Ethics statement

The study was approved by the Institutional Animal Care and Use Committee of the University of California, Davis (Protocol number 23106).

Animal source, care, use, and sample collection

This study used 12 healthy adult Indian origin rhesus macaques (RM, Macaca mulatta), including nine males and three females, aged 3.7 to 6.6 years. All animals were from a type D retrovirus-, simian immunodeficiency virus-, and simian lymphocyte tropic virus type 1-free colony at the California National Primate Research Center (CNPRC). All animals were previously exposed to the flavivirus dengue virus, which is not known to impact MAYV infection outcomes. All RM passed a physical exam before study enrollment. The RM were all born and raised at CNPRC in outdoor environments, Although the alphavirus Western equine encephalitis virus has historically circulated in California, no environmental activity in California mosquito pools has been detected since 2007, despite ongoing annual surveillance [31]. All experimental procedures were conducted at CNPRC, which is accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International (AAALAC). Animal care followed the guidelines outlined in the 2011 Guide for the Care and Use of Laboratory Animals by the Institute for Laboratory Animal Research. RM were housed indoors in stainless steel cages (Lab Products Inc., Seaford, DE), with cage sizes scaled to each animal according to national standards. Housing conditions included a 12-hour (h) light/dark cycle, temperatures of 64–84°F, and room humidity of 30–70%. Animals that were previously bonded were housed together. They had free access to water and were fed a commercial high-protein diet (Ralston Purina Co., St. Louis, MO) supplemented with fresh produce. Animals were monitored daily for clinical signs, including behavior changes, rash or redness, rectal temperature, and joint swelling at time of sedation. The cumulative number of clinical observations from 1-7 dpi are reported, with several caveats. Day 0 was excluded since observations were made prior to inoculation and were therefore likely not related to MAYV treatment. Observations past 7 dpi were also excluded as they were not available for all animals. Any observation that was related to a procedure (razor burn, injection) or mosquito presentation (e.g., rash at feeding site) on the day of or 1 day after were also excluded from totals. For MAYV inoculation, blood collections, and mosquito feedings, RM were immobilized with ketamine HCl (Dechra Veterinary Products, Overland Park, KS) at 10 mg/kg that was injected intramuscularly in the left leg after overnight fasting. Blood samples were collected using venipuncture. A portion of ethylenediaminetetraacetic acid (EDTA)-anticoagulated whole blood was used for complete blood counts, performed by the CNPRC Clinical Laboratories. The remaining blood was processed by centrifugation at 900 x g for 10 minutes (m) to separate plasma from cells. Plasma was then centrifuged again at 900 x g for 10 m to remove residual cells, after which aliquots were immediately frozen at -80°C. Blood collected without anticoagulant was processed via centrifugation at 900 x g for 10 m to obtain serum. Cerebrospinal fluid was collected from the cervical region. Half of the RM in each cohort were selected for euthanasia on 10 or 12 dpi. Euthanasia was performed with an overdose of pentobarbital, followed by necropsy and tissue collection. Samples were collected from all major organ systems, with a focus on lymphoid tissues, joints, and muscles. A skin biopsy, referred to as the “skin inoculation site” was collected from the middle back region, corresponding to the location where the SC-inoculated animals were injected with MAYV. Tissues were processed using three methods: (i) snap-frozen in liquid nitrogen and stored at -80°C; (ii) preserved in RNALater (ThermoFisher, Waltham, MA) followed by incubation overnight at 4°C then transfer to -80°C for long-term storage, and (iii) fixed in 4% paraformaldehyde.

