Non-Random Directed Mutations Confirmed

The neo-Darwinian theory of evolution by natural selection of random mutations should be consigned to history where it belongs; electromagnetic intercommunication and resonance may be involved in activating and mutating just the right genes Dr. Mae-Wan Ho

An obsolete theory challenged by directed mutations

Conventional neo-Darwinian theory of evolution is firmly based on the natural selection of random mutations plus the ‘central dogma’ assumption that environmental influences cannot change nucleic acids or become inherited. The central dogma has been invalidated at least since the early 1980s concomitantly with the emergence of the new genetics of the fluid genome [1, 2] (Living with the Fluid Genome, ISIS publication). Similarly, the randomness of mutations has been called into question since the 1970s in experiments demonstrating that cells subject to non-lethal selection come up repeatedly with just the  right ‘adaptive’ or ‘directed’ mutations in specific genes that enable the cells to grow and multiply [3] (see [4] To Mutate or Not to Mutate, SiS 24).

Note that mutations adaptive for cells in enabling them to grow and multiply are not so for the organism as a whole, as they give rise to tumours and cancers; hence it is more appropriate to refer to them as ‘directed’ mutations. In fact, it was the idea that cancers may involve directed mutations that prompted John Cairns, then at Harvard School of Public Health in Boston Massachusetts, to study the phenomenon in bacteria [5].

In one of the first experiments demonstrating directed mutations, a strain of E. coli bacterium - with a mutation in the lacZ gene that terminates the polypeptide prematurely and hence unable to use lactose - only mutated back to the wild-type when lactose is present in the medium, not when lactose was absent [3]. Even more remarkably, another strain with the lacZ completely deleted was able to mutate a cryptic gene ebgA coding for an enzyme that could hydrolyse lactose and the regulator ebgR that normally represses ebgA. During growth each of these point mutations occurs at a frequency of less than 10^-8, neither on its own would allow the lacZ-deleted strain to use lactose [6].

Cairns and colleagues concluded their 1988 review on the origin of mutants as follows [3]: “Curiously, when we come to consider what mechanism might be the basis for the forms of mutation described in this paper, we find that molecular biology has, in the interim, deserted the reductionists. Now, almost anything seems possible In certain systems, information freely flows back from RNA to DNA, genomic instability can be switched on under conditions of stress, and switched off when the stress is over, and instances exist where cells are able to generate extreme variability in localized regions of their genome. The only major category of informational transfer that has not been described is between proteins and the messenger RNA (mRNA) molecules that made them. If a cell discovered how to make that connection, it might be able to exercise some choice over which mutations to accept and which to reject.”

As it happens, reverse translation as suggested by Cairns is not needed, and the same mechanisms for directed mutations could apply from E. coli bacteria to humans, as recent findings indicate.

The unity of biochemical reactions

Mutagenesis mechanisms in vivo are essentially the same in all living cells, which is yet another instance of the unity of biochemistry.

The unity of biochemistry has been a guiding principle in the study of chemical processes in living organisms; as for example, the core energy metabolism in living organisms is reduction and oxidation of carbohydrates and water [7] (Living Rainbow H₂0, ISIS publication). But studying biochemical reactions in the test-tube can be very misleading as the biochemical reactions in vivo typically happen within an environment crowded with enzymes where substrates are limiting. To adequately represent biochemical reactions in vivo, Henrik Kacser (1918-1995) and colleagues pioneered metabolic control analysis [8] (see also [9], a kinetic framework for analysing rates and fluxes due to changes in enzymes and metabolites at steady state. 

