How much mutation is too much?

How much mutation is too much?

Newsflash - Human mutation rates are much too high.
By Dr. John Sanford.

For many decades geneticists have been worried about the impact of mutation on the human population (Muller 1950, Crow, 1997). When these concerns first arose, they were based upon an estimated rate of deleterious mutation of 0.12 to 0.30 mutations per person per generation (Morton, Crow and Muller, 1956). Since that time there have persisted serious concerns about accumulating mutations in man leading to a high "genetic load" - and a generally degenerating population. There has also been a long-standing belief that if the rate of deleterious mutations approached one deleterious mutation per person per generation, long-term genetic deterioration would be a certainty (Muller, 1950). This would be logical, since selection must eliminate mutations as fast as they are occurring. We need to prevent mutant individuals from reproducing, but we also need to leave enough remaining people to procreate and produce the next generation. By this thinking, deleterious mutations in man must actually be kept below one mutation for every three children - if selection is to eliminate all the mutations and still allow the population to reproduce. This is because global fertility rates are now less than 3 children for every 2 adults - so only one child in three could theoretically be selectively eliminated. For these reasons, geneticists have been naturally very eager to discover what the human mutation rate really is!

One of the most astounding recent findings in the world of genetics is that the human mutation rate (just within our reproductive cells) is at least 100 nucleotide substitutions (misspellings) per person per generation (Kondrashov, 2002). Other geneticists would place this number at 175 (Nachman and Crowell, 2000). These high numbers are now widely accepted within the genetics community. Furthermore, Dr. Kondrashov, the author of the most definitive publication, has indicated to me that 100 was only his lower estimate - he believes the actual rate of point mutations (misspellings) per person may be as high as 300 (personal communication). Even the lower estimate, 100, is an amazing number, with profound implications. When an earlier study revealed that the human mutation rate might be as high as 30, the highly distinguished author of that study, concluded that such a number would have profound implications for evolutionary theory (Neel et al., 1986). But the actual number is now known to be 100-300! Even if we were to accept the lowest estimate (100 mutations), and further assumed that 97 % of the genome is perfectly neutral junk, this would still mean that at least 3 additional deleterious mutations are occurring per person per generation. So every one of us is a mutant, many times over! What type of selection scheme could possibly stop this type of loss of information? As we will see - given these numbers, there is no realistic method to halt genomic degeneration. Since the portion of the genome that is recognized as being truly functional is rapidly increasing, the number of mutations recognized as being actually deleterious is also rapidly increasing. If all the genome proves functional, then every one of these 100 mutations per person is actually deleterious. Yet even this number is too small, firstly because it is only the lowest estimate, and secondly because it only considers point mutations (misspellings). Not included within this number are the many other types of common mutations - such as deletions, insertions, duplications, translocations, inversions, and all mitochondrial mutations.

To appreciate to what extent we are still underestimating the mutation problem, we should first consider the types of mutation which fall outside the scope of the normal 'point mutation' counts. Then we need to consider what portion of the whole genome is really functional, and not "junk".

Within each cell are sub-structures called mitochondria, which have their own small internal genome (about 16,500 nucleotides), which is inherited only through the mother. However, because the mitochondrial genome is highly polyploid (hundreds of copies per cell), and because the mitochondrial mutation rate is extremely high, there are still a large number of mitochondrial mutations that must be eliminated each generation - to halt degeneration. The human mitochondrial mutation rate has been estimated to be about 2.5 mutations, per nucleotide site, per million years (Parsons et al, 1997). Assuming a generation time of 25 years and a mitochondrial genome size of 16,500 - this approaches one mitochondrial mutation per person per generation within the reproductive cell line. Mitochondrial mutations, just by themselves, probably put us over the theoretical limit of one mutation per three children! Even if the mutation rate is only 0.1 per person, we would have to select away a very substantial portion (10%) of the human population, every generation, just trying to halt mitochondrial genetic degeneration. Yet this would still leave the 100-300 nuclear mutations per person per generation (as discussed above) accumulating - unabated. High rates of mitochondrial mutation are especially problematic in terms of selection (Chapters 4 and 5), because of lack of recombination ("Muller's ratchet" - Muller, 36 The Mystery of the Genome 1964), and lower effective population size (only women pass on this DNA, so selection can only be applied to half the population).

The most rapidly mutating regions of the human genome are within the very dynamic micro-satellite DNA regions. These unique regions mutate at rates nearly 1 million-fold above normal, and are not included in normal estimates of mutation rate. Yet these sequences are found to have biological impact, and their mutation results in many serious genetic diseases (Sutherland and Richards, 1995). It is estimated that for every "regular" point mutation, there is probably at least one micro-satellite mutation (Ellegren, 2000). This effectively doubles the mutation count per person per generation, from 100-300 to 200-600.