MAYV source and inoculations

All 12 RM were inoculated with MAYV strain IQT4235. The virus was derived from an infectious clone plasmid created from the sequence of strain IQT4235, a genotype D lineage virus which predominates in Latin America [6]. The plasmid was derived from the sequence of a MAYV strain that was originally isolated from the serum of a febrile human in Iquitos, Peru in 1997 (GenBank: MK070491) [27]. Before in vitro transcription, the complete MAYV cDNA sequence in the plasmid was verified (Plasmidsaurus, Eugene, OR). The sequence in the plasmid was identical to the GenBank sequence except for 2 non-coding mutations (T6298A and G10797A). To generate infectious MAYV, the cDNA clone was linearized using the PacI restriction endonuclease (New England Biolabs, Ipswich, MA), transcribed in vitro into mRNA using the mMessage mMachine SP6 transcription kit (ThermoFisher, Waltham, MA), and electroporated into Vero cells using a BioRad Gene Pulser XCell (Hercules, CA) with the following settings: 125 Volts, 10 millisecond pulse length, 1 second intervals, 3 pulses. After in vitro transcription, the RNA was deep sequenced using the methodologies described below. Infectious MAYV was harvested from Vero cell culture supernatant 2 days post-electroporation, titrated, and stored at -70°C until use. To evaluate effects of dose and inoculation route, male RM were randomly divided into 3 groups. Since only 3 females were available, 1 female was assigned to each group. Each group received a single MAYV inoculation on day 0 with one of the following: 7 log₁₀ IV, 7 log₁₀ PFU SC, or 3 log₁₀ PFU SC. The IV route and 7 log₁₀ dose were chosen to ensure infection, as these studies were initiated before publication of recent MAYV NHP studies [11,12], when infectivity of MAYV IQT4235 for RM was not certain. The SC route was selected to mimic mosquito-borne transmission [32]. The 3 log₁₀ dose was intended to simulate a mosquito-delivered dose, where our prior studies with CHIKV [33] and studies by us [27] and others [28,29] for MAYV show that Ae. aegypti salivate between 1.0-2.5 log₁₀ PFU or 1.0-4.0 log₁₀ genomic RNAs into capillary tubes. IV injections were administered in the saphenous vein, while SC injections were administered in the upper back between the shoulder blades. Inocula were back-titrated by plaque assay to verify the administered dose. Blood samples were collected daily from 1-7 days post-inoculation (dpi) and again at euthanasia, 10 or 12 dpi.

MAYV RNA isolation from macaque plasma and tissues

Frozen plasma and tissues were thawed at room temperature. For tissue processing, each sample was individually weighed, then combined with 500 µl Dulbecco’s modified eagle medium (DMEM) (GenClone, Waltham, MA) and a 4 mm glass bead (ThermoFisher, Waltham, MA). Homogenization was performed using a Retsch MM400 Tissue Lyser (Haan, Germany) at 30 shakes/s for 5 min, followed by a 10 min rest. Samples were then rotated and homogenized again using the same conditions. To clarify the supernatant, homogenized tissues were centrifuged at 3000 g for 2 min. Thawed plasma or supernatant from tissues was used immediately after homogenization for plaque assays or RNA extractions.

Tissue processing and histopathologic analyses

RM tissues were fixed in 4% paraformaldehyde and then paraffin embedded, sectioned, and stained with hematoxylin and eosin. A pathologist, blinded to treatment, evaluated joints, tendons, and muscles for pathologic changes using a quantitative scoring metric (S1 Table). Tissues from age- and sex-matched colony control animals not part of this study that were not MAYV exposed served as comparators.

Mosquito source, presentation to rhesus macaques, sampling, and processing

Adult Ae. aegypti mosquitoes aged 3–9 days from the 33^rd generation of a colony originally collected in Los Angeles, CA, and identified morphologically in the field [34], were used in this study. Mosquitoes were reared in an insectary under controlled conditions (28^oC, 75% humidity, 12hr light/dark cycle) with free access to 10% sucrose in water. All mosquitoes presented to RM were from the same cohort. Twenty-four h prior to RM presentation, sucrose was replaced with sterile water. Mosquitoes were housed in pint sized containers and transported inside plastic shoeboxes from a rearing insectary to the CNPRC animal rooms. On 2, 3, 5, and 7 days post inoculation (dpi) of the RM, cohorts of ~100 mixed-sex Ae. aegypti were presented to the shaved abdomens of anesthetized RM and allowed to feed for 10–15 m. Feeding was confirmed by visual observation of red abdomens. After feeding, mosquitoes were transported in shoeboxes placed inside sealed bags into an arthropod containment level 3 (ACL-3) facility. Mosquitoes were anesthetized with CO₂ and bloodfed females were separated from males and non-fed females on a cold table (Bioquip, Compton, CA). Bloodfed females were incubated at 28^oC with 75% humidity with 12-h light/dark cycle for 10 days post feed (dpf), with continuous access to 10% sucrose. Nine to 20 mosquitoes per RM per feeding day were incubated. At 10 dpf, mosquitoes were anesthetized with CO₂ for 10 s, immobilized on a chill table, and their legs and wings were removed. Saliva was collected using expectoration assays, where the proboscis was inserted into a capillary tube containing DMEM with 5% heat-inactivated fetal bovine serum (FBS) (56^oC, 30 min) (GenClone, Waltham, MA) and 1% penicillin-streptomycin (P/S) (Genesee Scientific, El Cajon, CA) for 30 m to stimulate salivation. Mosquito tissues were placed in 500 µl DMEM supplemented with 5% FBS and 1% P/S, then homogenized in a TissueLyser II (Qiagen, Hilden, Germany) at 20 shakes/s for 2 m and stored at -70^oC until further use.

