better understanding of the kind of challenges nature would face in getting anywhere close to the cell as we know it.
Everyone deserves to know what those hurdles are and whether nature really can overcome them on its own – especially atheists, because atheism is committed to the assumption that life arose by itself, without any supernatural interference.
I think it’s also a good idea, at least once in a lifetime, to carefully examine the basis of our own worldview and assumptions, to see if they stand up to deeper scrutiny. This is true just as much for believers as it is for atheists; and I think now is the perfect time to carefully examine the assumption that life came about on its own.
In our thought experiment, we will create a simple “protocell,” a very basic version of a cell with just a few critical features. We will start by giving it a simple membrane that keeps any inner machinery separate from its environment.
Inside our protocell, we will place a molecule that has the potential to store information. It will probably be an RNA molecule, but it might be DNA or maybe something more primitive, so we will simply call it the “master molecule” – for reasons that will hopefully become clear in a short while. To begin with, the master molecule doesn’t contain any useful information.
In a real cell, each unit of a DNA or RNA strand is called a “nucleotide.” We could think of a strand of RNA as a string of beads on a necklace, each individual bead representing a nucleotide. To help us shift away from our thought experiment and back to real biology later on, I will also stick with the term “nucleotide” to describe an individual unit making up the master molecule. In other words, the master molecule can be thought of as a string of nucleotides.
We’ll also assume that each nucleotide can be one of four bases, just as RNA uses the four bases adenine (A), cytosine (C), guanine (G) and uracil (U) – although there’s no reason to suppose a primitive protocell would use the same ones. In any case, if the master molecule is like a string of beads, then we could think of each individual bead as containing one of the letters A, C, G or U. We will start with a “blank” master molecule, which I suppose means that each nucleotide is the same letter, so there isn’t any useful information in the initial sequence of letters, just as blank digital media might start out full of 0’s before any meaningful information is stored in it.
We will also grant our protocell the ability to replicate – that is, to duplicate itself in its entirety. This feature isn’t coded for in the master molecule, which we have assumed to be blank. The ability to replicate and build a membrane need to be intrinsic chemical features of the protocell somehow. We will leave these processes unexplained for now, and just assume the protocell can do them.
Furthermore, we’ll assume that during replication, errors sometimes occur. As a result, the sequence of nucleotides in the master molecule of a replicated protocell might differ slightly from that of the original. In evolutionary terms, this would be the equivalent of a mutation.
Finally, let’s assume we have just the right environment for a virtually unlimited number of these protocells to come about through replication. We don’t want our thought experiment to hit any space or resource constraints.
I have made a lot of important assumptions here, and I haven’t explained how these features came about. In a sense, I have already given the race to life an unfair advantage. I have done this for the sake of pressing on with my main arguments. For the moment, let’s just assume the protocell has these features, and get on with unleashing it into the environment of our thought laboratory, so we can watch as it replicates to become thousands, then millions, then billions, and then trillions of protocells. What happens? Do we get anything interesting?
What we find is, all of the protocells look essentially the same. This is because, to evolve into anything even slightly different, the master molecule must become more than just a place to store information. It also has to become a laboratory of innovation, and a primitive factory where new machinery can emerge. This is why biologists favor RNA molecules in the origin of life. RNA sequences have the potential to store information, and certain sequences can also perform other useful functions.
For our growing colony of protocells, a slightly better membrane or a more efficient replication process might evolve. However, we have assumed these things are purely chemical in nature to begin with, rather than being controlled by the master molecule, for reasons that will become clear in a little while. This means there aren’t any blueprints stored away in the protocell for building a membrane or anything else. To begin with, the protocell must somehow manage without them.
Innovations requiring a blueprint therefore need to evolve in the master molecule, because as well as being the source of any new machinery that can help the protocell to survive, the molecule can also act as the protocell’s memory bank, the place where blueprints can be stored. This is why I called it the “master molecule.” It’s where the real evolutionary action needs to happen.
Now, we have already granted our protocells one tool of evolution – the ability of its master molecule to mutate whenever replication happens. Incidentally, it’s worth keeping in mind that mutation isn’t really a tool or feature as such, but rather, it’s the result of occasional copying errors in the replication process.
Another useful tool would be natural selection. If mutations can eventually produce a useful new function from the master molecule, one which can give a protocell a survival edge, the offspring of that protocell could perhaps inherit that advantage, and eventually their offspring could become dominant in the colony as a whole. But in order to find such an advantage in the first place, our colony of protocells has to perform what I call a “natural search.”
To explain what I mean by this, let’s suppose the master molecule in each new protocell is able to store a number in digital form, but it only has room for a sequence of six digits from 0 to 9. In other words, it can store a number ranging from 000000 to 999999. Each time a protocell replicates, the number is copied over to the new protocell, with one digit changed, the equivalent of a mutation.
