Outbreaks of human respiratory syncytial virus in wild gorillas highlight the importance of prevention measures and integrated surveillance for risk mitigation

Abstract

Transmission of human respiratory pathogens to wild, human-habituated great apes has been repeatedly documented within research and tourism projects. While the implementation of hygiene measures has significantly reduced the risk of pathogen introduction, vigilant surveillance remains essential to evaluate the effectiveness of the adopted measures and identify additional steps for risk reduction. Here, we combined behavioral observations and pathogen genomic surveillance in non-invasive samples to investigate three outbreaks of respiratory disease in human-habituated western lowland gorillas (Gorilla gorilla gorilla) across four sites within the Sangha Trinational Protected Area Network in the northwestern Congo Basin. Clinical signs of respiratory disease were recorded in three groups of monitored gorillas at two neighboring National Parks in the Central African Republic and Republic of Congo. Human respiratory syncytial viruses were identified as the causative agent for all three documented outbreaks. Genomic analyses revealed two distinct viral types suggesting independent introduction rather than intergroup transmission. All symptomatic individuals recovered. These findings highlight the importance of stringent prevention measures at great ape research sites and the need for addressing the burden of respiratory disease in neighboring human communities. The evolving integrated approach centered on the One Health concept in the Sangha Trinational Protected Area Network is proving beneficial to great ape conservation, the preservation of this high-biodiversity landscape and the public health of local communities.

1. Introduction

Common human respiratory pathogens (e.g., pneumoviruses, human coronavirus OC43, rhinovirus C and human respirovirus 3) have caused severe outbreaks in wild non-human great apes (hereafter great apes) habituated to human presence at research sites in sub-Saharan Africa [1], [2], [3], [4], [5]. Recent re-investigations of these outbreaks using genomic approaches have shown that transmission of human viruses to great apes, and subsequent spread among them, can occur without the need for adaptive mutations [6], underlining the susceptibility of these endangered species and the high risk of transmission at this interface. While the habituation of great apes to human presence may create opportunities for such transmissions, it remains an essential tool for the conservation of these species and their habitat, generating knowledge on their behavior, and offering local opportunities of employment and revenue through tourism [7]. The transmission risk is amplified through the high disease burden of lower respiratory tract infections of the human population in sub-Saharan Africa. With over 650,000 deaths annually caused by lower respiratory tract infections, sub-Saharan Africa accounts for over 25% of the global mortality [8].

In an effort to maximize the advantages of great ape habituation to human presence for research and tourism with respect to the risk of pathogen transmission, specific prevention measures have been adopted to protect great apes from human pathogens [9]. These include individual hygiene measures for staff and visitors, such as wearing a surgical face mask while observing the apes, maintaining a minimal distance of 7–10 m, using designated field clothes, and providing health checkups for staff, including vaccinations. Site-specific actions include sanitation points for hand and footwear disinfection at the entrance of great ape habitats, undergoing a 3–5-day quarantine, and testing for pathogens before visitation (e.g., COVID-19 rapid test). The implementation of these measures however varies broadly depending on each site's activity and available infrastructure [10]. Furthermore, touristic visitors' compliance with the hygiene measures poses additional challenges [11].

The longitudinal monitoring of great ape health has been fundamental to detecting illness and identifying the causative agent – whether human introduced or naturally occurring [1], [2], [3], [6], [12], [13], [14], [15]. This includes daily collection of health-related behavioral data and non-invasive samples (e.g., feces, urine and fruit wedges) for pathogen detection and genomic analysis. Such surveillance is also instrumental in evaluating the effectiveness of hygiene measures on a site-specific basis and guiding feasible improvements.

