One of the ancient DNA finds that continues to intrigue me is the discovery of Borrelia burgdorferi, the agent of Lyme disease, in Ötzi the 5300-year-old ice mummy from the Italian Alps. As far as I know, this is the only finding of B. burgdorferi in ancient remains of any date. I discussed the initial report of these findings back in the summer of 2012.
A reconstruction of Oetzi the iceman. (Thilo Parg / Wikimedia Commons; License: CC BY-SA 3.0)
Ixodes ricinus (starved) (Photo: James Lindsey, CC by SA 3.0 via Wikipedia)
The more we learn about Ötzi’s environment and lifestyle, the less mysterious it seems. There are no signs of human habitation or land management in these high Alpine regions. Indicators of deforestation, farming, and pasture maintenance are lacking from lake sediment and pollen studies. Festi, Putzer and Oeggl (2013) found the first signs of human land management in the Ötztal Alps to began about 1000 years after Ötzi’s time. During the Copper Age, subsistence occupation of the valley floor was sufficient for the population of Ötzi’s time. They did minimal farming, and breeding of caprines (sheep, goats, and ibex). Festi, Putzer and Oeggl (2013) note that Ötzi’s mummy is the only piece of evidence for humans that high in the Otztal Alps before the Bronze Age.
Before Ötzi’s time, landscape management in the Mesolithic was to support red deer herds that were “in a state of semi-domestication by means of active hunting” (Rollo et al, 2002). (Native Americans managed deer populations in similar ways by promoting a landscape where deer thrive near their hunting grounds.) The importance of deer to Ötzi is underscored by everything about him from the red deer meat in his stomach to the roe deerskin that made up his quiver and antler in some of his tools (Rollo et al, 2012). Two different species of deer have been confirmed by genetic analysis. Most of his clothing was made of sheep and goat skins (O’Sullivan et al, 2016).
The agent of Lyme disease, B. burgdorferi, is transmitted primarily by the tick Ixodes ricinus, common on deer, sheep, cattle, humans and dogs as adults and feed on rodents and small mammals as nymphs. Ticks often thrive at the forest edge where there are grasses for them to climb up to catch passing deer. It seems likely that they would also thrive in along upland forest edges as well. Ixodes ricinus is found throughout the Alps. It is feasible that Lyme disease was a greater problem for humans when we relied on deer as a staple food.
Ötzi’s B. burgdorferi has yet to be confirmed by a second group. Interestingly, a recent study of B.burgdorferi’s phylogeny suggests that it originated in Europe and later spread to ‘post-Columbian’ North America (Margos et al, 2008). Although Lyme disease was only recognized in the 20th century, it is apparently an ancient disease caused by multiple Borrelia species. And Ötzi’s sequence has not been added to any phylogeny I’ve found, odd. Overlooked, or a problematic sequence?
Festi, D., Putzer, A., & Oeggl, K. (2013). Mid and late Holocene land-use changes in the Otztal Alps, territory of the Neolithic Iceman “Otzi”. Quaternary International, 353, 1–18. http://doi.org/10.1016/j.quaint.2013.07.052
Margos, G., Gatewood, A. G., Aanensen, D. M., Hanincová, K., Terekhova, D., Vollmer, S. A., et al. (2008). MLST of housekeeping genes captures geographic population structure and suggests a European origin of Borrelia burgdorferi. Proceedings of the National Academy of Sciences, 105(25), 8730–8735. http://doi.org/10.1073/pnas.0800323105
O’Sullivan, N. J., Teasdale, M. D., Mattiangeli, V., Maixner, F., Pinhasi, R., Bradley, D. G., & Zink, A. (2016). A whole mitochondria analysis of the Tyrolean Iceman’s leather provides insights into the animal sources of Copper Age clothing. Scientific Reports, 6, 1–9. http://doi.org/10.1038/srep31279
Rollo, F., Ubaldi, M., Ermini, L., & Marota, I. (2002). Otzi’s last meals: DNA analysis of the intestinal content of the Neolithic glacier mummy from the Alps. Proceedings of the National Academy of Sciences of the United States of America, 99(20), 12594–12599. http://doi.org/10.1073/pnas.192184599
The explosion of Zika-related birth defects this past year came out of the blue. Zika has been known since the 1940s but was seen as a mild dengue-like illness (Fauci & Morens, 2016). Leaving aside how and why microcephaly has appeared so dramatically, it is undeniable that Zika’s emergence and transmission in the Americas have been unusually rapid and extensive.
