Serological evidence for potential yellow fever virus infection in non-human primates, southeastern Mexico

Background Arthropod-borne flaviviruses like dengue virus (DENV) and yellow fever virus (YFV) are major human pathogens. In Latin America, YFV is maintained in sylvatic cycles involving non-human primates (NHP) and forest-dwelling mosquitos. YFV supposedly does not circulate north of Panama. Methods We conducted a serologic study for flaviviruses and other emerging viruses in NHP from southeastern Mexico. A total of thirty sera of black-handed spider monkeys (Ateles geoffroyi, n = 25), black howler monkeys (Alouatta pigra, n = 3), and mantled howler monkeys (Al. palliata, n = 2) sampled in 2012 and 2018 were screened by an indirect immunofluorescence assay (IFA) to detected IgG antibodies against DENV, YFV, Zika virus (ZIKV), West Nile virus (WNV), Rift Valley fever virus, Crimean-Congo hemorrhagic fever virus, Middle East respiratory syndrome coronavirus, and Zaire Ebola virus, and confirmed by plaque reduction neutralization tests (PRNT90) representing all mosquito-borne flavivirus serocomplexes circulating in the Americas. Results A total of 16 sera (53.3%; 95% CI, 34.3–71.7) showed IFA reactivity to at least one tested flavivirus with end-point titers ranging from 1:100 to 1:1000. No serum reacted with other viruses. Monotypic and high mean PRNT90 endpoint YFV titers of 1:246 were found in 3 black-handed spider monkey sera (10.0%; 95% CI, 2.1–26.5) sampled in 2018 in Tabasco, compared to all other flaviviruses tested. Monotypic endpoint PRNT90 titers of 1:28 for Ilheus virus and 1:22 for WNV in serum of black howler monkeys sampled in 2018 in Tabasco suggested additional flavivirus exposure. Conclusions Our findings may suggest unnoticed YFV circulation. Intensification of YFV surveillance in NHP and vectors is warranted in Mexico and potentially other areas considered free of yellow fever.

In Mexico, dengue fever is endemic, with 25,226 cases reported in 2020 and 6746 in 2021 [9].WNV was first reported in Mexico in 2002 in birds, horses, humans and Culex mosquitos at different times and places [7].ZIKV was introduced in the country in 2015, and 11,869 cases have been reported since [9].In contrast to those three endemic flaviviruses, YFV was present in the country from the XIX century until 1963 only, when the country was considered free of yellow fever thanks to extensive campaigns to eliminate the urban vector Aedes aegypti [8].In 2021, according to the Pan American Health Organization (PAHO), four countries reported yellow fever cases in South America (659 reported / 31 confirmed), and 13 countries in Central and South America were considered at risk, ending northwards with Panama (Fig. 1B) [10,11].Typically, only severe cases of yellow fever are identified, and given that approximately 90% of cases are mild or exhibit no symptoms, this may contribute to the lack of reported cases [12].
With circa 200,000 cases and 30,000 deaths annually, YFV is currently endemic in 47 countries in Africa, Central and South America [13,14].Urban and sylvatic flavivirus transmission cycles can occur independently or in parallel.Sylvatic cycles of YFV involving non-human primates (NHP) nowadays prevail in the New World [15,16].YFV infection in New World NHP is very severe and can be fatal, however disease severity is variable among different NHP species [17].Sylvatic YFV cycles involve mosquitos of the genera Haemagogus and Sabethes which have been reported in Southeastern Mexico [18,19].Because the urban YFV vectors, Aedes mosquitoes, are widely distributed throughout the country, the re-emergence of YFV in urban regions could Animal handling was performed by specialized and experienced staff, following protocols of safety, care, and health of each institution.All animals were in good overall conditions at the time of blood sampling (2 -3 mL from the coccygeal vein for each specimen) and samples were stored on ice and immediately transported to the laboratory.Serum samples from all NHP that were available at the two locations were tested, without application of additional inclusion or exclusion criteria.

