Mayaro virus detection in patients from rural and urban areas in Trinidad and Tobago during the Chikungunya and Zika virus outbreaks

ABSTRACT

Mayaro virus, which can often go undetected due to its clinical manifestations and intimate alignment with dengue and chikungunya viruses, is one of the most neglected arboviruses. The virus has been found in several outbreaks, where a moderate-to-severe and potentially incapacitating joint disease has been observed. MAYV usually circulates in a sylvan cycle of forest mosquitoes and vertebrates, causing sporadic sylvatic infections to humans, and some outbreaks in sub-urban areas. This study focuses on the demonstration of the possible co-circulation of Mayaro virus with chikungunya virus and Zika virus during the outbreaks that occurred in Trinidad and Tobago in 2014 and 2016, respectively. Acute samples from patients who previously tested negative for chikungunya, dengue, and Zika, and specifically exhibiting joint pain were selected and investigated for the presence of Mayaro virus genome using real-time RT-PCR techniques. Nine persons were shown to be positive for Mayaro virus during the chikungunya outbreak of 2014, while no one during the Zika outbreak in 2016. Five results correspond to persons living in highly urbanized areas across Trinidad. These findings provide evidence that multiple arboviral circulations are possible and could easily go undetected especially during outbreak situations. Our study is the first to demonstrate the possible co-circulation of Mayaro and chikungunya viruses and the occurrence of human cases for both diseases during an outbreak in the Caribbean. A possible change in the pattern of distribution of human cases to more urbanized areas is also discussed.

Introduction

Mayaro virus (MAYV) was first isolated in 1954 from blood samples of five rural workers in the Mayaro District of Trinidad and Tobago [1] and later isolated from the Mansonia venezuelensis mosquito [2]. The main vector of MAYV belongs to the Haemagogus genus with the Aedes spp. and the Culex spp. Possibly acting as unorthodox secondary vectors [3] Figure 1, Figure 2.

Figure 1.

Areas within Trinidad and Tobago where MAYV infection occurred

Figure 2.

Location of Trinidad and Tobago in the Caribbean

MAYV is an enveloped virus of the Togaviridae family and Alphavirus genus. Its genome is a positive-sense single-stranded RNA [4]. MAYV has been classified into three major genotypes. Genotype D has been noted to be widely dispersed within areas of Trinidad and Tobago and South America [5,6]. Genotype L is restricted to Brazil and Haiti [7] while genotype N has been isolated in Peru [8].

MAYV belongs to the Semliki Forest virus complex which comprises several other viruses including chikungunya virus (CHIKV), O’nyong nyong virus, Semliki Forest virus, and getah virus. This complex within the Alphavirus genus forms a serological group that shares a few common antigenic sites. This results in cross-reactivity [9] that may cause misdiagnoses and underreport of Mayaro virus disease (MAYVD) in endemic areas where multiple arboviruses circulate [10].

MAYVD is an acute, non-fatal febrile illness of 3–5 days in duration, characterized by headache, retroorbital pain, arthralgias, arthritis, myalgias, vomiting, diarrhea, and rash [11]. It has been reported that symptoms of MAYVD including distal arthralgia, myalgia, and a skin rash with joint pain can last for months [12,13], and recurrent arthralgias after the infection have also been described [14]. Clinical signs and symptoms of the MAYVD are very similar to those observed during the dengue virus (DENV) and CHIKV diseases. This can also often lead to misdiagnosis of MAYVD and in turn under-reporting of the virus [15,16].

Thus far, MAYV has been restricted to subtropical and tropical climates of the Americas. There have been reports of MAYV infections occurring in South and Central America with a few cases in the Caribbean [17]. Sporadic outbreaks of MAYV have occurred in Trinidad, Venezuela, Brazil, and Ecuador [18]. While not frequent, imported cases of MAYV to North America and Europe have been reported due to persons visiting countries in South America, where MAYV is considered endemic [19].

The purpose of this study was to detect the presence of MAYV in acute serum samples of patients exhibiting arthralgias who tested negative for CHIKV, DENV, and Zika Virus (ZIKV) during the CHIKV and ZIKV outbreaks in 2014 and 2016, respectively. Additionally, it is also aimed at highlighting the importance of proper surveillance programs for the detection of opportunistic vector-borne viruses that may normally go undetected as these can later cause epidemics.