Two papers published in December focusing on the Aedes mosquito vectors begin to shed light on how Zika was able to be established so quickly and pervasively. Zika utilizes the same tropical mosquito Aedes aegypti as dengue; it wasonce known as the yellow fever mosquito. It is also the vector of the chikungunya virus.
As first observed in West Africa many years ago, Zika epidemics followed a chikungunya epidemic by a couple years. Chikungunya was the emerging infectious disease of 2013, the year that Zika is believed to have arrived in South America (Fauci & Morens, 2016). Unrecognized by public health workers at the time, a Chikungunya epidemic was simultaneously chugging along under the radar in at least Salvador, the capital of the Bahai state of Brazil, during the peak of Zika epidemic of 2015 (Cardoso et al, 2017).
El Niño 2015-2016
In the first study by Cyril Caminade and colleagues at the University of Liverpool modeled Zika transmission in the two critical vector species in the Americas, the tropical Aedes aegypti found primarily in South America and the temperate Aedes albopictus found in the southern United States. It is thought that Zika transmits better from A. aegytpi but more research is needed to fully understand the differences. They developed a two vector, one host model where the climate is a variable to compare the effect of climate patterns on Zika transmission. They ran these simulations for each vector individually and together against historic climate data sets.
When they compared the worldwide distribution of the vectors and climate, they were able to show that all of the countries where Zika has been reported were predicted in their model. Ominously, South America was the most friendly region in the world for Zika (Caminade et al, 2016). The model for Zika produced a map that correlates extremely well with the global distribution of dengue. Due to the overlap of A. aegypti and A. albopictus territory, they found a high probability that Zika would transmit well in most of the southern United States.
The global climate anomaly known as El Niño is known to impact mosquito-transmitted diseases, so they had a particular interest in comparing the 2015-2016 El Niño to historic data sets. The map shows the predicted Ro (reproduction number) for Zika around the world in 2015-2016 and in the bar graph compared to the last 50 years. The conditions for Zika were the best for the last 50 years. Other hot spots that did not experience a Zika epidemic, like India, did have a record year for dengue. They also note that the African hot spot for ideal transmission conditions corresponds and to Angola where there was a Yellow Fever outbreak. In short, it was a very good year for Andes aegypti! And now, as of January 2017, Yellow Fever had added to their misery in a Brazil.
A Sylvatic Reservoir?
Understanding if Zika will establish a sylvatic reservoir in South America is of vital importance for projections and mitigation of future Zika epidemics in Brazil and elsewhere in South America. Zika was initially detected in a sentinel monkey in Uganda and has since been detected in a wide variety of smaller primates in Africa and Asia. Using a model originally proposed for dengue they were able to show that primates with rapid birth rates and short lifespans are ideal for establishing sylvatic Zika. In primates with short life span, five years or less, and rapid birth rates, the establishment of a sylvatic reservoir is “nearly assured” (Althouse et al, 2016). They predict that a primate population as small as 6,000 members with 10,000 mosquitoes could support a sylvatic reservoir (Althouse et al, 2016). Ironically, since infection rate is dependent upon bites per primate, a small primate population with a large mosquito population is better at maintaining the reservoir than a large primate population. Old World monkeys like the African Green Monkey, a known African host of Zika, are already established in free-living troops in South American forests. While A. aegypti favors human environments, A. albopictus prefers forested environments and has been spreading in Brazil. It could be a prime candidate for a bridging vector between a sylvatic and domestic Zika cycle. Studies on Zika vulnerability and incidence in all South American primates has to be a priority. Our ability to manage Zika in the future depends on it.