For example, let’s start with a protocell storing the number 000000, which we will assume contains no useful information, so is “blank.” After replicating, the protocell copy might be storing the number 020000, because one number has mutated in the replication process. In turn, a copy of the copy might have the number 027000. A copy of the copy of the copy might have 027300, and so on.
Now, let’s say that the number 123456 would somehow give the protocell a survival advantage. How long would it take our colony of protocells to produce one with the number 123456 in it? It might happen very fast, but it could also take a while. However, as our colony grew larger and produced a greater variety of numbers, we could expect the 123456 sequence to be found more quickly.
This is a very simple example of what I mean by a “natural search.” The colony of protocells isn’t actively searching for the number 123456. It doesn’t particularly care for our thought experiments, and individuals in the colony are just getting on with the business of being imaginary protocells. But collectively, the colony is engaging in the equivalent of a search for the number, new generations producing mutated numbers until one protocell happens to hit on the sequence that will help it to survive better.
This is why I call it a “natural search.” I’m using the word “search” in the same sense that evolutionary theorists use the word “selection” in “natural selection.” Individuals aren’t deliberately searching or selecting. It’s a natural process that can take place in large populations.
For natural selection to apply to our colony of protocells, they first have to perform some natural searches, to find new functions that give some protocells an advantage. If the master molecule consists of RNA, then the protocell colony isn’t searching for numbers like 123456, but for RNA sequences that can perform some useful function, either on the master molecule itself, or in the protocell. Instead of the numbers 0 to 9, we can think of RNA sequences as made up of the “letters” A, C, G and U which represent the chemical bases used in RNA.
Incidentally, how many nucleotides in length do we make the master molecule in each protocell? What is its storage capacity? We didn’t specify this at the beginning of our experiment. For example, if we make it just 10 nucleotides in length, it can store the equivalent of 10 letters made up of A, C, G and U, ranging from AAAAAAAAAA to UUUUUUUUUU and every variation in-between, such as CAUGGCAAUC. There are over a million possible variations.1
However, this tiny length probably isn’t enough to store or produce any useful new functions. As we increase the storage capacity of our protocells, we also increase the number of variations available to our colony in a natural search. Each time we add one nucleotide of storage capacity, we multiply the number of potential variations fourfold. For example, if we go from 10 to 11 nucleotides, we go from about one million to about four million possible variations. If we add another nucleotide we have about 16 million possible variations.2
The problem here is, as we continue to add more nucleotides to the master molecule, the number of possible variations quickly gets out of hand, so that by the time we reach a length of just 150 nucleotides, there are more variations available than there are atoms in the universe! 3
This means, if we wanted to find a specific RNA sequence that was 150 letters long using a natural search, even if we were somehow able to stuff every atom in the universe with a protocell, we would still have a very hard time finding the sequence.
To put our protocell storage capacity into some kind of perspective, there are about 3 billion “letters” in a human DNA molecule, and several million in the DNA of a typical bacterium. Therefore even a 150 nucleotide storage capacity for our protocells would be pitifully small compared to a real cell.
However, storage size isn’t our main concern, because we can always add more nucleotides. What matters is, for our colony to come up with the first useful new function by itself, the length of the nucleotide sequence which makes up the function needs to be small enough that our colony doesn’t have to search for it using all the atoms in the universe – which, for obvious reasons, might be a bit impractical. Even though this is simply a thought experiment, we still want it to have some relevance to the real world.
A natural search is probably the only way our colony of protocells can acquire many new functions and RNA sequences, assuming we don’t put them in ourselves, since the process of natural selection can’t really begin until nature has something useful to select from. In other words, some basic functions probably need to come into existence from scratch, rather than evolve out of already existing functions, because in the primitive RNA World, there aren’t many functions around in the first place!
Now, in the language used by biologists, the complete genetic sequence of an organism is called its “genome,” and this is stored mainly in the DNA molecule. But at the moment, the nucleotides in each master molecule of our rapidly growing protocell colony don’t actually code for anything. Therefore, the protocells don’t really have a genome as such. At best, all they have is a potential one.
Until they evolve the kind of machinery I’ll discuss shortly, all that can happen to our protocells is, certain bits and pieces of the master molecule might possibly mutate to become little RNA machines. But to even begin getting close to cells as we know them, our protocells need the equivalent of a language or code.
In a real cell, the language of DNA is called the “genetic code,” and it is quite elegant. Three nucleotides in a row (such as AAA or AGC) make up one “codon,” and one codon usually represents an amino acid. For example, the AAA codon translates into an amino acid called lysine, and GGG codes for glycine.
