Biology › Evolution and speciation › Resistance, drift and the evidence: selection where you can watch it happen

Resistance, drift and the evidence: selection where you can watch it happen

Antibiotic resistance is the case where the whole mechanism runs fast enough to observe, which makes it the case worth working through in detail. Then two things that are not selection — genetic drift and the founder effect — and the several independent lines of evidence that put common ancestry beyond serious argument.

Before this Natural selection and variation · Population genetics and the Hardy-Weinberg principle · Pathogens, disease and immunity

What you should be able to do

• Explain the origin of antibiotic resistance in a bacterial population without implying the antibiotic caused it.
• Distinguish vertical from horizontal transfer of a resistance gene, and say why the second is faster.
• Give the biological reason a prescribed course of antibiotics is finished.
• Describe pesticide resistance and explain why a refuge of untreated crop slows it.
• Set out the fossil, anatomical, biochemical and biogeographical evidence for common ancestry.
• Distinguish genetic drift and the founder effect from natural selection, and say when each matters most.

Antibiotic resistance, in the order it actually happens

Bacteria divide quickly — Escherichia coli can manage it in twenty minutes under good conditions. Even a low mutation rate per base per division, multiplied by a population of 10⁹ cells in a single infection, guarantees that mutations of every kind are present in that population before any drug is given, and a few of them will affect a protein the drug needs.

The mechanisms are worth naming, because a question asking 'suggest how this mutation causes resistance' expects one of them. A mutation may change the shape of the drug's target protein so the drug no longer binds — this is how streptomycin and rifampicin resistance work, by altering the ribosome and RNA polymerase. It may produce an enzyme that destroys the drug: β-lactamase hydrolyses the β-lactam ring of penicillins. It may increase production of an efflux pump that removes the drug faster than it enters, or reduce the permeability of the cell envelope.

Now run the selection. The antibiotic is a selection pressure of unusual severity: for a susceptible cell, relative fitness is close to zero. The susceptible cells die and the resistant ones do not, and since the susceptible cells were also their competitors for nutrients, the survivors grow into space they previously had to share. Within hours the population has been rebuilt from the survivors and essentially every cell carries the resistance allele — and no cell anywhere changed its genotype in response to anything.

Antibiotic

A substance produced by a microorganism, or a synthetic derivative of one, that kills or inhibits the growth of bacteria.

Antibiotic resistance

The ability of a bacterium to survive and reproduce in a concentration of an antibiotic that would kill or inhibit others of its species.

Vertical gene transfer

Passing genes from a parent cell to its daughter cells by binary fission.

Horizontal gene transfer

Passing genes between bacteria that are not parent and offspring, most often on a plasmid transferred by conjugation.

The mutations are not free. A ribosome altered so streptomycin cannot bind is usually a slightly worse ribosome, and a cell running efflux pumps is spending ATP on them. Without the drug, resistant strains are often at a small disadvantage, which is why resistance frequencies can fall when use of an antibiotic is reduced across a population — a biological argument for reducing unnecessary prescribing. Do not overstate it: compensatory mutations often restore the lost efficiency while keeping the resistance.

What 'resistance is spreading' actually means

This phrase appears in every newspaper article on the subject and is almost never unpacked. Two entirely different processes hide inside it, and a good answer distinguishes them.

The first is vertical transfer: a resistant cell divides and both daughters are resistant. This makes the frequency of a resistance allele rise within a population, and it is ordinary natural selection running at the speed bacteria divide.

The second is horizontal transfer, which has no equivalent in the inheritance you met in B16. Resistance genes are very often carried on plasmids — small circular loops of DNA separate from the main chromosome — and a plasmid can be copied and passed to another bacterium through a conjugation tube, including one of a different species. A plasmid may carry resistance to several antibiotics at once, which is how multiple resistance appears in one step. Bacteria also take up free DNA by transformation, and bacteriophages move DNA between cells by transduction.

So 'resistance is spreading' means both that resistant lineages are becoming commoner and that the genes themselves are moving sideways between lineages and species. The second is why a resistance gene first seen in a harmless gut bacterium can turn up years later in a dangerous pathogen. MRSA — meticillin-resistant Staphylococcus aureus — rests on an acquired gene, mecA, encoding a penicillin-binding protein that β-lactam antibiotics cannot inhibit.