Barbara Wright and colleagues at the University of Montana, Missouri, in the United States are among the few biologists who see the need for such kinetic models as biochemical knowledge grows and the relevance of in vitro data to metabolism in vivo becomes increasingly questionable [10]. The usual hierarchical schemes of control based on any single cellular event of “gene activation” or “enzyme-induction” is simply not adequate to account for the biological effects that necessarily involve complex relationships among numerous molecular entities and biochemical pathways. However, models are only valid if they have predictive value. Using their computer programme mfg, the team has indeed created kinetic models that successfully predict directed mutations in three very different systems: E. coli under phosphate starvation, oxidative stress and osmotic stress; and in humans, the p53 tumour suppressor gene activated by genotoxic stress, and somatic hypermutation during the immune response to foreign antigens, where transcription frequency increases ten thousand-fold and mutation frequency a million fold.

They start from the observation that metabolic reactions unique to a particular environmental stress apparently target specific genes for increased rates of transcription and mutation, “resulting in higher mutation rates for those genes most likely to solve the problem.” Stressors typically lead to activating ~ 1% of the genome to transcription and mutation, thereby directing and selecting those mutations that correct the problem (overcome the stress). However, they are silent on how this might occur, I shall return to this important point later.

For example, in starving (stressed) E. coli bacteria presented with lactose, the obvious limiting enzyme is b-galactosidase, encoded by the lacZ gene, which can split lactose into glucose and galactose, simple sugars that can be metabolized by downstream enzymes to provide energy and material for growth. Hence bacteria with defective lacZ gene is directed to mutate the gene until the normal functioning enzyme is restored; or in the case of bacteria with lacZ completely deleted, a new cryptic gene and its regulator are mutated until the functioning new enzyme that can break down lactose is created and become expressed.

Similarly, in the systems analysed by Wright and colleagues [10], E. coli starved for inorganic phosphate directs mutations to de-repress the pho regulon (phosphate regulated suite of genes) resulting in a new high-affinity phosphate transport system that gets phosphate into the cell at much lower concentrations, and also activates a hydrolytic enzyme able to get phosphate from new sources. In humans, genotoxic stress activates transcription of the p53 gene, resulting in mutations that inactivate the gene. And foreign antigen stress activates the transcription and mutation of hypervariable regions of immunoglubins in B cells to produce the antibodies that can bind to and neutralize the foreign antigens.

Predicted mutagenesis hotspots based on secondary stem loop structures of ssDNA

In all cases, transcription provides the single stranded DNA (ssDNA being transcribed) that exposes unpaired, intrinsically mutable bases to mutation; and the mutable bases are guanines (Gs) and cytosines (Cs). Why they are mutable depends on the secondary stem loop structures (SLSs) the transcribed ssDNA adopts: sequences complementary to each other pair up to form stems leaving the unpaired bases as loops, and it is the unpaired Gs and Cs in loops that are vulnerable to mutations.

To simulate in vivo conditions in response to increased rates of transcription, the computer algorithm mfg was developed. During transcription, the vulnerability of a base to mutation depends on the stability of the secondary SLSs adopted by the ssDNA, and on the extent to which the base is unpaired. The mfg programme interfaces with the mfold programme that folds single-stranded segments of a specified length and sequence and report all possible secondary structures that can form from each folded segment in descending order of stability. Mfg reports the stability of the most stable secondary structure in which a mutable base is unpaired, and also the percent of total folds in which it is unpaired. The Mutability Index of each unpaired base is the product of the two variables.

The mfg-identified highly mutable Gs and Cs in p53 for example, are actually located in ssDNA loops of predicted SLSs, as confirmed by analysis for codon 175 in exon 5 using S1 endonuclease, which cuts at single stranded DNA and RNA. Similar analysis demonstrated that the hypermutable codons 245, 248, 273 and 282 of p53 are also located in single stranded loops.

The unpaired Gs and Cs are intrinsically mutable; unpaired Gs primarily mutate to As (adenines) and unpaired Cs mutate to Ts (thymines). Compelling evidence for the underlying instability of unpaired Gs and Cs is shown by examples of silent mutations that do not change amino acid sequence, and hence not subject to any selection. In all cases examined, the mutable bases are located in loops of identified secondary structures.

