|Home | About | Journals | Submit | Contact Us | Français|
The taxonomy of the family Filoviridae (marburgviruses and ebolaviruses) has changed several times since the discovery of its members, resulting in a plethora of species and virus names and abbreviations. The current taxonomy has only been partially accepted by most laboratory virologists. Confusion likely arose for several reasons: species names that consist of several words or which (should) contain diacritical marks, the current orthographic identity of species and virus names, and the similar pronunciation of several virus abbreviations in the absence of guidance for the correct use of vernacular names. To rectify this problem, we suggest (1) to retain the current species names Reston ebolavirus, Sudan ebolavirus, and Zaire ebolavirus, but to replace the name Cote d'Ivoire ebolavirus [sic] with Taï Forest ebolavirus and Lake Victoria marburgvirus with Marburg marburgvirus; (2) to revert the virus names of the type marburgviruses and ebolaviruses to those used for decades in the field (Marburg virus instead of Lake Victoria marburgvirus and Ebola virus instead of Zaire ebolavirus); (3) to introduce names for the remaining viruses reminiscent of jargon used by laboratory virologists but nevertheless different from species names (Reston virus, Sudan virus, Taï Forest virus), and (4) to introduce distinct abbreviations for the individual viruses (RESTV for Reston virus, SUDV for Sudan virus, and TAFV for Taï Forest virus), while retaining that for Marburg virus (MARV) and reintroducing that used over decades for Ebola virus (EBOV). Paying tribute to developments in the field, we propose (a) to create a new ebolavirus species (Bundibugyo ebolavirus) for one member virus (Bundibugyo virus, BDBV); (b) to assign a second virus to the species Marburg marburgvirus (Ravn virus, RAVV) for better reflection of now available high-resolution phylogeny; and (c) to create a new tentative genus (Cuevavirus) with one tentative species (Lloviu cuevavirus) for the recently discovered Lloviu virus (LLOV). Furthermore, we explain the etymological derivation of individual names, their pronunciation, and their correct use, and we elaborate on demarcation criteria for each taxon and virus.
Virus taxonomy is the practice of describing, classifying, and naming viruses. Description involves the genetic, biological, and morphological characterization of viruses and their virions, and is typically done by laboratory virologists. Classification is the process of sorting related viruses into groups, the so-called taxa, which ideally reflect phylogeny. This is typically done by virus taxonomists of the International Committee on Taxonomy of Viruses (ICTV), the body tasked by the International Union of Microbiological Societies (IUMS) to make decisions on matters of virus classification and nomenclature, with the help of expert groups—the ICTV Study Groups. Naming involves the issuing of unique designations for taxa (names) and viruses (names and abbreviations) according to nomenclature. Nomenclature is the set of rules or process for naming as established by the ICTV in the case of taxa and the ICTV Study Groups or other expert groups in the case of taxa, viruses, strains, variants, and isolates.
The taxonomy of marburgviruses and ebolaviruses has changed several times since the discovery of the type marburgvirus in 1967  and the type ebolavirus in 1976 [3, 21]. The development of their taxonomy has been reviewed in detail  and does not need to be repeated here. Key taxonomic changes were approved by the ICTV and published in the eight ICTV Reports [9, 13, 14, 30, 31, 37, 62, 72]. These changes over the years are summarized in Table 1 and are contrasted with the terminology used by laboratory virologists. The table emphasizes several key points addressed in this manuscript. First, laboratory virologists have been using the terms “Marburg virus” and “Ebola virus” for the type viruses of marburgviruses and ebolaviruses, respectively, for decades and have not accepted the novel names for these agents (Lake Victoria marburgvirus and Zaire ebolavirus) listed in the latest (Eighth) ICTV Report . Second, laboratory virologists have embraced the fact that there are several ebolaviruses, rather than one ebolavirus with several subtypes. However, they usually do not use the names listed in the Eighth ICTV Report . Third, laboratory virologists have not yet made the distinction between species (names italicized) and viruses (names not italicized) [6, 8, 26, 63-65], a fact that, as described previously by two of the authors , is in part due to the identical spelling of currently approved marburgvirus and ebolavirus species and virus names . We address these and other problems by proposing an improved nomenclature for the family Filoviridae that reflects terminology usage by laboratory scientists while upholding the rules and regulations of the ICTV. Furthermore, we suggest an updated classification based on newly obtained data.