In addition to nuclear point mutations, mitochondrial mutations, and micro-satellite mutations, there are a wide variety of more severe chromosomal mutations - called macro-mutations. These include deletions and insertions. According to Kondrashov (2002), such mutations, when combined, add another 4 macro-mutations for every 100 point mutations (this estimate appears to consider only the smallest of macro-mutations, and excludes the insertions/ deletions affecting larger regions of DNA). Although there may be relatively few such mutations (only 4-12 per person per generation), these "major" mutations will unquestionably cause much more genomic damage, and so would demand higher priority if one were designing a selection scheme to stop genomic degeneration. Macro-mutations can affect any number of nucleotides - from one to a million - even as we might accidentally delete a letter, a word, or even an entire chapter from this book. These relatively few macro-mutations are believed to cause 3 to 10-fold more sequence divergence than all the point mutations combined (Britton, 2002; Anzai, 2003). This brings our actual mutation count per person per generation up to about 204 - 636. But if we factor in the fact that macro-mutations can change 3-10-fold more nucleotides than all point mutations combined, our final tally of nucleotide changes per person could come up to as high as 612-6,360 per person per generation! These numbers are mind-boggling! Yet even these numbers may still be too low - we have not yet considered inversions and translocations. Furthermore, evolutionary theorists are now invoking extremely high inter-genic conversion rates, which could double these numbers again. Wow! Do you recall the beginning of this chapter, where we learned that the famous geneticist Muller considered that a human mutation rate of 0.5 per person or higher, would doom man to rapid genetic degeneration? Although we do not know the precise human mutation rate, there is good reason to believe that there are more than 1,000 nucleotide changes in every person, every generation (see Table 1). To be exceedingly generous, for the rest of this book I will use the most conservative number being referred to in the literature today - 'just' 100 mutations per person per generation (except where otherwise specified). However, please note that this is only a fraction of the true number, and this number excludes the most destructive classes of mutations.

Of all these mutations - what percent are truly neutral? In the last few years there has been a dramatic shift in the perceived functionality of most components of the genome. The concept of "junk DNA" is quickly disappearing. In fact, it is the "junk DNA" (non-protein-coding DNA), which appears to be key to encoding biological complexity (Taft and Mattick, 2003). The recent Taft and Mattick study strongly suggest that the more "junk" - the more advanced is the organism. So mutations within "junk DNA" can hardly be assumed to be neutral!

Approximately 50% of the human genome is now known to be transcribed into RNA (Johnson et al., 2005). At least half of all this transcribing DNA appears to be transcribed in both directions (Yelin et al., 2003)! So all of this DNA is not only functional - but much of it may be doubly functional. While only a small fraction of the genome directly encodes for proteins, every protein-encoding sequence is embedded within other functional sequences that regulate the expression of such proteins. This includes promoters, enhancers, introns, leader sequences, trailing sequences, and sequences affecting regional folding and DNA architecture. I do not believe any serious biologist now considers introns (which comprise most of a typical genic region) as truly neutral "junk". In fact, many of the most strongly conserved (essential and invariant) sequences known, are found within introns (Bejerano et al., 2004). While a typical protein-coding sequence may only be 3,000 nucleotides long or less, the typical "whole gene" that controls the expression of that protein can be in the range of 50,000 nucleotides long. Since there are 20,000 - 40,000 proteinencoding genes (estimates greatly vary), if we include all their associated nucleotides (50,000 per gene), the true complete genes could easily account for over 1.5 billion nucleotides. This is fully half the genome. In addition, a whole new class of genes has been discovered which do not encode proteins, but encode functional RNAs. Such genes have escaped recognition in computer searches for protein-coding sequences, and so have been overlooked as true genes. But they are true genes, and they probably comprise a large part of the genome (Mattick, 2001; Dennis, 2002; Storz, 2002). They are just now being discovered - within DNA regions that were previously dismissed as "junk". In addition, two independent studies have shown extensive sequence functionality within the large regions between genes (Koop and Hood, 1994; Shabalina et al., 2001). Previously, such regions had also been assumed to be junk. Pseudogenes, long considered dead duplicated genes, have recently been shown to be functional (Hirotsune et al., 2003; Lee, 2003). Pseudogenes seem to be designed to make regulatory RNA molecules (see Chen et al. 2004), rather than proteins, so they are not "dead fossils". As I will discuss in more detail elsewhere, there even appear to be diverse cellular functions for the much-maligned "selfish genes", sometimes called "parasitic DNA sequences", also called "transposable elements". These elements appear to have multiple, and extremely important functions within the cell, including the control of chromosome pairing (Hakimi et al. 2004), and DNA repair (Morrish, et al., 2002). Repetitive DNA , including satellite DNA, long considered junk, has been shown to be essential to genome function, and comprise such essential genomic structures as centromeres and telomeres (Shapiro and Sternberg, 2005). Lastly, there are fundamental genome-wide structural patterns, which virtually permeate every portion of the genome - such as isochores (GC rich areas -Vinogradov, 2003), genomewide 'word' patterns (Karlin, 1998) and nucleosome binding sites (Tachida, 1990). These genome-wide patterns appear crucial to cell function, and suggest functionality throughout the entire genome. For example, nucleosome binding (crucial to chromosome structure and gene regulation) appears to be specified by di-nucleotide patterns that repeat every 10 nucleotides (Sandman et al., 2000). This means that one-fifth of the genome appears functional and essential - just for the purpose of specifying nucleosome binding sites! It is becoming increasingly clear that most, or all, of the genome is functional. Therefore, most, or all, mutations in the genome must be deleterious.

