Many epidemics of fever ravaged Europe from ancient times through the early 20th century. But one disease stands out in historical accounts because authors describe patients appearing to recover before relapsing into fever again and again. This disease has been around for so long that Hippocrates described a series of such fevers that struck the city of Thasos in the wake of an especially harsh winter, and outbreaks have persisted through last century.
The disease tended to show up when times were hardest. Over the centuries, records describe epidemics of a nearly identical illness, usually on the heels of war or famine, with isolated cases popping up between times among the poor. One such epidemic struck during the Great Irish Famine of 1846 to 1852. Another ravaged Central Europe and Russia in the aftermath of World War I, killing at least five million people.
For years, historians have blamed those epidemics, termed louse-borne relapsing fever (LBRF), on Borrelia recurrentis, a twisting, spiral-shaped bacterium transmitted only by the human body louse. The logic was simple: B. recurrentis causes the only relapsing fever we know of thats carried by lice and capable of spreading fast enough to cause an epidemic. Although it seems to make frequent and horrible appearances in the historical record, LBRF has been totally invisible in the archaeological record. A new study changes that and provides evidence that B. recurrentis is indeed at fault.
An isolated case
Paleopathologist Meriam Guellil of the University of Oslo and her colleagues managed to assemble a nearly complete B. recurrentis genome from sequences of DNA recovered from the skeleton of a woman buried in a medieval graveyard in Oslo. She had been buried with her child, who was between seven and nine years old, near the southern edge of the graveyard—the one farthest from the church. This revealed something about her socioeconomic status even before anyone found evidence of a louse-borne pathogen on her remains.
She must have had a difficult life, and radiocarbon dating indicates that it came to an end around 1430 to 1465 CE. Based on how much bacterial DNA showed up in the shotgun sequencing performed by Guellil and her colleagues, the woman probably died with a lot of B. recurrentis in her system, so its likely that the fever killed her.
“Although at the time of the burial… the town was still affected by the economic decline caused by the Black Death in the mid-1300s, which probably left parts of the population vulnerable to disease and malnourishment,” wrote Guellil and her colleagues, “the results reported in this study probably represent an isolated case of the disease.”
Thats not enough to confirm that louse-borne relapsing fever caused all those historical epidemics, but it at least proves that the disease was present in medieval Europe. The real surprise, however, is how the medieval European version of LBRF differs from the modern strains that still impact people living in Ethiopia, Eritrea, Somalia, and Sudan. At some point in its evolutionary history, B. recurrenti appears to have split into two lineages—and theyve evolved different adaptive strategies, the medieval DNA suggests.
Different ways to downsize a genome
B. recurrentis causes relapses because its genetic code enables the bacteria to run through a random sequence of different variants of two proteins, called antigens, on its surface. The variations change the shape of its antigens, allowing it to keep dodging the hosts immune responses. The result is a fever that seems to clear up, then flares up again as the immune system tries to respond to a new set of antigens.
The genes that code for the antigen variants are stored on several plasmids, strands of DNA that can replicate or move around separately from the bacteria's chromosome. The more places in the bacterial genome that encode an antigen variant, the larger the number of them it can randomly produce to thwart a hosts immune system—at least in theory. But compared to modern African strains, the medieval strain is missing copies of six loci, mostly at a particular site on some plasmids.
But this method of hiding from the immune system is only one possible evolutionary solution. Other pathogens facing similar challenges tend to evolve smaller genomes and greater virulence. Guellil and her colleagues compared the genomes of the medieval strain of B. recurrenti to the modern African strains, as well as DNA sequences from a close relative, a tick-borne pathogen called B. duttonii that also causes a relapsing fever. Compared to its tick-borne cousin, all the louse-borne relapsing fever strains had fewer intact copies of the antigen variation genes.
“But this difference is even more pronounced in the medieval strain,” wrote the researchers. Modern strains, meanwhile, have also downsized their genomes over the countless generations since their last common ancestor with the medieval European strain, but theyve done it differently.
One of the major differences between the medieval strain and the modern strains is a genetic sequence called OppA-1, which is involved in metabolism—or would be, if it worked. In modern strains, it has been reduced to an imperfect, nonfunctional copy called a pseudogene, thanks to a stop codon in the middle of its sequence. In the medieval strain and in its close relative B. duttonii, OppA-1 still works.
Adaptive trade-offs
But everything has consequences, and for the medieval European strain of C. recurrenti, the price of a smaller genome was probably fewer antigen variants. That may have meant that the medieval European version of LBRF brought fewer relapses, on average, than the modern African versions. But Guellil and her colleagues say its impossible to be sure, because theres not much modern data on untreated cases of the fever, and historical sources arent always specific about the number of relapses patients suffered.
One clue, however, may lie in the LBRFs tick-borne cousin, B. duttoni, which has more intact loci for antigen variation than modern B. recurrenti. It tends to produce more relapses in untreated patients.
As for OppA-1, Guellil and her colleagues wrote that “We can only speculate about the effect of this mutation on the ecological life cycle of the disease.”
Based on their comparison of the medieval bacterial genome with modern strains, the researchers suggest that the two lineages have adapted to different environments. Humans may have helped shape that adaptation by putting pressure on the vector—maybe by changing hygiene or housing practices—or its host, but paleopathologists dont have enough data yet to be sure.
PNAS, 2018. DOI: 10.1073/pnas.1807266115 (About DOIs).
[contf] [contfnew] 
Ars Technica
[contfnewc] [contfnewc]