Results
Initial IFA screening showed that 53% (16/30) of sera were reactive for at least one of the flavivirus antigens tested (WNV, ZIKV, DENV, and YFV), with end-point titers ranging from 1:100 to 1:1000 serum dilution, not allowing unambiguous identification of flavivirus infection histories (Table 1; Fig. 2).No sample was positive against other viruses included in the IFA Biochip, suggesting absence of antigenically related phlebo-, filo or coronaviruses in those Mexican NHP and all samples tested negative for flaviviral RNA in the RT-PCR.
Because antibody cross-reactivity is very common among flaviviruses, we tested all IFA-positive sera by highly specific PRNT against at least one flavivirus of each serocomplex.Both PRNT 50 and PRNT 90 titers were calculated, but we selected PRNT 90 to achieve higher specificity levels by reducing the impact of antibody cross-reactivity leading to inconclusive PRNT 50 results in most sera (Fig. 3A; Table 2).High and monotypic PRNT 90 titers were found against YFV (mean, 245.9;SD, 142.9; range 82.6-348.0) in 12.5% (3/24; 95% CI, 2.1-26.5) of the captive black-handed spider monkeys sampled in Tabasco in 2018 (Table 2; Fig. 3B).The presence of neutralizing antibodies against YFV in NHP in Mexico is unexpected since, to date, YFV has only been reported outside South America in the 1970ies in Panama and Trinidad and Tobago (Fig. 1B), despite a large outbreak in humans and NHP affecting many countries in South America in 2016-2018 [15,16,21].On the other hand, YFV seropositive NHP lived all their lives or for many years in the same location: animal 879 arrived in the park in 1997 and was sampled after 21 years; animal 881 was born in the park in 2007 and sampled after 11 years and 882 arrived in the park in 2009 and was sampled after 9 years.Moreover the ability of Ateles species to elicit an immune response against YFV is supported by previous finding in the 1950s in free-living animals showing the presence of antibodies in black-handed spider monkeys in Panama [22] and in spider monkeys in Brazil [23].The absence of antibodies in black howler monkeys (genus Alouatta) must be interpreted with caution because of the low number of animals tested.However, absence of antibodies in those animals is consistent with high mortality of yellow fever in NHP belonging to Alouatta species exceeding 60%, in contrast to NHP species of the genus Ateles such as spider monkeys which seem to be more resistant to infection and clinical disease.
No ZIKV, DENV-1, DENV-3, and DENV-4-specific neutralizing antibodies could be confirmed by PRNT 90 .One black howler monkey serum previously IFA-reactive for all tested flavivirus showed monotypic PRNT 90 endpoint titers for WNV (16.7%; 95% CI, 0.4-64.1),while another serum previously reactive in the IFA at the dilution 1:100 for ZIKV showed monotypic PRNT 90 endpoint titers for ILHV (16.7%; 95% CI, 0.4-64.1),whereas one black-handed spider monkey serum that had previously been IFA-reactive for all tested flaviviruses presented monotypic endpoint titers for DENV-2 (4.2%; 95% CI, 0.1-21.1)(Table 2).The difference in the results of these two tests (IFA and PRNT) is likely due to the lower IFA specificity due to flavivirus antibody cross-reactivity [24].The absence of ZIKV-specific antibodies is in agreement with previous studies from Brazil showing the absence of ZIKV in NHP in areas with reported high ZIKV dispersion in the human population [21,25].However, recent reports of potential African lineage ZIKV in Brazilian NHP may suggest that continuous surveillance for ZIKV in NHP is required [26].The presence of DENV-specific antibodies in New World NHP has been reported in different countries, such as Brazil, Argentina, and Puerto Rico, without evidence of sylvatic transmission cycles [27].The low DENV seroprevalence and low endpoint titer of 38.2 (Table 2) was compatible with sporadic infection of a captive NHP in a highly endemic area [27].The presence of neutralizing antibodies against WNV and ILHV in this study should be interpreted cautiously because of cross-reactivity against potentially co-circulating flaviviruses from the same serogroups, such as Saint Louis encephalitis virus (SLEV) and Rocio virus (ROCV) (Fig. 1A), which cannot be excluded even when using highly specific PRNT 90 .
Our serological data may support the occurrence of YFV in NHP in Mexico.On the one hand, we cannot exclude that antigenically related flaviviruses could also have elicited cross-reactive antibodies.This hypothesis is compatible with the detection of a monotypic neutralizing antibody response against Wesselsbron virus in a single Brazilian cow during a sero-epidemiological investigation of several thousand sera from livestock and pets [28].On the other hand, no flavivirus belonging to the yellow fever serogroup beyond YFV (e.g., Banzi, Bouboui, Jugra, Potiskum, Saboya, Sepik and Wesselsbron viruses) has ever been reported in the Americas, the aforementioned viruses are all endemic to Africa and Asia [28][29][30][31].Additionally, the three YFV-seropositive individuals presented neutralizing antibodies titers either monotypic or higher than fourfold against the other flavivirus serocomplexes, which is deemed highly robust and decisive for flavivirus serological diagnosis [32].Moreover, the seropositive animals lived in captivity for years within a zoo in Tabasco City within a colony established 30 years ago that was founded by animals seized in the same area.Therefore, despite the fact that the distribution of black-handed spider monkey extends from southeastern Mexico to Panama in which yellow fever may occur, this colony has not been subject to introduction of animals from other countries potentially infected with yellow fever prior to arrival in Mexico.This is relevant because, if confirmed, the presence of YFV would complete the ecological determinants necessary for establishing a sylvatic cycle of yellow fever in southeastern Mexico, i.e., the presence of susceptible wild reservoirs and the presence of invertebrate vectors.
The apparent absence of yellow fever northwards of Panama is not entirely understood.In contrast to other Latin American countries, such as Brazil, in which yellow fever outbreaks often occur, wild NHP populations in Mexico are geographically limited and in small numbers.That is probably one of the reasons because, at least so far, no recent human cases have been reported in Mexico.If YFV was indeed present in Mexico, it would be possible that at any time, YFV may emerge in humans or NHPs.
Our study is chiefly limited by the small number of NHP that were available for testing and the absence of testing for all potentially co-circulating flaviviruses.However, the latter is most critical for WNV, ILHV and BSQV (Fig. 3A) due to likely SLEV co-circulation in Mexico [33] and potential circulation of other flaviviruses reported in Central and South America [32,34,35].In contrast, we covered all  circulating serocomplexes and high neutralizing antibody titers and the absence of co-circulating viruses of the same serogroup support robustness of the data for YFV.