The classification of marburgviruses and ebolaviruses as sister groups in a unique family related to the families Bornaviridae, Paramyxoviridae, and Rhabdoviridae and placed in a common order has been unchallenged since the ICTV approved it [39, 44, 45, 47]. Likewise, it is uncontroversial that there are several different ebolaviruses that obviously should be placed in different species [39, 47]. Article 3 Rule II-3.9 of the International Code for Virus Classification and Nomenclature (ICVCN) states that “[e]xisting names of taxa and viruses should be retained whenever feasible” . We accept and uphold this rule for the names of the order, family, and genera harboring marburgviruses and/or ebolaviruses, as well as for three ebolavirus species. However, because of the lack of acceptance of two species names and several virus names by the laboratory virologist community (Table 1), and because of the confusion of species names with virus names [6, 8, 26, 63-65], we find the retention of several names not feasible and here propose changes.
Pringle et al.  introduced the order Mononegavirales in 1991 as a taxon “to embrace families of [certain] viruses with similar genomic organization and replicative strategies”. These viruses, the mononegaviruses, are enveloped and contain a linear, nonsegmented, single-stranded RNA genome with the characteristic general gene order 3′-UTR —core proteins genes—envelope protein genes—polymer-ase gene—5′-UTR . Marburgviruses and ebolaviruses fulfill these and other criteria for member inclusion [5, 10, 18, 23, 24, 36, 40, 50, 54, 68] and consequently have been assigned to this order .
Etymology of Mononegavirales: derived from Gre. adj. μóνος [monos]—alone or single, referring to the single-stranded [RNA] genome of order members; Lat. v. negare—to negate, referring to the negative polarity of the single-stranded [RNA] genomes of order members; and suff. –virales—ending denoting a virus order  → Neo-Lat. n. masc. pl. Mononegavirales—the order of [RNA] viruses with single-stranded genomes of negative polarity.
Marburgviruses and ebolaviruses differ from other mononegaviruses. They have longer genomes (≈19 kb) than most other members of the order: the only other mononegaviruses with similarly long genomes are assigned to the paramyxovirus genus Henipavirus (Hendra and Nipah viruses)  and the proposed paramyxovirus genus “Jeilongvirus” (Beilong and J viruses) [19, 28]. Marburgvirus and ebolavirus genomes encode two unusual proteins: VP30 is exceptional in that only pneumoviruses (family Paramyxoviridae) encode a protein with possibly similar function (M2-1) [4, 60], and VP24 is unique in that no other similar protein has yet been identified . Marburgviruses and ebolaviruses are the only mammal-infecting members of the order Mononegavirales forming truly filamentous virions in the near absence of nonspherical forms [16, 53]. Finally, marburgviruses and ebolaviruses are the only known mononegaviruses that cause viral hemorrhagic fever in primates [3, 21, 56]. Consequently, marburgviruses and ebolaviruses have been assigned to their own family, Filoviridae .
Etymology of Filoviridae: derived from: Lat. n. neut. sg. filum—thread, referring to the unique filamentous morphology of virions produced by family members; and suff. –viridae—ending denoting a virus family [13, 72] → NeoLat. n. fem. pl. Filoviridae—the family of thread-like viruses.
Marburgviruses are endemic in arid woodlands in eastern, south-central, and western Africa, whereas ebolaviruses are endemic in the humid rain forests of central and western Africa [42, 43]. Marburgviruses may be adapted to chiropteran (bat) reservoirs , and at least one ebolavirus may infect porcids (pigs) in nature . Marburgvirus and ebolavirus genomes differ from one another by ≥50% at the nucleotide level . Marburgvirus genomes differ from ebolavirus genomes in that they have only one, rather than several gene overlaps . Marburgvirus gene four (GP) expresses only one protein, the spike glycoprotein GP1,2 , whereas ebolavirus gene four expresses four proteins (sGP, Δ-peptide, GP1,2, ssGP) via transcriptional editing that results in open reading frame shifts and, in the case of sGP/Δ-peptide, proteolytic processing [55, 67, 69]. Marburgvirus spike proteins are highly N- and O-glycosylated but lack sialic acids, whereas ebolavirus spike proteins are highly N- and O-glycosylated and may contain sialic acids [11, 17, 52, 73]. There is minimal to no serological cross-reactivity between marburgvirions and ebolavirions [11, 51, 71]. Marburgvirions are shorter (average of ≈665 nm in length) than ebolavirions (≈805 nm) . Consequently, marburgviruses and ebolaviruses have been assigned to two different genera, Marburgvirus and Ebolavirus, respectively [12, 32, 39, 47].