Since three nucleotides make up one codon, and DNA uses four bases (A, C, G and T), there are 64 possible codon variations, ranging from AAA to TTT.4 This means the genetic code has room to code for 64 different amino acids. However, only 20 are actually coded for directly in a DNA sequence, with a few more included later in the system. Many amino acids are represented by more than one codon. For example, the codons GGG, GGC and GGA all code for glycine.
As you may recall, a DNA sequence is copied, or “transcribed,” into an mRNA strand. The mRNA strand is then read, or “translated,” by a ribosome, turning the codons into a chain of amino acids – so if the ribosome encounters a GGG, GGC or GGA codon, glycine gets added to a chain of amino acids that will make up a protein.
Why does our genetic code only use 20 amino acids, when it has room for 64? At first, biologists thought this was inefficient, and referred to it as “degenerate.” But it turns out to be a very clever system. It reduces the likelihood of producing the wrong amino acid. Coders and engineers call this “redundancy.”
Let’s say the third nucleotide in a GGG codon is accidentally mutated and becomes GGC. The redundancy of the genetic code means the same amino acid will still be produced, because both GGG and GGC code for glycine. In other words, the redundancy is a feature, meant to reduce errors. Furthermore, research has shown that, compared to a million other coding schemes, our genetic code is optimal for minimizing errors in many ways.5
Of course, our hypothetical protocell couldn’t start out with a genetic code, because that would be cheating even more than we have already done by giving it the ability to replicate. However, at some point it will need some kind of genetic code, if it’s going to begin storing information in the master molecule that can then be read and translated into something else.
Incidentally, I will use the words “language” and “code” interchangeably from now on, because both contain the properties we need, to turn data into something useful.
In a living cell, ribosomes and “transfer RNA” (tRNA) molecules do the job of turning coded information in an mRNA strand into a protein. If our protocells are going to evolve into anything that resembles a cell, they will need machinery equivalent to a ribosome. However, ribosomes as we know them are very complex machines, built out of multiple RNA strands and proteins that, in the language of DNA, require tens of thousands of nucleotides worth of data to construct.
One of the reasons biologists think some kind of RNA World must have existed in the past, is because the core of a ribosome seems to be made primarily out of RNA, rather than proteins. Biologists interpret this as evidence that the primitive ribosome was made entirely of RNA. But is there an alternative explanation? For example, could this be a design feature? If so, why would it be this way?
The ribosome is the machine that makes proteins. If ribosomes stopped working in a cell, proteins couldn’t be made; and if ribosomes were made out of proteins, the cell would die, since it couldn’t make any new ribosomes.
However, since they are made primarily out of RNA, in an emergency scenario where the old ribosomes stopped working, the cell may still be able to put together new ones, and thus continue with protein production. In other words, that the core functions of a ribosome are made primarily out of RNA is perhaps an emergency fail-safe system to prevent the cell from dying if old ribosomes stopped working.
Basing the ribosome on RNA also solves a chicken and egg problem. If ribosomes were built out of proteins, and proteins built by ribosomes, which does the cell build first? It couldn’t build proteins without ribosomes, but it couldn’t build ribosomes without proteins. Therefore, it builds the core of a ribosome out of RNA strands instead, avoiding the dilemma.
In addition, if certain amino acids aren’t available from the environment, this could disrupt the availability of proteins, but would have less effect on the construction of ribosomes, which are essential for making any proteins at all.
In other words, the construction of ribosomes primarily out of RNA could be a smart feature, designed to reduce the possibility of a cell dying, and to give it more flexibility at times when certain resources aren’t available.
Whatever the case, in our thought experiment we can’t simply gift ribosomes to our protocells. This would be cheating. However, this is what some researchers do. When creating “artificial” cells, they add in components such as ribosomes that have been extracted from living systems, because they know cells need to be able to read and translate information.6
The same is true for our protocells. If they are going to be able to read and translate information from the master molecule, they will need something equivalent to a ribosome. Of course, it would be much more primitive than a ribosome as we know it, but it must still be able to read the master molecule, and translate the information into something useful for the protocell.
However, this creates several dilemmas for our protocell experiment, and by extension, for any naturalistic theories on the origin of life.
First of all, if master molecules don’t contain any coded information, there would be no need for a ribosome to evolve to do any translating. The master molecules might contain sequences of RNA that do interesting things, but these sequences wouldn’t be coded proteins.
Second, if master molecules did contain information worth reading, if our protocells lacked the machinery to read and use the information, it wouldn’t give the protocell any survival advantage and so would be mutated away.
We might choose to insert the complete works of Shakespeare into one particular master molecule, but unless the protocell can access the information and do something useful with it, Shakespeare’s works would gradually mutate away into oblivion over many replication cycles, because our protocells don’t care about great literature. It has no survival value to them.