Caminade, C., Turner, J., Metelmann, S., Hesson, J. C., Blagrove, M. S. C., Solomon, T., et al. (2016). Global risk model for vector-borne transmission of Zika virus reveals the role of El Niño 2015. Proceedings of the National Academy of Sciences of the United States of America, 201614303–28. http://doi.org/10.1073/pnas.1614303114
Cardoso, C. W., Kikuti, M., Prates, A. P. P. B., Paploski, I. A. D., Tauro, L. B., Silva, M. M. O., et al. (2017). Unrecognized Emergence of Chikungunya Virus during a Zika Virus Outbreak in Salvador, Brazil. PLoS Neglected Tropical Diseases, 11(1), e0005334–8. http://doi.org/10.1371/journal.pntd.0005334
Althouse, B. M., Vasilakis, N., Sall, A. A., Diallo, M., Weaver, S. C., & Hanley, K. A. (2016). Potential for Zika Virus to Establish a Sylvatic Transmission Cycle in the Americas. PLoS Neglected Tropical Diseases, 10(12), e0005055–11. http://doi.org/10.1371/journal.pntd.0005055
Fauci, A. S., & Morens, D. M. (2016). Zika virus in the Americas—yet another arbovirus threat. New England Journal of Medicine, 374(7), 601–604. http://doi.org/10.1056/nejmp1600297
The key to understanding plague — past, present, and future — has always been understanding its vector dynamics. By the latest tally, there are 269 known flea species, plus a small collection of ticks and lice, that can be infected with Yersinia pestis. With this many infected parasites, it’s not a surprise that 344 hosts have been identified(Dubyanskiy & Yeszhanov, 2016), but this list is still incomplete. (It does not include all of the minor hosts in North America.) Regardless, this is not a description of a picky pathogen! Unfortunately, it is far easier to identify infected hosts and potential vectors than to determine which of these insects are effective vectors and their transmission dynamics.
Numerous species of fleas have been identified as plague vectors in specific localities. However, only five infected potential human vectors are possibly involved in a wide distribution of human plague cases — the rat flea Xenopsylla cheopis, the cat flea Ctenocephalides felis, the human flea Pulex irritans, the human body louse Pediculus humanus humanus, and the human head louse Pediculus humanus capitis. The rat flea and the cat flea are known vectors, but they are unlikely to account for the full transmission of the massive first and second pandemics. I recently discussed the possible role of the human flea. Given the worldwide distribution of human lice, they are attractive vectors but there is still work to do before they can be considered likely primary vectors for human to human transmission during the first two pandemics.
Raoult Makes his Case…again
Didier Raoult and his team have been working on plague and their louse transmission hypothesis for a long time. It has already been ten years since they had enough information to write their first review article putting forth their human ectoparasite theory of plague transmission (Drancourt, Houhamdi, & Raoult, 2006).
At this point, their primary supporting evidence was some experiments with human louse plague transmission in rabbits (Houhamdi et al, 2006) and they thought they could associate louse transmission with an “Orientalis-like” biovar of Yersinia pestis they identified in the first two pandemics (Drancourt et al, 2004). However, later ancient DNA work showed that the first two pandemics were caused by strains of Yersinia pestis that emerged before the Orientalis biovar. Genetic reclassification of Yersinia pestis has also made the biovars largely obsolete. Tensions between groups working on ancient plague DNA developed quickly, and have been documented by historians Lester Little (2011) and Jim Bolton (2013).
This summer Didier Raoult (2016) restated his “personal view” on the role of lice in plague transmission. This essay is unusual not only as a first person narrative in science, including individual claims of discovery, but also for being so vindictive in its attack on his rivals. Again, see Little (2011) and Bolton (2013) for less biased accounts. His team has done very impressive work.
His team has continued to assemble much of the work needed to argue that human lice were instrumental in at least some of the major human outbreaks of plague during the first two pandemics. Combing primarily French medical reports in North Africa, they were able to identify observations that suggest that lice were involved in some mid-twentieth century outbreaks (Raoult, 2016; Malak, Bitam, & Drancourt, 2016). One of their most interesting findings in 2011 was the discovery of co-infection with Yersinia pestis and Bartonella quintana (trench fever) in late medieval French remains (Tran et al, 2011). Trench fever is well known to be transmitted by the human louse. Both B. quintana and Y. pestis have been found in contemporary lice taken from plague patients in regions of endemic plague in the Congo (Piarroux et al, 2013; Drali et al, 2015). Unfortunately, neither of these studies mention the presence or absence of fleas. That blood feeding lice would be infected is not a surprise, but the question of transmission stubbornly remains. There has yet to be a contemporary outbreak where all potential vectors, fleas and lice, were investigated. On a side note, the finding of widespread B. quintana is interesting, and perhaps a proxy for heavy lice infestation.