Etymology of Marburgvirus: derived from geo. Marburg— short form of Marburg an der Lahn, the city in Hessen [Hesse], Germany, where the type virus of this genus was first isolated; and –virus—ending denoting a virus genus  → Neo-Lat. N. neut. sg. Marburgvirus—the genus of marburgviruses.
Etymology of Ebolavirus: derived from geo. Ebola—name of the headstream of the Mongala River, a tributary of the Zaire River (today Congo River) in Zaire (today the Democratic Republic of the Congo), where the type virus of this genus was thought to be first encountered; and–virus—ending denoting a virus genus  → Neo-Lat. N. neut. sg. Ebolavirus—the genus of ebolaviruses.
At least five lineages of marburgviruses exist according to the most recent phylogenetic data. Virus genomes of four of these lineages differ from each other only by 0.0–7.4%. Genomes from viruses of the fifth lineage reach 21% nucleotide difference compared to the four other lineages . Representatives of all lineages are identical in gene order, number and position of gene overlaps, and other structural and organizational features. Representatives of all lineages cross-react serologically. Genomes of even the most divergent marburgvirus lineage do not reach the 30% nucleotide difference cut-off established below for differentiating members of different ebolavirus species. We suggest extrapolating the same cut-off criterion to marburgvirus species. Consequently, there is currently the need for recognition of only one marburgvirus species.
Etymology of Marburg marburgvirus: derived from geo. Marburg—short form of Marburg an der Lahn, the city in Hessen [Hesse], Germany, where the type virus of this species was first isolated; and marburgvirus—the genus of marburgviruses → the Marburg species of marburgviruses.
The first ebolaviruses were discovered in 1976, when simultaneous viral hemorrhagic fever outbreaks occurred in Zaire (today Democratic Republic of the Congo) and Sudan [3, 21, 41]. In 1983, convincing data were published demonstrating that the viruses causing the two outbreaks were antigenically related, but not identical [7, 33, 51]. In the following years, two additional ebolaviruses were discovered that, while antigenically cross-reactive with the Zaire and Sudan viruses, were unique: the first in 1989 in the USA and the second in 1994 in the Republic of Côte d'Ivoire, [11, 15, 27]. Today, full-length genomic sequences are available for isolates of all of these viruses. Their comparison reveals that the genomes of the four viruses differ from each other by 36.7–42.3% . ICVCN Article 3 Rule III-3.21 states that “[a] virus species is defined as a polythetic class of viruses that constitutes a replicating lineage and occupies a particular ecological niche” . Due to the genomic sequence diversity and the fact that the four viruses are endemic to different geographic areas and possibly different reservoir hosts, the creation of several different ebolavirus species, first accepted by the ICTV in 1995 , has retrospectively been justified. In 2008, a fifth ebolavirus was described whose genomic sequence differed from previously recognized viruses by 31.7–42.4% . Thus, far, a formal description of a new species for this virus has not yet been forwarded to the ICTV. Here we describe all five ebolavirus species.
Etymology of Bundibugyo ebolavirus: derived from geo. Bundibugyo—name of the chief town of Bundibugyo District in the Republic of Uganda, where members of this species were first encountered; and ebolavirus—the genus of ebolaviruses → the Bundibugyo species of ebolaviruses.
Etymology of Reston ebolavirus: derived from geo. Reston—the town in Virginia, USA, where members of this species were first encountered; and ebolavirus—the genus of ebolaviruses → the Reston species of ebolaviruses.
Etymology of Sudan ebolavirus: derived from geo. Sudan—English conventional short form of the Republic of Sudan, where members of this species were first encountered; and ebolavirus—the genus of ebolaviruses → the Sudan species of ebolaviruses.
Etymology of Taï Forest ebolavirus: derived from geo. Parc National de Taï [Taï National Park]—the place in the Republic of Côte d'Ivoire, where members of this species were first encountered; and ebolavirus—the genus of ebolaviruses → the Taï Forest species of ebolaviruses.
Etymology of Zaire ebolavirus: derived from geo. Zaire—English conventional short form of the Republic of Zaire (today the Democratic Republic of the Congo), where members of this species were first encountered; and ebolavirus—the genus of ebolaviruses → the Zaire species of ebolaviruses.