Third, a protocell can’t store or translate coded information in its master molecule without a genetic code, a language in which to read the information. But it wouldn’t evolve a genetic code unless there was useful information written down; and if it’s not useful, it would be mutated away. Without a genetic code, a GGG sequence in the master molecule wouldn’t code for anything. It wouldn’t be a codon for the glycine amino acid, it would just be three guanine (G) nucleotides.
In our own cells, tRNA molecules are the little machines that store the genetic code. Each of the amino acids used to make proteins has a corresponding tRNA molecule that carries it to the ribosome. For example, one particular tRNA molecule has, at one end, an “anticodon” that matches up with a GGG nucleotide sequence, and at the other end, it carries a glycine amino acid.
If a ribosome encounters a GGG codon while reading an mRNA strand, this type of tRNA molecule finds its way into a binding site in the ribosome, pairs up with the GGG codon, and the glycine amino acid it carries gets added to a chain. This is how GGG codes for glycine. Codons are simply instructions letting the ribosome know which amino acids are to be added to a chain. But without something like a ribosome and a genetic code stored in tRNA molecules, GGG is just GGG. It doesn’t code for anything.
To sum up these three dilemmas: in the hypothetical RNA World, information can’t be encoded in the master molecule, since there is no genetic code in which to encode it. RNA sequences have no reason to evolve protein-building equipment such as a ribosome, since there is no encoded information to extract from the master molecule. And even if there was, without the equivalent of a ribosome and a genetic code to translate the encoded information into something useful, the information would be mutated away.
This is why many theorists tend to skip over the precise details of how something like a master molecule, complete with the language to encode information, and the equipment to decode that information, actually came to be. They chalk it up to incredible luck, or brush it off with a little storytelling.
Fortunately, we don’t need to rely on theories or stories. There is a way to put all of this to the test. To begin with, an engineer could design a minimally functional equivalent of a primitive ribosome. We don’t want evolutionary theorists saying it evolved out of an even simpler ribosome, so ask the engineer to make it as simple as possible.
Presumably it would be very crude in comparison to a real ribosome. It just has to be capable of reading a sequence of nucleotides, and building something physical based on the sequence, in a consistent manner. We won’t quibble if there is a slightly simpler design available. We just want something that consistently translates information in nucleotides into something physical and potentially useful for the protocell.
For example, let’s say our primitive ribosome can read a guanine (G) nucleotide, and then somehow attract a specific amino acid to itself, which begins to form a chain. It can also read an adenine (A) nucleotide, attracting a different amino acid which can be added to the chain as well. Keeping it fairly simple, let’s say it completely ignores uracil (U), another base used by RNA, and it breaks the chain of amino acids when it encounters a cytosine (C) nucleotide.
In other words, this primitive ribosome uses a very simple genetic code. It can only process a couple of amino acids, while the human genetic code handles at least twenty. The machinery for this primitive ribosome may be crude, but the important thing is, it does something roughly similar to a real ribosome.
An engineer can design it and tell us the minimum number of parts required, and what those parts would need to do. Biologists can then tell us the precise RNA strands that would be needed for the whole thing to actually work in a protocell.
I suspect this is vastly more difficult to achieve in real life than it might sound on paper. Think about what we need our primitive ribosome to do. First it has to somehow connect to the master molecule. Then it has to move along the molecule one nucleotide at a time. It has to know which amino acid corresponds to which nucleotide base. It has to attract the relevant amino acid somehow, then add it to a growing chain of amino acids, and then break the chain when it encounters a cytosine (C) nucleotide. Think of the chemistry and engineering required to get our primitive ribosome to do any one of these things, let alone all of them together!
An intelligent engineer may be able to design this, but what we’re really interested in is whether nature could do it by itself. Evolutionary theorists might say that natural selection could allow it to be built in a gradual manner, but nature can only select from what is available at the time. If our protocells don’t yet have even the crudest ribosomes, then protein-based machinery can’t be built for natural selection to test out. All nature can do is wait around until a new function mutates in and then somehow comes out of the master molecule.
As our imaginary colony grows to include enormous numbers of protocells, useful RNA sequences might perhaps mutate in a few master molecules now and then. The critical question is: could they eventually form a primitive ribosome, given enough time? I would suggest the answer is no. There are three reasons why I would argue it’s actually impossible for them to do this.
The first reason is the set of dilemmas I posed earlier. Without coded information, there would be no need for a ribosome to evolve. If RNA sequences are already acting directly like primitive proteins, they wouldn’t need to invent anything like a ribosome. And with no ribosome, there would be no need for a genetic code. And with no genetic code there would be no coded information.
In other words, coded information, a ribosome and a genetic code all need to be in place for proteins to be made, and proteins are a critical part of cellular life as we know it.
In the primitive ribosome I described a few moments ago, I bypassed the need for codons altogether, and had one nucleotide such as guanine (G) represent one amino acid. This presumably makes the whole thing easier to evolve early on, but it creates an enormous problem later, because how do we switch from a system where a single nucleotide represents an amino acid, to one where three nucleotides do?
