- Email: [email protected]
Ancient common ancestry of Crimean-Congo hemorrhagic fever virus
Molecular Phylogenetics and Evolution 55 (2010) 1103–1110
Contents lists available at ScienceDirect
Molecular Phylogenetics and Evolution journal homepage: www.elsevier.com/locate/ympev
Ancient common ancestry of Crimean-Congo hemorrhagic fever virus Serena A. Carroll, Brian H. Bird, Pierre E. Rollin, Stuart T. Nichol * Special Pathogens Branch, Division of Viral and Rickettsial Diseases, Centers for Disease Control and Prevention, 1600 Clifton Road, MS G-14, Atlanta, GA 30333, USA
a r t i c l e
i n f o
Article history: Received 18 September 2009 Revised 31 December 2009 Accepted 6 January 2010 Available online 13 January 2010 Keywords: Crimean-Congo hemorrhagic fever CCHF Bunyaviridae Molecular evolution Bayesian Coalescent
a b s t r a c t Crimean-Congo hemorrhagic fever (CCHF) is a tick-borne RNA virus responsible for outbreaks of severe hemorrhagic fever in humans. Although CCHF was first detected in the 1940s, high levels of genomic diversity argue against a recent origin. Here, Bayesian coalescent analyses were used to estimate the rate of evolution and relative age of the virus. A total of 43 S, 34 M, and 23 L segment sequences from samples collected between 1956 and 2005 were analyzed from across the broad geographic range of the virus. Using a relaxed molecular clock model, nucleotide substitutions were estimated to have occurred at a rate of 1.09 10 4, 1.52 10 4, and 0.58 10 4 substitutions/site/year for the S, M, and L segments, respectively. The most recent common ancestor of the viruses existed approximately 3100–3500 years before present, or around 1500–1100 BC. Changes in agricultural practices and climate occurring near the time of the most recent common ancestor of CCHFV may have contributed to its emergence and spread. Published by Elsevier Inc.
1. Introduction Crimean-Congo hemorrhagic fever virus (CCHFV), genus Nairovirus, family Bunyaviridae, is responsible for outbreaks of severe hemorrhagic fever in humans with case fatalities as high as 30% or more in some instances (Swanepoel et al., 1987; Whitehouse, 2004; Ergonul, 2006; Schmaljohn and Nichol, 2007). The disease was initially recognized by Russian scientists in the 1940s (Chumakov, 1945), and the virus was first isolated from the Democratic Republic of Congo some years later (Woodall et al., 1965, 1967; Simpson et al., 1967). CCHFV is now known to be distributed throughout broad regions of Africa, Europe, the Middle East, and Asia. The virus is transmitted primarily by ticks in the genus Hyalomma; however, several vertebrate and tick species are involved in the natural transmission cycle (Hoogstraal, 1979; Swanepoel et al., 1983, 1987). Human exposures can occur via tick bites, direct contact with blood or tissues of infected animals, person-to-person spread, or by nosocomial infection (Burney et al., 1980; Whitehouse, 2004; Ergonul, 2006; Schmaljohn and Nichol, 2007; Gurbuz et al., 2009). CCHFV is a negative sense, single stranded, tripartite RNA virus with a genome size of approximately 19 kb (Whitehouse, 2004; Schmaljohn and Nichol, 2007). The small (S) segment encodes the nucleocapsid protein whereas the medium (M) segment contains a highly variable region known as the mucin domain (Sanchez et al., 2002) and encodes the envelope glycoproteins Gn and Gc. The large (L) segment encodes the virus RNA-dependent RNA
* Corresponding author. Fax: +1 404 639 1509. E-mail addresses: [email protected] (S.A. Carroll), [email protected] (B.H. Bird), [email protected] (P.E. Rollin), [email protected] (S.T. Nichol). 1055-7903/$ - see front matter Published by Elsevier Inc. doi:10.1016/j.ympev.2010.01.006
polymerase. While arthropod-borne RNA viruses generally show low levels of genetic diversity, perhaps due to a hypothesized ‘‘double-filter” selection mechanism that allows for the maintenance of high fitness in both arthropod and vertebrate hosts (Weaver, 2006), previous studies have demonstrated surprisingly high levels of diversity for CCHFV (Papa et al., 2002, 2005; Hewson et al., 2004a; Chamberlain et al., 2005; Deyde et al., 2006; Burt et al., 2009). Additionally, both reassortment and recombination appear to have contributed to the high amount of genetic variation present in the virus (Chare et al., 2003; Hewson et al., 2004b; Lukashev, 2005; Deyde et al., 2006; Burt et al., 2009). Despite the apparent emergence of CCHFV in the 1940s (Chumakov, 1945), the large amount of sequence diversity found within the species argues against a recent origin. In the past few years, advances in Bayesian coalescent phylogenetic analyses have resulted in more sophisticated methods for determining the relative ages of emerging pathogens (Drummond et al., 2006; Drummond and Rambaut, 2007). The purpose of this study was to apply these tools to address the question of CCHFV ancestral origins. If the most recent common ancestor (MRCA) of the extant CCHFVs existed in the relatively recent past (i.e. near the initial date of disease recognition) then it could be inferred that the observed high viral genomic diversity is a result of rapid evolution. Alternatively, a more ancient origin would suggest that the virus has had sufficient time to generate high levels of genomic diversity without the need for a rapidly evolving virus population. A secondary goal of the study was to compare the molecular evolutionary rate of this tick-borne bunyavirus with that of a mosquito-borne bunyavirus (i.e. Rift Valley fever virus, RVFV) and with other unrelated tick and mosquito-borne RNA viruses.
1104
S.A. Carroll et al. / Molecular Phylogenetics and Evolution 55 (2010) 1103–1110
of 10,000,000 generations. An analysis of marginal likelihoods (Suchard et al., 2001) indicated that the relaxed uncorrelated lognormal clock (Drummond et al., 2006) and constant population size model was most appropriate for each of the three segments (log10 Bayes Factors > 2 in all cases). Final Bayesian coalescent analyses used the GTR + G substitution model, a relaxed uncorrelated lognormal clock, and a constant population size. The S and L segment datasets were run for 20,000,000 generations whereas the M segment dataset was run for 115,000,000 generations to ensure effective sample sizes (ESSs) of at least 200. Maximum clade credibility trees were summarized using TreeAnnotator and depicted using FigTree (Drummond and Rambaut, 2007). Further comparative analyses were conducted to investigate sources that might have biased the estimated rates and/or time to MRCAs. Given that reassortment can potentially affect coalescent analysis, all possible reassortant viruses were removed from the datasets. In addition, the highly variable mucin domain was removed from the M segment analysis to determine its influence on the overall rate estimate. Finally, to investigate the effects of taxon sampling, the most densely sampled dataset (S segment) was pared down to match the least densely sampled dataset (L segment). These additional analyses followed the same methods as described above.
2. Materials and methods CCHFV sequences, representing the broad geographic range of the virus, were obtained from GenBank (Fig. 1); only whole S, M, or L segment sequences with associated dates of collection were retrieved (Table 1). A total of 43 S segment sequences spanning the years 1956–2005, 34 M segment sequences from 1956–2003, and 23 L segment sequences from 1956–2003 were examined. Multiple sequence alignments were generated using the MAFFT function (Katoh et al., 2005) in SeaView (Galtier et al., 1996). ModelTest version 3.7 (Posada and Crandall, 1998) examined 56 possible nucleotide substitution models and was used to select the most appropriate model for each dataset based on Akaike’s Information Criterion (AIC). The General Time Reversible model with a gamma distribution (GTR + G) was selected for all three segments. Rates of evolution and estimates of the time to MRCA were generated via Bayesian coalescent phylogenetic analysis using the BEAST software package version 1.4.7 and Tracer version 1.4 (Drummond and Rambaut, 2007). Preliminary analyses were conducted to determine which clock (strict versus relaxed) and demographic (constant versus Bayesian Skyline population size) models were most appropriate for the datasets, and runs initially consisted
Russia
Kosovo Bulgaria Greece
Uzbekistan
Turkey
Tajikistan
China
Iraq Pakistan
Oman
Mauritania Senegal
Nigeria Cen. Af. Rep.
Democratic Republic of Congo
Uganda
South Africa
Fig. 1. Geographic distribution of CCHF virus samples analyzed in this study, originating from 18 countries in Africa, Europe, Asia, and the Middle East.
1105
S.A. Carroll et al. / Molecular Phylogenetics and Evolution 55 (2010) 1103–1110 Table 1 CCHF viruses examined, with associated countries of origin, collection dates, and GenBank accession numbers. Sample
Country
Date
S
BulgariaHu517 CentralAfricanRepBT958 China66019 China7001 China75024 China7803 China79121 China8402 China88166 ChinaC68031 ChinaTI05146 ChinaTI05099 ChinaHY13 ChinaTI05035 CongoUG3010 GreeceAP92 IraqBaghdad12 Kosovo9553 KosovoHoti MauritaniaArD39554 NigeriaIbAr10200 Oman PakistanJD206 PakistanMatin PakistanSR3 RussiaASTR30908 RussiaDrosdov RussiaKashmanov RussiaSTVHU29223 RussiaTI28044 RussiaVLGTI29414 RussiaVLV100 SouthAfricaSPU10387 SouthAfricaSPU128817 SouthAfricaSPU481 SouthAfricaSPU41585 SouthAfricaSPU4184 SouthAfricaSPU9785 SenegalArD15786 SenegalArD8194 TajikistanHU8966 TajikistanHU8975 TajikistanHU8978 Turkey200310849 UgandaSemunya UzbekistanHodzha UzbekistanTI10145 UzbekistanU22002U6415
Bulgaria Central African Republic China China China China China China China China China China China China Democratic Rep. of Congo Greece Iraq Kosovo Kosovo Mauritania Nigeria Oman Pakistan Pakistan Pakistan Russia Russia Russia Russia Russia Russia Russia South Africa South Africa South Africa South Africa South Africa South Africa Senegal Senegal Tajikistan Tajikistan Tajikistan Turkey Uganda Uzbekistan Uzbekistan Uzbekistan
1978 1975 1966 1970 1975 1978 1979 1984 1988 1968 2005 2005 1968 2005 1956 1975 1979 2001 2001 1984 1966 1997 1965 1976 2000 2002 1967 1967 2000 2000 2000 2003 1987 1981 1981 1985 1984 1985 1972 1969 1990 1990 1991 2003 1958 1967 1985 2002
AY277676 EF123122 AJ010648 AF415236 AF362080 AF354296 AF358784 AJ010649 AY029157 DQ211642 DQ227496 DQ227495 U88413 DQ217602 DQ211650 DQ211638 AJ538196
3. Results 3.1. Phylogeny Bayesian phylogenetic analyses recovered highly supported tree topologies consistent with those previously generated using a maximum likelihood algorithm (Hewson et al., 2004a; Chamberlain et al., 2005; Deyde et al., 2006). Genetic diversity was relatively high with 20%, 31%, and 23% nucleotide variation for the S, M, and L segments, respectively. Following the classification scheme of Deyde et al. (2006), six main lineages were recovered from S segment analysis (Fig. 2): I, West Africa 1; II, Democratic Republic of the Congo (DRC); III, South Africa and West Africa 2; IV, Asia and the Middle East; V, Europe and Turkey; and VI, Greece. Interestingly, a single sample from the Central African Republic grouped with the lineage IV viruses from the Middle East. The Bayesian tree generated from the M segment (Fig. 3) showed a signature typical of reassortment, and the clades described above, based on the S segment analysis, failed to cluster together. Instead, viruses from each lineage were scattered throughout the tree with the exception of those in lineage V (Russia, Turkey, and Kosovo).
