Global distribution and genetic diversity of orthoebolaviruses: Mapping and evolutionary analysis

Abstract

Orthoebolavirus genus (Family: Filoviridae) poses an ongoing threat to both human and animal health. To better understand and address this persistent threat, we conducted a comprehensive analysis. Through a systematic investigation of orthoebolaviruses' global geographic distribution, phylogenetic relationships, and detailed examination of nonsynonymous amino acid mutations in Zaire ebolavirus (EBOV) from both human and non-human primate (NHP) hosts, we have made several key findings. Our results demonstrate that the viruses in the genus Orthoebolavirus primarily circulate in Africa and Asia, with a possible trend of geographic expansion to other regions. Phylogenetic analysis revealed significant evolutionary divergence among the six viral species across different hosts and temporal scales. Most notably, we identified 73 statistically significant amino acid substitution sites (all p < 0.0001) in EBOV from both human and NHP hosts. Collectively, this multifaceted investigation offers valuable insights that will inform future research on orthoebolaviruses' pathogenesis and intervention strategies.

1. Introduction

Six species within the Orthoebolavirus genus (Family: Filoviridae) have been identified to date, namely Orthoebolavirus zairense (Zaire ebolavirus, EBOV), Orthoebolavirus sudanense (Sudan ebolavirus, SUDV), Orthoebolavirus bundibugyoense (Bundibugyo ebolavirus, BDBV), Orthoebolavirus taiense (Taï Forest ebolavirus, TAFV), Orthoebolavirus restonense (Reston ebolavirus, RESTV), Orthoebolavirus bombaliense (Bombali ebolavirus, BOMV). EBOV and SUDV emerged nearly simultaneously in 1976, occurring in Zaire (currently named the Democratic Republic of the Congo) and Sudan, respectively [1,2]. EBOV poses the most severe threat to humans; the outbreak of Ebola virus disease (EVD) in West Africa from 2013 to 2016 led to nearly 30,000 cases [3]. The case fatality rate of EVD is known to be up to 40 % to 50 % in absence of treatment [4]. In 2008, a new species of the Orthoebolavirus genus known as BDBV was identified during a hemorrhagic fever outbreak in Uganda [5]. TAFV was first identified in chimpanzees in Côte d'Ivoire and caused non-fatal human infections [6]. RESTV typically induces fatal disease in non-human primates (NHPs), specifically cynomolgus monkeys (Macaca fascicularis) originating in the Philippines [7]. Subsequently, RESTV was detected in pigs with porcine reproductive and respiratory disease syndrome in the Philippines [8] and China [9]. Although not pathogenic in humans, a study indicated RESTV could cause asymptomatic infections in humans [10]. BOMV was initially detected in bats (Chaerephon pumilus and Mops condylurus) from Sierra Leone in 2018 [11]. There has not been a report of a human case, but studies have identified BOMV antibodies in human serum samples from the Democratic Republic of the Congo and Guinea [12,13]. Critically, orthoebolaviruses are classified as Biosafety Level 4 (BSL-4) pathogens. Consequently, the severe threat they pose necessitates comprehensive research spanning geographic distribution, host range, evolutionary dynamics, and non-synonymous substitutions.

However, the global distribution, host specificity, and evolutionary relationship among orthoebolaviruses remain unclear; these uncertainties significantly impede risk prediction of potential zoonotic spillover and the development of targeted surveillance strategies.

Therefore, this study characterized the geographic and host distribution of orthoebolaviruses, conducted a comprehensive phylogenetic analysis, and profiled non-synonymous substitutions in EBOV to elucidate its genetic diversity. These variation sites underpin future advances in vaccine design, evolutionary research, and antibody therapeutics.

2. Methods

2.1. Literature search and data extraction regarding orthoebolaviruses

In this study, we searched PubMed, Web of Science (WoS), Scopus, China National Knowledge Infrastructure (CNKI) and WanFang Data for literature before February 19, 2024, containing the terms “ebola virus” or “ebolavirus” and its Chinese equivalents. Articles were excluded based on the following criteria: duplicate articles from different databases, languages other than Chinese and English, literature review and meta-analysis, serological analysis of survivors, and lack of specific samples collected information. Relevant data were then extracted from all the eligible articles.