Materials and methods

Clinical samples

In this study, clinical samples from the Trinidad and Tobago Public Health Laboratory (TPHL) were sent to the Caribbean Public Health Agency (CARPHA) during the chikungunya and Zika outbreaks in 2014 and 2016, respectively. Samples used for this study were taken less than 7 days after the onset of symptoms, from patients who presented symptoms compatible with an arboviral disease (fever, headache, retroorbital pain, malaise) and the presence of arthralgias. A total of 65 samples were identified to meet those sample selection criteria.

Viral RNA extraction

The isolation of viral RNA was performed using 140 µL of the previously stored serum samples. The QIAamp Viral RNA Mini Kit (Qiagen, USA) was used to isolate viral RNA as described by the manufacturer’s instructions. One negative extraction control (nuclease free water) was included during each extraction step.

RT-PCR assay

All samples which met the selection criteria were tested for DENV, CHIKV, and ZIKV using primers and probes for the Trioplex real-time RT-PCR assay (rRT-PCR) for these viruses (CDC, San Juan, Puerto Rico). The samples selected for MAYV testing were further tested using the MAYV Genesig® Easy Kit (Primerdesign LTD, UK). For this assay, 5 µL of template was mixed with 15 µL of a master mix containing Genesig® primers and probes. 5 µl of negative extraction control template and 5 µl of nuclease free water was used as a negative control for the assay.

Limitations

A relatively small number of samples were obtained for testing after the application of the selection criteria. Due to the low viremia titer in the blood beyond 7 days, samples taken beyond this period could not be used for MAYV testing by PCR. Furthermore, some specimens did not have sufficient volumes of serum for extractions while some laboratory investigation forms did not have any signs and symptoms recorded. Although these samples would have been tested for either CHIKV or ZIKV (and also for DENV) during their respective outbreaks, they were not selected for this study as they did not meet the study’s sample selection criteria for MAYV testing (no records for arthralgias). Due to the limited volume of the original sample after validation and testing, there was no more sample left to proceed to sequencing.

Results

The following results are summarized in Table 1. During 2014, the CARPHA Medical Microbiology Laboratory (CMML) received 666 acute samples from Trinidad and Tobago for CHIKV investigation. According to the CMML’s testing algorithms, all samples were tested for CHIKV and DENV. In total, 278 samples (41.74%) were confirmed positive for CHIKV and 258 (38.73%) for DENV. From the remaining 170 negative samples (25.5% of the total), 17 samples (2.55%) were selected for MAYV detection.

Table 1.

Results for the detection of the different arbovirus from the 2014 and 2016 outbreak in Trinidad and Tobago. Italics: this study. N/A: Not applicable.

	Samples received for CHIKV or ZIKV testing	CHIKV rRT-PCR ±	ZIKV rRT-PCR ±	DENV rRT-PCR ±	No. of negative samples and arthralgias	MAYV rRT-PCR positive results	No. of results from urban areas	No. of results from rural areas	Unknown location
2014 CHIKV outbreak	666	278/378	-	258/408	17	9	5	3	1
2016 ZIKV outbreak	2,259	15/2,244	719/1,540	9/2,250	48	0	N/A	N/A	N/A

Nine (1.35%) out of the 17 samples tested positive for MAYV. Of those nine samples, three corresponded to individuals that resided in rural areas (The Whim and Patience Hill in Tobago and Las Cuevas in Trinidad), whereas five samples were from patients who resided in urban areas in Trinidad (San Juan, Diego Martin, Barataria, Morvant, and San Fernando) (Figure 1). The residence for one patient was not specified on the laboratory investigation form (Figure 2).

In 2016, 2,259 acute samples were received for ZIKV testing. In total, 719 samples (31.82%) were confirmed positive for ZIKV, 15 (0.66%) for CHIKV, and 9 (0.40%) were positive for DENV. Based on the selection criteria of samples, 48 samples were selected for MAYV testing. All samples selected from the 2016 ZIKV outbreak tested negative for MAYV. Table 2 shows the demographics and location of the patients whose samples were selected for this study.