Cells

Vero cells (ATCC CCL-81, Manassas, VA) were used for titration and neutralization assays. Vero cells are African green monkey kidneys cells were maintained in DMEM with 5% FBS and 1% P/S and incubated at 37^oC with 5% CO₂.

Infectious MAYV detections in inocula, RM, and mosquito samples

Infectious MAYV was quantified from RM inocula, plasma, and mosquito tissues using plaque assays on Vero cells. Mosquito bodies were initially tested for infection. If virus was detected, legs and wings were subsequently tested, following the route of virus dissemination in a mosquito [35,36]. If infectious virus was detected in legs and wings, saliva samples were passaged in Vero cells. Saliva was passaged to maximize virus detection as our prior CHIKV studies [33] show transmitted doses are low. For saliva, 50–75 µl of sample in DMEM was inoculated onto a single well of 60% confluent Vero cells in a 24-well plate. An additional 25–50 µl DMEM was added, and the plate was incubated for 1 h, with rocking every 10 m to facilitate viral attachment and maintain cell hydration. After incubation, 1 ml of DMEM was added to each well, and plates were incubated at 37^oC with 5% CO₂ for 3 days. If cytopathic effects (CPE) were observed 3 dpi in the test wells but not in the negative control wells, the supernatant was harvested and stored at -80^oC. If no CPE was observed, the supernatant was transferred to a 1.5 ml tube, vortexed for 10 s, and then 100 µl was inoculated into fresh Vero cells for a second passage under identical conditions to the first passage. If CPE were detected in either passage, the saliva sample was considered positive for infectious MAYV. For titrations, Vero cells were seeded into 12 (RM)- or 24 (mosquito)-well plates and incubated until 100% confluent. Plasma and homogenized tissue samples were diluted in ten-fold series in DMEM in technical duplicates, with dilution ranges of 10^-1-10^-11 (plasma), and 10^-1-10^-12 (mosquitoes). For infection assays, 125 µl (12-well) or 100 µl (24-well) of each diluted sample was inoculated per well and incubated for 1 h, with rocking every 10 m. After incubation, 2 ml (12-well) or 1 ml (24-well) 0.4% agarose (Genesee Scientific, El Cajon, CA) -DMEM at 42^oC was added. Plates were incubated for 3 days, then fixed with 2% formalin for 1 h. The agarose overlay was removed and plates were stained with 0.05% weight/volume crystal violet (Millipore-Sigma, Burlington, MA) for 5 m, rinsed with deionized (DI) water, and dried before reading on a light box. Viral titers were calculated using equation 1:

(1)

The infectious MAYV titer of a sample is presented as the mean of duplicate titrations. The limit of detection of titrations was 80 PFU/ml (12-well plate) or 100 PFU/ml (24-well). To convert PFU/ml to PFU/mosquito tissue, the PFU/ml titer was adjusted to account for the volume of media the mosquito tissue was homogenized in (500 µl) and the fraction of that homogenate that was used for serial dilutions (12 µl for 24-well plates, 2.4% of total volume) using equation 2:

(2)

Samples with no detectable infectious MAYV are reported below the LOD lines on graphs and are excluded from calculations of means.