Now the question about finishing the course. The cells in an infection are not equally susceptible, and the ones killed first are the most susceptible. Stop when the symptoms go and the cells left alive are a biased sample — the least susceptible ones, including any carrying partial resistance. They regrow, and the population that comes back is less susceptible than the one that started. Finishing the prescribed course removes the tail of that distribution rather than just its head.

One honest caveat, since this lesson is about the biology rather than the slogan. The blanket claim that stopping early 'causes resistance to evolve' is weaker than it is usually stated, and since around 2017 a number of clinicians have argued that unnecessarily long courses do their own harm by selecting on the bacteria you are carrying rather than the ones making you ill; shorter evidence-based courses are now prescribed for several infections. None of that changes the exam answer or the clinical advice, which is to complete the course you were prescribed.

The same argument, in a field instead of a patient

Nothing about the antibiotic story is special to bacteria or to medicine. Every case has the same four parts: variation already present, a severe and uniform selection pressure applied by people, a short generation time, and a huge population.

Insecticide resistance in mosquitoes is the most consequential example. Anopheles carrying a mutation in the voltage-gated sodium channel — the target of pyrethroids, and the change is called knockdown resistance — survive treated bed nets that kill their neighbours. Pyrethroid resistance is now widespread across sub-Saharan Africa, within a couple of decades of nets being distributed at scale.

Warfarin resistance in rats is a neater teaching case. Warfarin blocks the enzyme VKORC1, which recycles vitamin K, so poisoned rats bleed to death, and a resistance allele spread rapidly through British rat populations after warfarin was introduced in the 1950s. But homozygous resistant rats need far more vitamin K than they can normally obtain, so where warfarin is used the heterozygote is the fittest genotype and the population settles at an intermediate allele frequency rather than going to fixation. That is heterozygote advantage, the pattern you met with the sickle-cell allele.

Herbicide resistance follows from the same arithmetic, and glyphosate-resistant weed populations are established on farmland in North America and Australia. The agricultural response is designed around the biology: rotating chemicals with different modes of action so no single pressure is applied for long, and — for insect-resistant Bt crops — planting a refuge of non-Bt plants alongside. The refuge keeps susceptible insects breeding, so a rare resistant survivor is likely to mate with a susceptible one and produce heterozygous offspring, which the toxin still kills if resistance is recessive. It is selection theory used as engineering.

Four independent lines of evidence

Selection observed in a Petri dish shows that the mechanism works. The claim that it produced the diversity of life is supported separately, by evidence from fields that could have contradicted one another and do not.

Palaeontology. The fossil record puts organisms in an order: simple prokaryote-like forms in the oldest rocks, and major groups appearing in a sequence matching the order predicted by their anatomy. Transitional fossils exist where the argument most needs them — Tiktaalik between lobe-finned fish and tetrapods, Archaeopteryx between theropod dinosaurs and birds, and a series of whale ancestors with progressively reduced hind limbs. The record is incomplete, because fossilisation needs quick burial and soft tissues rarely survive, and a good answer says so.

Comparative anatomy. The pentadactyl limb is the standard case. The forelimb of a human, a bat, a whale and a horse contains the same bones in the same arrangement, doing entirely different jobs. If each had been designed for its job independently there would be no reason for the shared plan; if all four are modifications of one ancestral limb there is every reason. Structures like this are homologous: same origin, different function, evidence of divergent evolution. Compare them with analogous structures — the wing of a bird and the wing of an insect — which do the same job with entirely different underlying structures, and are evidence of convergent evolution rather than of relatedness.

Comparative biochemistry. All life uses the same genetic code, the same twenty amino acids and ATP as an energy currency, which is hard to explain except by common ancestry. More precisely, conserved molecules can be compared between species. Cytochrome c is identical in humans and chimpanzees, differs by about a dozen amino acids between humans and horses, by around twenty between humans and a tuna, and by more than forty between humans and yeast — a pattern matching the branching order the fossils and the anatomy already gave. Ribosomal RNA does the same job over greater distances, and it was rRNA comparison that separated the archaea from the bacteria. Where mutations accumulate at a steady rate, the number of differences estimates the time since two lineages diverged: a molecular clock.