Beyond the primary risk of pathogen transmission from humans, great ape social behavior is also of consideration in disease transmission dynamics [16], [17], [18]. Within-group transmission dynamics may be a primary driver in disease spread based on mountain gorilla (Gorilla beringei beringei) research [19]. While multi-male groups are common in mountain gorillas [20], [21], western lowland gorilla sociality is largely centered on a single dominant adult male that if lost due to an infectious disease, results in the group disintegration, which could lead to further disease transmission from dispersing members to other groups. Although the spread within the group is the most frequent route for transmission, evidence suggests some western lowland gorilla populations feature group members that are socially connected with individuals via familiarity or even relatedness in neighboring conspecific social groups [17], [22], [23], [24], [25]. Nearly a decade of monitoring social interactions of four western lowland gorilla groups inhabiting the Nouablé-Ndoki National Park in the Republic of Congo, found individuals maintained extended social networks outside of their immediate group [26]. While such encounters may be generally rare (and even less likely to occur when individuals are sick) compared to within-group interactions, they still can increase the number of associates and should be considered as a potential pathway for group-to-group disease transmission.

In this study, we investigated respiratory disease outbreaks in three groups of wild, human habituated western lowland gorillas living in the World Heritage site of the Sangha Trinational Protected Area Network in the northwestern Congo Basin. We combined behavioral monitoring with PCR diagnostics and genomic analysis of fecal samples to characterize the outbreaks and the causative agents, and discuss future strategies to further tackle respiratory disease transmission at this interface.

2. Methods

2.1. Field sites

The Sangha Trinational Protected Area Network (TNS) is comprised of the Dzanga-Sangha Protected Areas (DSPA) in the Central African Republic (CAR), Lobéké National Park in Cameroon, and Nouabalé-Ndoki National Park (NNNP) in the Republic of Congo (see Fig. 1).

Fig. 1.

Map of the Sangha Trinational Protected Area Network, including the Dzanga-Sangha Protected Areas (DSPA, Central African Republic), the Nouabalé-Ndoki National Park (NNNP, Republic of Congo), and the Lobéké National Park (Cameroon). Show are the locations of the Mayele (DSPA), Kingo and Loya (NNNP) habituated gorilla groups. Red dots indicate villages.

The DSPA is home to WWF's long-term Primate Habituation Program, which at the time of this study monitored three groups of western lowland gorillas (Mayele, Mata and Makumba) from adjacent sites (Bai Hokou and Mongambe, approximately 10 km apart) (see Fig. 1). People employed by the Primate Habituation Program live within the DSPA, either in Bayanga or in neighboring villages.

The NNNP on the Republic of Congo side of the TNS was home to three habituated western lowland gorilla groups at the time of this study. The Goualougo Triangle was the location of the Loya group situated approximately 30 km away from the other two habituated western lowland gorilla groups (Kingo and Buka) located at the Mondika research station (see Fig. 1). The area around Mondika was originally part of the Kabo Forestry Management until its protected status was elevated with inclusion into the NNNP [27]. Many people employed at the Mondika site of the NNNP live in or around Bayanga in CAR, and there is regular movement of staff (e.g., trackers and field assistants) from Bayanga via boat on the Sangha River crossing the border to the Republic of Congo to the field sites of the NNNP.

All visitors and staff members of these TNS research sites are subjected to standard hygiene measures: proof of vaccination against certain pathogens (e.g., measles virus), denial of entry into great ape habitats with signs of respiratory illness, testing for SARS-CoV-2 using lateral flow quick tests, use of sanitation points for hand and footwear disinfection when entering and exiting great ape habitats, wearing a surgical face mask and keeping a minimum distance of 7–10 m when observing the animals.

2.2. Behavioral monitoring

Clinical signs were recorded on a daily basis by trained Research Assistants at DSPA and NNNP and included overt detectable signs associated with a potential respiratory illness (e.g., coughing, sneezing) at all study sites [28], [29], [30] or behaviors associated with declining health (e.g., decreased play behavior, a decline in food intake, lethargy, etc.). The Animal Observer configuration used at Goualougo and Mondika research stations includes a focal observational health assessment form, which is completed at the end of each daily follow on all individual gorillas present and adequately viewed. We defined intergroup encounters as when the focal group detected one or more gorillas from outside the study group through visual or auditory cues or when a non-group member directly approached a member of the focal group [31].