MAYV RNA quantitation by reverse transcription quantitative polymerase chain reaction (RT-qPCR)

MAYV RNA was extracted from RM plasma and homogenized tissues using a viral RNA isolation kit (Thermo Fisher Scientific, #AMB18365) and machine (MagMA Express 96 #4400076, Waltham, MA). Two hundred microliters of sample were mixed with 10 µl RNA binding beads, 10 µl lysis binding enhancer, 120 µl 100% isopropanol (Geel, Antwerp, Belgium) and 120 µl lysis binding solution, which was loaded into the machine in a 96 deep well plate. Extracted RNA was eluted in 65 µl of Elution Buffer (ThermoFisher, Waltham, MA) for storage at -80°C prior to quantification. Into each well of the RT-qPCR plate, 10 µl of extracted RNA, 0.8 µl each of the 100mM forward and reverse primers, 0.2 µl of 100mM probe, 5 µl of TaqMan Fast Advanced Master Mix 4x (Applied Biosystems, Waltham, MA), and 3.2 µl of nuclease-free water (Waltham, MA) were added. Each RT-qPCR plate included 2 wells of negative controls consisting of nuclease-free water (Waltham, MA), and 2 wells of each serial 10-fold positive control dilution that comprised standards of in vitro transcribed MAYV RNA derived from the infectious clone, ranging from 8.1 x 10⁹ to 8.1 x 10^-2 genomes/ml, measured using a Qubit (Waltham, MA) prior to serial dilution. The RT-qPCR conditions on the Applied Biosystem ViiA 7 real-time PCR machine (Waltham, MA) were as follows: 52^oC 15 s, 94^oC 2 m, then 40 cycles of 94^oC 15 s, 55^oC 40 s, 68^oC 20 s. The following primers were used: Forward: GGTAATGATCCACAGTCCATGC (5156 → 5177), Reverse: CGGGTTGAACAACCGGTTC (5237 → 5219), Probe: FAM-CTGACAACCAACAACCCATATCAACGG-NFQ-MGB (5190 → 5216).

Every sample was tested using technical triplicates. The primers and probe were designed de novo and purchased from ATCC (Manassas, VA). The primers all bind in the MAYV nonstructural protein 3 (nsp3) gene. The limit of detection (LOD), determined by averaging the lowest RNA level detected in the standards of all RT-qPCR assays, was 1.0 x 10^3.5 genomes/gram. The MAYV RNA level in a sample is presented as the mean of triplicate measurements. The number of samples with no detectable RNA levels are reported below the LOD lines on graphs. Samples with no detectable MAYV RNA are excluded from means.

Neutralizing antibody quantification by plaque reduction neutralization test (PRNT)

MAYV neutralizing antibodies in RM sera were quantified using a PRNT₈₀ assay in a 12-well format using MAYV strain IQT4235, the homologous strain used for RM inoculation. The PRNT₈₀ titer was defined as the serum dilution that neutralized 80% of plaques compared to a virus-only control sample, which consisted of approximately 40 PFU of MAYV without RM serum. Sera were heat inactivated at 56^oC for 30 m to destroy complement proteins. Controls included a positive serum from a MAYV-inoculated RM in this study that showed neutralization in a preliminary screening and a negative control consisting of serum from a RM with no alphavirus exposure history in a 1:100 final dilution. Each of these serum controls was tested with and without MAYV. A negative control of DMEM only was also included. All samples and controls were tested in technical duplicates. Serum was diluted 1:10 in technical duplicates, then serially two-fold diluted from 1:10–1:5,120. Serum dilutions were mixed 1:1 with virus to make a final series of 1:20–1:10,240. Serum-virus samples and controls were incubated at 37^oC with 5% CO₂ for 1 h. Following incubation, serum-virus mixtures or controls were inoculated onto plates of Vero cells at a volume of 125 µl/well and incubated for 1 h with rocking every 10 m. After incubation, a 0.4% agarose-DMEM overlay at 42^oC was applied to wells. Plates were incubated for 3 days at 37^oC with 5% CO₂, then fixed with 2% formalin for 1 h. After removal of the agarose overlay, plates were stained with 0.05% weight/volume crystal violet for 5 m, washed with DI water, and dried before counting on a light box. PRNT₈₀ titers are reported as the mean inverse serum dilution that resulted in 80% plaque reduction compared to virus-only control wells. If duplicate tests yielded different PRNT₈₀ titers, the sample was re-tested, and all re-tested samples showed concordance in both replicates. The limit of detection for this assay was 20.