2.2. Geographical distribution of orthoebolaviruses

Thematic maps showing the geographical distribution of each virus species in humans and animal hosts were produced with software ArcGIS 10.8. The original collection locations of samples from patients and animals infected with various virus species were used for mapping, and the administrative region centroids were used when exact locations were not available. For sequence records containing country/province information in their “Geo_Location” metadata, the latitude/longitude of the corresponding geographic centroid was similarly adopted as a positioning reference. All redundant records from overlapping sources were removed prior to final mapping.

2.3. Collection of orthoebolaviruses sequences

All viral sequences classified under the genus “Orthoebolavirus” were systematically retrieved from the NCBI Taxonomy database in NCBI Taxonomy database (https://www.ncbi.nlm.nih.gov/taxonomy/) with a collection cutoff date of July 2, 2024. The dataset encompassed complete genomic sequences from all six currently recognized species: EBOV, SUDV, BDBV, TAFV, RESTV, and BOMV. To ensure data quality, we excluded three categories of sequences from downstream analyses: (1) partial genes, (2) externally derived sequences including imported samples, and (3) sequences lacking critical metadata (host, collection date, and geographic location). The curated sequences were processed using Geneious Prime version 2024.0.5 (https://www.geneious.com/), where genomic annotations were carefully reviewed and, when necessary, manually corrected by alignment with reference sequences from NCBI. This re-annotated process guaranteed accurate identification and consistent labeling of all genomic features across the dataset.

2.4. Phylogenetic analysis of orthoebolaviruses

Phylogenetic analysis was performed based on the coding sequences (CDS) of viral conserved proteins, including RNA-dependent RNA polymerase (L), glycoprotein (GP), nucleoprotein (NP) and viral matrix proteins (VP24, VP30, VP35, and VP40). All sequences were first aligned to related viral sequences within the Orthoebolavirus genus using the program MAFFT (version v7.505) [14] with the E-INS-i algorithm. All regions of ambiguous alignment were then trimmed with trimAl program (version 1.4.rev15) [15]. Maximum likelihood (ML) trees were constructed using IQ-TREE program (version 2.2.0.3) [16], with 1000 bootstrap replicates. The evolutionary trees were then visualized using the ggtree package (version v3.8.0) [17], treeio package (version 1.24.0) [18], and ggplot2 package (version 3.4.2) [19] in R software (version 4.3.0).

2.5. Non-synonymous substitutions detection of Orthoebolavirus zairense

Using MAFFT, we aligned the sequences of RdRP, GP, NP, VP24, VP30, VP35, and VP40 proteins from EBOV. For our analysis, we included only sequences derived from full-length genomes possessing complete sequence information, including detailed genomic annotation. Subsequently, we systematically observed and compared the non-synonymous amino acid substitutions between humans and non-human primate hosts relative to the reference sequence (GenBank accession no. NC_002549). To quantify the significance of inter-group variation at mutation sites, Chi-squared tests were applied with Fisher's exact test employed when theoretical frequencies were low.

3. Results

3.1. Geographical distribution of orthoebolaviruses

As of February 19, 2024, a comprehensive database search across WoS, Scopus, PubMed, CNKI, and WanFang Data yielded 42,826 publications. Based on predefined inclusion and exclusion criteria, we excluded seven articles reporting imported cases and three studies unable to confirm the viruses within the Orthoebolavirus genus. However, studies involving suspected cases or case reports were retained for geographic analysis. The literature screening and data extraction workflow is detailed in Supplementary Materials Fig. S1. Up to July 2, 2024, 4159 sequences downloaded from NCBI were processed by removing non-naturally infected and unverified sequences, along with those from imported cases. After eliminating geographic redundancies through literature and sequence validation, 495 records were retained. Orthoebolaviruses have been documented in at least 26 countries worldwide. Among these, EBOV exhibited the highest number of records in every host category. (Fig. 1).

Fig. 1.

Literature and sequence retrieval workflow for Orthoebolaviruses. EBOV, Orthoebolavirus zairense; SUDV, Orthoebolavirus sudanense; BDBV, Orthoebolavirus bundibugyoense; TAFV, Orthoebolavirus taiense; RESTV, Orthoebolavirus restonense; BOMV, Orthoebolavirus bombaliense.