2.3. Non-invasive sample collection

Fecal samples were routinely collected as part of the longitudinal health monitoring programs in place at both DSPA and NNNP sites, which entail a monthly sample collection from each habituated individual and intensified sampling from all group members during disease outbreaks. Samples were collected in Nucleic Acid Preservation buffer and stored at room temperature until storage at −20 °C was available, depending on infrastructure at each site. During the respiratory disease outbreaks, the following fecal samples were collected: 10 from the Mayele group (DSPA), 14 from the Kingo group (Mondika, NNNP) and nine from the Loya group (Goualougo, NNNP). Samples were exported to Germany for laboratory analyses.

2.4. Laboratory investigations and genomic analyses

Testing of fecal samples from the Mayele group (DSPA) for common human respiratory pathogens was initially conducted in a field laboratory at the WWF project site in Bayanga, where a conventional PCR system is in place and used for rapid identification of pathogens infecting wildlife [2]. Nucleic acids were extracted using the Roboklon Stool DNA Kit according to the manufacturer's instructions. Synthesis of cDNA was performed by using Superscript IV reverse transcriptase (Thermo Fisher) or Lunascript (New England Biolabs, UK). All samples were tested for the following viruses using a combination of conventional and real-time PCRs: pneumoviruses, coronaviruses, paramyxoviruses, rhinoviruses, orthomyxoviruses and orthopoxviruses. Positive PCR products were analyzed by Sanger sequencing (see detailed analysis conditions in the Supplementary material).

To obtain more comprehensive viral genomic information, we implemented a target enrichment step using an RNA-bait kit designed to target the genome of major human respiratory viruses [6] (synthesized by a service provider, MYBaits®, see supplementary material) coupled with high-throughput sequencing. Dual-index libraries were built from double stranded cDNA using the NEBNext Ultra II DNA library preparation kit (New England Biolabs). Libraries underwent two rounds of overnight hybridization at 65 °C with the above-mentioned baits. Enriched libraries were then diluted to the recommended concentration and sequenced on an Illumina MiSeq.

2.5. Bioinformatic and phylogenetic analyses

Reads were trimmed and low-quality reads removed using Trimmomatic [32] and read deduplication was performed with Picard [33]. Filtered reads were mapped to a RSV A (accession number MZ515572.1) or RSV B (accession number OR326772.1) reference genome using BWA-MEM2 [34]. Consensus sequences were generated using Geneious Prime v11.0.15 + 10 by calling bases at positions covered by at least five unique reads for low-coverage genomes or 10 unique reads for high-coverage genomes, and with agreement of 95% on nucleotide frequency among reads. The Nextclade web-based tool [35] was used for initial phylogenetic placement and genotyping. For whole-genome phylogenetic analyses, we used a dataset of Nextclade comprising a selection of global representatives of each RSV clade, containing 88 RSV B and 125 RSV A genomes [36]. We aligned genome sequences using MAFFT [37]. Identical sites and sites containing ambiguities (Ns) were stripped using Geneious; therefore, only variable sites were included in the analysis. Maximum-likelihood phylogenetic analyses were performed using IQ-tree [38] with ModelFinder [39]. The resulting tree was visualized and annotated in iTOL [40].

3. Results

3.1. Behavioral monitoring

The first manifestation of respiratory disease was observed at the end of August 2019 in the Mayele group (DSPA) and included coughing, sneezing, presence of secretion from nose and eyes, and fatigue. In total, eight out of 11 individuals of the group became symptomatic. The majority was affected in a three-week span in September but few individuals were symptomatic until late October. Similar clinical signs were observed after a few weeks in the Kingo and Loya groups within the NNNP. The outbreak in the Kingo group started in October 2019 and affected nine out of 12 individuals. The Loya group began showing clinical signs at the end of November 2019, with five out of six individuals being affected (Table 1). All individuals in the three outbreaks recovered fully by the end of December 2019. No intergroup encounters between known groups were observed during the disease outbreaks.

Table 1.

Prevalence and time periods of respiratory disease signs among affected western lowland gorilla groups and results of genomic analysis including PCR-based methods and whole genome sequencing. *: This individual did not show overt clinical signs. These may have been missed by observers due to a milder form of the infection. ND: Not detected; samples from the Mayele group were positive in a semi-nested conventional PCR targeting the gene coding for the large (L) protein of pneumoviruses but negative in the RSV qPCR, therefore no ct value is available.