MAYV deep sequencing from in vitro transcribed MAYV RNA, RM, and mosquito saliva samples

MAYV genomes were sequenced from in vitro transcribed RNA that was used to generate the virus stock that was inoculated into RM, plasma samples from all 12 RM at 2 dpi, and from all Vero passaged Ae. aegypti saliva samples that contained infectious MAYV. To generate material for sequencing, viral RNA from plasma was extracted a second time from the same samples previously tested by RT-qPCR following the same extraction protocol. Viral RNA from passaged saliva samples was extracted from Vero supernatants that had been thawed once. The extracted RNA was sequenced using ClickSeq [37]. To enrich and capture MAYV genomic RNA, we designed set of tiled MAYV primers (S2 Table) that match with 100% identity to the MAYV genome used in these studies (GenBank: MK070491). 10 ul of extracted RNA was used as input in reverse transcription reactions, together with 10 pM of MAYV-specific tiled primers and a mixture of dNTPs and 3′-azido-2′,3′-dideoxynucleotides (AzNTPs) (nucleotide analogues with modified 3′ groups that serve as chain terminators) at a ratio of 35:1, per the standard protocol for ClickSeq [38]. Primers were annealed by heating the extracted RNA to 95^oC for 2 m, followed by slow cooling to 50^oC. Reverse transcription components (Superscript IV, ThermoFisher, Waltham, MA) were added at 50^oC to initiate cDNA synthesis. The reverse transcription reaction was incubation for 10 m at 50^oC and then 10 m at 80^oC, per the manufacturer’s standard protocol. 2.5 units of RNaseH were then added and incubated at 37^oC for 20 m and then heat-denatured at 80^oC for 10 m. The 3′-azido-blocked cDNA fragments were then “click-ligated” to the 5′-hexynyl adaptor containing the full Illumina i5 adaptor with i5 indexes via copper(I)-catalyzed azide-alkyne cycloaddition (CuAAC). Click-ligated cDNA was used as input to a PCR reaction that provided i7 indexes for 21 cycles. cDNA library fragment sizes of 300–600 nt were selected using double-sided SPRI bead purification and sequenced on an Illumina NovaSeq X, yielding > 10M 150x2 paired end reads per sample. Raw FASTQ data were preprocessed (adaptor trimming and quality filtering) using fastp [39] and cutadapt [40], using batch scripts previously published for analysis Tiled-ClickSeq data [41]. Reads were then mapped to the MAYV reference genome with bowtie2 [42] and single nucleotide variants and minority variants were called using pilon [43].

Statistical Analyses

GraphPad Prism 10.3.1 was used for all statistical analyses, which are enumerated in the results. Analyses of infectious MAYV titers and RNA levels were performed on log-transformed data.