According to ICVCN Article 3 Rule I-3.3, “[t]he ICTV is not responsible for classification and nomenclature of virus taxa below the rank of species. The classification and naming of serotypes, genotypes, strains, variants and isolates of virus species is the responsibility of acknowledged international specialist groups. It is the responsibility of ICTV Study Groups to decide if an isolate or a group of isolates should constitute a species” . Traditionally, filoviruses were named by their discoverers in original publications, and these names were then accepted, and sometimes changed, by the ICTV Filoviridae Study Group. The names recommended and recognized by the Study Group and those used by laboratory virologists are depicted in Table 1. ICVCN Article 3 Rule I-3.3 states that “[t]he policy of the ICTV is that as far as is possible, decisions on questions of taxonomy and nomenclature should reflect the majority view of the appropriate virological constituency” [emphasis added by the authors] . Rule II-3.19 states that “…[w]hen names are universally used by virologists in published work, these or derivatives shall be the preferred basis for creating names, irrespective of national origin” . This rule was followed for most of the history of filovirology. However, in 2002/2005, the Filoviridae Study Group followed ICTV guidance and recommended to rename filoviruses and to create virus names that are identical in spelling to species names [12, 32], a possible violation of ICVCN Article 3 Rule III-3.24 (“A species name must provide an appropriately unambiguous identification of the species. [It] should not be in a form that could be easily confused with the names of other taxa”) . Five to eight years have passed since the introduction of the name Lake Victoria marburgvirus. Yet, most laboratory scientists still do not use this term either in publications or in seminars. Instead, the overwhelming majority of publications refer to “Marburg virus” (Table 6), a preference that is also followed by the public press. ICVCN Article 2.1 emphasizes that “[t]he essential principles of virus nomenclature are (1) to aim for stability; (2) to avoid or reject the use of names which might cause error or confusion; (3) to avoid the unnecessary creation of names” . Introducing the name Lake Victoria marburgvirus was a mistake, as it contradicts principle (1), while the introduction of virus names identical in spelling to species names contradicts principle (2), consequently contradicting principle (3). Here, we rectify this situation by recommending that the traditional virus name (“Marburg virus”) be used. Retrospectively, the virus nomenclature in most published articles will then be correct. Likewise, press articles, which almost invariably refer to “Marburg virus” will be correct retrospectively and prospectively. As the traditional name is different from the species name, confusing species and virus names will be much more difficult, even in the absence of taxonomic education.
As mentioned above, five lineages of marburgviruses are currently recognized. The genomes of representative marburgvirus variants of one of these lineages differs from all others by up to 21.3% in nucleotide sequence, whereas the genomes of variants from the other four lineages differ from each other only by as much as 0.0–7.8% . To reflect the clear dichotomy of marburgvirus variants in formal classification, we suggest the existence of two distinct viruses belonging to the same species. Consequently, variants of the four related lineages continue to represent Marburg virus, the type virus of the species Marburg marburgvirus. We suggest the name Ravn virus and the abbreviation RAVV for a second virus in the same species to represent the divergent variants 02Uga2007, 09DRC1999, 44Bat2007, 188Bat2007, 982Bat2008, and Ravn (see ). Such a solution to this problem is hardly radical, since similar decisions have been made elsewhere, for instance in the case of Amur, Da Bie Shan, and Hantaan viruses (all of which are members of the species Hantaan virus).
Etymology of Marburg virus: derived from geo. Marburg—short form of Marburg an der Lahn, the city in Hessen [Hesse], Germany, where this virus was first isolated; and Lat. n. neut. sg. virus—poison, slime, venom.
Etymology of Ravn virus: derived from nom. Ravn—last name of the Danish patient from whom this virus was first isolated; and Lat. n. neut. sg. virus—poison, slime, venom.
Five to eight years have passed since the introduction of the names Cote d'Ivoire ebolavirus [sic], Reston ebolavirus, Sudan ebolavirus, and Zaire ebolavirus for the members of the four recognized ebolavirus species. Instead of using these names, the overwhelming majority of publications refer to “Ebola virus” instead of Zaire ebolavirus (Table 6), a preference that is also followed by the public press. The remainder of the viruses are usually referred to as Côte d'Ivoire/Ivory Coast/Taï Forest, Reston, and Sudan in the context of “Ebola virus.” Worse, in the few cases in which the recommended names were used, they almost invariably were confused with species names (virus names italicized), an error that is understandable because discernment of the identically spelled taxa, a violation of ICVCN Article 3 Rule III-3.24, requires a thorough grasp of the difference between species and viruses . Introducing the name “Zaire ebolavirus” was a mistake, as it contradicts ICVCN Article 2.1 (described above). Here, we rectify this situation by recommending that the traditional virus name (“Ebola virus”) be used. Retrospectively, the virus nomenclature in most published articles will then be correct. Likewise, press articles, which almost invariably refer to “Ebola virus,” and usually with that term aim at referring to the virus that is currently officially named “Zaire ebolavirus,” will be correct retrospectively and prospectively. As the traditional names are different from the species names, confusing species and virus names will be much more difficult, even in the absence of taxonomic education.