The somatic hypermutation (SHM) in pre-B cell involves a 10 000-fold increase in transcription, which is linked to a million-fold increase in mutation frequency, especially during phase 1 of the SHM. Activation induced deaminase (AID), an RNA-editing enzyme, is implicated in the mutational mechanism in SHM as well as in all the other systems. In liver cancers induced by genotoxins and involving p53, the availability of unpaired intrinsically mutable Gs in ssDNA is rate-limiting for mutation frequency. Circumstances in vivo at low endogenous levels of transcription show that the majority of intrinsic G mutations are to A, and that the availability of Gs in ssDNA is rate-limiting for mutation frequency. The dual effects of oxyradicals (reactive oxygen species arising from incomplete oxidation (see [12] The Body Does Burn Water, SiS 43) which both activate transcription (about 4-fold) and increase G to T mutations (to 85.8 %) is accompanied by corresponding decreases in G to A mutations. Thus oxyradicals compete for the fate of rate-limiting directed mutations.

Other evidence of non-random mutations

In humans, SHM and class switch recombination (CSR, a genetic rearrangement) result in distinct genetic alterations at different regions of the immunoglobulin genes in B lymphocytes in generating antibody diversity: point mutations in variable regions and large deletions in S (switch) regions respectively [12]; yet both depend on AID. B cell stimulation that induces CSR but not SHM leads to AID-dependent accumulation of SHM-like point mutations in the switch mu region independently of CSR. These findings strongly suggest that AID itself or some single molecule generated by RNA editing function of AID may mediate a common step of SHM and CSR, which is likely to be involved in DNA cleavage.

Another study suggests that DNA double-strand breaks (DSBs) are responsible for mutation hotspots in stress-induced mutation in E. coli by means of two mechanisms [13]. The first involve mutations occurring maximally within the first 2 kb and decrease logarithmically to ~60 kb. The second involves a weak mutation tail extending to 1 Mb from the double-strand break. Hotspots occur independently upstream and downstream in the replication path. The enzyme Rec D which allows DSB-exonuclease activity is required for strong local but not long distance hotspot mutations, indicating that double-strand resection (cutback) and gap-filling synthesis underlie local hotspot mutations. Hotspots near DSBs open the possibility that specific genomic regions could be targeted for mutagenesis, and could also promote concerted (simultaneous) evolution within genes/gene clusters.

Finally, a study combining phylogenetic and population genetic techniques to compare 34 E. coli genomes carried out by researchers at the European Bioinformatics Institute, Welcome Trust Genome Campus, Cambridge UK found that the rate of neutral mutations –neither advantageous nor deleterious - varies by more than an order of magnitude across 2 659 genes, with mutational hot and cold spots spanning several kilobases (entire operons) [14]. The variation is not random; a lower rate in highly expressed genes and in genes undergoing stronger “purifying selection” (which implies they are preferential protected from mutation or by repair mechanisms). According to the researchers, the findings suggest that mutation rate has been “evolutionarily optimized to reduce the risk of deleterious mutations.”

However, current knowledge of factors influencing the mutation rate – including transcription-coupled repair and context-dependent mutagenesis – do not explain these observations, indicating additional mechanisms must be involved.

More than 12 000 single-nucleotide polymorphisms have been examined. Given that transcription is mutagenic (as found in the other studies described earlier), the negative association between expression and mutation rate is unexpected, indicating that there are indeed repair pathways coupled to transcription. But this mechanism alone is insufficient, as the non-transcribed strand also displays dependence between expression and mutation rate, and molecular experiments have reported that transcription-induced mutagenesis occurs in the presence of transcription-coupled repair. Therefore additional mechanisms that generally target highly expressed genes but are not directly coupled with the transcriptional machinery, must exist.

How does the cell know which genes to mutate?

