Field Isolation and Laboratory Vector–Host Studies of Brazoran Virus (Peribunyaviridae: Orthobunyavirus) from Florida

ABSTRACT.

Brazoran virus was first isolated from Culex mosquitoes in Texas in 2012, yet little is known about this virus. We report the isolation of this virus from Culex erraticus from southern Florida during 2016. The Florida strain had a nucleotide identity of 96.3% (S segment), 99.1% (M segment), and 95.8% (L segment) to the Texas isolate. Culex quinquefasciatus and Aedes aegypti colonies were subsequently fed virus blood meals to determine their vector competence for Brazoran virus. Culex quinquefasciatus was susceptible to midgut infection, but few mosquitoes developed disseminated infections. Aedes aegypti supported disseminated infection, but virus transmission could not be demonstrated. Suckling mice became infected by intradermal inoculation without visible disease signs. The virus was detected in multiple mouse tissues but rarely infected the brain. This study documents the first isolation of Brazoran virus outside of Texas. Although this virus infected Ae. aegypti and Cx. quinquefasciatus in laboratory trials, their vector competence could not be demonstrated, suggesting they are unlikely vectors of Brazoran virus.

The genus Orthobunyavirus, family Peribunyaviridae, represents a diverse group of RNA viruses with more than 100 described species and new viruses being continually discovered.¹ Members of this genus are transmitted by mosquitoes or biting midges and include emerging human and livestock pathogens such as La Crosse virus, Jamestown Canyon virus, Oropouche virus, Cache Valley virus, and Schmallenberg virus.² The viral genome is comprised of three negative-sense RNA molecules termed the S, M, and L segments, which encode the nucleocapsid, envelope glycoproteins, and RNA polymerase, respectively.³

During 2012, a novel orthobunyavirus (Brazoran virus) was first isolated from three pools of mixed-species Culex mosquitoes collected in Brazoria, Orange and Montgomery counties, Texas.⁴ All virus isolations were made in a span of a few days, but never during the 12 years prior to the arbovirus surveillance program, despite relatively consistent sampling and testing procedures in the state. The S segment sequence of Brazoran virus was shown to be highly divergent and nearly twice the length of those from other orthobunyaviruses, whereas the M and L segments grouped within known serogroups within the genus. The virus replicated within mosquito and mammalian cell lines, but its ability to be vectored by mosquitoes or infect vertebrate hosts in vivo remains unknown.

We report the isolation of Brazoran virus from mosquitoes collected in the Florida Everglades during 2016. The virus genome was sequenced and compared with available sequences on GenBank. We then assessed the vector competence of Aedes aegypti and Culex quinquefasciatus for Brazoran virus, as well as the susceptibility of suckling mice to virus infection.

More than 650,000 mosquitoes were collected from the Florida Everglades region during 2016 and 2017 and tested for virus infection by cell culture assay as described elsewhere.⁵ Virus isolations were identified using standard polymerase chain reaction (PCR) and sequencing assays as reported previously,⁵ except for one virus isolated from a pool of 50 female Culex erraticus collected near a coastal lake (latitude 25.9590°N, longitude 81.5149°W) in Collier County on December 12, 2016. The virus culture was then sequenced using direct RNA sequencing on a flongle with v9 chemistry on a MinIon per the manufacturer’s instructions (Oxford Nanopore Technology, Oxford, UK). MinIon results were confirmed using an established protocol for RNA viruses with the Illumina platform (San Diego, CA), as described elsewhere.⁶ Sequence data were analyzed and annotated using CZ-ID and Geneious software, version 2023.2 (Boston, MA).⁷

Edited nucleotide sequences were searched for matching sequences in GenBank using the Blastn search algorithm. The Florida strain was most similar to the prototype strain of Brazoran virus and had a nucleotide identity of 96.0% (S segment), 99.1% (M segment), and 95.7% (L segment) to the 2012 Texas isolate.