Table 2.

Demographics and results for samples selected for MAYV testing

Sample ID	Demographics	Coordinates	PCR Ct Value	Results
2014 CHIKV outbreak
001000	San Fernando	10° 16ʹ 52” N 61° 27ʹ 1” W	No Ct	Negative
001001	The Whim, Tobago	11° 12ʹ 0” N 60° 45ʹ 0” W	34.53	Positive
001002	San Juan*	10° 38ʹ 59” N 61° 26ʹ 59” W	35.52	Positive
001003	Patience Hill, Tobago	11° 11ʹ 0” N 60° 46ʹ 0” W	34.84	Positive
001004	Unknown	N/A	35.17	Positive
001005	Unknown	N/A	No Ct	Negative
001006	Diego Martin*	10° 42ʹ 2” N 61° 32ʹ 3” W	34.32	Positive
001007	Barataria*	10° 38ʹ 59” N 61° 27ʹ 59” W	35.09	Positive
001008	Las Cuevas	10° 47ʹ 0” N 61° 23ʹ 0” W	34.51	Positive
001009	Moruga	10° 06ʹ 42” N 61° 17ʹ 9” W	No Ct	Negative
001010	Barataria	10° 38ʹ 59” N 61° 27ʹ 59” W	No Ct	Negative
001011	Penal	10° 09ʹ 60” N 61° 27ʹ 59” W	No Ct	Negative
001012	Morvant*	10° 38ʹ 59” N 61° 27ʹ 59” W	35.00	Positive
001013	Maracas	10° 45ʹ 0” N 61° 26ʹ 0” W	No Ct	Negative
001014	Patience Hill, Tobago	11° 11ʹ 0” N 60° 46ʹ 0” W	No Ct	Negative
001015	Arima	10°37ʹ59” N 61°16ʹ59” W	No Ct	Negative
001016	San Fernando*	10° 16ʹ 52” N 61°27ʹ1” W	34.08	Positive
2016 ZIKV outbreak
001017	Chaguanas	10° 31ʹ 2ʹ’ N 61° 24ʹ 41ʹ’ W	No Ct	Negative
001018	Port of Spain	10° 39ʹ 50” N 61° 31ʹ 0” W	No Ct	Negative
001019	Arima/Wallerfield	10° 37ʹ 0” N 61° 13ʹ 0” W	No Ct	Negative
001020	Freeport	10° 27ʹ 0” N 61° 25ʹ 0” W	No Ct	Negative
001021	New Settlement	10° 31ʹ 39” N 61°24ʹ58” W	No Ct	Negative
001022	Chaguanas	10° 31ʹ 2ʹ’ N 61° 24ʹ 41ʹ’ W	No Ct	Negative
001023	Gulf View	10° 15ʹ 36ʹ’ N 61° 28ʹ9’’ W	No Ct	Negative
001024	Barataria	10° 38ʹ 59” N 61° 27ʹ 59” W	No Ct	Negative
001025	New Grant	10° 17ʹ 0” N 61° 19ʹ 0” W	No Ct	Negative
001026	Marabella	10° 18ʹ 22” N 61° 26ʹ 48” W	No Ct	Negative
001027	San Fernando	10° 16ʹ 53” N 61° 27ʹ 1” W	No Ct	Negative
001028	Port of Spain	10° 39ʹ 50” N 61° 31ʹ 0” W	No Ct	Negative
001029	San Fernando	10° 16ʹ 53” N 61° 27ʹ 1” W	No Ct	Negative