DQ133507 DQ211641 U88410 DQ211645 U88414 AF527810 AJ538198 DQ211643 DQ211644 AF481802 AY277672 DQ206447 DQ211647 DQ076415 DQ076416 DQ211648 DQ211646 DQ211640 DQ211639 AY049083 AY297692 AY297691 DQ211649 DQ076413 AY223475 AF481799
M
L
AB069669 AB069670 AB069671 AB069672 AB069673 AB069674 AB069675 DQ211629
DQ211616
AY900145 DQ211637 DQ211625 AJ538197 AY675511 EU037902 DQ211628 AF467768 DQ211632
DQ211624 DQ211612 AY947890
AF467769 AJ538199
AY422208
DQ211630 DQ211631
AY179961 DQ206448 DQ211634 DQ157174 DQ211635 AY900142 DQ211633 DQ211627 DQ211626 AY179962
DQ211636
EU044832 DQ211615 AY389508 DQ211619
AY675240 DQ211617 DQ211618
AY995166 DQ211621 DQ076414 DQ076417 DQ211622 DQ211620 DQ211614 DQ211613 AY720893
DQ211623 DQ076412
AY223476 AY900144
Similar to the findings of Deyde et al. (2006), a seventh lineage, Mauritania, could be detected based only on the M segment analysis. The tree topology based on the L segment analysis (Fig. 4) was similar to that generated by the S segment, with the exception of the group I (West Africa 1) viruses from Senegal. Here, the viruses clustered within group III (South Africa and West Africa 2) as opposed to representing a separate lineage in the S segment analysis. 3.2. Evolutionary rates and time to MRCA Mean molecular evolutionary rate estimates and 95% highest posterior density (HPD) intervals calculated from the Bayesian coalescent analyses were S = 1.09 10 4 (0.17 10 4–2.09 10 4), M = 1.52 10 4 (0.62 10 4–2.40 10 4) and L = 0.58 10 4 (0.15 10 4–1.03 10 4) nucleotide substitutions/site/year. Given that reassortment is most frequently observed among CCHFV M segments, 6 viruses appearing to be reassortants were removed from the dataset and the analysis was repeated. The mean evolutionary rate (1.71 10 4 substitutions/site/year; 95% HPD = 0.81 10 4–
1106
S.A. Carroll et al. / Molecular Phylogenetics and Evolution 55 (2010) 1103–1110
Fig. 2. Bayesian coalescent analysis of CCHF viruses based on the S segment. Maximum clade credibility tree is shown with time to MRCA estimates (bold text) and posterior probability values depicted at the nodes. Virus lineages are labeled with roman numerals: I, West Africa 1; II, Democratic Republic of the Congo (DRC); III, South Africa and West Africa 2; IV, Asia and the Middle East; V, Europe and Turkey; and VI, Greece.
2.58 10 4) was similar to that obtained from the original dataset including the reassortant viruses and established that the analysis was not biased by their inclusion. An analysis was also conducted to investigate the effect of the highly variable mucin domain on the overall substitution rate of the M segment. Although a lower evolutionary rate might be expected in the absence of the domain, a similar value was obtained when it was excluded from the analysis (1.43 10 4 substitutions/site/year; 95% HPD = 0.76 10 4–2.11 10 4) and indicated no discernable effect on the overall substitution rate for the whole M segment. Given the difference in S and L segment rate estimates, and to investigate the possible effects of taxon sampling, the most densely sampled dataset (S segment) was pared down to match the least densely sampled dataset (L segment). This resulted in an S segment dataset consisting of 23 sequences for reanalysis. Nearly all of the excluded sequences (18 of 20) were from lineage IV (Asia and the Middle East), with most of these samples originating from China. The estimated nucleotide substitution rate shifted from 1.09 10 4 substitutions/site/year to 0.58 10 4 substitutions/ site/year (95% HPD = 6.21 10 8–1.55 10 4). This rate was identical to that obtained when analyzing the less densely sampled L segment dataset (0.58 10 4) and demonstrated a significant effect of taxon sampling on the molecular evolutionary rate estimate. In addition, the large spread in the 95% HPD indicated that taxon sampling can influence the amount of uncertainty in the data. The time to MRCA of the viruses identified today as CCHFV is estimated to have occurred 3138 (95% HPD = 710–6911) years
before present based the S segment analysis and 3560 (95% HPD = 1604–6216) years before present based on the M segment analysis (Figs. 2 and 3). The estimate obtained from the L segment analysis was older at 7358 (95% HPD = 2435–14801) years before present and was likely biased by taxon sampling issues (Fig. 4). Individual virus lineages shared ancestral genotypes between 134 and 1177 years before present based on the S segment and between 517 and 1849 years before present based on the L segment. Of the lineages with multiple sequence representatives, lineage IV (Asia and the Middle East) can be traced back farthest in time. Lineage VI, which consists of the Greek sample AP92, occupies a basal position on the tree and appears to represent one of the earliest offshoots of CCHFV. 4. Discussion Similar time to MRCA estimates were obtained by the S and M segment analyses, with the L segment appearing to evolve more slowly and resulting in a significantly older time to MRCA. This was surprising given previous suggestions that the CCHFV S and L segments coevolve (Chamberlain et al., 2005). In fact, an initial hypothesis was that the M segment analysis would produce an inconsistent evolutionary rate and time to MRCA due to the inclusion of the highly variable mucin domain. Anagnostou and Papa (2009) recently examined the evolution of CCHFVs and indicated that the mucin domain evolved 2.3 times faster than the conserved region of the M segment. In the current study, however, removal of
S.A. Carroll et al. / Molecular Phylogenetics and Evolution 55 (2010) 1103–1110
1107
Fig. 3. Bayesian coalescent analysis of CCHF viruses based on the M segment. Maximum clade credibility tree is shown with time to MRCA estimates (bold text) and posterior probability values depicted at the nodes. Virus lineages are labeled with roman numerals as in the S segment analysis, however, reassortment is evident with this analysis: I, West Africa 1; II, Democratic Republic of the Congo (DRC); III, South Africa and West Africa 2; IV, Asia and the Middle East; V, Europe and Turkey; VI, Greece; and VII, Mauritania.
the mucin domain resulted in similar nucleotide substitution rates (exclusion of mucin domain = 1.43 10 4 substitutions/site/year; inclusion of mucin domain = 1.52 10 4 substitutions/site/year). The difference between the two studies is likely due to the methodologies that were employed. Anagnostou and Papa (2009) utilized maximum likelihood rather than Bayesian inference and a constant molecular clock for their analysis. Conversely, results from this study indicated that a relaxed clock model was more appropriate for the data. A relaxed clock can account for potential rate variation within a segment or among lineages and explains why the overall rate estimation was not greatly influenced by the inclusion of the mucin domain. Additional inquiry into the disparity between S and L segment estimates led to the finding that taxon sampling can influence evolutionary rate approximations in the same way that it is known to affect phylogenetic reconstruction (Hillis, 1996; Pollock et al., 2002; Zwickl and Hillis, 2002). The calculated rates of evolution of 1.09 10 4, 1.52 10 4 and 0.58 10 4 substitutions/site/year for the S, M and L genome RNA segments of CCHFV, a tick-borne member of the family Bunyaviridae, were similar to the rates of 3.9 10 4, 3.6 10 4 and 2.8 10 4 substitutions/site/year obtained from a corresponding analysis of Rift Valley fever virus (RVFV), a mosquito-borne member of the same family (Bird et al., 2008). Comparison of the estimates for the virus S and M segments (excluding the CCHFV L segment on the basis of insufficient taxon sampling) would
indicate that RVFV is evolving approximately 2–4-fold faster than CCHFV. This could be considered consistent with the general view that mosquito-borne RNA viruses evolve more rapidly than tickborne viruses within the same family (Zanotto et al., 1996; Gould and Solomon, 2008). However, the differences in rates reported here are rather modest and may not be noteworthy. Similar studies of RNA viruses in the family Flaviviridae have reported the molecular evolutionary rate of the tick-borne Kyasanur Forest disease virus (KFDV) as 6.4 10 4 substitutions/site/year (Mehla et al., 2009) whereas somewhat lower rates have been reported for yellow fever (4.2 10 4 substitutions/site/year; Bryant et al., 2007), and Saint Louis encephalitis (2.17 10 4 substitutions/site/year; Baillie et al., 2008), both mosquito-borne viruses. The inclusion of data from both the Flaviviridae and Bunyaviridae families provides comparative examples of RNA viruses with either positive or negative sense genomes, and with either non-segmented or segmented genomes. Taken together, these data would indicate that no clear, general pattern exists with regard to RNA virus evolution in tick versus mosquito vectors. It is unclear why CCHF was only first recognized in the 1940s. In all instances, Bayesian coalescent analysis clearly demonstrated that CCHF represents an old virus that has been circulating for centuries. In fact, a disease consistent with CCHF, linked to a louse or tick that parasitized blackbirds, was described in the literature as early as the 12th century by a physician in Tajikistan (Hoogstraal,
1108
S.A. Carroll et al. / Molecular Phylogenetics and Evolution 55 (2010) 1103–1110
Fig. 4. Bayesian coalescent analysis of CCHF viruses based on the L segment. Maximum clade credibility tree is shown with time to MRCA estimates (bold text) and posterior probability values depicted at the nodes. Virus lineages are labeled with roman numerals as in the S segment analysis, but lineage I samples group within lineage III in this analysis: I, West Africa 1; II, Democratic Republic of the Congo (DRC); III, South Africa and West Africa 2; IV, Asia and the Middle East; V, Europe and Turkey; and VI, Greece.