Rhesus macaques develop a 3-day MAYV infectious viremia

This study aimed to characterize infection dynamics and disease outcomes in RM following inoculation of infectious clone derived MAYV IQT4325 from Iquitos, Peru, using 3 dose and route combinations (Fig 1A). Back-titration of the inocula confirmed that the administered doses were within 10% of the target doses. Clinical signs of disease were monitored, and blood samples were collected daily from 1-7 dpi and at necropsy, 10 or 12 dpi. Infectious MAYV and MAYV RNA were quantified from plasma, while tissues were harvested and analyzed for MAYV RNA levels. Histopathological evaluation of muscle and joint tissues was conducted, with inflammation scored using quantitative metrics. None of the MAYV-inoculated RM exhibited signs of behavioral changes or joint swelling, but 8 animals exhibited other signs of mild clinical disease between 1 and 4 dpi. Three animals had mildly elevated temperatures, 3 animals had nausea, 5 animals developed mild erythema at the inoculation site, and one animal developed an abdominal rash prior to mosquito feeding. Some of the rashes were detected on a day of mosquito feeding and on the shaved area of the abdomen where mosquitoes were presented. Eight of 12 RM also developed rashes or skin redness in a region distant from the abdomen and not on a day of mosquito presentation (S3 Table). All MAYV-inoculated RM developed detectable viremia, with infectious MAYV in plasma observed from 1-2 (N = 1) or 1–3 (N = 11) dpi, peaking at 2–6 log₁₀ PFU/ml. No infectious MAYV was detectable in plasma after 3 dpi (Fig 1B). Peak infectious viremia occurred at 1 (N = 4) or 2 (N = 8) dpi and mean infectious viremia between the 1 female and 3 male RM in each group were similar, although statistical analyses were not performed given only 1 female was in each group. RM that received 7 log₁₀ PFU IV had significantly higher mean viremia areas under the curve (AUC) compared to those receiving 3 log₁₀ PFU SC (Kruskal Wallis, Tukey’s post-hoc, p = 0.005) (Fig 1C). However, no significant difference in AUC was observed between the 2 SC groups (Fig 1C). The mean number of cumulative clinical observations from 1-7 dpi was not different across groups of RM that received different MAYV doses and routes, and there was no correlation with peak viremia or AUC (Kruskal-Wallis, p > 0.05) (S1A-C Fig). The detection pattern of MAYV RNA in plasma mirrored that of infectious virus, with RNA detected from 1-3 (N = 8) and 1–4 dpi (N = 4) animals, peaking at 7–9 log₁₀ genomes/ml (Fig 1D). MAYV RNA levels peaked 1 (N = 1) or 2 (N = 11) dpi. In most samples, MAYV RNA levels were typically several logs higher than infectious MAYV levels. There were no significant differences in mean genome AUC levels across cohorts (Kruskal-Wallis, p > 0.05) (Fig 1E). Together, these results demonstrate that RM are susceptible to infection with MAYV infectious clone IQT4235 and that some infected animals develop mild clinical disease and focal rash. Viremia was short-lived, with infectious virus becoming undetectable after 3 dpi and viral RNA levels undetectable after 4 dpi.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Rhesus macaques develop short MAYV viremias.

(A) Study design in which 12 rhesus macaques (9 males and 3 females) were inoculated with 3 or 7 log₁₀ PFU MAYV IQT4235 from an infectious clone by intravenous (IV) or subcutaneous (SC) routes on day 0, followed by blood collections 1-7 days post-inoculation (dpi) and blood collections and euthanasia 10 or 12 dpi. (B) MAYV infectious viremia in plasma 1-12 dpi measured as log₁₀ plaque forming units (PFU)/ml. Each line shows kinetics from 1 RM and bars report standard deviations from duplicate titrations. (C) Area under the viremia curve (AUC). Horizontal lines denote averages. D) MAYV RNA in plasma measured as log₁₀ genomes/ml 1-7 and 10 or 12 dpi. Each line shows kinetics from 1 RM and bars report standard deviations from triplicate RT-qPCR measurements. (E) Area under the MAYV RNA curve. Horizontal lines denote averages. Data points outlined in black represent female animals. P values are based on Kruskal-Wallis with Tukey’s post-hoc tests, ns = not significant (p > 0.05). The colors and shape identifiers denote individual RM, which are also identified with a unique number. LOD is the limit of detection. Images created using BioRender.

https://doi.org/10.1371/journal.pntd.0013061.g001

MAYV-infected rhesus macaques have detectable neutralizing antibody

We measured endpoint neutralizing antibody titers in RM serum upon euthanasia, 10 or 12 dpi, against the homologous strain of MAYV used for RM inoculation. All 12 animals had detectable PRNT₈₀ titers, ranging from 80 to 1280 (Fig 2A), with a mean titer of 160. There was no significant difference in mean titers between the different dose and route groups (Kruskal-Wallis, p > 0.05).These data demonstrate RM have MAYV-specific neutralizing antibody detectable 10–12 days after MAYV infection.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Rhesus macaques have MAYV neutralizing antibodies in serum and MAYV RNA in multiple tissues 10 or 12 dpi.

