New weapons in an ancient battle

Karl Phillips and colleagues explore how the role of novel immune genes in the never-ending struggle against invasion by pathogens.

Check out the full paper here

A brief history of the MHC

This story has an arguable start point: does it begin with the first parasitic interaction, the first host response to a parasitic interaction, or the first counter-response by the parasite? In any case, fast-forward to a time when vertebrates are hosts, and pause to note this. Then fast-forward again to the 1930s, and to scientists studying tissue graft rejection in laboratory mice. Rejection was found to be determined by group of genes that came to be called the major histocompatibility complex, or MHC – ‘histo-’ coming from the Ancient Greek for ‘tissue’. The MHC was found to play a crucial role in vertebrate immune systems – specifically, in telling ‘self’ molecules from those of pathogens, cancers, or other ‘foreign’ tissue, and passing on the message to the cells that mount an immune response. There was immediate medical potential: when you hear talk of ‘tissue typing’ for organ donation, you are hearing about the MHC. If the donor and recipient don’t have sufficiently similar MHCs (in people, also known as ‘HLA’ – human leukocyte antigen), the recipient’s immune system will recognise the organ as ‘foreign’ and start a fight.

Why are there so many versions of the MHC?

Researchers were quickly struck by the MHC’s extreme polymorphism (variability) – indeed, it’s by far the most polymorphic gene family in the vertebrate genome, with dozens to hundreds of alleles (versions of genes) routinely present in populations of pretty much any species. By comparison, most genes are much less variable, with only one or a few alleles. ‘How’, asked evolutionary biologists, ‘does this extraordinary variation persist?’, ‘Why has natural selection not fixed a single, ‘best’ MHC variant or set of variants?’ One of the earliest and most influential hypotheses was proposed by Walter Bodmer in 1972. He suggested that pathogens will evolve to make their peptides (small amino acid chains) hard for the host’s MHC protein to bind to. This should, in turn, give an advantage to novel or rare variants of the host’s MHC, to which pathogens have not had a chance to adapt. Crucially, this process never ends, and the shifting targets promote MHC diversity. This perpetual scrambling between hosts and pathogens, each trying to get ahead of the other, extends beyond the MHC and gives rise to what is arguably evolutionary biology’s best literary allusion: so-called ‘Red Queen’ processes, named after the character in Lewis Carroll’s Through The Looking-Glass who tells Alice “It takes all the running you can do, to keep in the same place”.Sir-John-Tenniels-drawing-of-Alice-and-the-Red-Queen-Running

While the core prediction of Bodmer’s hypothesis is simple – novel or rare alleles should be associated with lower numbers of pathogens in a host – it was difficult to test. Typically, a novel allele will occur in too low a frequency to allow meaningful statistical analyses to separate its effect from a number of other selection pressures that the MHC experiences, and thus test if it is indeed advantageous. Moreover, relying on snap-shot data (from a single time period) on allele frequencies in wild populations (determined from DNA samples) is unlikely to give a representative picture. These systems are constantly changing, and a currently rare allele could have been common in the recent past and still under negative selection, and vice versa for common alleles (the Red Queen is still running!).

Assembling a research team – scientists and study species

Our project set out to test Bodmer’s hypothesis. The project’s theoretical and logistical gametes fused when Jacek Radwan, of Adam Mickiewicz University in Poznań, and CEEC’s Cock van Oosterhout, both professors of evolutionary biology, discussed their respective ideas for such a test. All of these ideas involved guppies: small, tropical, freshwater fish, and something of a celebrity model species in experimental evolutionary biology. Although guppies are a popular aquarium species, Jacek and Cock, together with Cock’s long-term collaborator Jo Cable, professor of parasitology at Cardiff University, designed an experiment that would use wild guppies and a common skin parasite they suffer from. This parasite is a relative of flatworms and tapeworms called Gyrodactylus turnbulli, dubbed ‘gyros’, that eats the fish’s protective slime. This species of gyro is guppy-specific and exerts an important selection pressure – heavy infections can kill fish outright, reduce swimming efficiency, and make a fish a less attractive sexual partner. Importantly, gyros are superb experiment subjects. Their simple life cycle and rapid reproduction rate allow precisely controlled experimental infections (anaesthetise a donor and recipient fish and carefully bring them together) that can be easily monitored without harming host or parasite (anaesthetise the host and count the gyros on its skin).