1979; Whitehouse, 2004; Ergonul, 2006). In Central Asia, hemorrhagic diseases compatible with CCHF were known for centuries by a variety of names (Chumakov, 1974; Chumakov et al., 1976; Hoogstraal, 1979; Whitehouse, 2004), and a disease consistent with CCHF described as a ‘‘haemorrhagic purpura [due] to the toxic effect of tick bites” was identified from a forest officer on safari in Tanzania some 20 + years before its initial recognition (Parry, 1925). Our analysis indicated that CCHFV, or a similar virus that gave rise to what is currently recognized as CCHFV, was in circulation as early as 1500 to 1100 BC. This is in striking contrast to the mosquito-borne RVFV, which while having a slightly faster rate of evolution was found to have a much more recent common ancestor. RVFV was first isolated in 1930 (Findlay and Daubney, 1931) but was suspected to have occurred in Africa since the early 1900s (Kabete Veterinary Laboratories, 1910). Similar Bayesian analyses estimated the MRCA of extant RVFVs to have existed sometime in the late 1800s (Bird et al., 2007, 2008). It was suggested that changes in agricultural practices and the importation of European cattle and sheep around this time may have led to conditions favorable for the emergence of RVFV. Why might CCHFV have a much more ancient common ancestor than RVFV? As others have noted, with the exception of Madagascar, the range of CCHFV falls completely within the distribution of its principal vector, ticks in the genus Hyalomma. (Hoogstraal, 1979; Swanepoel et al., 1987; Whitehouse, 2004; Burt and Swane-
poel, 2005). Hyalomma ticks evolved from an ancestor that lived in the Oriental region in the early Miocene, approximately 19 million years ago, with dispersal into Africa likely around 14 million years ago (Balashov, 1994; Murrell et al., 2001; Barker and Murrell, 2002). These dates indicate that the ticks are significantly older than the MRCA of CCHFV, and they would have already been widespread by the time that the ancestral CCHFV emerged. It is interesting to speculate that with ticks already in place, changes in agricultural practices and climate may have contributed to the initial emergence of a CCHF-like virus. Sheep, goats, and cattle were domesticated in Eurasia prior to 6000 BC (Christian, 1998) and appeared in Africa shortly thereafter (Maddox, 2006). Between 4000 and 2000 BC, agricultural and livestock rearing communities expanded along with an overall increase in human population density. Around that same time (approximately 2500 BC), warmer and drier climatic conditions began to occur throughout the Old World (Bell et al., 2001; Maddox, 2006). As new diseases affected the humans and their livestock, likely due to the alteration of the landscape as well as close contact with animals, pastoralists tried to adapt. While cattle herdsmen moved to drier lands to avoid tsetse flies (Maddox, 2006), they may have inadvertently contributed to creating favorable conditions for CCHFV emergence as drier conditions are preferred by Hyalomma ticks. In fact, their abundance can be tied to both warm temperatures and high levels of aridity (Linthicum and Bailey, 1994). Recently, Randolph and Rogers
S.A. Carroll et al. / Molecular Phylogenetics and Evolution 55 (2010) 1103–1110
(2007) found that the top predictor variable for occurrence of CCHF was the minimum land surface temperature, with the presence of CCHF cases consistently associated with warmer temperatures. Despite the more detailed understanding of the genetic diversity and evolutionary history of CCHFV, it is difficult to interpret how the distribution of CCHFV may have changed over time. The broad geographic distribution of some viral clusters (e.g. lineage III in South Africa, Mauritania and Nigeria), and the finding of reassortment between viruses of different lineages, suggests that CCHFV can be translocated over considerable distances. Some have speculated that migrating birds could have dispersed CCHFV-infected ticks (Hoogstraal, 1979; Hewson et al., 2004a) as has been hypothesized for KFDV (Mehla et al., 2009); however, no direct evidence of this has been demonstrated for CCHFV (Randolph and Rogers, 2007). Tick movement could also be due to attachment to wild animals, livestock, or even humans (El-Azazy and Scrimgeour, 1997; Khan et al., 1997; Rodriguez et al., 1997; Yashina et al., 2003; Hewson et al., 2004a; Chamberlain et al., 2005). Regardless, the long evolutionary history of CCHFV and Hyalomma ticks has provided ample opportunity for expansion to their current geographic ranges throughout Africa, Europe, the Middle East, and Asia. More recently, dramatic increases in unregulated wildlife trade, livestock import and export, and global movement of humans are likely to be creating additional opportunities for spread of this severe disease. This potential is clearly illustrated by the report of ticks infesting 97 shipments, containing 54,376 animals, imported into the United States in 1995 alone (Karesh et al., 2007). Increased surveillance and livestock inspection/quarantine procedures should be considered to help prevent the establishment of ticks and their associated pathogens, especially in previously CCHFV-free countries. Acknowledgments The authors thank R. Ryan Lash for providing the distribution map and Cesar G. Albarino for assistance with the final versions of the figures. The findings and conclusions reported here do not necessarily represent the views of the Centers for Disease Control and Prevention or the Department of Health and Human Services. References Anagnostou, V., Papa, A., 2009. Evolution of Crimean-Congo hemorrhagic fever virus. Infect. Genet. Evol. 9, 948–954. Baillie, G.J., Kolokotronis, S.O., Waltari, E., Maffei, J.G., Kramer, L.D., Perkins, S.L., 2008. Phylogenetic and evolutionary analyses of St. Louis encephalitis virus genomes. Mol. Phylogenet. Evol. 47, 717–728. Balashov, Y.S., 1994. Importance of continental drift in the distribution and evolution of ixodid ticks. Entomol. Rev. 73, 42–50. Barker, S.C., Murrell, A., 2002. Phylogeny, evolution, and historical zoogeography of ticks: a review of recent progress. Exp. Appl. Acarol. 28, 55–68. Bell, P.A., Greene, T., Fisher, J., Baum, A.S., 2001. Weather, climate, and behavior. In: Environmental Psychology, fifth ed. Harcourt College, Ft. Worth, pp. 167–204. Bird, B.H., Githinji, J.W.K., Macharia, J.M., Kasiiti, J.L., Muriithi, R.M., Gacheru, S.G., Musaa, J.O., Towner, J.S., Reeder, S.A., Oliver, J.B., Stevens, T.L., Erickson, B.R., Morgan, L.T., Khristova, M.L., Hartman, A.L., Comer, J.A., Rollin, P.E., Ksiazek, T.G., Nichol, S.T., 2008. Multiple virus lineages sharing recent common ancestry were associated with a large Rift Valley fever outbreak among livestock in Kenya during 2006–2007. J. Virol. 82, 11152–11166. Bird, B.H., Khristova, M.L., Rollin, P.E., Ksiazek, T.G., Nichol, S.T., 2007. Complete genome analysis of 33 ecologically and biologically diverse Rift Valley fever virus strains reveals widespread virus movement and low genetic diversity due to recent common ancestry. J. Virol. 81, 2805–2816. Bryant, J.E., Holmes, E.C., Barrett, A.D., 2007. Out of Africa: a molecular perspective on the introduction of yellow fever virus into the Americas. PLoS Pathog. 3, e75. doi: 10.1371/journal.ppat.0030075. Burney, M.I., Ghafoor, A., Saleen, M., Webb, P.A., Casals, J., 1980. Nosocomial outbreak of viral hemorrhagic fever caused by Crimean hemorrhagic feverCongo virus in Pakistan, January 1976. Am. J. Trop. Med. Hyg. 29, 941–947. Burt, F.J., Paweska, J.T., Ashkettle, B., Swanepoel, R., 2009. Genetic relationship in southern African Crimean-Congo haemorrhagic fever virus isolates: evidence for occurrence of reassortment. Epidemiol. Infect. 137, 1302–1308.