If we alter the genetic code, we substantially change the meaning of any coded information previously stored away. For example, let’s say we have a GGG sequence stored away in a master molecule. In our primitive ribosome, one G represents glycine, so the GGG sequence would code for three glycine amino acids.
But in the human genetic code, a GGG sequence represents one glycine amino acid. In other words, the different genetic codes would produce very different proteins. Altering the code would substantially change the meaning of any genes stored in the master molecule. It can’t evolve without essentially breaking the proteins that have already evolved. This is why Francis Crick called the genetic code a “frozen accident.” Nature must have hit upon a good code early on, since most organisms use roughly the same genetic code, with some fairly minor variations.
The second reason I don’t think RNA sequences could form into ribosomes by themselves is an even more formidable obstacle. In an animal or plant cell, ribosomes are assembled in a giant manufacturing center called the “nucleolus.” The pieces are assembled in the right order at the direction of control mechanisms and sequences that are written into the cell’s blueprint.
Our protocells don’t have a ribosome manufacturing center. Even if the necessary RNA pieces to make a ribosome somehow all evolved, were available in a protocell, and luckily assembled themselves into the equivalent of the first primitive ribosome (despite not having any reason to, since there was no information that needed decoding), what would hold all of the pieces together in this arrangement for its offspring? How would future generations know how to build the ribosome? There would be no blueprint, no “How To Make a Ribosome” manual.
Many theorists suggest that the primitive ribosome was self-replicating and self-assembling. It could somehow make a copy of and assemble itself. I suppose if we could get one or more of these into a protocell, then assuming the protocell was able to divide, and the ribosome was also able to self-replicate at roughly the same rate, new protocells would presumably also contain these ribosomes.
This would solve the problem of how our protocell puts together a ribosome. In short, it doesn’t. The ribosome puts itself together all by itself.
The idea of a self-replicating, self-assembling ribosome sounds great in theory, but I think it would be very difficult and perhaps even impossible to achieve in reality. It would be the equivalent of designing a complex machine comprised of many parts, that could replicate itself using only its own parts as a blueprint, that was then able to assemble those parts into the same form after each replication.
If theorists think this is easy, maybe they should first ask engineers to design a machine that not only acts like a primitive ribosome, but is also self-replicating and self-assembling, so we at least have proof of concept. Then biologists can tell us the exact RNA sequences that would be needed to make this machine at the molecular level. Personally, I suspect a self-replicating, self-assembling ribosome is actually impossible.
But wait a minute. Living cells are self-replicating, so why can’t this be true of primitive ribosomes? The critical difference is, cells have access to detailed blueprints encoded in the DNA molecule, and they have all of the machinery they need to make copies of themselves. The cell’s ability to self-replicate requires a lot of equipment and information! This is why the genomes of self-replicating organisms are typically millions or even billions of nucleotides in length.
In other words, a self-replicating, self-assembling ribosome would be an engineering miracle, which I suspect even human engineers may find impossible to design. But if they could actually achieve this, they would be showing us the exact level of engineering knowledge required, and the size of the blueprints needed, to create such a machine. They will have also created something that nature itself isn’t using, since ribosomes in real cells are built by machinery other than itself, and so they are neither self-replicating nor self-assembling.
But if the first ribosome was neither of these, and instead came about through a lucky arrangement of RNA sequences, this arrangement would break down in later generations, because the blueprint for putting it together couldn’t be written down. It would be like one of those good ideas you’ve had, that you forgot the next day because you didn’t write it down.
However, the third reason why RNA sequences couldn’t build a sustainable ribosome is perhaps the most formidable challenge of all. It’s to do with the process of mutation itself. Mutations are necessary for new RNA sequences and useful functions to evolve, but in the RNA World, mutations would also destroy those things.
Cells as we know them employ several error correcting strategies when dividing, to minimize copying errors and therefore bring the rate of mutation down substantially. But we can’t just give our protocells this ability, for the same reason we can’t just gift them with ribosomes. If we did, that would be intelligent design, the opposite of the idea that life arose by itself.
A useful RNA sequence might evolve, and give a protocell an immediate survival advantage, but without error correcting, this advantage will be mutated away in later generations. Natural selection might help, but if the mutation rate is too high, which is very likely without error correcting to reduce it, any small innovations will be swept away faster than natural selection can keep up.
This paradox is acknowledged by biologists, and is often called the “error catastrophe” problem. It is usually framed in terms of the length of a functional nucleotide sequence. Without error correcting, its length would be very limited, because in larger molecules, mutations would eventually destroy the information content for subsequent generations. But paradoxically, the maximum size available without error correcting would be too small to encode an error correcting process.
