Transcription is mutagenic so that over the life of an individual transcription-associated mutations should accumulate in the highly expressed genes whose products carry out translation and protein/RNA degradation.
This targeted accumulation of mutations should ultimately compromise translation and protein/RNA degradation and hence the cellular steady-state.
This may set an intrinsic limit to the number of times these crucial genes can express functional gene product, which can explain straightforwardly the classic comparative-biological trade-offs between fecundity, metabolic rate, longevity, etc.
and the need to evade these limits may be behind the evolution of the metazoan germ-line and of the transcriptionally inactive micro-nucleus of protists.
In other words the exertion and mutational exhaustion of these crucial genes could be the basic mechanism underlying adaptive trade-offs throughout the biological world, one that can shed light on what is and what is not, functionally and structurally, at every level of biological organization.
Transcription is well known to be mutagenic to the gene being transcribed and to its immediate surroundings (point mutations, rearrangements, etc; most mutagens are more effective on naked ssDNA), and impressive DNA repair machinery is mobilized around the transcription bubble.
To a first approximation, cells are machines to express the genes whose products are necessary to carry out transcription, translation, and orderly protein and RNA degradation.
In fact, the bulk of transcription, translation, and protein/RNA degradation in the cell is directed at synthesizing and degrading ribomosomal-protein mRNAs, ribomosomal RNAs (rRNAs), ribomosomal proteins, and the RNAs (?) and proteins needed for transcription and protein/RNA degradation.
Therefore, most transcription-associated mutations that manage to evade the surveillance of transcription-associated repair must alter the primary-structure of these crucially important highly transcribed genes.
These genes should therefore accumulate mutations over the active life of a cell, and so should the genes in charge of containing transcription-associated mutagenicity if these genes are highly expressed.
The accuracy and speed of transcription, translation, and protein/RNA degradation and possibly the containment of transcription-associated mutagenicity depend on the primary-structural integrity of these highly transcribed genes, genes that non-surprisingly are much conserved phylogenetically.
The accumulation of mutations in these genes must reduce progressively the accuracy and speed of transcription, translation, and protein/RNA degradation in the cell, and may increase transcription-associated mutagenicity (in yeast most ribosomal-protein mutations do not affect viability severely, unlike rRNA ones, a robustness possibly evolved to warrant some degree of cell viability even when such mutations occur; this subtler mode of action is consistent with the slow deterioration typical of aging).
In particular, mutations in these highly transcribed genes must also impair the accuracy and speed of the transcription and translation of these genes themselves altering additionally epigenetically the primary structure of the gene products expressed by the mutated genes.
Over the metabolic life of a cell the accuracy and speed of transcription, translation, and protein/RNA degradation, and possibly transcription-associated mutagenicity, should worsen inexorably compromising quantitatively and qualitatively the steady-state
Aging is likely to be due to a progressive, "targeted" accumulation of mutations in the crucially important genes whose products carry out transcription, translation, protein/RNA degradation, and, possibly, in those whose products contain transcription-associated mutagenicity, a preferential accumulation ensuing from the huge transcription rates of these genes.
Such mutations should impair the above functions not only directly, by modifying the primary structure of these genes and thus altering the intrinsic capability of the products of these mutated genes to function correctly, but also indirectly, by decreasing the accuracy and the speed of the transcription and translation of these genes themselves, mutated or not, which would alter their gene products additionally, compounding their loss of functionality.
Aging may not require that most genes mutate over time since the transcription-associated mutations that massively expressed genes are expected to accumulate, should suffice.
Indeed such mutations must be much more frequent than any others and should lead to deleterious effects throughout the proteome, transcriptome, and ultimately the genome.
Aging could be prevented and/or stopped by enhancing the containment of transcription-associated mutation, and it may be reversible by supplying the cell with intact (non-mutated) gene products to carry out the above functions and/or by administering drugs that restore the functionality of the mutated gene products.
Even if adaptation by natural selection were to neutralize fully the damage caused by transcription-associated mutation to highly expressed genes by means of some very effective adaptation, the adaptation itself is likely to be less and less effective late in life (since then it is not favored by natural selection) allowing transcription-associated mutation to start damaging highly expressed genes.