In the meantime, while Raoult’s latest summary was in press, additional evidence was beginning to be revealed. Graduate student Katharine Dean of the MedPlag project in Oslo was modeling past epidemics for transmission by rat fleas, human lice, and pneumonic transmission. In her master’s thesis, she showed that lice transmission fits the second pandemic epidemics at Givry in 1348, London in 1563-64, and Florence in 1630-31 better than rat fleas or pneumonic transmission (Dean 2015). At the most recent Yersinia meeting in October, Dean presented a poster with expanded data finding outbreaks that fit each of these three modes of transmission (pneumonic in Manchuria, rat fleas in Sydney and Hong Kong, and lice in many locations) (Dean et al, 2016). Their work is still in progress and I’m sure many will be eager to see their results in due course.
There are still a few lingering things to nail down. A modern outbreak investigation that looks at all ectoparasites, fleas and lice, in the region that suggests lice are involved. It would be good to find lice (or fleas for that matter) in a plague burial that yields Y pestis aDNA. Alternatively, detection of more coinfection of Y. pestis with a louse-transmitted infection like B. quintana would lend additional support. These findings will will require some good fortune. To differentiate between the human flea and lice, a better understanding of the pathology of a Y. pestis infection in the potential vector and its transmission dynamics is really needed. The models can’t differentiate human ectoparasites without more information.
Human ectoparasites are beginning to look much more likely especially for northern epidemics (Hufthammer & Walløe, 2013). More information is still needed to distinguish between human fleas and lice, although they may be both involved in different outbreaks. We need to be ready for yet another paradigm change in plague history. Looking at the overall plague vector dynamics of the great pandemics, from sylvatic reservoir to distant human populations, is going to get a whole lot more complicated but also more interesting.
Bolton, J. L. (2013). Looking for Yersinia pestis: Scientists, Historians, and the Black Death. The Fifteenth Century, XII, 15–38.
Drali, R., Shako, J. C., Davoust, B., Diatta, G., & Raoult, D. (2015). A New Clade of African Body and Head Lice Infected by Bartonella quintana and Yersinia pestis–Democratic Republic of the Congo. American Journal of Tropical Medicine and Hygiene, 93(5), 990–993. http://doi.org/10.4269/ajtmh.14-0686
Drancourt, M., Roux, V., Dang, L. V., Tran-Hung, L., Castex, D., Chenal-Francisque, V., Ogata, H., Fournier, P-E., Crubezy, E, and Raoult, D. (2004). Genotyping, Orientalis-like Yersinia pestis, and plague pandemics. Emerging Infectious Diseases, 10(9), 1585–1592. http://doi.org/10.3201/eid1009.030933
Dubyanskiy, V. M., & Yeszhanov, A. B. (2016). Ecology of Yersinia pestis and the Epidemiology of Plague. Yersinia Pestis: Retrospective and Perspective, 918(Chapter 5), 101–170. http://doi.org/10.1007/978-94-024-0890-4_5
Houhamdi, L., Lepidi, H., Drancourt, M., & Raoult, D. (2006). Experimental model to evaluate the human body louse as a vector of plague. The Journal of Infectious Diseases, 194(11), 1589–1596. http://doi.org/10.1086/508995
Hufthammer, A. K., & Walløe, L. (2013). Rats cannot have been intermediate hosts for Yersinia pestis during medieval plague epidemics in Northern Europe. Journal of Archaeological Science, 40(4), 1752–1759. http://doi.org/10.1016/j.jas.2012.12.007
Piarroux, R., Abedi, A. A., Shako, J. C., Kebela, B., Karhemere, S., Diatta, G., et al. (2013). Plague epidemics and lice, Democratic Republic of the Congo. Emerging Infectious Diseases, 19(3), 505–506. http://doi.org/10.3201/eid1903.121329
Tran, T.-N.-N., Forestier, C. L., Drancourt, M., Raoult, D., & Aboudharam, G. (2011). Brief communication: co-detection of Bartonella quintana and Yersinia pestis in an 11th-15th burial site in Bondy, France. American Journal of Physical Anthropology, 145(3), 489–494. http://doi.org/10.1002/ajpa.21510