On The Origin of Collectives -- Bacterial Evolution

Bacteria evolved for low encounter-rate environments do not move quickly relative to their sizes. High fertility with low mobility would mean that a bacterium is surrounded by copies of itself. Competition by aggression is pointless. To divide, then kill its sibling would just waste resources compared to a strategy of not dividing.

Bacterial species also tend to be fairly specialized to a few metabolic pathways consuming external materials. With energy harvesting processes commonly taking place on their membranes, multi-species metabolic chains not only efficiently utilize the local resources; they remove potentially toxic intermediate byproducts.

Bacteria evolved toward vertically integrated metabolic chains must prevent undesired interactions between intermediate byproducts; where distance suffices for metabolically specialized organisms.

These circumstances favor evolution of metabolite specialists and reproductive controls among, not only individuals, but multi-species bacterial communities.

Mutation Rates for Coding DNA

Given a mutation probability R = 1 × 10-9 per basepair per generation, a population of N survivors after each generation, a genome size of S basepairs, and a generation latency L; after one generation we expect an instance of a member with k mutations with probability Rk · S · N.

To first order we expect to wait
Rk · S · N
for such a mutated individual to appear.

If a minority M<<N has mutation probability Q = a · R, then we expect an instance of an individual with k mutations with probability Qk · S · M.

The appearance of an individual with those k mutations is expected after a wait:

Qk · S · M

Equating the expected waits:

Rk · S · N
 = L
Qk · S · M

Rk · N = ak · Rk · M

 = ak

So normal (non-mutator) population must be ak times as large as a mutator population in order to produce as many k-fold mutations. The E. coli rpsL- allele increases replication errors by a factor of 50 [2]. Thus a rpsL- mutator produces 6-fold mutations at a rate (506 = ) 15.6 × 109 times higher than a rpsL+.

With the rpsL- allele occuring with a frequency 1 × 10-9, this minority of mutators will dominate production of 6-fold and larger mutations.

Mutation Rates for Non-Coding DNA

Because its content is largely unconstrained, the population of plasmid DNAs is subject to two processes:

Bacterial Mutators

DNA repair mechanisms are degraded far more easily than new repair mechanisms arise through random mutations. So some fraction of a prokaryote population will always have higher mutation rates than the norm. These cells are termed "mutators" and have a range of mutation rates depending on the particular DNA-polymerase alleles they carry [1].

Single point and knockout mutations are plentiful when DNA repair genes are compromised. But larger changes require many more mutations. These cannot be serially accumulated in expressed genes without impairing cell functions. The mutators are the source of these larger genetic jumps.

The DNA where bacteria do accumulate mutations are plasmids. Non-functioning plasmid genes require only insignificant amounts of nucleotides for sustenance --- millions of generations would be required to naturally select against this small burden.

Each healthy member of the community accumulates point mutations in its plasmid at the glacial rate of one per 104 generations. The mutators with their 50-fold mutation rate will always push the gene-space envelope. Whatever the plasmid is mutating to, the mutator will tend to get it there first.

Recombination, viruses, and transposable elements perform the major shifts, duplication, and deletions in developing protogenes. Even viruses can be helpful in transferring genetic material between organisms.

Bacteria trade plasmids between individuals of different species. Many bacteria are not in direct competition with other species of bacteria; moreover, many belong to multi-species symbiotic communities. This sharing of genetic material is an essential mechanism for gene test and development.

Bacterial communities essentially have a social structure where a small portion of that community (the mutators) sacrifice themselves performing multi-mutation experiments on their genomes. Conjugation within the community transfers portions of the modified genome to peers; allowing that genetic material to be tried outside of the genetically debilitated mutator. Conjugation can also transfer undamaged genes into the mutators; extending their useful lives. A mutator's utility does not die with it; bacteria can absorb DNA from their environment.

When one of these mutations confers selective advantage, for instance toxin tolerance, one or more of the species in the community can gain vigor; and the gene can eventually be incorporated into their chromosomes.


Robert de Marrais suggests that these principals might also apply to human communities. An exceptionally creative shaman, leader, or scientist can bring change to an entire society.
In the first part, the low motility of bounded communities led to cooperation with neighbors, metabolic specialization among species, and reproductive controls.


But in unbounded, feedstock-rich environments many bacteria are motile, and this changes the game. Metabolic relationships with other bacteria would not be stable. So the cell must potentially be able utilize a variety of feedstocks; although it need not tolerate its own waste.

The ability to move allows higher fecundity with the farthest wanderers able to escape the consequences of periodic over-population.

Motile mutators will suffer earlier mortality than stationary ones because motile cells tend to have longer metabolic chains; there is more machinery to break. But with promiscuous genetic exchange, the slow-moving mutators can supply motile cells with mutations.

The next chapter is Bacterial Metabolism, Neurotransmitters, Migraine Headaches, and Birth Defects

Copyright © 2003 Aubrey Jaffer

I am a guest and not a member of the MIT Computer Science and Artificial Intelligence Laboratory.  My actions and comments do not reflect in any way on MIT.
The Game Theory of Life
agj @
Go Figure!