1109
Burt, F.J., Swanepoel, R., 2005. Molecular epidemiology of African and Asian Crimean-Congo haemorrhagic fever isolates. Epidemiol. Infect. 133, 659– 666. Chamberlain, J., Cook, N., Lloyd, G., Mioulet, V., Tolley, H., Hewson, R., 2005. Coevolutionary patterns of variation in small and large RNA segments of CrimeanCongo hemorrhagic fever virus. J. Gen. Virol. 86, 3337–3341. Chare, E.R., Gould, A., Holmes, E.C., 2003. Phylogenetic analysis reveals a low rate of homologous recombination in negative-sense RNA viruses. J. Gen. Virol. 84, 2691–2703. Christian, D., 1998. A History of Russia, Central Asia, and Mongolia, vol. I: Inner Eurasia from Prehistory to the Mongol Empire. Blackwell, Malden. Chumakov, M.P., 1945. A new tick-borne virus disease—Crimean hemorrhagic fever. In: Sokolov, A.A., Chumakov, M.P., Kolachev, A.A. (Eds.), Crimean Hemorrhagic Fever (Acute Infectious Capillary Toxicosis). Izd. Otd. Primorskoi Armii, Simferopol, Moscow, pp. 13–45. Chumakov, M.P., 1974. On 30 years of investigation of Crimean hemorrhagic fever. Tr. Inst. Polio Virusn. Entsefalitov Akad. Med. Nauk SSSR 22, 5–18. Chumakov, M.P., Smirnova, S.E., Shalunova, N.Y., Mart’yanova, L.I., Fleer, G.P., Zgurskaya, G.N., Maksumov, S.S., Kasymov, K.Y., Pak, T.P., 1976. Proofs of etiological identity to Crimean hemorrhagic fever in Central Asian hemorrhagic fever. In: IX International Congress on Tropical Medicine and Malaria, Athens, vol. 1, pp. 33–34. Deyde, V.M., Khristova, M.L., Rollin, P.E., Ksiazek, T.G., Nichol, S.T., 2006. CrimeanCongo hemorrhagic fever virus genomics and global diversity. J. Virol. 80, 8834– 8842. Drummond, A.J., Ho, S.Y.W., Phillips, M.J., Rambaut, A., 2006. Relaxed phylogenetics and dating with confidence. PLoS Biol. 4, e88. doi: 10.1371/ journal.pbio.0040088. Drummond, A.J., Rambaut, A., 2007. BEAST: Bayesian evolutionary analysis by sampling trees. BMC Evol. Biol. 7, 214. El-Azazy, O.M., Scrimgeour, E.M., 1997. Crimean-Congo haemorrhagic fever virus infection in the western province of Saudi Arabia. Trans. R. Soc. Trop. Med. Hyg. 91, 275–278. Ergonul, O., 2006. Crimean-Congo haemorrhagic fever. Lancet Infect. Dis. 6, 203– 214. Findlay, G.M., Daubney, R., 1931. The virus of Rift Valley fever or enzootic hepatitis. Lancet ii, 1350–1351. Galtier, N., Gouy, M., Gautier, C., 1996. SeaView and Phylo_win, two graphic tools for sequence alignment and molecular phylogeny. Comput. Appl. Biosci. 12, 543–548. Gould, E.A., Solomon, T., 2008. Pathogenic flaviviruses. Lancet 371, 500–509. Gurbuz, Y., Sencan, I., Ozturk, B., Tutuncu, E., 2009. A case of nosocomial transmission of Crimean-Congo hemorrhagic fever from patient to patient. Int. J. Infect. Dis. 13, e105–e107. Hewson, R., Chamberlain, J., Mioulet, V., Lloyd, G., Jamil, B., Hasan, R., Gmyl, A., Gmyl, L., Smirnova, S.E., Lukashev, A., Karganova, G., Clegg, C., 2004a. CrimeanCongo haemorrhagic fever virus: sequence analysis of the small RNA segments from a collection of viruses world wide. Virus Res. 102, 185–189. Hewson, R., Gmyl, A., Gmyl, L., Smirnova, S.E., Karganova, G., Jamil, B., Hasan, R., Chamberlain, J., Clegg, C., 2004b. Evidence of segment reassortment in CrimeanCongo haemorrhagic fever virus. J. Gen. Virol. 85, 3059–3070. Hillis, D.M., 1996. Inferring complex phylogenies. Nature 383, 130–131. Hoogstraal, H., 1979. The epidemiology of tick-borne Crimean-Congo haemorrhagic fever in Asia, Europe and Africa. J. Med. Entomol. 15, 307–417. Kabete Veterinary Laboratories, 1910. Diseases of sheep. Kenya Veterinary Department Annual Report—1910. Kenya Dept. of Vet. Services, Nairobi, Kenya. Karesh, W.B., Cook, R.A., Gilbert, M., Newcomb, J., 2007. Implications of wildlife trade on the movement of avian influenza and other infectious diseases. J. Wildl. Dis. 43, 55–59. Katoh, K., Kuma, K., Toh, H., Miyata, T., 2005. MAFFT Version 5: improvement in accuracy of multiple sequence alignment. Nucleic Acids Res. 33, 511–518. Khan, A.S., Maupin, G.O., Rollin, P.E., Noor, A.M., Shurie, H.H., Shalabi, A.G., Wasef, S., Haddad, Y.M., Sadek, R., Ijaz, K., Peters, C.J., Ksiazek, T.G., 1997. An outbreak of Crimean-Congo hemorrhagic fever in the United Arab Emirates, 1994–1995. Am. J. Trop. Med. Hyg. 57, 519–525. Linthicum, K.J., Bailey, C.L., 1994. Ecology of Crimean-Congo hemorrhagic fever. In: Sonenshine, D.E., Mather, T.N. (Eds.), Ecological Dynamics of Tick-Borne Zoonoses. Oxford Univ. Press, New York, pp. 392–437. Lukashev, A.N., 2005. Evidence for recombination in Crimean-Congo hemorrhagic fever virus. J. Gen. Virol. 86, 2333–2338. Maddox, G., 2006. Sub-Saharan Africa: An Environmental History. ABC-CLIO, Santa Barbara. Mehla, R., Kumar, S.R.P., Yadav, P., Barde, P.V., Yergolkar, P.N., Erickson, B.R., Carroll, S.A., Mishra, A.C., Nichol, S.T., Mourya, D.T., 2009. Recent ancestry of Kyasanur Forest disease virus. Emerging Infect. Dis. 15, 1431–1437. Murrell, A., Campbell, N.J.H., Barker, S.C., 2001. A total-evidence phylogeny of ticks provides insights into the evolution of life cycles and biogeography. Mol. Phylogenet. Evol. 21, 244–258. Papa, A., Bozovic, B., Pavlidou, V., Papadimitriou, E., Pelemis, M., Antoniadis, A., 2002. Genetic detection and isolation of Crimean-Congo hemorrhagic fever virus, Kosovo, Yugoslavia. Emerging Infect. Dis. 8, 852–854. Papa, A., Papadimitriou, E., Bozovic, B., Antoniadis, A., 2005. Genetic characterization of the M RNA segment of a Balkan Crimean-Congo hemorrhagic fever virus strain. J. Med. Virol. 75, 466–469. Parry, J.H., 1925. Haemorrhagic purpura due to tick bites. Tanganyika Territory, Annual Medical Report for the Year Ending 31st December, 1925, 101–102.
1110
S.A. Carroll et al. / Molecular Phylogenetics and Evolution 55 (2010) 1103–1110
Pollock, D.D., Zwickl, D.J., McGuire, J.A., Hillis, D.M., 2002. Increased taxon sampling is advantageous for phylogenetic inference. Syst. Biol. 51, 664–671. Posada, D., Crandall, K.A., 1998. MODELTEST: testing the model of DNA substitution. Bioinformatics 14, 817–818. Randolph, S.E., Rogers, D.J., 2007. Ecology of tick-borne disease and the role of climate. In: Ergonul, O., Whitehouse, C.A. (Eds.), Crimean-Congo Hemorrhagic Fever: A Global Perspective. Springer, Dordrecht, pp. 167–186. Rodriguez, L.L., Maupin, G.O., Ksiazek, T.G., Rollin, P.E., Khan, A.S., Schwarz, T.F., Lofts, R.S., Smith, J.F., Noor, A.M., Peters, C.J., Nichol, S.T., 1997. Molecular investigation of a multisource outbreak of Crimean-Congo hemorrhagic fever in the United Arab Emirates. Am. J. Trop. Med. Hyg. 57, 512–518. Sanchez, A.J., Vincent, M.J., Nichol, S.T., 2002. Characterization of the glycoproteins of Crimean-Congo hemorrhagic fever virus. J. Virol. 76, 7263–7275. Schmaljohn, C.S., Nichol, S.T., 2007. Bunyaviridae. In: Knipe, D.M., Howley, P.M. (Eds.), Fields Virology, fifth ed. Lippincott, Williams, and Wilkins, Philadelphia, pp. 1741–1789. Simpson, D.I.H., Knight, E.M., Courtois, G., Williams, M.C., Weinbren, M.P., Kibukamusoke, J.W., 1967. Congo virus: a hitherto undescribed virus occurring in Africa, Part 1: human isolations-clinical notes. E. Afr. Med. J. 44, 86–92. Suchard, M.A., Weiss, R.E., Sinsheimer, J.S., 2001. Bayesian selection of continuoustime Markov chain evolutionary models. Mol. Biol. Evol. 18, 1001–1013. Swanepoel, R., Shepherd, A.J., Leman, P.A., Shepherd, S.P., McGillivray, G.M., Erasmus, M.J., Searle, L.A., Gill, D.E., 1987. Epidemiologic and clinical features
of Crimean-Congo hemorrhagic fever in southern Africa. Am. J. Trop. Med. Hyg. 36, 120–132. Swanepoel, R., Struthers, J.K., Shepherd, A.J., McGillivray, G.M., Nel, M.J., Jupp, P.G., 1983. Crimean-Congo hemorrhagic fever in South Africa. Am. J. Trop. Med. Hyg. 32, 1407–1415. Weaver, S.C., 2006. Evolutionary influences in arboviral disease. Curr. Top. Microbiol. Immunol. 299, 285–314. Whitehouse, C.A., 2004. Crimean-Congo hemorrhagic fever. Antiviral Res. 64, 145– 160. Woodall, J.P., Williams, M.C., Simpson, D.I.H., Ardoin, P., Lule, M., West, R., 1965. The Congo group of agents. East Afr. Virus Res. Inst. Rep. 14, 34–36. Woodall, J.P., Williams, M.C., Simpson, D.I.H., 1967. Congo virus: a hitherto undescribed virus occurring in Africa, Part 2: identification studies. East Afr. Med. J. 44, 93–98. Yashina, L., Vyshemirskii, O., Seregin, S., Petrova, I., Samokhvalov, E., Lvov, D., Gutorov, V., Kuzina, I., Tyunnikov, G., Tang, Y.-W., Netesov, S., Petrov, V., 2003. Genetic analysis of Crimean-Congo hemorrhagic fever virus in Russia. J. Clin. Microbiol. 41, 860–862. Zanotto, P.M. de A., Gould, E.A., Gao, G.F., Harvey, P.H., Holmes, E.C., 1996. Population dynamics of flaviviruses revealed by molecular phylogenies. Proc. Natl. Acad. Sci. USA 93, 548–553. Zwickl, D.J., Hillis, D.M., 2002. Increased taxon sampling greatly reduces phylogenetic error. Syst. Biol. 51, 588–598.