The findings indicate that there are numerous mechanisms for the cell to direct mutations to specific genes and specific sites in these genes, but they give no indication as to how the cell is able to do that. I suggest that electromagnetic signals are involved. There are good reasons to suspect that molecules intercommunicate by electromagnetic signals, and molecules that interact share common frequencies so they can attract one another through resonance (see [15] The Real Bioinformatics Revolution, SiS 33). If that is the case, lactose supplied during starvation for example, will send strong electromagnetic signals to its normal metabolic enzyme, b-galactosidase, as well as to its gene, lacZ causing it in turn to respond by transcription and to attract the requisite mutagenic machinery until the gene can be transcribed and translated into the enzyme that breaks down lactose, thereby restoring normal metabolic flux. The same applies to other situations of stress and stress relief. Resonance to electromagnetic signals is very precise, and will have all the appearance of being directed, particularly if the cell and organism is quantum coherent, as there are also strong reasons to suspect [7, 16] (The Rainbow And The Worm, ISIS Publication). This is a testable hypothesis, as the signals could be revealed by appropriately sensitive detectors and analyzers.

To conclude

Mutations are highly non-random and directed; numerous mechanisms for generating mutations are involved that appear to be under the control of the cell or organism as a whole in different environmental contexts, leading to repeatable mutations in specific genes. These results are contrary to the fundamental neo-Darwinian tenet that evolution depends on the natural selection of random genetic mutations. I suggest that specific electromagnetic signals emitted by key molecules that can relieve the stress are communicated directly to activate the transcription and mutation of the requisite gene(s).

Article first published 07/10/13

---

References

1. Dover G and Flavell D. Genome Evolution, Oxford University Press, Oxford, 1982.
2. Ho MW. Living with the Fluid Genome, ISIS/TWN London/Penang, 2003. https://www.i-sis.org.uk/fluidGenome.php
3. Cairns J, Overbaugh J and Miller S. The origin of mutants. Nature 1988, 335, 142-
4. Ho MW. To mutate or not to mutate. Science in Society 24, 9-10, 2004.
5. Cairns J. Mutation and cancer: the antecedents to our studies of adaptive mutation. Genetics 1998, 148, 1433-40.
6. Hall BG and Hartl DL. Regulation of newly evolved enzymes 1. Selection of a novel lactase regulated by lactose in Escherichia coli. Genetics 1975, 76, 391-400.
7. Ho MW. Living Rainbow H2O, World Scientific and Imperial College Press, Singapore and London, 2012. https://www.i-sis.org.uk/Living_Rainbow_H2O.php
8. Kacser H and Burns JA. The control of flux. Symp Soc Exp Biol 1973, 27, 65-104.
9. Ho MW ed. Living Processes Book 2, Bioenergetics, pp. 121-137, Open University Press, Milton Keynes, 1995.
10. Wright BE, Schmidt KH and Minnick MF. Kinetic models reveal the in vivo mechanisms of mutagenesis in microbes and man. Mutation Research 2013, 752, 129-37.
11. Ho MW. The body does burn water. Science in Society 43, 14-16, 2009.
12. Naqaoka H, Muramatsu M, Yamamura N, Kinoshita K and Honjo T. Activation-induced deaminase (AID)-directed hypermutation in the immunoglobulin Smu region: iplication of AID involvement in a common step of class switch recombination and somatic hypermutation J Exp Med 2002, 195, 529-34.
13. DNA double-strand breaks provoke mutation hotpots via stress-induced mutation in E. coli. Shee C, Gibson JL and Rosenberg SM. Two mechanisms produce mutation hotspots at DNA breaks in E. coli. Cell Reports 2012, 2, 714-21.
14. Martincorena H, Seshasayee ASN and Luscombe NM. Evidence of non-random mutation rates suggests an evolutionary risk management strategy. Nature 2013, 485, 95-98.
15. Ho MW. The real bioinformatics revolution – proteins and nucleic acids singing to one another? Science in Society 33, 42-45, 2007.
16. Ho MW. The Rainbow and the Worm, The Physics of Organisms, 3^rd ed., World Scientific and Imperial College Press, Singapore and London, 2008. https://www.i-sis.org.uk/rnbwwrm.php