On a per person basis, 100 mutations represent a loss of only a miniscule fraction of the total information in our genome (the 40 The Mystery of the Genome genome is huge). However, the real impact of such a high mutation rate will be at the population level, and is primarily expressed with the passage of time. Since there are six billion people in the world, and each person has added an average of 100 new mutations to the global population, our generation alone has added roughly 600 billion new mutations to the human race. If we remember that there are only three billion nucleotide positions in the human genome, we see that in our lifetime there have been about 200 mutations for every nucleotide position within the genome. Therefore, every possible point mutation that could happen to the human genome has happened many times over - just during our lifetime! Because of our present large population size, humanity is now being flooded by mutations like never before in history. The consequences of most of these mutations are not felt immediately, but will become manifested in coming generations.

As we will be seeing, there is no selection scheme that can reverse the damage that has been done during our own generation - even if further mutations could be stopped. No amount of selection can prevent a significant number of these mutations from drifting deeper into the population and consequently causing permanent genetic damage to the population. Yet our children's generation will add even more new mutations - followed by the next and the next. This degenerative process will continue into the foreseeable future. We are on a downward slide

that cannot be stopped.

When selection is unable to counter the loss of information due to mutations, a situation arises called "error catastrophe". If not rapidly corrected, this situation leads to the eventual death of the species - extinction. In its final stages, genomic degeneration leads to declining fertility, which curtails further selection (selection always requires a surplus population - some of which can then be eliminated each generation). Inbreeding and genetic drift must then take over entirely - rapidly finishing off the genome. When this point is reached, the process becomes an irreversible downward spiral. This advanced stage of genomic degeneration is called "mutational meltdown" (Bernardes, 1996). Mutational meltdown is recognized as an immediate threat to all of today's endangered species. The same process appears to potentially be a theoretical threat for mankind. What can stop this process?

About the Author

Dr. John Sanford has been a Cornell University Professor for more than 25 years (being semi-retired since 1998). He received his Ph.D. from the University of Wisconsin in the area of plant breeding and plant genetics. While a professor at Cornell he trained graduate students and conducted genetic research at the New York State Agricultural Experiment Station in Geneva, NY. During this time John bred new crop varieties using conventional breeding, and became heavily involved in the newly emerging field of plant genetic engineering. While at Cornell, John published over 70 scientific publications, and was granted over 25 patents. His most significant scientific contributions involved three inventions - the biolistic (“gene gun”) process, pathogen-derived resistance, and genetic immunization. Most of the transgenic crops grown in the world today were genetically engineered using the gene gun technology developed by John and his collaborators. John also started two successful biotech businesses deriving from his research - Biolistics, Inc. and Sanford Scientific, Inc. John still holds a position at Cornell (Courtesy Professor), but has largely retired from Cornell and has started a small non-profit organization - Feed My Sheep Faundation.

NOTES

a Mitochondrial mutation rate estimates vary, but can approach 0.5 per person (Parsons et al., 1997).

b Nuclear substitutions are hard to measure, but Kondroshov (2002) has estimated 100 per person. In personal communication he has indicated this may actually be 300.

c Normal estimates of nucleotide substitutions would not include mutational hotspots such as microsatellites. Microsattelite mutation rates have been estimated to be roughly equal to all other point mutation rates.

d,e Kondrashov (2002) estimated that deletions plus insertions occur at a combined rates of about 4-12% of the point mutations - or about 2-6% each. However, he seemed to limit his estimate to only small inserts and deletions, so the actual number may be higher. Because mutations and insertions can be very large, their total effect is believed to be 3-10 fold greater than all point mutations, in terms of total nucleotides changed.

f The actual rate of chromosomal rearrangements is unknown. Evolutionary assumptions about the recent divergence of chimp and man require high rates of such changes. These changes can affect very large pieces of DNA, and so for the evolutionary scenario to work, many thousands of nucleotides on average, must move in this way every generation.

g The actual rate of inter-genic conversion is unknown, but evolutionary assumptions require extremely high rates of gene conversion between different loci - many thousands per person per generation.

h The total number of mutations can only be estimated in a very crude way, but it should be very clear that the number of all types of new mutations, including conversions, must be over 1,000 per person. These mutations, which include many macro-mutations, must clearly change many thousands of nucleotides per person per generation.

Go back to home page

If you have any comments or questions
contact: bhaktivedanta_108@yahoo.com