In other words, the error correcting code would take up more nucleotides than would be available to encode this information in a stable manner. Ideas have been suggested to help solve the dilemma, but they are theoretical, and other biologists have pointed out that the mechanisms proposed still suffer from loss of information.
Furthermore, in real cells, error correcting is performed by machinery made out of proteins. But these protein machines can’t begin to evolve without ribosomes and a genetic code to make the proteins. All of these emerging functions will be mutated away without error correcting to preserve them.
Many origin of life theorists recognize that error correcting is one of the first problems nature has to solve, because without it, protocells can’t hold on to any newly evolved advantages. For this reason, researchers hope to find a suite of RNA sequences that could perhaps perform error correction quickly and efficiently, without any need for complex proteins. I suppose they might exist, and come to be intelligently designed in a laboratory setting.
But whatever the case, without error correcting functions, emerging RNA sequences that could potentially be useful would be destroyed by mutations in later generations. This would make the evolution and long term sustainability of a complex machine such as a ribosome impossible.
Furthermore, until researchers have found this magical suite of RNA sequences that provide error correcting, we can’t know how likely or unlikely the whole suite is to emerge without somehow mutating away.
Let me sum up the three reasons why I think a ribosome would never emerge out of RNA sequences from the master molecule, no matter how many protocells we had or how much time we allowed. First, RNA sequences have no need for a ribosome or a genetic code, and they don’t have foresight, so they wouldn’t even begin to assemble a ribosome.
Second, even if protocells did assemble a primitive ribosome, they would have no way to pass on the blueprint to future generations. The idea that they might be self-replicating and self-assembling sounds great on paper, but would be an engineering miracle in real life.
Third, there would be no way to keep any of these things around for the long haul without error correcting functions. Mutations would rip them apart faster than natural selection can preserve them.
Now, if a ribosome can’t emerge on its own out of RNA sequences in the master molecule, could it have perhaps evolved out of the protocell’s replication process? In our thought experiment, we granted protocells the ability to replicate, mainly to get the show started. But now it’s time to give more serious consideration to how they actually do this.
If replication involves fairly simple chemistry that is somewhat independent from the rest of the protocell, then the process probably wouldn’t have much to offer in the evolution of a ribosome, which would need specific functions and machinery.
However, if the process of protocell replication is controlled somehow by the master molecule, this is a very different story. The control sequence would be complex, and take up at least several thousand nucleotides worth of information. But where did this information come from?
This is why I was only willing to grant protocells the ability to replicate if it was a simple chemical process not controlled by the master molecule. If the process is governed by machinery and some kind of blueprint, then their existence would need explaining. If we just granted our protocells these things, this would clearly be intelligent design. But without them, how would the protocell replicate in the first place, to allow even a crude form of evolution to get started?
I suppose we might argue that the replication process could be entirely controlled by RNA sequences and machines. However, this presents us with two critical dilemmas. First, how would the RNA sequences know how, where and when to perform their functions? Would they be working to some kind of a blueprint? If so, how is the blueprint read and interpreted? Or do they all just happen to perform the same cascade of activity every time?
Again, this would need to be tested scientifically. Biologists and engineers could team up to tell us the precise RNA sequences that could perform this magical cascade which enables them to replicate themselves and build a protocell every time.
Personally, I think designing such a protocell is an incredibly tall order, and may even be impossible without endowing it with lots of machinery and blueprints, which would be evidence that intelligent design was required. After all, nature itself tells us that a self-replicating cell requires a genome, ribosomes and a genetic code. There is nothing in nature that equates to a protocell. Even a virus, which is perhaps closer to a protocell in some respects, isn’t truly self-replicating, because it needs to hijack the equipment from a cell.
Either way, this is still nothing compared with the second dilemma for a replication process controlled by RNA sequences: what happened before the necessary RNA sequences evolved? How did the protocell replicate before this?
Evolutionary theorists argue that new functions usually evolve out of previously existing functions, but there aren’t many functions around in RNA World, and besides, there can’t be an infinite regress of simpler replication processes. That would be just as absurd as an infinite regress of creator gods. But without a self-replicating protocell, there can be no evolution. Indeed, the only way to evolve the RNA sequences needed to make a replicating protocell is to have lots of replicating protocells, so that RNA sequences can evolve. Clearly this is a problem!
What I have shown so far in this chapter is that naturalistic theories about the origin of life face several enormous dilemmas, and have to make several assumptions. If we wish to believe life arose on its own, we first need to assume that just the right environment existed, and that the raw materials for life were available. It’s one thing to create a chemical reaction in a laboratory under highly controlled conditions, but quite another to achieve it in a natural environment.
We need to assume that some kind of protocell existed, the forerunner of a cell, and that it had some information storage capacity, such as an RNA or DNA “master molecule,” having the potential to contain genetic information.