We generated global distribution maps for six species within the Orthoebolavirus genus across four host categories (Fig. 2A). The results demonstrate that EBOV exhibits the broadest geographical range, with documented occurrences spanning multiple African and Asian countries. SUDV displays the second-most extensive distribution, primarily concentrated in Central African nations and parts of West Africa, with additional serological detections in NHPs and Chiroptera across Southeast Asia. BDBV outbreaks in Uganda are confined to Central Africa. TAFV has the fewest records but includes confirmed human infections. RESTV primarily infects NHPs, Chiroptera, and other mammals with rare human cases, however, seropositive individuals have been identified in Germany [20], the Philippines [21], and Guinea [13]. Finally, BOMV circulates predominantly in bats but has been detected in humans via serological surveys [12,13]. Collectively, these findings indicate widespread global distribution of orthoebolaviruses, with all six species documented in both human populations and bat reservoirs.

Fig. 2.

Global distribution of orthoebolaviruses. (A) Worldwide occurrence of six viruses within the Orthoebolavirus genus, (B) geographic distribution in Africa, and (C) habitats of Primates and Chiroptera associated with orthoebolaviruses, color intensity represents species richness within each category.

Orthoebolaviruses exhibit their broadest distribution across Africa, with all six species within the Orthoebolavirus genus documented on the continent. Africa represents the primary endemic zone for orthoebolaviruses, particularly for EBOV, SUDV, and BDBV (Fig. 2B). Viral records span 20 African nations, with the highest concentrations in Sierra Leone, Liberia, and the Democratic Republic of Congo. Additionally, we mapped the suitable habitats for Primates and Chiroptera species reported as orthoebolaviruses carriers using the IUCN Red List of Threatened Species (https://www.iucnredlist.org/) (Fig. 2C). Spatial analysis reveals substantial overlap between observed infection records and predicted reservoir habitats. However, significant sampling gaps persist within suitable ecological zones, underscoring the need for enhanced surveillance in these understudied regions to preempt potential emergence events.

3.2. Phylogenetic analysis of orthoebolaviruses

A total of 4159 sequences within the Orthoebolavirus genus were obtained from NCBI. Following rigorous quality control and metadata curation (including host species, collection date, and geographical location), sequences linked to imported cases were excluded based on predefined criteria, yielding a final dataset of 745 sequences for subsequent phylogenetic analyses (see Supplementary Materials Table S1). We consolidated data from both the complete sequences we collected and information gathered from the literature. (Fig. 3) EBOV exhibited the highest number of records, predominantly from the 2010s, whereas sequences for orthoebolaviruses spanned at least one decade. To assess evolutionary divergence among these species, phylogenetic analyses were conducted using nucleotide sequences of full-length genomes, the L, GP, and NP genes, along with their corresponding amino acid sequences.

Fig. 3.

Sankey diagram of six viruses within the Orthoebolavirus genus. This plot visualizes integrated metadata from literature-sourced records and viral sequences across orthoebolaviruses in this study, with linked dimensions of collection year, host origin, and geographic location.

Phylogenetic analysis based on full-length genome (Fig. 4A) and the L gene (Fig. 4B and C) indicated that EBOV circulating between the 1970s and 2020s primarily infected humans and non-human primates, forming distinct phylogenetic clades between two host groups, while exhibiting L gene nucleotide (nt) and L protein amino acid (aa) identities of 96.110 %–100 % and 97.509 %–100 %, respectively. SUDV and TAFV were predominantly detected in humans with full-genome nt identities of 94.684 %–100 % and 99.947 %–100 %, respectively. BDBV sequences formed two distinct clades corresponding to the 2000s (solely human hosts) and the 2010s (human and non-human primate hosts), yet exhibited high identities with full-genome nt, L gene nt, and L protein aa identities of 99.969 %–99.996 %, 99.985 %–100 %, and 99.005 %–100 %, respectively. RESTV infected only non-human primates (primarily Macaca fascicularis) and other mammals (primarily Sus scrofa), showing full-genome nt, L gene nt, and L protein aa identities of 95.219 %–100 %, 96.746 %–100 %, and 97.694 %–100 %. BOMV sequences were exclusively detected in Chiroptera, exhibiting full-genome nt identity of 97.495 %–100 % and L gene nt/aa identities of 98.161 %–100 %/98.643 %–100 %. Phylogenies for the GP gene (Supplementary Materials Figs. S2-S3) and NP gene (Supplementary Materials Figs. S4-S5) nucleotide and amino acid sequences were congruent with these findings regarding host associations and sequence information patterns.