Many mutations in humans that lead to accelerated aging, affect transcription-associated DNA repair, the migration speed of the transcription bubble, the nuclear structures that make the nucleus non-permeable to fully assembled ribosomes yet permeable to ribosomal subunits, etc, and they often result in altered ribosomes, nucleolus, etc.
Decreasing artificially translation accuracy in fibroblasts was shown to lead to premature senescence and a contrast of the translation accuracy of old and young fibroblasts showed the former to have about one order of magnitude lower accuracy.
So-far documented comparative-biological trade-offs between early and late organismic performance (metabolic rate, reproductive output, longevity, etc) may be simple reflections of the limited rounds of transcription a cell can mobilize when deploying genetic programs that are translationally onerous, a limit set by the accumulation of mutations that its highly expressed genes must experience when such genetic programs are expressed among these programs is also somatic cell division because this entails a massive degradation of ribosomes before cell division that compels daughter and mother cells after dividing to restock their ribosome populations by expressing heavily ribosomal-protein genes and rRNA genes.
The separation between germ line and soma may have been evolved to protect the DNA of gamete-producing cells from the mutagenic effects of transcription efforts and of other biosynthetic efforts that do not need to take place in gamete-producing cells.
Similar pressures may explain the decreased transcription and biosynthetic activity of plants meristems, the transcriptionally silent protistan micronucleus, and the early shutting-down of X-chromosome transcription in the germline of mammalian males.
Transcription-associated mutation should damage the highly transcribed genes active in the soma compromising the quality of the RNA they produce and the primary-structure of these genes when inherited by mitotically produced descendant cells, i.e.
transcription-associated mutation should damage the performance of the soma.
For this reason, there should be a strong adaptive pressure to minimize this damage not only by repairing such mutations but also by modulating their nature (e.g. by tuning their context-dependence and/or base-to-base mutation probabilities in order to make the produced mutations less consequential).
Thus, mutation rates may be modulated strongly because of their somatic consequences in highly transcribed genes and not only because of their germ-line consequences for the genes to be inherited by sexually produced progeny.
Having a second template appears to be crucial for template-assisted DNA repair during transcription.
Thus no fully active cell may ever be really haploid in the sense that in most organisms heavy transcription may require the presence of at least a second haploid genome in order for template-assisted DNA repair to be possible, a presence that can be created by using opportunistically cell-division-related DNA replication (as haploid yeast's extended G2 and E.coli's growth-phase partial diploidy appear to do).
Recombination itself may have been evolved to reduce the huge costs of transcription-associated mutagenicity rather than to reduce DNA damage in gametes or to produce genetically variable progeny.
Therefore segregation, fission, syngamy, etc, may simply have taken off from there, i.e. from the state of chronic semi-orderly non-haploidy required for efficient template-assisted recombinational DNA repair of genes that because heavily transcribed, would be otherwise seriously damaged by transcription-associated mutation.
Has genetic analysis biased our view of cells as biomachines?
To a first approximation, cells are machines to synthesize and degrade ribosomes.
To a second approximation, cells are machines to express and degrade the products of highly expressed genes.
Even the nuclear membrane appears to be little more than a device to export from the nucleus rRNA bound to ribosomal subunits and to keep fully assembled ribosomes outside the nucleus, and yet for most of us who have done empirical molecular-biology work, ribosomal proteins and rRNAs are rather nuisances that hamper our attempts to study truly interesting (sic) genes and gene products.
This is probably due to epistemological biases created by the limits that history imposed on genetic analysis, e.g. by the biased nature of the phenotypes early geneticists were compelled to study in order to understand heredity, phenotypes that because somewhat understood transmission-genetically acquired afterwards an unjustified legitimacy as phenotypes tout court receiving more and more epigenetic attention and ending up shaping in a fully undeserved way our view of cells as biomachines.
Importantly, this has resulted in a remarkable neglect of the overwhelming functional and thus cyto-architectural, cyto-metabolic, and hence evolutionary centrality of those cellular functions that require extremely high levels of gene expression and gene-product turnover.