Conclusions
Our findings warrant monitoring of potential YFV infections in free-living and captive NHP as well as mosquito surveillance in Mexico and neighboring countries.Additional confirmation, including direct viral detection, is needed to provide definite assessments of the potential occurrence of YFV in areas north of Panama considered free of yellow fever.
• fast, convenient online submission • thorough peer review by experienced researchers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations maximum visibility for your research: over 100M website views per year

•
At BMC, research is always in progress.

Learn more biomedcentral.com/submissions
Ready to submit your research Ready to submit your research ?Choose BMC and benefit from: ?Choose BMC and benefit from:

Fig. 3
Fig. 3 Reciprocal plaque reduction neutralization (PRNT) endpoint titers against selected flaviviruses in sera from Mexican non-human primates.A PRNT 50 and B PRNT 90

Table 1
Flavivirus antibody detection using an indirect immunofluorescence assay in sera from non-human primates, Southeastern Mexico "-" not reactive, IFA indirect immunofluorescence assay, YFV yellow fever virus, WNV West Nile virus, ZIKV Zika virus, DENV dengue virus

Table 2
Neutralizing antibody endpoint titers against different flaviviruses in sera from non-human primates, Southeastern Mexico PRNT plaque reduction neutralization test, "Inc" Inconclusive, "-" no neutralization observed, JE Japanese encephalitis, SPO Spondweni, YF yellow fever, YFV yellow fever virus, WNV West Nile virus, ZIKV Zika virus, DENV dengue virus, ILHV Ilheus virus, BSQV Bussuquara virus a Although low titers were observed for BSQV, they were not considered positive because the titers obtained were below the entry dilution (1:80) in both PRNT 90 and PRNT 50 b Considered monotypic for ILHV because the titer was higher than the entry dilution (1:80) in the PRNT 50