Fig. 4.

Phylogenetic analysis of orthoebolaviruses. (A) Nucleotide phylogeny based on complete genomes, (B) Nucleotide phylogeny of RNA-dependent RNA polymerase (RdRP), and (C) Amino acid phylogeny of RdRP.

3.3. Non-synonymous substitutions comparison of Orthoebolavirus zairense between human and non-human primates

Given its status as a particularly representative and harmful virus within the Orthoebolavirus genus with abundant available sequences, EBOV was selected to investigate non-synonymous substitutions between human and non-human primate hosts. We compared the protein sequences of L, GP, NP, VP24, VP30, VP35, and VP40 against the reference sequence (GenBank accession no. NC_002549). Our analysis utilized the comprehensive set of human and non-human primate sequences included in this study. For clarity in presentation, each variant type is illustrated by a single representative sequence in Fig. 5 (Detailed in Supplementary Materials Tables S2-S8 and Supplementary Materials Texts S1-S7).

Fig. 5.

Host-specific non-synonymous substitutions between human and non-human primates in Orthoebolavirus zairense. (A) Non-synonymous substitutions in L (B) GP (C) NP (D) VP24 (E) VP30 (F) VP35 and (G) VP40 proteins. The left column indicates host types. Sites marked in red text denote human-specific occurrences, while those in black represent primate-specific occurrences. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Analysis revealed a total of 73 distinct substitution sites across the proteins. Of these, 41 substitutions were specific to human-derived sequences, while 32 were specific to NHP-derived sequences. The L protein (Fig. 5A) exhibited the highest number of substitutions (29 sites), comprising 17 human-specific sites and 12 NHP-specific sites. Notably, the substitutions L-D759G (Fig. 5A), GP-A82V (Fig. 5B), and NP-R111C (Fig. 5C) identified exclusively in human sequences have been previously associated with potentially significant functional implications [[22], [23], [24], [25], [26]]. Host-specific substitution patterns were also observed: substitutions in VP24 (Fig. 5D) and VP30 (Fig. 5E) occurred only in NHPs, while substitutions in VP35 (Fig. 5F) occurred only in humans, while substitutions in VP40 (Fig. 5G) were found in both humans and NHPs. The observed differences in substitution profiles between human and NHP sequences were statistically significant (p < 0.0001).

4. Discussion

This study presents a comprehensive analysis of the orthoebolaviruses by integrating three key dimensions: (1) geographical distribution patterns, (2) phylogenetic analysis of orthoebolaviruses, and (3) host-specific variation sites in EBOV. Geographical analysis revealed that orthoebolaviruses infections are predominantly distributed across Africa, with limited positive detections in Asia and sporadic occurrences in Germany [20]. Phylogenetic analysis using full-length annotated sequences from the GenBank revealed significant divergence among the six species within the Orthoebolavirus genus. Notably, host-specific phylogenetic clustering was observed within individual virus species. Concurrently, variation analysis across all seven proteins of EBOV identified 73 host-associated variation sites specific to human or NHPs, among these, L-D759G (Fig. 5A), GP-A82V (Fig. 5B), and NP-R111C (Fig. 5C) identified exclusively in human sequences, have been previously associated with potentially significant functional implications [[22], [23], [24], [25], [26]]. The GP-A82V substitution has been associated with increased mortality [22], supporting the hypothesis that its enhanced infectivity contributes to greater disease severity [23]. The L-D759G mutation has been shown to promote viral transcription and replication [24], while the NP-R111C substitution plays a multifaceted role in modulating nucleoprotein structure, viral budding, transcription, and replication [25]. These three single amino acid substitutions influence EBOV virulence in animal models and may inform the rational design of effective antiviral therapies [26].