We have to assume the protocell was somehow able to replicate. But the process couldn’t be controlled by instructions or coded information in the master molecule, because that would require the existence of machinery to read and interpret the information.
We can assume that in the process of replication, mutations to the master molecule happened, allowing for variation and therefore the possibility of natural selection. We also need to assume that some parts of the master molecule emerged to act somewhat independently, and become little machines, but without any complex machinery like a ribosome to manufacture them.
These assumptions are necessary just to get things started. After that, we arrive at the real challenges, although each of the prior assumptions are huge challenges in themselves.
Without some kind of error correcting process while the cell is replicating, mutations will make the protocells unstable. Useful RNA sequences might emerge, assuming the master molecule is made of RNA, but the high mutation rate means they would be destroyed in later generations. Error correcting functions can’t be arrived at either, because they will also be mutated away beforehand.
An origin of life theorist can make up a story about how “a simple error correcting process evolved and was able to confer an advantage on the early protocell,” but this is devoid of any actual details. Engineers are better suited to tell us what such a “simple” process would actually involve, and what functions and equipment would be needed to do it. Then biologists can tell us which collection of RNA sequences could do the job, and demonstrate it in the real world.
Incidentally, in most cells as we know them, the separation between DNA and RNA provides the cell with an additional layer of stability. True, DNA can actually be altered by the cell, so it isn’t just a read-only system as biologists initially thought. But deliberate changes made by the cell are highly regulated. Thus, the DNA molecule provides the cell with a stable blueprint for the equipment it needs, while disposable mRNA transcripts of DNA sequences are used to make proteins.
But protocells can’t be granted this luxury. They aren’t anywhere near as sophisticated. The entire system is unstable. If RNA machines evolve directly from the master molecule, then when they come out of it to perform their functions, they may end up breaking the molecule. If they somehow cut themselves out of the master molecule before the replication process begins, the duplicated master molecule won’t have them in! But if they are separate from it, or self-replicating, then what mechanism enables them to later create a blueprint for their own construction that can be written into the master molecule?
Another enormous challenge for our protocells is to evolve something like a ribosome. Without these, proteins can’t be made and our protocells wouldn’t even evolve into anything resembling bacteria, let alone creatures like birds and butterflies.
But how could a ribosome come about? Without a ribosome and a genetic code, there can’t be coded information. But without coded information and a genetic code, there is no need for a ribosome to evolve.
Even if a ribosome fortuitously emerged in a lucky protocell, how would future protocells know how to build it? Cells rely on blueprints to build machinery, but our protocell would have no way to write up the blueprint for a ribosome to pass on to its offspring.
If the ribosome was self-replicating, it would be a miraculous invention. The machine would need to move along a sequence of nucleotides one at a time, attract a specific amino acid corresponding to the nucleotide, build a chain of amino acids, and break the chain when encountering another specific nucleotide. Furthermore, it would need to be able to build an exact duplicate of itself, without referring to a blueprint.
Origin of life researchers have made self-replicating RNA strands, but this is still a world away from designing a self-replicating protocell with a primitive ribosome that hasn’t been extracted from actual cells, which would be cheating. If and when they finally achieve this, it will demonstrate the sheer amount of chemical, biological, engineering and coding knowledge needed to design even a “simple” self-replicating protocell.
But it’s much easier just to assume the protocell was able to replicate, store information, read that information and translate it into something useful, without giving the details of how it acquired all of these abilities, other than glib stories. All of these things are assumed to have happened because, after all, life exists.
Of course life exists. The critical question for all of us is whether it arose by itself, or whether some kind of intelligence designed it. The dilemmas faced in the origin of life would be much more easily solved if there was a chemist, physicist, engineer and computer programmer behind them. It’s easy to say that life came about on its own, but far from easy to demonstrate how the complex machinery for even the simplest cell came together.
I have also bypassed other critical features of the protocell, which I will briefly mention here. For example, how does it get access to the raw materials needed to build the things we have been talking about?
Humans have it easy. We only have to eat food, drink water and breath air, and that is the job of survival sorted out, at least in a basic sense. Our cells, organs, and the bacteria that live in our body all do the hard work of turning those inputs into useful energy and components. What they do every moment of every day makes humans look like lazy sloths by comparison.
Obviously a protocell couldn’t be anywhere near as complex as our own cells, but it would still need processes to access raw materials, and turn them into energy or building blocks. Little RNA machines may be able to do some pretty interesting things, but could enough of them evolve quickly enough to create a system of metabolism, before being mutated away? And how do they even begin to do any of this, without the ability to convert raw materials into energy in the first place?
Furthermore, many chemical processes in living cells require a catalyst, something that dramatically increases the rate of a chemical reaction. These are often in the form of proteins called “enzymes” (from the Greek word meaning “leavened”), but some RNA sequences can also act as catalysts, and this class of sequences are given a special name, called “ribozymes.”