Site	Group	Number of symptomatic individuals	Proportion of symptomatic individuals sampled	Proportion of positive samples from symptomatic individuals	Outbreak time period	Individuals with positive PCR test	Sample collection date	Ct value of RSV qPCR	Sanger sequencing	Coverage of reference sequence	G-clade (Joya et al)	Clade (RGCC nomenclature)
									L-Gene, 264 bp
Dzanga-Sangha Protected Areas (CAR)	Mayele	8 out of 11	75% (6/8)	50% (3/6)	August–September 2019	Mayele	01-09-2019	ND	RSV A	86.30%	GA2.3.5	A.D.2.2
						Lungu	05-09-2019	ND		95.60%	GA2.3.5	A.D.2.2
						Duma	17-09-2019	ND	inconclusive	no genome recovered
Mondika,	Kingo	9 out of 12			October–November 2019	Ike	31-10-2019	38.94	RSV B	no genome recovered
Nouabalé-Ndoki National Park			44.4% (4/9)	75% (3/4)		Kingo	11–11-2019	38.74		78.30%	GB5.0.5a	B.D.4.1.1
(ROC)						Ekendi	12-11-2019	39.12		12.70%	GB5.0.5a	B.D.4.1.1
Goualougo,	Loya	5 out of 6	20% (1/5)	100% (1/1)	November–December 2019	Loya	10-12-2019	39.49	RSV B	no genome recovered
Nouabalé-Ndoki National Park
(ROC)						Kao*	10-12-2019	36.33		no genome recovered

3.2. Laboratory investigations and genomic analyses

Preliminary analyses of fecal samples collected from the Mayele group (DSPA) performed in the field laboratory of Bayanga indicated the presence of a human respiratory syncytial virus (RSV) in three samples collected from symptomatic individuals. These results were later independently confirmed in Germany. Laboratory testing of the samples from NNNP led to the identification of three RSV-positive samples from the Kingo group, and two RSV-positive from the Loya group. In 2016, the human respiratory syncytial virus was renamed human orthopneumovirus [41], but the previous naming is still widely used and accepted. The virus is classified in two antigenic subgroups: RSV A and RSV B, which are further divided into various genotypes and clades defined by variation in the sequence coding for the glycoprotein [36]. Sanger sequencing of a 264 bp fragment of the L-gene revealed two distinct viral types: RSV A in the Mayele group (DSPA) and RSV B in the Kingo and Loya groups (NNNP) (Table 1).

Target enrichment coupled with high-throughput sequencing allowed us to reconstruct two nearly complete RSV A genomes from the outbreak in the Mayele group, and two partial RSV B genomes from the Kingo group (Table 1). In both instances, viral sequences sampled from different individuals were identical. For an initial phylogenetic placement and genotyping of the identified RSVs, we used the Nextclade web-based tool [35], which assigned the RSV A found in Mayele (DSPA) to the GA2.3.5 (G-Clade, [42]) and A.D.2.2 [36] clades and the RSV B found in Kingo (Mondika, NNNP) to the GB5.0.5a (G-Clade, [42]) and B.D.4.1.1 [36] clades (Table 1). Maximum-likelihood phylogenetic analyses placed the viruses found in the gorillas within the diversity of global representatives of human strains, confirming the initial placement obtained with Nextclade (Fig. 2).

Fig. 2.

Maximum-likelihood phylogenetic trees of the human respiratory syncytial virus (RSV) A and B. Sequences of the viruses found in the gorillas are highlighted in bold. Black circels indicate the clades to which the viruses identified in the gorillas belong to. Triangles indicate clades that were collapsed for a better visualization of the trees. Black dots indicate branch support values. The scale bars are expressed in substitutions per variable site.