This study demonstrates that the majority of orthoebolaviruses detections occur in Africa, consistent with the African origins of EBOV [1], SUDV [2], BDBV [6], TAFV [7], and BOMV [11]. The distribution in Asia is primarily attributed to RESTV, first identified in cynomolgus monkeys (Macaca fascicularis) imported to the US from the Philippines [8]. Critically, our findings suggest a possible expansion of the geographical range of orthoebolaviruses to Germany [20], indicating persistent viral circulation and potential expansion patterns. However, this interpretation should be made with caution. The human sera analyzed in the report were obtained from unidentified donors, some of whom had known contact with Marburg patients. Since the exact origin and exposure history of these individuals were not documented, it is not possible to confirm that the infections occurred within Germany. Moreover, when combined with the habitat distribution maps of non-human primates and bats, the suitable habitat ranges of these key reservoir hosts extend far beyond the currently documented orthoebolavirus detection zones. These findings may reflect limitations in current surveillance coverage in certain regions, highlight significant monitoring gaps, and suggest the potential for undetected viral transmission in under-surveilled areas. This underscores the need for more comprehensive and systematic surveillance efforts.

Our phylogenetic analysis of fully annotated orthoebolaviruses genomes from GenBank revealed distinct evolutionary patterns among species in terms of host specificity, geographic distribution, and temporal dynamics, suggesting ongoing evolutionary risks. BOMV primarily infects bats with high L protein amino acid identity (98.643–100 %), indicating conserved adaptive sites, while recent Tanzanian detections further expand the geographic range of BOMV and support Mop condylurus' role as a natural BOMV host [27]. RESTV infected non-human primates (Macaca fascicularis) and mammals (Sus scrofa) with >95 % whole-genome identity, including asymptomatic human cases demonstrating cross-species transmission [21]. BDBV formed temporally distinct clades (human-only in the 2000s; human/non-human primate in the 2010s), suggesting host adaptation. SUDV and TAFV showed high intraspecies conservation with predominant human infections. EBOV displayed distinct phylogenetic clades between humans and non-human primates, likely due to 73 nonsynonymous mutations accumulated during cross-species transmission, including the GP-A82V mutation that enhances human dendritic cell infection [22] and other GP substitutions increasing human tropism while decreasing bat tropism [28], indicating these mutations may potentially influence viral host tropism. These findings confirm GP's critical role as the primary vaccine target [29], while L protein mutations were shown to prolong viral shedding in animal models [26], suggesting potential associations with both prolonged transmission periods and expanded geographical distribution. The identified variations highlight the importance of integrated surveillance strategies combining genomic monitoring, computational prediction, and targeted intervention to address viral evolution.

Nevertheless, this study has two limitations. First, only English and Chinese literature were included in our study, and articles in other languages may have been missed. This limitation may have led to an incomplete representation of geographical distribution or host diversity, as potentially relevant data from non-English/Chinese sources were not captured. Secondly, inconsistent nomenclature for Ebola virus in earlier literature (e.g., variations like ZEBOV, EBOV, and Orthoebolavirus zairense) may have led to the inadvertent exclusion of some relevant studies or sequence data during our literature review and database searches.

5. Conclusion

In summary, this study integrated analyses of orthoebolaviruses' geographic distribution, phylogenetic relationships, and amino acid variations in EBOV, revealing a possible geographic expansion trend and differential host tropism among species. Through evolutionary and mutation profiling, we elucidated inter-species divergence and intra-species variations across different hosts and time periods, identifying 73 amino acid substitutions potentially associated with humans and NHPs. These findings provide valuable references for the development of future therapeutics and vaccines against orthoebolaviruses. Given the persistent threats of orthoebolaviruses to human and animal health, we emphasize implementing sustained integrated surveillance and prioritizing targeted prevention strategies.

CRediT authorship contribution statement

Nuo Cheng: Writing – original draft, Formal analysis, Data curation. Run-Ze Ye: Methodology, Investigation. Yu-Yu Li: Formal analysis, Data curation. Kandeh Bassie Kargbo: Data curation. Li-Li Ren: Writing – review & editing, Validation. Wu-Chun Cao: Writing – review & editing, Validation, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplementary material

Data availability

Data will be made available on request.