Without catalysts, many chemical processes would be painfully slow, and in many cases the cell would die long before the process was completed. Many of these processes require more than one ribozyme to act as the catalyst, so what would happen to the process before the required catalysts evolved? It would be mutated away, since only the addition of catalysts would make it useful.
The problem of explaining the origin of life was summed up by one synthetic biologist as being about finding the solution to four paradoxes.7 The first is the “tar” paradox, the tendency of organic matter to devolve into tar, or asphalt. The second is the “water” paradox, that every interesting chemical bond is unstable with respect to water. The third is the “entropy” paradox, that nature likes to break things up, but any theory on the origin of life has to assemble larger and larger building blocks that fight entropy. The fourth is the “destruction” problem, that RNA sequences acting as catalysts are more likely to be active in a destructive sense, rather than a creative one.
Most or even all of the dilemmas and paradoxes faced by the RNA World hypothesis are also faced by the alternative ideas put forward, with the exception of one – that is, the idea that the cell was actually designed.
In this chapter, I haven’t disproved the idea that life arose by itself, and neither have I proved that it came about through design. What I hope to have demonstrated is that the many dilemmas only exist when we insist on looking at life from the assumption that it arose by itself.
Origin of life scientists often like to give the impression that they are on the verge of solving life’s deepest mysteries. This certainly helps with publicity and funding. In reality, the reason there are so many hypotheses about life’s origins is precisely because of the dilemmas I’ve highlighted. Even the “RNA World” hypothesis has been described by one biochemist, slightly tongue-in-cheek, as “the worst theory of the early evolution of life (except for all the others).” 8
Since no human was around to witness the origin of life, we shouldn’t be surprised that different ideas exist. But then, why shouldn’t design also be a valid idea? The fact is, many complex processes are required in order to put together even the simplest living cell: replication, metabolism, error correcting, a genetic code, the ability to store, read and interpret information using that code, encoded proteins, blueprints for machinery, and so on.
If humans were to try and put together such things from scratch, it would require a high degree of physics, chemistry, engineering and coding knowledge; and even our smartest and best scientists would struggle to emulate these things.
Yet even before they were able to break open the cell and explore its contents, many biologists already assumed they knew how it came about. It must have arisen by itself. Wasn’t that a little presumptuous, more akin to a religious belief or ideology, rather than actual science? But we already know part of the reason why. At least when working for a scientific establishment, biologists have to accept the Naturalistic Assumption, which excludes a designer by default.
The English clergyman William Paley argued that if we came across a watch on a beach, we would assume it had not arisen by chance, because we would notice how complex it was, and how its parts worked together purposefully. He used this as an argument for the existence of God. He reasoned that, since the world exhibits greater complexity than the watch, life also cannot have arisen by chance.
Then naturalists like Charles Darwin came along and argued that complexity can be explained by natural selection and large amounts of variation, which later biologists attributed to mutations in the DNA molecule of organisms. They argued that what we see is merely the illusion of design.
But in the origin of life after it supposedly emerged from non-life, mutations and natural selection aren’t adequate explanations. Mutations would break things before anything beneficial could gain a foothold, and nature can’t select from things that don’t exist. New functions can’t be preserved unless they are stored in blueprints, but the blueprints can’t survive unless the emerging cell can read and do something useful with the information. Evolution can’t even get started, without multiple complex mechanisms already being in place.
On the other hand, intelligence combined with a high degree of physics, chemistry, biology, engineering and coding knowledge could conceive, design and create a complete working cell, with all of its metabolizing, replicating, coding, decoding and manufacturing features. Indeed, this is partly what origin of life scientists, using their knowledge and intelligence, are trying to do in their labs, even though they often use parts extracted from real cells.
In other words, at least in the origin of life, perhaps the best explanation is not that design is an illusion and life merely looks designed, but that life actually was designed after all.
1 4 to the power of 10, which mathematicians write as 4^10 or 410, is 1,048,576. 2 411 = 4,194,304. 412 = 16,777,216. 3 There are an estimated 1080 atoms in the universe, and about 1090 possible permutations of 150 letters in an RNA sequence. I use the mathematical term “permutations” here because order matters in an RNA sequence. 4 4x4x4 or 43 = 64. 5 Freeland, Hurst, “The Genetic Code Is One in a Million”, Journal of Molecular Evolution, 1998. 6 As an example, see “Biologists create the most lifelike artificial cells yet”, Sciencemag.org, November 19, 2018. 7 These paradoxes are discussed by Steve Benner in an interview with Suzan Mazur, the author of “The Origin Of Life Circus: A How To Make Life Extravaganza”. Page 151. 8 Bernhardt, “The RNA world hypothesis: the worst theory of the early evolution of life (except for all the others)”, Biology Direct, 2012.