How to test if hosts with novel MHC variants fare better against pathogens

Our experimental design was to mate fish from two guppy populations and allow the children of this cross to mate among themselves (see figure 1 below). An important requirement was that these ‘parent’ populations share no MHC alleles. They would then take the grandchildren of the original cross and infect them with gyros from one of the original source populations. If Bodmer’s hypothesis was correct, experimental fish with MHC alleles from parent population A should experience infections of higher intensity when infected with gyros that also came from population A, as the parasites should be adapted to the local host immune responses. This crossing design allowed separation of the effects of MHC from background host genetic effects (because the two generations of breeding will ‘mix’ the rest of the genome), and using multiple crossed population pairs allowed testing for benefits from novelty as a general property of MHC.


Figure 1. Schematic of the experiment. (A) Breeding design. Wild fish from two wild guppy populations (‘P-generation’) were crossed to produce F1 ‘children’ that will have exactly half of their total genetic makeup from each parental population. We allowed these to mate at random to produce F2 ‘grandchildren’. At the MHC class II (the subset of MHC that deals with pathogens that live outside cells) these grandchildren segregate into three groups, but will be similarly ‘mixed’ across the rest of their genomes. This design helps us to separate effects of MHC from other genes that may be have population-specific variants. (B) Controlled experimental infections. Two gyros from one of the P-generation source streams were inoculated on to the tail fin of each F2 fish. Each infected fish was then kept in isolation and its infection monitored every other day for 17 days.

A lesson in persistence!

Jacek, Jo and Cock wrote this up as a grant proposal to the Polish National Science Centre. The proposal, won the grant, and employed me as a postdoc. The project was to take place at a small house in a rural corner of Tobago, using locally caught wild guppies and guppies from Trinidad. By the time the experiment had concluded, I had spent 15/17 consecutive months in that house. Instrumental to the project’s success was Ryan Mohammed, a PhD student of the University of the West Indies, and a regular collaborator with Cock and Jo.


At the beginning, though, everything seemed to be going wrong. Our fish were dying or not reproducing, and our gyros were also dying. No fish or no gyros = no experiment! Added to that, logistics and sourcing materials were routinely problematic, power cuts were regular, and I suffered a number of mysterious illnesses. I was also 30cm away from being killed by falling branch. When the experimental infections were finally in full swing, they were also destructive to mind and body for all of those involved – intensive animal husbandry coupled with hours counting hundreds and hundreds of tiny mobile worms under a microscope. But, with often painful persistence, these obstacles were overcome and the work paid off. We tested the novel MHC allele hypothesis, found support for it, and published the work in the prestigious PNAS journal (see figure 2 below).


Figure 2. Infection trajectory graphs for guppies infected with gyros during our experiment, with boxplots showing cumulative infection intensity (“worm days” – the area under the trajectories). Local/local = all MHC variants from the same wild guppy population as the worms; novel/novel = all variants from the “other” population; local/novel = mixed. Pie charts show the proportion of fish of each MHC group that are dead (black), still infected (coloured), or have cleared their infection (grey) at each infection day (trajectories and boxplots exclude fish that died).

Jacek, Cock and Jo wrote the original research grant with a simple question: does MHC novelty, in general, confer selective advantage. After having painted the project as something between martyrdom and masochism, it feels almost anticlimactic to summarise the final paper simply as ‘yes’. A softly spoken and open-minded ‘yes’ – but a ‘yes.

There is a lot more to MHC evolution than just novel allele advantage. But, literally through blood, sweat, toil and tears, we hope to have contributed an important piece to this gene family’s fascinating story. The Red Queen keeps running!


Snell, G.D. & Higgins, G.F. 1951. Alleles at the histocompatibility-2 locus in the mouse as determined by tumor transplantation. Genetics, 36, 306-310.

Bodmer, W.F. 1972. Evolutionary significance of the HL-A system. Nature, 237, 139-145.

Cable, J. & van Oosterhout, C. 2007. The impact of parasites on the life history evolution of guppies (Poecilia reticulata): The effects of host size on parasite virulence. Int. J. Parasitol., 37, 1449-1458.

Carroll, L. 1872. Through the Looking-Glass, and What Alice Found There. Macmillan, UK.

About the author

IMG_0342Karl is an evolutionary biologist and molecular ecologist with research interests spanning animal mating systems, host-pathogen coevolution, and the impacts that domestic animals have when they escape and breed with their wild relatives. Karl is part of the CEEC diaspora, having studied for his PhD under BIO’s David S Richardson. He worked on the guppy project while a postdoc with Jacek Radwan’s Evolutionary Biology Group at Adam Mickiewicz University in Poznań, Poland ( He is currently a postdoc with Phil McGinnity’s FishEyE group at University College Cork (; @FishEvoEco), studying the effects of escaped farmed salmon on wild salmon.