(A) Plaque reduction neutralization test 80% (PRNT₈₀) titers in serum. PRNT₈₀ titers are based on duplicate titrations. Numbers above bars show day of euthanasia on which PRNT₈₀ titers were measured. The solid black line shows the average titer for all RM. The colors and shape identifiers denote individual RM, which are also identified with a unique number. The dotted line shows the LOD. MAYV RNA levels in (B) central nervous system (CNS), quadriceps, and reproductive organs, (C) joints and skin, and (D) lymphoid tissues, heart, and liver. The colors and shapes identify individual RM. Data points with black outlines represent female animals. MAYV RNA levels were measured by RT-qPCR with an average LOD of 3.5 log₁₀ genomes/gram, indicated by the horizontal dotted line. Cerebrospinal fluid data is graphed in genomes/ml with the same LOD. Horizontal lines show mean titers. Values from animals without detectable MAYV RNA are not included in means but the number of animals with no detectable RNA (ND) are shown below the LOD line. “ns” indicates no statistically significant difference by Kruskal-Wallis test, p > 0.05. No statistical analyses were performed on heart, cerebellum, ovary, and CSF tissues due to sample numbers with measurable RNA levels. CSF = cerebrospinal fluid; Q. Muscle/Tendon = quadricep; JC = joint capsule; inoc. = inoculation; LN = lymph node.

https://doi.org/10.1371/journal.pntd.0013061.g002

MAYV RNA tropism is diffuse and includes muscle, lymphoid, central nervous, and circulatory systems of rhesus macaques

We evaluated whether infectious MAYV or MAYV RNA could be detected in various tissues, including muscle and joint, as well as other systems previously identified as targets in male rhesus [12] and female cynomolgus [11] macaques. We measured both infectious MAYV and MAYV RNA levels in 14 non-reproductive tissues for each RM and in testes and ovaries. No infectious MAYV was detected in any of the RM tissues. However, MAYV RNA was detected in every tissue type for at least one RM (Fig 2B–2D). We observed considerable variability in MAYV RNA detection both within and across cohorts and tissue types. No statistically significant differences in mean tissue titers were observed across different dose and route groups (Kruskal-Wallis, p > 0.05).

The mean MAYV RNA level across all tissues ranged from 4.1-9.0 log₁₀ genomes/gram (Fig 2B–2D). Most RM in each group had detectable MAYV RNA in the majority of their tissues (Fig 3). For the reproductive tissues, all but one male and one female had detectable MAYV RNA in their testes and ovaries, respectively. These data show that MAYV infects multiple RM tissue systems, as evidenced by viral RNA detections in most animals at 10 or 12 dpi, nearly a week after MAYV RNA clearance from the blood. In general, the highest MAYV RNA levels were detected in lymphoid tissues including lymph nodes and spleen.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. MAYV RNA is detectable in multiple tissues of rhesus macaques 10 or 12 dpi.

The heatmap shows the magnitude of MAYV RNA detection. The colors and shape identifiers denote individual RM, which are also identified with a unique number. Female animals (4, 8, and 12) are represented as diamonds. MAYV RNA levels were measured by RT-qPCR with an average limit of detection (LOD) of 3.5 log₁₀ genomes/gram. Boxes with X symbols show tissues that were not available due to sex.

https://doi.org/10.1371/journal.pntd.0013061.g003

MAYV infection of rhesus macaques produces histopathologic signs of inflammation in in joints and muscles

After establishing the susceptibility, viremia kinetics, and tissue tropism of MAYV IQT4235 in RM, we focused on identifying histopathologic changes in target tissues. Since alphaviruses like MAYV target joints and muscles, causing muscle pain and arthritic disease, we concentrated on these tissue types. Using our quantitative histopathologic scoring scale (S1 Table), we identified and quantified inflammatory changes in hematoxylin and eosin-stained joint and muscle tissues. Inflammation was considered present if any of the 12 tissues evaluated showed had a score greater than 0, where 0 represents the baseline for colony control animals not exposed to MAYV. All 12 MAYV inoculated RM showed histopathologic signs of inflammation, with joint scores ranging from 6 to 14 and muscle scores from 1 to 10 (Fig 4A). No significant differences were found in mean pathology scores in joint, muscle, or composite (joint & muscle) tissues across any treatment groups (Kruskal-Wallis, p > 0.05). Additionally, there was no significant difference in mean pathology scores between RM euthanized at 10 or 12 dpi (Kruskal-Wallis, p > 0.05). Representative images of the inflammatory responses show immune cells (dark purple, marked with asterisks*) in the joint synovium for a finger (Fig 4B), toe (Fig 4C), and elbow (Fig 4D). There was no correlation between the joints & muscles pathology score and the cumulative clinical score (S1D Fig). These data demonstrate that MAYV caused histopathologic signs of muscle and joint inflammation in all infected RM at 10 or 12 dpi. Furthermore, the route of inoculation and dose did not affect the magnitude of these histopathologic changes at these sites.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. MAYV infection of rhesus macaques produces mild inflammatory responses in joints and muscles.