001030	St Ann’s	10° 40ʹ 59” N 61° 28ʹ 0” W	No Ct	Negative
001031	St Ann’s	10° 40ʹ 59” N 61° 28ʹ 0” W	No Ct	Negative
001032	Glencoe	10° 40ʹ 53” N 61° 34ʹ 26” W	No Ct	Negative
001033	Glencoe	10° 40ʹ 53” N 61° 34ʹ 26” W	No Ct	Negative
001034	Lange Park	10° 31ʹ 2ʹ’ N 61° 24ʹ 41ʹ’ W	No Ct	Negative
001035	Tunapuna	10° 38′ 0” N 61° 23′ 0” W	No Ct	Negative
001036	Point-a-Pierre	10° 18ʹ 59” N 61° 27ʹ 59” W	No Ct	Negative
001037	Macoya	10° 38′ 30″ N 61° 23′ 5″ W	No Ct	Negative
001038	Diego Martin	10° 42ʹ 2” N 61° 32ʹ 3” W	No Ct	Negative
001039	Penal	10° 09ʹ 59” N 61° 27ʹ 59” W	No Ct	Negative
001040	Fyzabad	10° 10ʹ 35” N 61° 32ʹ 48” W	No Ct	Negative
001041	Princes Town	10° 16ʹ 3” N 61° 22ʹ 55” W	No Ct	Negative
001042	Penal	10° 09ʹ 59” N 61° 27ʹ 59 “ W	No Ct	Negative
001043	Fyzabad	10° 10ʹ 35” N 61° 32ʹ 48” W	No Ct	Negative
001044	Penal	10° 09ʹ 59” N 61° 27ʹ 59” W	No Ct	Negative
001045	Penal	10° 09ʹ 59” N 61° 27ʹ 59 W	No Ct	Negative
001046	Westmoorings	10° 40ʹ 31” N 61° 33ʹ 27” W	No Ct	Negative
001047	Penal	10° 09ʹ 59” N 61° 27ʹ 59” W	No Ct	Negative
001048	Unknown	N/A	No Ct	Negative
001049	El Dorado	10° 38ʹ 0” N 61° 23ʹ 0” W	No Ct	Negative
001050	Longdenville	10° 31ʹ 0” N 61° 23ʹ 0” W	No Ct	Negative
001051	Union Hall	10° 15ʹ 10” N 61° 27ʹ 18”W	No Ct	Negative
001052	Belmont	10° 40ʹ 00ʹ’ N 61° 30ʹ 00ʹ’ W	No Ct	Negative
001053	Rio Claro	10° 19ʹ 9” N 61° 7ʹ 48” W	No Ct	Negative
001054	Barataria	10° 38ʹ 59” N 61° 27ʹ 59” W	No Ct	Negative
001055	Diego Martin	10° 42ʹ 2” N 61° 32ʹ 3” W	No Ct	Negative
001056	Siparia	10° 8ʹ 4ʹ’ N 61° 29ʹ 46ʹ’ W	No Ct	Negative
001057	Matura	10° 40ʹ 0” N 61° 4ʹ 0” W	No Ct	Negative
001058	Carenage	10° 41ʹ 0” N 61° 36ʹ 0” W	No Ct	Negative
001059	Diego Martin	10° 42ʹ 2” N 61° 32ʹ 3” W	No Ct	Negative
001060	Unknown	N/A	No Ct	Negative
001061	Unknown	N/A	No Ct	Negative
001062	Siparia	10° 07ʹ 59” N 61° 29ʹ 59” W	No Ct	Negative
001063	La Romaine	10° 14ʹ 4” N 61° 29ʹ 46” W	No Ct	Negative
001064	Maraval	10° 41ʹ 59” N 61° 30ʹ 59” W	No Ct	Negative