To assess the vector competence of two anthropophilic mosquito species for Brazoran virus, we exposed Ae. aegypti and Cx. quinquefasciatus to virus blood meals and assayed them for infection. Mosquitoes were reared to adulthood from laboratory colonies of Ae. aegypti (Orlando strain) and Cx. quinquefasciatus (Benzon Research Inc., Carlisle, PA) that originated from Florida, as described elsewhere.⁸ The Florida strain of Brazoran virus was passaged twice on Vero cells and titrated on BHK-21 cells. Concentrated stocks of virus were also prepared by passing virus supernatants through a 100-kDa filter (MilliporeSigma, Burlington, MA). Female mosquitoes were provided an infectious blood meal consisting of a 1:1 mixture of defibrinated sheep blood and Brazoran virus by means of a water-jacketed membrane feeder. Blood-engorged mosquitoes were held for 7 and 14 days at 28°C under a 14-/10-hour light/dark cycle and maintained on a diet of 10% sucrose solution. Mosquito bodies and legs were harvested at 7 and 14 days and tested separately for Brazoran virus by quantitative reverse transcription PCR. Viral RNA was extracted and amplified using procedures described elsewhere,⁹ with primer set BRAZfwd (CTTATCTGTTACTGCCCATCTTC), BRAZrev (GCTCCGCATACACAGTTTTC), and BRAZprobe ([6∼FAM]CGCATACTGTGGTCTAGCATATCATCCATT[BHQ1a∼Q]).

Aedes aegypti were susceptible to Brazoran virus infection (18.9–21.6%) when exposed to blood meals at a concentration of 1.0 × 10⁴ PFU/mL (Table 1). Of these, 2.7% and 9.1% developed disseminated infections, identified by testing mosquito legs at days 7 and 14, respectively. Culex quinquefasciatus also became infected with Brazoran virus (14.7–17.4%) at this concentration, and 0.9% to 2.4% of these mosquitoes developed disseminated infections. Infection rates increased to 95.4% when Cx. quinquefasciatus were exposed to virus bloodmeals at 3.8 × 10⁵ PFU/mL, but relatively few mosquitoes developed disseminated infections (7.2%), suggesting a strong midgut escape barrier.

Table 1.

Infection and dissemination rates of Aedes aegypti and Culex. quinquefasciatus exposed to Brazoran virus^*

Species	Viral Dose (PFU/mL)	Days Postinfection	No. Tested	% Infected (95% CI)	% Disseminated (95% CI)
Aedes aegypti	1.0 × 104	7	112	18.8 (12.0–27.2)	2.7 (0.6–7.6)
Ae. aegypti	–	14	88	26.1 (17.3–36.6)	9.1 (4.0–17.1)
Culex quinquefasciatus	–	7	170	14.7 (9.8–20.9)	2.4 (0.6–5.9)
Cx. quinquefasciatus	–	14	109	17.4 (10.8–25.9)	0.9 (<0.1–5.0)
Cx. quinquefasciatus	3.8 × 105	14	153	95.4 (90.8–98.1)	7.2 (3.6–12.5)

^*Percentages are based on the number of mosquitoes tested.

We showed previously¹⁰ that giving mosquitoes a second noninfectious blood meal increased virus escape from the mosquito midgut by altering the permeability of the surrounding basal lamina layer. Given that mosquitoes may frequently blood-feed in nature,¹¹ we modified vector competence experiments by allowing some mosquitoes to feed again on uninfected blood 7 days after the initial infectious blood meal. Aedes aegypti became readily infected with Brazoran virus (76.7–83.0%) (Table 2) when exposed to a greater concentration of virus (3.8 × 10⁵ PFU/mL) than in previous experiments. Administering a second noninfectious blood increased disseminated infection rates from 46.7% for the single-feed group to 67.9% in the double-feed group. Likewise, 87.5% to 88.8% of Cx. quinquefasciatus became infected with Brazoran virus, but relatively few developed disseminated infection rates in the single-feed (5.7%) or double-feed (12.5%) groups.

Table 2.

Infection and dissemination rates of Aedes aegypti and Culex quinquefasciatus exposed to Brazoran virus followed by a second noninfectious blood meal^*

Species	Viral Dose (PFU/mL)	Treatment†	No. Tested‡	% Infected (95% CI)	% Disseminated (95% CI)
Aedes aegypti	3.8 × 105	SF	120	76.7 (68.1–83.9)	46.7 (37.5–56.0)
Ae. aegypti	–	DF	53	83.0 (70.2–91.9)	67.9 (53.7–80.1)
Culex quinquefasciatus	–	SF	35	88.6 (73.3–96.8)	5.7 (0.7–19.2)
Cx. quinquefasciatus	–	DF	32	87.5 (71.0–96.5)	12.5 (3.5–29.0)

^*Percentages are based on the number of mosquitoes tested.

^† The single-feed (SF) group was given a single infectious blood meal. The double-feed (DF) group was given an infectious blood meal followed by a second noninfectious blood meal at day 7.

^‡ Mosquitoes were tested for virus infection on day 14.