Etymology of Bundibugyo virus: derived from geo. Bundibugyo—name of the chief town of Bundibugyo District in the Republic of Uganda, where members of this species were first encountered; and Lat. n. neut. sg. virus—poison, slime, venom.
Etymology of Ebola virus: derived from geo. Ebola—name of the headstream of the Mongala River, a tributary of the Zaire River (today Congo River), in Zaire (today the Democratic Republic of the Congo), where this virus was thought to be first encountered; and Lat. n. neut. sg. virus— poison, slime, venom.
Etymology of Reston virus: derived from geo. Reston—the town in Virginia, USA, where this virus was first encountered; and Lat. n. neut. sg. virus—poison, slime, venom.
Etymology of Sudan virus: derived from geo. Sudan—English conventional short form of the Republic of Sudan, where this virus was first encountered; and Lat. n. neut. sg. virus—poison, slime, venom.
Etymology of Taï Forest virus: derived from geo. Parc National de Taï [Taï National Park]—the place in the Republic of Côte d'Ivoire, where members of this species were first encountered; and Lat. n. neut. sg. virus—poison, slime, venom.
A new virus, named Lloviu virus (LLOV), was recently detected in Schreiber's long-fingered bats (Miniopterus schreibersii Kuhl, 1817) in Cueva del Lloviu, Principality of Asturias, Spain. The virus has not yet been isolated in tissue culture (primarily because of the lack of maximum-containment laboratories in the region). However, RNA isolation from tissues of individual bats allowed repeated determination of the full-length genomic sequence of the virus, thereby fulfilling the prerequisites of the ICTV for classification of an uncultured virus . Genomic analysis revealed that the virus fulfills the criteria characteristic for filoviruses as outlined in the Eighth ICTV Report  and this publication. Phylogenetic analyses using full-length genomic sequence demonstrated that LLOV is roughly equally distant from both marburgviruses and ebolaviruses (≈56 and ≈51%, respectively, as determined by the p-distance method using whole-genome nucleotide sequences). Its genomic organization is more reminiscent of that of ebolaviruses than that of marburgviruses: LLOV gene four (GP) possesses three overlapping ORFs coding for sGP/Δ-peptide, GP1,2, and ssGP analogs. However, the LLOV genome contains a rather truncated 5′-UTR compared to known ebolaviruses . Together, these data justify the creation of a new tentative filovirus genus and tentative species for LLOV.
Etymology of Cuevavirus: derived from Spa. n. fem. sg. i—cave, referring to the fact that members of this genus were first identified in bats located in a cave; and – virus—ending denoting a virus genus  → Neo-Lat. N. neut. sg. Cuevavirus—the genus of cave viruses.
Etymology of Lloviu cuevavirus: derived from geo. Lloviu, referring to the name of the cave in the Kingdom of Spain where members of this species where first encountered; and Cuevavirus—the genus of cave viruses → the Lloviu species of cave viruses.
Etymology of Lloviu virus: derived from geo. Cueva del Lloviu—the cave in the Kingdom of Spain where this virus was first encountered; and Lat. n. neut. sg. virus—poison, slime, venom.
A short summary of all proposed changes contrasted with the filovirus taxonomy as described in the latest (Eighth) ICTV Report is shown in Table 8.
We would like to thank Anna N. GerasimovaClawson (Logos Consulting, Seattle, WA, USA) for help with the International Phonetic Alphabet and English Phonetic Notations. We are also indebted to Thomas S. Postler (New England Primate Research Center, Southborough, MA, USA), Philip J. Kranzusch (Harvard Medical School, Boston, MA, USA), Sheli R. Radoshitzky (United States Army Medical Research Institute of Infectious Diseases, Fort Detrick, Frederick, MD, USA) and Victoria M. Wahl-Jensen (Integrated Research Facility at Fort Detrick, Frederick, MD, USA) for their very useful editorial comments and suggestions.
Publisher's Disclaimer: Disclaimer The content of this publication does not necessarily reflect the views or policies of the US Department of Health and Human Services or of the institutions and companies affiliated with the authors.