4. Discussion

Here, we described three respiratory disease outbreaks caused by different human RSV types that occurred at two field stations closely in time in neighboring gorilla groups. Given the frequent movement of staff and occasionally of tourists between these stations, we initially hypothesized that the same virus may have been involved. This hypothesis was discarded by the identification of at least two distinct viral types (RSV A and RSV B). The comparison of the RSV A genomes in two individuals of the Mayele group (DSPA) showed they were identical, suggesting that no mutation occurred during spread within the group. The same was observed for the RSV B in the Kingo group (Mondika), although only 12% of the genome could be compared. These findings are in line with previous genomic studies of pneumovirus outbreaks in wild chimpanzees [6], confirming that spread within great ape groups occurs without requiring adaptive mutations. Due to the lack of genomic information from the RSV B in Loya (Goualougo), we could not determine whether the virus involved was the same as in Kingo (Mondika). Since no intergroup encounters were observed during the outbreak period, and given the considerable distance between the Kingo and Loya groups (> 30 km) we hypothesize that the most plausible scenario is that these infections were the result of independent transmission from humans. Phylogenetic analyses showed that the two RSVs found in the gorillas cluster within the known diversity of human strains, supporting the hypothesis of human-to-gorilla transmissions. The absence of viral genomic data from the studied regions remains an important limitation which prevented a direct comparison with local patterns of RSV circulation in humans.

Despite observing clinical signs in almost all individuals of the affected groups, we detected RSV in only a small fraction of the collected samples, and we were not able to obtain complete viral genomic information from all positive samples. This discrepancy reflects the partial limitations of using fecal samples for non-invasive molecular diagnostics of respiratory diseases. The probability of detecting viral RNA in fecal samples collected opportunistically from wild animals is influenced number of factors, including viral shedding patterns, which might vary among sampled individuals, and opportunity of sampling an individual at an ideal time (e.g., when clinical signs are more pronounced). A number of additional factors may have hampered the detection of the virus in more samples and obtaining a better genome coverage. For example, storage and cooling of samples in remote field locations can be logistically challenging, and several days/weeks may pass between collection and refrigeration, potentially leading to RNA/DNA degradation [43]. Although these limitations do not allow for a precise estimate of disease prevalence in the groups, our results confirm once again that non-invasive sampling remains valuable for diagnosing infectious diseases in wildlife without interfering with their behavior and wellbeing. Collecting additional non-invasive sample types, such a food remains that contain saliva, may increase the chances of detecting viral presence in a greater number of group members, especially from younger individuals (i.e., infants) whose feces are difficult to obtain.

In humans, RSV infections are the predominant cause of childhood acute lower respiratory infections [18], highlighting the potential for transmission to great apes, as shown by the many studies reporting infections in wild chimpanzees, gorillas and bonobos with this virus in the past [2], [3], [44], [45]. While seasonality in temperate regions is thoroughly researched and understood, the patterns in tropical countries remain less clear [19], [20]. A previous RSV outbreak in a distinct habituated gorilla group in DSPA in 2012, associated with respiratory illness and RSVA detection in the local human population, also occurred in the August–September months [3]. A recent study of acute RSV cases in children in the Central African Republic showed a significantly higher prevalence between August and December [46]. Together with our findings, this may be suggestive of a local seasonal pattern of RSV circulation towards the transition to the heavy rainy season. Identifying such local and seasonal peaks could inform timely public health responses and temporary adjustments of prevention measures at research sites, such as quarantine protocols or targeted pathogen testing. Depending on available infrastructure, both PCR-based methods and easy-to-use multi viral pathogens lateral flow tests could be implemented. While hunting and gathering activities are not allowed in National Parks, these are occasionally observed, therefore human presence other than staff and visitors of the primate habituation programs in the gorilla habitat cannot be excluded. Although RSV survival on surfaces is limited to a few hours [47], and thus acquiring the infection through a contaminated environmental source seems unlikely, persistence of the virus across seasons and varying rainforest temperatures and humidity should be assessed.

Given the frequent human-to-great ape transmission of common respiratory viruses reported over the past 20 years [1], [2], [3], [6], [48], supporting national health authorities in establishing surveillance for respiratory infections among people living near great ape habitats has the potential to benefit both local communities and wildlife. A recent study performed in the communities bordering Kibale National Park in Uganda, showed that human respiratory viruses causing outbreaks in local chimpanzee groups were commonly found to cause illness in children, while adults often carried the same viruses asymptomatically [20]. The initiative also identified school reopening as periods of increased viral detection, highlighting a seasonal peak likely due to the aggregation of children and identifying a clear window where strengthened surveillance among great ape research staff could be implemented.