Contents lists available at ScienceDirect
Molecular Phylogenetics and Evolution journal homepage: www.elsevier.com/locate/ympev
Ancient common ancestry of Crimean-Congo hemorrhagic fever virus Serena A. Carroll, Brian H. Bird, Pierre E. Rollin, Stuart T. Nichol * Special Pathogens Branch, Division of Viral and Rickettsial Diseases, Centers for Disease Control and Prevention, 1600 Clifton Road, MS G-14, Atlanta, GA 30333, USA
a r t i c l e
i n f o
Article history: Received 18 September 2009 Revised 31 December 2009 Accepted 6 January 2010 Available online 13 January 2010 Keywords: Crimean-Congo hemorrhagic fever CCHF Bunyaviridae Molecular evolution Bayesian Coalescent
a b s t r a c t Crimean-Congo hemorrhagic fever (CCHF) is a tick-borne RNA virus responsible for outbreaks of severe hemorrhagic fever in humans. Although CCHF was first detected in the 1940s, high levels of genomic diversity argue against a recent origin. Here, Bayesian coalescent analyses were used to estimate the rate of evolution and relative age of the virus. A total of 43 S, 34 M, and 23 L segment sequences from samples collected between 1956 and 2005 were analyzed from across the broad geographic range of the virus. Using a relaxed molecular clock model, nucleotide substitutions were estimated to have occurred at a rate of 1.09 10 4, 1.52 10 4, and 0.58 10 4 substitutions/site/year for the S, M, and L segments, respectively. The most recent common ancestor of the viruses existed approximately 3100–3500 years before present, or around 1500–1100 BC. Changes in agricultural practices and climate occurring near the time of the most recent common ancestor of CCHFV may have contributed to its emergence and spread. Published by Elsevier Inc.
1. Introduction Crimean-Congo hemorrhagic fever virus (CCHFV), genus Nairovirus, family Bunyaviridae, is responsible for outbreaks of severe hemorrhagic fever in humans with case fatalities as high as 30% or more in some instances (Swanepoel et al., 1987; Whitehouse, 2004; Ergonul, 2006; Schmaljohn and Nichol, 2007). The disease was initially recognized by Russian scientists in the 1940s (Chumakov, 1945), and the virus was first isolated from the Democratic Republic of Congo some years later (Woodall et al., 1965, 1967; Simpson et al., 1967). CCHFV is now known to be distributed throughout broad regions of Africa, Europe, the Middle East, and Asia. The virus is transmitted primarily by ticks in the genus Hyalomma; however, several vertebrate and tick species are involved in the natural transmission cycle (Hoogstraal, 1979; Swanepoel et al., 1983, 1987). Human exposures can occur via tick bites, direct contact with blood or tissues of infected animals, person-to-person spread, or by nosocomial infection (Burney et al., 1980; Whitehouse, 2004; Ergonul, 2006; Schmaljohn and Nichol, 2007; Gurbuz et al., 2009). CCHFV is a negative sense, single stranded, tripartite RNA virus with a genome size of approximately 19 kb (Whitehouse, 2004; Schmaljohn and Nichol, 2007). The small (S) segment encodes the nucleocapsid protein whereas the medium (M) segment contains a highly variable region known as the mucin domain (Sanchez et al., 2002) and encodes the envelope glycoproteins Gn and Gc. The large (L) segment encodes the virus RNA-dependent RNA
* Corresponding author. Fax: +1 404 639 1509. E-mail addresses: [email protected] (S.A. Carroll), [email protected] (B.H. Bird), [email protected] (P.E. Rollin), [email protected] (S.T. Nichol). 1055-7903/$ - see front matter Published by Elsevier Inc. doi:10.1016/j.ympev.2010.01.006
polymerase. While arthropod-borne RNA viruses generally show low levels of genetic diversity, perhaps due to a hypothesized ‘‘double-filter” selection mechanism that allows for the maintenance of high fitness in both arthropod and vertebrate hosts (Weaver, 2006), previous studies have demonstrated surprisingly high levels of diversity for CCHFV (Papa et al., 2002, 2005; Hewson et al., 2004a; Chamberlain et al., 2005; Deyde et al., 2006; Burt et al., 2009). Additionally, both reassortment and recombination appear to have contributed to the high amount of genetic variation present in the virus (Chare et al., 2003; Hewson et al., 2004b; Lukashev, 2005; Deyde et al., 2006; Burt et al., 2009). Despite the apparent emergence of CCHFV in the 1940s (Chumakov, 1945), the large amount of sequence diversity found within the species argues against a recent origin. In the past few years, advances in Bayesian coalescent phylogenetic analyses have resulted in more sophisticated methods for determining the relative ages of emerging pathogens (Drummond et al., 2006; Drummond and Rambaut, 2007). The purpose of this study was to apply these tools to address the question of CCHFV ancestral origins. If the most recent common ancestor (MRCA) of the extant CCHFVs existed in the relatively recent past (i.e. near the initial date of disease recognition) then it could be inferred that the observed high viral genomic diversity is a result of rapid evolution. Alternatively, a more ancient origin would suggest that the virus has had sufficient time to generate high levels of genomic diversity without the need for a rapidly evolving virus population. A secondary goal of the study was to compare the molecular evolutionary rate of this tick-borne bunyavirus with that of a mosquito-borne bunyavirus (i.e. Rift Valley fever virus, RVFV) and with other unrelated tick and mosquito-borne RNA viruses.
1104
S.A. Carroll et al. / Molecular Phylogenetics and Evolution 55 (2010) 1103–1110
of 10,000,000 generations. An analysis of marginal likelihoods (Suchard et al., 2001) indicated that the relaxed uncorrelated lognormal clock (Drummond et al., 2006) and constant population size model was most appropriate for each of the three segments (log10 Bayes Factors > 2 in all cases). Final Bayesian coalescent analyses used the GTR + G substitution model, a relaxed uncorrelated lognormal clock, and a constant population size. The S and L segment datasets were run for 20,000,000 generations whereas the M segment dataset was run for 115,000,000 generations to ensure effective sample sizes (ESSs) of at least 200. Maximum clade credibility trees were summarized using TreeAnnotator and depicted using FigTree (Drummond and Rambaut, 2007). Further comparative analyses were conducted to investigate sources that might have biased the estimated rates and/or time to MRCAs. Given that reassortment can potentially affect coalescent analysis, all possible reassortant viruses were removed from the datasets. In addition, the highly variable mucin domain was removed from the M segment analysis to determine its influence on the overall rate estimate. Finally, to investigate the effects of taxon sampling, the most densely sampled dataset (S segment) was pared down to match the least densely sampled dataset (L segment). These additional analyses followed the same methods as described above.
2. Materials and methods CCHFV sequences, representing the broad geographic range of the virus, were obtained from GenBank (Fig. 1); only whole S, M, or L segment sequences with associated dates of collection were retrieved (Table 1). A total of 43 S segment sequences spanning the years 1956–2005, 34 M segment sequences from 1956–2003, and 23 L segment sequences from 1956–2003 were examined. Multiple sequence alignments were generated using the MAFFT function (Katoh et al., 2005) in SeaView (Galtier et al., 1996). ModelTest version 3.7 (Posada and Crandall, 1998) examined 56 possible nucleotide substitution models and was used to select the most appropriate model for each dataset based on Akaike’s Information Criterion (AIC). The General Time Reversible model with a gamma distribution (GTR + G) was selected for all three segments. Rates of evolution and estimates of the time to MRCA were generated via Bayesian coalescent phylogenetic analysis using the BEAST software package version 1.4.7 and Tracer version 1.4 (Drummond and Rambaut, 2007). Preliminary analyses were conducted to determine which clock (strict versus relaxed) and demographic (constant versus Bayesian Skyline population size) models were most appropriate for the datasets, and runs initially consisted
Russia
Kosovo Bulgaria Greece
Uzbekistan
Turkey
Tajikistan
China
Iraq Pakistan
Oman
Mauritania Senegal
Nigeria Cen. Af. Rep.
Democratic Republic of Congo
Uganda
South Africa
Fig. 1. Geographic distribution of CCHF virus samples analyzed in this study, originating from 18 countries in Africa, Europe, Asia, and the Middle East.
1105
S.A. Carroll et al. / Molecular Phylogenetics and Evolution 55 (2010) 1103–1110 Table 1 CCHF viruses examined, with associated countries of origin, collection dates, and GenBank accession numbers. Sample
Country
Date
S
BulgariaHu517 CentralAfricanRepBT958 China66019 China7001 China75024 China7803 China79121 China8402 China88166 ChinaC68031 ChinaTI05146 ChinaTI05099 ChinaHY13 ChinaTI05035 CongoUG3010 GreeceAP92 IraqBaghdad12 Kosovo9553 KosovoHoti MauritaniaArD39554 NigeriaIbAr10200 Oman PakistanJD206 PakistanMatin PakistanSR3 RussiaASTR30908 RussiaDrosdov RussiaKashmanov RussiaSTVHU29223 RussiaTI28044 RussiaVLGTI29414 RussiaVLV100 SouthAfricaSPU10387 SouthAfricaSPU128817 SouthAfricaSPU481 SouthAfricaSPU41585 SouthAfricaSPU4184 SouthAfricaSPU9785 SenegalArD15786 SenegalArD8194 TajikistanHU8966 TajikistanHU8975 TajikistanHU8978 Turkey200310849 UgandaSemunya UzbekistanHodzha UzbekistanTI10145 UzbekistanU22002U6415
Bulgaria Central African Republic China China China China China China China China China China China China Democratic Rep. of Congo Greece Iraq Kosovo Kosovo Mauritania Nigeria Oman Pakistan Pakistan Pakistan Russia Russia Russia Russia Russia Russia Russia South Africa South Africa South Africa South Africa South Africa South Africa Senegal Senegal Tajikistan Tajikistan Tajikistan Turkey Uganda Uzbekistan Uzbekistan Uzbekistan
1978 1975 1966 1970 1975 1978 1979 1984 1988 1968 2005 2005 1968 2005 1956 1975 1979 2001 2001 1984 1966 1997 1965 1976 2000 2002 1967 1967 2000 2000 2000 2003 1987 1981 1981 1985 1984 1985 1972 1969 1990 1990 1991 2003 1958 1967 1985 2002
AY277676 EF123122 AJ010648 AF415236 AF362080 AF354296 AF358784 AJ010649 AY029157 DQ211642 DQ227496 DQ227495 U88413 DQ217602 DQ211650 DQ211638 AJ538196
3. Results 3.1. Phylogeny Bayesian phylogenetic analyses recovered highly supported tree topologies consistent with those previously generated using a maximum likelihood algorithm (Hewson et al., 2004a; Chamberlain et al., 2005; Deyde et al., 2006). Genetic diversity was relatively high with 20%, 31%, and 23% nucleotide variation for the S, M, and L segments, respectively. Following the classification scheme of Deyde et al. (2006), six main lineages were recovered from S segment analysis (Fig. 2): I, West Africa 1; II, Democratic Republic of the Congo (DRC); III, South Africa and West Africa 2; IV, Asia and the Middle East; V, Europe and Turkey; and VI, Greece. Interestingly, a single sample from the Central African Republic grouped with the lineage IV viruses from the Middle East. The Bayesian tree generated from the M segment (Fig. 3) showed a signature typical of reassortment, and the clades described above, based on the S segment analysis, failed to cluster together. Instead, viruses from each lineage were scattered throughout the tree with the exception of those in lineage V (Russia, Turkey, and Kosovo).