We then evaluated whether suckling mice are susceptible to Brazoran virus infection and could serve as recipient hosts for vector competence experiments. A total of 18 four-day old mice, strain CD-1 (Charles River, Wilmington, MA), were inoculated intradermally by 0.05 mL Brazoran virus (4.2 × 10² or 4.2 × 10³ PFU/mouse). None of the mice showed visible signs of illness associated with arbovirus infection such as weight loss, weakness, tremors, or partial paralysis. Mice were euthanized 5 days postinoculation and tissues were then assayed for virus infection. Brazoran virus was detected most frequently in the liver (88.9–100%), followed by the front limb (55.6–66.7%), spleen (44.4–66.7%), kidney (33.3–44.4%), and brain (0–11.1%) (Table 3). All mice except one had detectable virus in at least one tissue.

Table 3.

Infection of Brazoran virus in suckling mice by intradermal inoculation^*

Viral Dose (PFU/Mouse)	No. of Mice Tested	% Positive
		Brain	Limb	Spleen	Liver	Kidney	Any Tissue
4.2 × 102	9	11.1	66.7	44.4	88.9	44.4	88.9
4.2 × 103	9	0	55.6	66.7	100	33.3	100

^*Mice were euthanized on day 5 and individual tissues were assayed for virus infection.

To determine whether Ae. aegypti may transmit Brazoran virus, we allowed infected mosquitoes to feed on suckling mice. Aedes aegypti were infected orally with Brazoran virus and held for 14 days prior to feeding on mice. A total of 26 suckling mice were exposed to mosquito bite by placing each mouse on the upper screen of a cup for 30 minutes that contained two mosquitoes. Mice were restrained by draping a second piece of mesh netting over them and securing it with a rubber band to the cup. After exposure, mosquitoes were examined to determine whether there was visible blood in the gut, and their legs were removed for virus testing. All mosquitoes except two had fully or partially fed on the mice. A total of 14 mice were exposed to two infected mosquitoes, eight mice were exposed to one infected and one uninfected mosquito, and four mice were exposed to two uninfected mosquitoes. Mice were held for 5 days and then euthanized. None of their tissues (liver, limb, spleen, kidney, or brain) were positive for Brazoran virus, indicating that Ae. aegypti failed to transmit the virus.

Our study documents the first isolation of Brazoran virus outside of Texas from a pool of Cx. erraticus collected in southern Florida. Nucleotide identities of S, M, and L segments were most similar (96–99%), but not identical to the Texas isolate, suggesting independent evolution of these virus strains. The origin of Brazoran virus is unclear because it has only been isolated in Texas in 2012 and in Florida in 2016. These virus detections may be a result of recent independent introductions from endemic countries given the extent of viral genetic divergence among strains. Alternatively, the virus has persisted long term within the United States and has evolved into divergent strains that evaded detection until recently.

Vector competence trials showed that Ae. aegypti and Cx. quinquefasciatus were poor vectors of Brazoran virus as a result of failures in virus dissemination and transmission. Aedes aegypti were susceptible to Brazoran virus infection, and the virus disseminated readily to peripheral leg tissue. Furthermore, a second noninfectious blood meal enhanced virus dissemination within this species. Culex quinquefasciatus also acquired virus infection, but few mosquitoes developed disseminated infections. Despite high rates of disseminated infection within Ae. aegypti, we could not demonstrate Brazoran virus transmission to suckling mice. This may be the result of our choice of a host animal, which may require a greater inoculum to establish infection than that transmitted by Ae. aegypti. We opted for this approach because our prior study⁸ showed that using mice was superior to in vitro methods such as forced mosquito salivation in capillary tubes, which consistently underestimated arbovirus transmission.

Although we did not assess the vector competence of Cx. erraticus, this species deserves further scrutiny as a possible vector. Brazoran virus was isolated from this species in our study and from Culex mosquitoes (mixed-species pools) in eastern Texas, where Cx. erraticus also occurs.⁴ Culex erraticus feeds on a wide range of vertebrate hosts, including mammals, birds, reptiles, and amphibians, making it difficult to identify the host source of Brazoran virus.¹² Orthobunyaviruses usually involve small mammals or ungulates in their transmission cycles, which may pertain to Brazoran virus too.¹³ This is supported by the ability of Brazoran virus to replicate within mammalian cell cultures and infect laboratory mice, but more studies are needed to define the host range of this virus.⁴

The potential for Brazoran virus to cause disease in humans or domestic animals remains unknown. We found that newborn mice became infected by needle inoculation but did not develop neuroinvasive disease during a limited observation period. Suckling mice are highly sensitive to lethal infection by many but not all arboviruses.¹⁴ Moreover, these animals do not necessarily reproduce the pattern of disease found in humans.¹⁵ Further studies are needed to understand whether humans, domestic animals, or wildlife are exposed to this virus and whether infection is associated with any disease.