Based on the outbreaks documented in this and many other studies and the subsequent COVID-19 pandemic, one imperative has been to address the high health risk the daily teams and visitors pose to the habituated great apes [49]. Countering the risks included strengthening prevention measures such as strict adherence to facial masks and increased distancing to at least 10 m from gorillas. Adaptive management also included initiating off and on-site quarantine periods for staff, rapid testing for COVID-19 prior to entering the stations and vaccinations for staff as soon as they became available. In spite of the many COVID-19 cases detected worldwide, including in staff and visitors of great ape research sites, no outbreak of SARS-CoV-2 has been reported from habituated wild great apes, to the best of our knowledge. This highlights how these measures are contributing effectively to reducing the risk of SARS-CoV-2 transmission to human-habituated wild great apes.

Future tourism efforts in the region will benefit from recent standardization and implementation of protocols developed at these research sites. Considering the risk respiratory viruses have repeatedly played in great ape populations across Africa, addressing the pathways that allow emergence and spread is crucial for effective preventive responses. TNS field research stations are rather unique when compared to stations in west or east Africa through the support of larger numbers of local community members including indigenous peoples in essential front-line roles as trackers and auxiliary researchers. These jobs and the movement between distant communities across an international border are providing critical long-term local employment opportunities and positive relationships with the communities. To protect vulnerable great ape populations, habituation projects should continue observing and adjusting hygiene measures, and promoting One Health education among staff across regional partnering institutions and visitors. RSV vaccinations, which have been recently developed and are currently recommended for older, high-risk and pregnant individuals [50] could also be considered as additional prevention tool. Supporting the regional implementation of RSV vaccines in the recommended population could help reduce disease burden and may also decrease viral transmission among humans, thereby lowering the risk of spillover to great apes. These efforts, combined with community-based public health initiatives, can have positive impacts on safeguarding endangered species and improving the health of local communities that share their landscapes. Our findings underscore the urgent need for integrative surveillance systems targeting both wildlife and human health to better understand transmission dynamics and inform targeted prevention measures.

CRediT authorship contribution statement

Moritz J.S. Jochum: Writing – original draft, Visualization, Methodology, Investigation, Formal analysis, Data curation. Frédéric S. Singa-Niatou: Writing – review & editing, Resources, Project administration, Methodology, Investigation, Data curation. Crickette Sanz: Writing – review & editing, Supervision, Resources, Conceptualization. Sean Brogan: Writing – review & editing, Investigation, Data curation. Therese Löhrich: Writing – review & editing, Investigation, Data curation. Terrence Fuh Neba: Writing – review & editing, Resources, Investigation. Fabian H. Leendertz: Writing – review & editing, Supervision, Resources, Funding acquisition, Conceptualization. David Morgan: Writing – review & editing, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Conceptualization. Livia V. Patrono: Writing – review & editing, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Conceptualization.

Funding

This work was supported by the ARCUS Foundation grant No. 640 G-PGM-2107-3516. Additional funding was provided by the German Research Council Project 639 LE1813/11–1 and LE1813/14–1 (Great Ape Health in Tropical Africa), Ndoki Foundation, Lincoln Park Zoo, Indianapolis Zoo, Saint Louis Zoo, Woodland Park Zoo, Zoo Atlanta, Cincinnati Zoo and Botanical Garden, Washington University in Saint Louis, Margot Marsh Biodiversity Fund (MMBF) and the Wildlife Conservation Society.

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 1

Supplementary material 2

Data Availability

Raw reads from this study are deposited in the European Nucleotide Archive (ENA) under project accession number PRJEB106960 and sample accession numbers ERS28513856, ERS28513855, ERS28513854 and ERS28513853. Behavioral data and field records are available from the authors upon reasonable request.

Ethics and Declaration of interest

This research was strictly non-invasive and no ethical approval was required. The authors declare no conflict of interest.