DQ133507 DQ211641 U88410 DQ211645 U88414 AF527810 AJ538198 DQ211643 DQ211644 AF481802 AY277672 DQ206447 DQ211647 DQ076415 DQ076416 DQ211648 DQ211646 DQ211640 DQ211639 AY049083 AY297692 AY297691 DQ211649 DQ076413 AY223475 AF481799
M
L
AB069669 AB069670 AB069671 AB069672 AB069673 AB069674 AB069675 DQ211629
DQ211616
AY900145 DQ211637 DQ211625 AJ538197 AY675511 EU037902 DQ211628 AF467768 DQ211632
DQ211624 DQ211612 AY947890
AF467769 AJ538199
AY422208
DQ211630 DQ211631
AY179961 DQ206448 DQ211634 DQ157174 DQ211635 AY900142 DQ211633 DQ211627 DQ211626 AY179962
DQ211636
EU044832 DQ211615 AY389508 DQ211619
AY675240 DQ211617 DQ211618
AY995166 DQ211621 DQ076414 DQ076417 DQ211622 DQ211620 DQ211614 DQ211613 AY720893
DQ211623 DQ076412
AY223476 AY900144
Similar to the findings of Deyde et al. (2006), a seventh lineage, Mauritania, could be detected based only on the M segment analysis. The tree topology based on the L segment analysis (Fig. 4) was similar to that generated by the S segment, with the exception of the group I (West Africa 1) viruses from Senegal. Here, the viruses clustered within group III (South Africa and West Africa 2) as opposed to representing a separate lineage in the S segment analysis. 3.2. Evolutionary rates and time to MRCA Mean molecular evolutionary rate estimates and 95% highest posterior density (HPD) intervals calculated from the Bayesian coalescent analyses were S = 1.09 10 4 (0.17 10 4–2.09 10 4), M = 1.52 10 4 (0.62 10 4–2.40 10 4) and L = 0.58 10 4 (0.15 10 4–1.03 10 4) nucleotide substitutions/site/year. Given that reassortment is most frequently observed among CCHFV M segments, 6 viruses appearing to be reassortants were removed from the dataset and the analysis was repeated. The mean evolutionary rate (1.71 10 4 substitutions/site/year; 95% HPD = 0.81 10 4–
1106
S.A. Carroll et al. / Molecular Phylogenetics and Evolution 55 (2010) 1103–1110
Fig. 2. Bayesian coalescent analysis of CCHF viruses based on the S segment. Maximum clade credibility tree is shown with time to MRCA estimates (bold text) and posterior probability values depicted at the nodes. Virus lineages are labeled with roman numerals: I, West Africa 1; II, Democratic Republic of the Congo (DRC); III, South Africa and West Africa 2; IV, Asia and the Middle East; V, Europe and Turkey; and VI, Greece.
2.58 10 4) was similar to that obtained from the original dataset including the reassortant viruses and established that the analysis was not biased by their inclusion. An analysis was also conducted to investigate the effect of the highly variable mucin domain on the overall substitution rate of the M segment. Although a lower evolutionary rate might be expected in the absence of the domain, a similar value was obtained when it was excluded from the analysis (1.43 10 4 substitutions/site/year; 95% HPD = 0.76 10 4–2.11 10 4) and indicated no discernable effect on the overall substitution rate for the whole M segment. Given the difference in S and L segment rate estimates, and to investigate the possible effects of taxon sampling, the most densely sampled dataset (S segment) was pared down to match the least densely sampled dataset (L segment). This resulted in an S segment dataset consisting of 23 sequences for reanalysis. Nearly all of the excluded sequences (18 of 20) were from lineage IV (Asia and the Middle East), with most of these samples originating from China. The estimated nucleotide substitution rate shifted from 1.09 10 4 substitutions/site/year to 0.58 10 4 substitutions/ site/year (95% HPD = 6.21 10 8–1.55 10 4). This rate was identical to that obtained when analyzing the less densely sampled L segment dataset (0.58 10 4) and demonstrated a significant effect of taxon sampling on the molecular evolutionary rate estimate. In addition, the large spread in the 95% HPD indicated that taxon sampling can influence the amount of uncertainty in the data. The time to MRCA of the viruses identified today as CCHFV is estimated to have occurred 3138 (95% HPD = 710–6911) years
before present based the S segment analysis and 3560 (95% HPD = 1604–6216) years before present based on the M segment analysis (Figs. 2 and 3). The estimate obtained from the L segment analysis was older at 7358 (95% HPD = 2435–14801) years before present and was likely biased by taxon sampling issues (Fig. 4). Individual virus lineages shared ancestral genotypes between 134 and 1177 years before present based on the S segment and between 517 and 1849 years before present based on the L segment. Of the lineages with multiple sequence representatives, lineage IV (Asia and the Middle East) can be traced back farthest in time. Lineage VI, which consists of the Greek sample AP92, occupies a basal position on the tree and appears to represent one of the earliest offshoots of CCHFV. 4. Discussion Similar time to MRCA estimates were obtained by the S and M segment analyses, with the L segment appearing to evolve more slowly and resulting in a significantly older time to MRCA. This was surprising given previous suggestions that the CCHFV S and L segments coevolve (Chamberlain et al., 2005). In fact, an initial hypothesis was that the M segment analysis would produce an inconsistent evolutionary rate and time to MRCA due to the inclusion of the highly variable mucin domain. Anagnostou and Papa (2009) recently examined the evolution of CCHFVs and indicated that the mucin domain evolved 2.3 times faster than the conserved region of the M segment. In the current study, however, removal of
S.A. Carroll et al. / Molecular Phylogenetics and Evolution 55 (2010) 1103–1110
1107
Fig. 3. Bayesian coalescent analysis of CCHF viruses based on the M segment. Maximum clade credibility tree is shown with time to MRCA estimates (bold text) and posterior probability values depicted at the nodes. Virus lineages are labeled with roman numerals as in the S segment analysis, however, reassortment is evident with this analysis: I, West Africa 1; II, Democratic Republic of the Congo (DRC); III, South Africa and West Africa 2; IV, Asia and the Middle East; V, Europe and Turkey; VI, Greece; and VII, Mauritania.
the mucin domain resulted in similar nucleotide substitution rates (exclusion of mucin domain = 1.43 10 4 substitutions/site/year; inclusion of mucin domain = 1.52 10 4 substitutions/site/year). The difference between the two studies is likely due to the methodologies that were employed. Anagnostou and Papa (2009) utilized maximum likelihood rather than Bayesian inference and a constant molecular clock for their analysis. Conversely, results from this study indicated that a relaxed clock model was more appropriate for the data. A relaxed clock can account for potential rate variation within a segment or among lineages and explains why the overall rate estimation was not greatly influenced by the inclusion of the mucin domain. Additional inquiry into the disparity between S and L segment estimates led to the finding that taxon sampling can influence evolutionary rate approximations in the same way that it is known to affect phylogenetic reconstruction (Hillis, 1996; Pollock et al., 2002; Zwickl and Hillis, 2002). The calculated rates of evolution of 1.09 10 4, 1.52 10 4 and 0.58 10 4 substitutions/site/year for the S, M and L genome RNA segments of CCHFV, a tick-borne member of the family Bunyaviridae, were similar to the rates of 3.9 10 4, 3.6 10 4 and 2.8 10 4 substitutions/site/year obtained from a corresponding analysis of Rift Valley fever virus (RVFV), a mosquito-borne member of the same family (Bird et al., 2008). Comparison of the estimates for the virus S and M segments (excluding the CCHFV L segment on the basis of insufficient taxon sampling) would
indicate that RVFV is evolving approximately 2–4-fold faster than CCHFV. This could be considered consistent with the general view that mosquito-borne RNA viruses evolve more rapidly than tickborne viruses within the same family (Zanotto et al., 1996; Gould and Solomon, 2008). However, the differences in rates reported here are rather modest and may not be noteworthy. Similar studies of RNA viruses in the family Flaviviridae have reported the molecular evolutionary rate of the tick-borne Kyasanur Forest disease virus (KFDV) as 6.4 10 4 substitutions/site/year (Mehla et al., 2009) whereas somewhat lower rates have been reported for yellow fever (4.2 10 4 substitutions/site/year; Bryant et al., 2007), and Saint Louis encephalitis (2.17 10 4 substitutions/site/year; Baillie et al., 2008), both mosquito-borne viruses. The inclusion of data from both the Flaviviridae and Bunyaviridae families provides comparative examples of RNA viruses with either positive or negative sense genomes, and with either non-segmented or segmented genomes. Taken together, these data would indicate that no clear, general pattern exists with regard to RNA virus evolution in tick versus mosquito vectors. It is unclear why CCHF was only first recognized in the 1940s. In all instances, Bayesian coalescent analysis clearly demonstrated that CCHF represents an old virus that has been circulating for centuries. In fact, a disease consistent with CCHF, linked to a louse or tick that parasitized blackbirds, was described in the literature as early as the 12th century by a physician in Tajikistan (Hoogstraal,
1108
S.A. Carroll et al. / Molecular Phylogenetics and Evolution 55 (2010) 1103–1110
Fig. 4. Bayesian coalescent analysis of CCHF viruses based on the L segment. Maximum clade credibility tree is shown with time to MRCA estimates (bold text) and posterior probability values depicted at the nodes. Virus lineages are labeled with roman numerals as in the S segment analysis, but lineage I samples group within lineage III in this analysis: I, West Africa 1; II, Democratic Republic of the Congo (DRC); III, South Africa and West Africa 2; IV, Asia and the Middle East; V, Europe and Turkey; and VI, Greece.