Discussion

The initial discovery of MAYV in Trinidad was in the rural area of Mayaro district [1,8], a southeastern location in the country. From the results of this study, two symptomatic persons residing in different rural areas in Tobago and one person residing in a northern rural town in Trinidad were infected with MAYV. According to previous studies, MAYV infection in humans living in rural areas could be accidental and confined to forested environments. The occurrence of MAYVD in humans is often observed during outbreaks for other arboviruses, i.e. DENV [20]. Our findings indicate that MAYV was circulating during the 2014 chikungunya outbreak in Trinidad and Tobago, but not detected in our study during the 2016 Zika outbreak.

It has been proposed that Haemagogus mosquitoes are the main vectors for MAYV. A recent study [21] has shown the wide-spread distribution of Haemagogus species throughout the island of Trinidad. The authors noted that Haemagogus mosquito species which had restricted distribution in Trinidad up to 1995 are now found to be widespread on the island, being Haemagogus janthinomys the dominant species. The findings of that study also suggested that there may be alternative hosts and reservoirs of this virus in the sylvatic cycle in Trinidad, other than non-human primates. In the case of MAYVD, this could explain the expansion of the disease to areas where the virus was not previously isolated. Also, it has been established through proximity analysis that population settlements within a 1 km buffer of the forest peripherals may be at risk for any emerging arboviral disease associated with these mosquito vectors. Although the Haemagogus species are known to have diurnal habits and reside in tree canopies, some members of the Haemagogus species have demonstrated the ability to adapt to human environments [22]. All these data suggest that a combined territorial expansion of the virus and its vector has occurred and has favored the appearance of human infections by MAYV.

There has been a concern that MAYV (similarly to yellow fever virus) can shift its patterns of transmission from a sylvatic cycle to an urbanized transmission, borne by Aedes mosquitoes [15]. A report from 2016 [23] described a case of DENV and MAYV co-infection in Haiti from an 8-year-old boy who lived in a non-forest area; the fact of MAYV recovery in the context of a DENV outbreak from this patient suggested a possible role in the transmission of MAYV by A. aegypti.

In the Americas, urban-dwelling Aedes aegypti mosquitoes are the primary vectors for the more common arboviruses including CHIKV, Yellow Fever virus, DENV, and MAYV [24]. The anthropophilic nature of the Aedes aegypti mosquitoes as well as the movement of these mosquitoes and their host to urban environments and their ability to effectively carry the MAYV offers a great opportunity for the expansion of the virus into more developed areas. Such movement increases the probability of a new epidemic [25]. In vitro studies conducted by Pereira et al. [26], have demonstrated the vector competence of A. aegypti and A. albopictus in the transmission of MAYV. The spillover in transmission from rural forested habitats into urbanized areas could be possible due to the evolution of mosquito species within the Aedes and the Culex mosquitoes having the ability to bear the virus [13,27].

Our findings demonstrated the occurrence of five MAYVD cases in patients from urban zones on the island of Trinidad (Barataria, Morvant, Maraval, Diego Martin, and San Fernando). After a review of the laboratory investigation forms, there were no records of recent travel to sylvatic areas. Except for the city of San Fernando, all other four locations are in the northwestern region of the country, a highly urbanized area that includes the capital of the country, Port-of-Spain. All four locations are also urbanized and are part of the city’s metropolitan area. To our knowledge, this is the first report of MAYV infections in patients residing in urban areas in the Caribbean.

The urban cases of MAYVD described in this study could be a consequence of an increased competence to bear the virus by mosquito species other than Haemagogus spp. (e.g. Aedes spp.). More studies are needed to determine the possible role of Aedes spp. and Culex spp. as potential vectors for MAYV and the relationship to urban cases.

Although CHIKV, DENV, and ZIKV are endemic viruses in the Caribbean region, our study also suggests that other arboviruses can co-circulate during the outbreaks associated with these agents and that the human cases for other arboviral infections could be undetected and as such misdiagnosed. As has been suggested by others [28] the emergence of MAYD is a risk for Latin America and the Caribbean, and public health emergency preparedness programs for emergent and reemergent infectious diseases should consider MAYV (as well as other viral agents) among their priorities.

In conclusion, our results indicate a possible spread of MAYV from the rural areas where it was first isolated to more urbanized areas within Trinidad, probably due to either the spillover in the transmission of the virus, the presence of more competent urban mosquitoes (e.g. Aedes spp.) or the evolution of the Haemagogus mosquitoes to adapt to human environments. The implementation of robust health surveillance programs utilizing reliable diagnostic methods is critical for monitoring other arboviruses besides DENV, CHIKV and ZIKV, that have the potential to cause outbreaks.

Biography

GGE conceived the idea and made the general design of the research, analyzed the data, edited the manuscript, and reviewed the final version. CC performed the experimental work and drafted the manuscript. CR performed part of the experimental work and searched the samples on the database and in the repository. RS and SMN conducted the validation of the rRT-PCR for MAYV and reviewed the manuscript.

Disclosure statement

The authors declare no conflict of interest.

Disclaimer

The authors hold sole responsibility for the views expressed in the manuscript, which may not necessarily reflect the opinion or policy of the Caribbean Public Health Agency, or the Ministry of Health of the Republic of Trinidad and Tobago.