Biogeography. The distribution of species across the land masses follows their history rather than their climates. Australia's mammals are overwhelmingly marsupial, not because pouches suit Australia but because the continent separated before placental mammals arrived. Oceanic islands hold endemic species closely related to those on the nearest mainland — the Galápagos finches, the Hawaiian Drosophila — and, before humans arrived, characteristically lacked whole groups such as native land mammals and amphibians that cannot cross an ocean.

A board note. OCR Biology A names palaeontology, comparative anatomy and comparative biochemistry explicitly in module 4 and can ask for them by those names. AQA reaches the same material largely through DNA and amino acid sequence comparison in its classification content, and Cambridge International examines molecular evidence alongside classification. Whatever your board, the argument that scores is that the lines are independent and agree, not that any one of them is decisive alone.

Changes in allele frequency that selection did not cause

Not every change in a gene pool is adaptive, and treating every difference between populations as an adaptation is a habit worth breaking early.

Genetic drift is change in allele frequency caused by chance alone. In any finite population the alleles reaching the next generation are a sample of those in this one, and samples deviate from what they are drawn from. Which individuals happen to breed, which gametes happen to meet, which juvenile happens to be standing under the falling branch: none of it has anything to do with fitness, and all of it changes allele frequencies. The smaller the population, the larger the deviation, for the reason ten coin tosses stray further from half than a thousand do.

Two consequences matter. Drift eventually leads to fixation — one allele reaching a frequency of 1 while the other is lost — and a lost allele can only return by mutation or migration. And drift can fix a mildly harmful allele or lose a mildly beneficial one, because in a small population chance overwhelms a small fitness difference. Conservation genetics treats small population size as a threat for that reason.

Genetic drift

A change in allele frequencies from one generation to the next caused by chance sampling of gametes, rather than by selection. Its effect is greatest in small populations.

Genetic bottleneck

A sharp reduction in population size that leaves the survivors carrying only a fraction of the original genetic diversity.

Founder effect

The reduced and unrepresentative genetic diversity of a population established by a small number of individuals separated from a larger population.

A bottleneck is drift applied violently. Northern elephant seals were hunted down to a few dozen animals in the 1890s; the species has recovered to hundreds of thousands but shows remarkably little genetic variation, because the variation was not in the survivors and no amount of later breeding recreates it. Cheetahs show the same signature from an older event, to the point that unrelated individuals accept skin grafts from one another.

The founder effect is the same statistics at the start of a population rather than in the middle of one. When a few individuals colonise an island or found a settlement, the alleles they carry are a small sample of the parent gene pool, and any allele they happen to carry at unusual frequency will be at that frequency in all their descendants. Hence conditions far commoner in some human populations than in the world at large: Ellis-van Creveld syndrome among the Old Order Amish, descended from around two hundred eighteenth-century founders, and variegate porphyria among Afrikaners, traceable to a single seventeenth-century couple. Nobody was selected for anything.

In the exam

• For any resistance question, put the mutation before the drug in your sentence order. 'Some bacteria already had the allele' first, 'the antibiotic killed the rest' second.
• If the question mentions a plasmid, it wants horizontal gene transfer by conjugation, and it wants you to say that this can move the gene between species.
• The reason for finishing a course is that the least susceptible cells are the last to die. Write that rather than 'so all the bacteria are killed'.
• Evidence questions reward independence. Two named lines of evidence that agree are worth more than a long account of one.
• Drift is a chance change with no fitness difference involved, and its size depends on population size. If a question tells you the allele has no effect on survival, it has ruled selection out for you.
• Bottleneck and founder effect are both drift. The bottleneck happens to an existing population; the founder effect happens when a new one starts.

Questions

Written to the command words the boards use. Try them on paper before opening a scheme: the marks go to points made, not to length.

← Natural selection: what varies, what is selected, and what actually changes · Speciation: how one gene pool becomes two, and why the line is genuinely fuzzy →