1979; Whitehouse, 2004; Ergonul, 2006). In Central Asia, hemorrhagic diseases compatible with CCHF were known for centuries by a variety of names (Chumakov, 1974; Chumakov et al., 1976; Hoogstraal, 1979; Whitehouse, 2004), and a disease consistent with CCHF described as a ‘‘haemorrhagic purpura [due] to the toxic effect of tick bites” was identified from a forest officer on safari in Tanzania some 20 + years before its initial recognition (Parry, 1925). Our analysis indicated that CCHFV, or a similar virus that gave rise to what is currently recognized as CCHFV, was in circulation as early as 1500 to 1100 BC. This is in striking contrast to the mosquito-borne RVFV, which while having a slightly faster rate of evolution was found to have a much more recent common ancestor. RVFV was first isolated in 1930 (Findlay and Daubney, 1931) but was suspected to have occurred in Africa since the early 1900s (Kabete Veterinary Laboratories, 1910). Similar Bayesian analyses estimated the MRCA of extant RVFVs to have existed sometime in the late 1800s (Bird et al., 2007, 2008). It was suggested that changes in agricultural practices and the importation of European cattle and sheep around this time may have led to conditions favorable for the emergence of RVFV. Why might CCHFV have a much more ancient common ancestor than RVFV? As others have noted, with the exception of Madagascar, the range of CCHFV falls completely within the distribution of its principal vector, ticks in the genus Hyalomma. (Hoogstraal, 1979; Swanepoel et al., 1987; Whitehouse, 2004; Burt and Swane-
poel, 2005). Hyalomma ticks evolved from an ancestor that lived in the Oriental region in the early Miocene, approximately 19 million years ago, with dispersal into Africa likely around 14 million years ago (Balashov, 1994; Murrell et al., 2001; Barker and Murrell, 2002). These dates indicate that the ticks are significantly older than the MRCA of CCHFV, and they would have already been widespread by the time that the ancestral CCHFV emerged. It is interesting to speculate that with ticks already in place, changes in agricultural practices and climate may have contributed to the initial emergence of a CCHF-like virus. Sheep, goats, and cattle were domesticated in Eurasia prior to 6000 BC (Christian, 1998) and appeared in Africa shortly thereafter (Maddox, 2006). Between 4000 and 2000 BC, agricultural and livestock rearing communities expanded along with an overall increase in human population density. Around that same time (approximately 2500 BC), warmer and drier climatic conditions began to occur throughout the Old World (Bell et al., 2001; Maddox, 2006). As new diseases affected the humans and their livestock, likely due to the alteration of the landscape as well as close contact with animals, pastoralists tried to adapt. While cattle herdsmen moved to drier lands to avoid tsetse flies (Maddox, 2006), they may have inadvertently contributed to creating favorable conditions for CCHFV emergence as drier conditions are preferred by Hyalomma ticks. In fact, their abundance can be tied to both warm temperatures and high levels of aridity (Linthicum and Bailey, 1994). Recently, Randolph and Rogers
S.A. Carroll et al. / Molecular Phylogenetics and Evolution 55 (2010) 1103–1110
(2007) found that the top predictor variable for occurrence of CCHF was the minimum land surface temperature, with the presence of CCHF cases consistently associated with warmer temperatures. Despite the more detailed understanding of the genetic diversity and evolutionary history of CCHFV, it is difficult to interpret how the distribution of CCHFV may have changed over time. The broad geographic distribution of some viral clusters (e.g. lineage III in South Africa, Mauritania and Nigeria), and the finding of reassortment between viruses of different lineages, suggests that CCHFV can be translocated over considerable distances. Some have speculated that migrating birds could have dispersed CCHFV-infected ticks (Hoogstraal, 1979; Hewson et al., 2004a) as has been hypothesized for KFDV (Mehla et al., 2009); however, no direct evidence of this has been demonstrated for CCHFV (Randolph and Rogers, 2007). Tick movement could also be due to attachment to wild animals, livestock, or even humans (El-Azazy and Scrimgeour, 1997; Khan et al., 1997; Rodriguez et al., 1997; Yashina et al., 2003; Hewson et al., 2004a; Chamberlain et al., 2005). Regardless, the long evolutionary history of CCHFV and Hyalomma ticks has provided ample opportunity for expansion to their current geographic ranges throughout Africa, Europe, the Middle East, and Asia. More recently, dramatic increases in unregulated wildlife trade, livestock import and export, and global movement of humans are likely to be creating additional opportunities for spread of this severe disease. This potential is clearly illustrated by the report of ticks infesting 97 shipments, containing 54,376 animals, imported into the United States in 1995 alone (Karesh et al., 2007). Increased surveillance and livestock inspection/quarantine procedures should be considered to help prevent the establishment of ticks and their associated pathogens, especially in previously CCHFV-free countries. Acknowledgments The authors thank R. Ryan Lash for providing the distribution map and Cesar G. Albarino for assistance with the final versions of the figures. The findings and conclusions reported here do not necessarily represent the views of the Centers for Disease Control and Prevention or the Department of Health and Human Services. References Anagnostou, V., Papa, A., 2009. Evolution of Crimean-Congo hemorrhagic fever virus. Infect. Genet. Evol. 9, 948–954. Baillie, G.J., Kolokotronis, S.O., Waltari, E., Maffei, J.G., Kramer, L.D., Perkins, S.L., 2008. Phylogenetic and evolutionary analyses of St. Louis encephalitis virus genomes. Mol. Phylogenet. Evol. 47, 717–728. Balashov, Y.S., 1994. Importance of continental drift in the distribution and evolution of ixodid ticks. Entomol. Rev. 73, 42–50. Barker, S.C., Murrell, A., 2002. Phylogeny, evolution, and historical zoogeography of ticks: a review of recent progress. Exp. Appl. Acarol. 28, 55–68. Bell, P.A., Greene, T., Fisher, J., Baum, A.S., 2001. Weather, climate, and behavior. In: Environmental Psychology, fifth ed. Harcourt College, Ft. Worth, pp. 167–204. Bird, B.H., Githinji, J.W.K., Macharia, J.M., Kasiiti, J.L., Muriithi, R.M., Gacheru, S.G., Musaa, J.O., Towner, J.S., Reeder, S.A., Oliver, J.B., Stevens, T.L., Erickson, B.R., Morgan, L.T., Khristova, M.L., Hartman, A.L., Comer, J.A., Rollin, P.E., Ksiazek, T.G., Nichol, S.T., 2008. Multiple virus lineages sharing recent common ancestry were associated with a large Rift Valley fever outbreak among livestock in Kenya during 2006–2007. J. Virol. 82, 11152–11166. Bird, B.H., Khristova, M.L., Rollin, P.E., Ksiazek, T.G., Nichol, S.T., 2007. Complete genome analysis of 33 ecologically and biologically diverse Rift Valley fever virus strains reveals widespread virus movement and low genetic diversity due to recent common ancestry. J. Virol. 81, 2805–2816. Bryant, J.E., Holmes, E.C., Barrett, A.D., 2007. Out of Africa: a molecular perspective on the introduction of yellow fever virus into the Americas. PLoS Pathog. 3, e75. doi: 10.1371/journal.ppat.0030075. Burney, M.I., Ghafoor, A., Saleen, M., Webb, P.A., Casals, J., 1980. Nosocomial outbreak of viral hemorrhagic fever caused by Crimean hemorrhagic feverCongo virus in Pakistan, January 1976. Am. J. Trop. Med. Hyg. 29, 941–947. Burt, F.J., Paweska, J.T., Ashkettle, B., Swanepoel, R., 2009. Genetic relationship in southern African Crimean-Congo haemorrhagic fever virus isolates: evidence for occurrence of reassortment. Epidemiol. Infect. 137, 1302–1308.
1109
Burt, F.J., Swanepoel, R., 2005. Molecular epidemiology of African and Asian Crimean-Congo haemorrhagic fever isolates. Epidemiol. Infect. 133, 659– 666. Chamberlain, J., Cook, N., Lloyd, G., Mioulet, V., Tolley, H., Hewson, R., 2005. Coevolutionary patterns of variation in small and large RNA segments of CrimeanCongo hemorrhagic fever virus. J. Gen. Virol. 86, 3337–3341. Chare, E.R., Gould, A., Holmes, E.C., 2003. Phylogenetic analysis reveals a low rate of homologous recombination in negative-sense RNA viruses. J. Gen. Virol. 84, 2691–2703. Christian, D., 1998. A History of Russia, Central Asia, and Mongolia, vol. I: Inner Eurasia from Prehistory to the Mongol Empire. Blackwell, Malden. Chumakov, M.P., 1945. A new tick-borne virus disease—Crimean hemorrhagic fever. In: Sokolov, A.A., Chumakov, M.P., Kolachev, A.A. (Eds.), Crimean Hemorrhagic Fever (Acute Infectious Capillary Toxicosis). Izd. Otd. Primorskoi Armii, Simferopol, Moscow, pp. 13–45. Chumakov, M.P., 1974. On 30 years of investigation of Crimean hemorrhagic fever. Tr. Inst. Polio Virusn. Entsefalitov Akad. Med. Nauk SSSR 22, 5–18. Chumakov, M.P., Smirnova, S.E., Shalunova, N.Y., Mart’yanova, L.I., Fleer, G.P., Zgurskaya, G.N., Maksumov, S.S., Kasymov, K.Y., Pak, T.P., 1976. Proofs of etiological identity to Crimean hemorrhagic fever in Central Asian hemorrhagic fever. In: IX International Congress on Tropical Medicine and Malaria, Athens, vol. 1, pp. 33–34. Deyde, V.M., Khristova, M.L., Rollin, P.E., Ksiazek, T.G., Nichol, S.T., 2006. CrimeanCongo hemorrhagic fever virus genomics and global diversity. J. Virol. 80, 8834– 8842. Drummond, A.J., Ho, S.Y.W., Phillips, M.J., Rambaut, A., 2006. Relaxed phylogenetics and dating with confidence. PLoS Biol. 4, e88. doi: 10.1371/ journal.pbio.0040088. Drummond, A.J., Rambaut, A., 2007. BEAST: Bayesian evolutionary analysis by sampling trees. BMC Evol. Biol. 7, 214. El-Azazy, O.M., Scrimgeour, E.M., 1997. Crimean-Congo haemorrhagic fever virus infection in the western province of Saudi Arabia. Trans. R. Soc. Trop. Med. Hyg. 91, 275–278. Ergonul, O., 2006. Crimean-Congo haemorrhagic fever. Lancet Infect. Dis. 6, 203– 214. Findlay, G.M., Daubney, R., 1931. The virus of Rift Valley fever or enzootic hepatitis. Lancet ii, 1350–1351. Galtier, N., Gouy, M., Gautier, C., 1996. SeaView and Phylo_win, two graphic tools for sequence alignment and molecular phylogeny. Comput. Appl. Biosci. 12, 543–548. Gould, E.A., Solomon, T., 2008. Pathogenic flaviviruses. Lancet 371, 500–509. Gurbuz, Y., Sencan, I., Ozturk, B., Tutuncu, E., 2009. A case of nosocomial transmission of Crimean-Congo hemorrhagic fever from patient to patient. Int. J. Infect. Dis. 13, e105–e107. Hewson, R., Chamberlain, J., Mioulet, V., Lloyd, G., Jamil, B., Hasan, R., Gmyl, A., Gmyl, L., Smirnova, S.E., Lukashev, A., Karganova, G., Clegg, C., 2004a. CrimeanCongo haemorrhagic fever virus: sequence analysis of the small RNA segments from a collection of viruses world wide. Virus Res. 102, 185–189. Hewson, R., Gmyl, A., Gmyl, L., Smirnova, S.E., Karganova, G., Jamil, B., Hasan, R., Chamberlain, J., Clegg, C., 2004b. Evidence of segment reassortment in CrimeanCongo haemorrhagic fever virus. J. Gen. Virol. 85, 3059–3070. Hillis, D.M., 1996. Inferring complex phylogenies. Nature 383, 130–131. Hoogstraal, H., 1979. The epidemiology of tick-borne Crimean-Congo haemorrhagic fever in Asia, Europe and Africa. J. Med. Entomol. 15, 307–417. Kabete Veterinary Laboratories, 1910. Diseases of sheep. Kenya Veterinary Department Annual Report—1910. Kenya Dept. of Vet. Services, Nairobi, Kenya. Karesh, W.B., Cook, R.A., Gilbert, M., Newcomb, J., 2007. Implications of wildlife trade on the movement of avian influenza and other infectious diseases. J. Wildl. Dis. 43, 55–59. Katoh, K., Kuma, K., Toh, H., Miyata, T., 2005. MAFFT Version 5: improvement in accuracy of multiple sequence alignment. Nucleic Acids Res. 33, 511–518. Khan, A.S., Maupin, G.O., Rollin, P.E., Noor, A.M., Shurie, H.H., Shalabi, A.G., Wasef, S., Haddad, Y.M., Sadek, R., Ijaz, K., Peters, C.J., Ksiazek, T.G., 1997. An outbreak of Crimean-Congo hemorrhagic fever in the United Arab Emirates, 1994–1995. Am. J. Trop. Med. Hyg. 57, 519–525. Linthicum, K.J., Bailey, C.L., 1994. Ecology of Crimean-Congo hemorrhagic fever. In: Sonenshine, D.E., Mather, T.N. (Eds.), Ecological Dynamics of Tick-Borne Zoonoses. Oxford Univ. Press, New York, pp. 392–437. Lukashev, A.N., 2005. Evidence for recombination in Crimean-Congo hemorrhagic fever virus. J. Gen. Virol. 86, 2333–2338. Maddox, G., 2006. Sub-Saharan Africa: An Environmental History. ABC-CLIO, Santa Barbara. Mehla, R., Kumar, S.R.P., Yadav, P., Barde, P.V., Yergolkar, P.N., Erickson, B.R., Carroll, S.A., Mishra, A.C., Nichol, S.T., Mourya, D.T., 2009. Recent ancestry of Kyasanur Forest disease virus. Emerging Infect. Dis. 15, 1431–1437. Murrell, A., Campbell, N.J.H., Barker, S.C., 2001. A total-evidence phylogeny of ticks provides insights into the evolution of life cycles and biogeography. Mol. Phylogenet. Evol. 21, 244–258. Papa, A., Bozovic, B., Pavlidou, V., Papadimitriou, E., Pelemis, M., Antoniadis, A., 2002. Genetic detection and isolation of Crimean-Congo hemorrhagic fever virus, Kosovo, Yugoslavia. Emerging Infect. Dis. 8, 852–854. Papa, A., Papadimitriou, E., Bozovic, B., Antoniadis, A., 2005. Genetic characterization of the M RNA segment of a Balkan Crimean-Congo hemorrhagic fever virus strain. J. Med. Virol. 75, 466–469. Parry, J.H., 1925. Haemorrhagic purpura due to tick bites. Tanganyika Territory, Annual Medical Report for the Year Ending 31st December, 1925, 101–102.
1110
S.A. Carroll et al. / Molecular Phylogenetics and Evolution 55 (2010) 1103–1110
Pollock, D.D., Zwickl, D.J., McGuire, J.A., Hillis, D.M., 2002. Increased taxon sampling is advantageous for phylogenetic inference. Syst. Biol. 51, 664–671. Posada, D., Crandall, K.A., 1998. MODELTEST: testing the model of DNA substitution. Bioinformatics 14, 817–818. Randolph, S.E., Rogers, D.J., 2007. Ecology of tick-borne disease and the role of climate. In: Ergonul, O., Whitehouse, C.A. (Eds.), Crimean-Congo Hemorrhagic Fever: A Global Perspective. Springer, Dordrecht, pp. 167–186. Rodriguez, L.L., Maupin, G.O., Ksiazek, T.G., Rollin, P.E., Khan, A.S., Schwarz, T.F., Lofts, R.S., Smith, J.F., Noor, A.M., Peters, C.J., Nichol, S.T., 1997. Molecular investigation of a multisource outbreak of Crimean-Congo hemorrhagic fever in the United Arab Emirates. Am. J. Trop. Med. Hyg. 57, 512–518. Sanchez, A.J., Vincent, M.J., Nichol, S.T., 2002. Characterization of the glycoproteins of Crimean-Congo hemorrhagic fever virus. J. Virol. 76, 7263–7275. Schmaljohn, C.S., Nichol, S.T., 2007. Bunyaviridae. In: Knipe, D.M., Howley, P.M. (Eds.), Fields Virology, fifth ed. Lippincott, Williams, and Wilkins, Philadelphia, pp. 1741–1789. Simpson, D.I.H., Knight, E.M., Courtois, G., Williams, M.C., Weinbren, M.P., Kibukamusoke, J.W., 1967. Congo virus: a hitherto undescribed virus occurring in Africa, Part 1: human isolations-clinical notes. E. Afr. Med. J. 44, 86–92. Suchard, M.A., Weiss, R.E., Sinsheimer, J.S., 2001. Bayesian selection of continuoustime Markov chain evolutionary models. Mol. Biol. Evol. 18, 1001–1013. Swanepoel, R., Shepherd, A.J., Leman, P.A., Shepherd, S.P., McGillivray, G.M., Erasmus, M.J., Searle, L.A., Gill, D.E., 1987. Epidemiologic and clinical features
of Crimean-Congo hemorrhagic fever in southern Africa. Am. J. Trop. Med. Hyg. 36, 120–132. Swanepoel, R., Struthers, J.K., Shepherd, A.J., McGillivray, G.M., Nel, M.J., Jupp, P.G., 1983. Crimean-Congo hemorrhagic fever in South Africa. Am. J. Trop. Med. Hyg. 32, 1407–1415. Weaver, S.C., 2006. Evolutionary influences in arboviral disease. Curr. Top. Microbiol. Immunol. 299, 285–314. Whitehouse, C.A., 2004. Crimean-Congo hemorrhagic fever. Antiviral Res. 64, 145– 160. Woodall, J.P., Williams, M.C., Simpson, D.I.H., Ardoin, P., Lule, M., West, R., 1965. The Congo group of agents. East Afr. Virus Res. Inst. Rep. 14, 34–36. Woodall, J.P., Williams, M.C., Simpson, D.I.H., 1967. Congo virus: a hitherto undescribed virus occurring in Africa, Part 2: identification studies. East Afr. Med. J. 44, 93–98. Yashina, L., Vyshemirskii, O., Seregin, S., Petrova, I., Samokhvalov, E., Lvov, D., Gutorov, V., Kuzina, I., Tyunnikov, G., Tang, Y.-W., Netesov, S., Petrov, V., 2003. Genetic analysis of Crimean-Congo hemorrhagic fever virus in Russia. J. Clin. Microbiol. 41, 860–862. Zanotto, P.M. de A., Gould, E.A., Gao, G.F., Harvey, P.H., Holmes, E.C., 1996. Population dynamics of flaviviruses revealed by molecular phylogenies. Proc. Natl. Acad. Sci. USA 93, 548–553. Zwickl, D.J., Hillis, D.M., 2002. Increased taxon sampling greatly reduces phylogenetic error. Syst. Biol. 51, 588–598.