(A) Pathology scores in joints and muscles for MAYV infected RM based on quantitative histopathologic criteria (S1 Table). The colors and shape identifiers denote individual RM. Bars show means. ns denotes not significant (Kruskal-Wallis, p > 0.05). Representative histology images of (B) Joint synovium of finger, animal 2 from 7 log₁₀ PFU IV group, scale bar: 20 micrometers. (C) Joint synovium of toe, animal 10 from 3 log₁₀ PFU SC group, scale bar: 100 micrometers. (D) Joint synovium of elbow, animal 2 from 7 log₁₀ PFU IV group scale bar: 50 micrometers. Asterisks indicate immune cell(s).

https://doi.org/10.1371/journal.pntd.0013061.g004

MAYV-viremic rhesus macaques infect Ae. aegypti 2 days after macaque inoculation

To evaluate the ability of RM to transmit MAYV to a primary arbovirus vector in urban settings, we allowed female Ae. aegypti mosquitoes, sourced from a colony originating from Los Angeles, California, USA, to bloodfeed on each anesthetized RM 2, 3, 5, and 7 dpi (Fig 5A). These times were chosen based on our projection of when the animals would be viremic. Since we did not test RM plasma in real-time, we were unaware that animals cleared detectable viremia after 3 dpi, and neither of the 2023 macaque studies showing short viremias had been published at the time our study was conducted. All mosquitoes that bloodfed were incubated for 10 days, the period during which MAYV disseminates and reaches mosquito saliva [27–30]. We assessed MAYV infection in mosquitoes by testing their bodies, legs and wings for dissemination, and saliva, a proxy for transmission. All RM showed detectable viremias at 2 and 3 dpi, but none were viremic at 5 or 7 dpi (Fig 5B). The timing of mosquito presentations therefore covered both viremic (2 and 3 dpi) and aviremic (5 and 7 dpi) periods in RM. Except for one mosquito that fed 3 dpi, only RM 2 dpi were infectious to Ae. aegypti (Fig 5C). For mosquitoes that fed 2 dpi, 48% (104/217) of bodies tested positive for MAYV. Among these, 38% (39/104) of their legs and wings tested positive. These data demonstrate that MAYV-infected RM can infect Ae. aegypti but the window of infectiousness is narrow and does not extend past 2 dpi in this model. Dissemination rates from these mosquitoes were also low.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 5. MAYV viremic rhesus macaques are infectious to Ae. aegypti mosquitoes in a narrow window, 2 dpi.

(A) Experimental design showing the timing of mosquito exposure to MAYV-infected RM 2, 3, 5, and 7 dpi, followed by mosquito incubation and harvest of bodies, legs and wings, and saliva 10 days post-feed (dpf). (B) Mean viremia levels in RM on mosquito feeding days, represented as cohort log-transformed means from the data shown in Fig 1B. Error bars show standard deviations. (C) Rates of MAYV infection in Ae. aegypti (bodies), dissemination in legs and wings (L&W), and transmission in saliva from mosquitoes that fed on RM 2, 3, 5, or 7 dpi. Error bars show standard deviations and the fractions above each error bar represent the number of mosquitoes with detectable infectious MAYV divided by the number of mosquitoes tested. Images created using BioRender.

https://doi.org/10.1371/journal.pntd.0013061.g005

Few Ae. aegypti that feed on viremic rhesus macaques transmit infectious MAYV

To evaluate the transmissibility of MAYV by infected mosquitoes, we used detection of infectious virus in saliva as a surrogate for transmission to a vertebrate. Saliva samples from individual Ae. aegypti were tested using a Vero cell passaging approach. Saliva collected into capillary tubes from mosquitoes with detectable infectious MAYV in their legs and wings was passaged once or twice on Vero cells. Samples showing cytopathic effects at either passage in Vero cells were considered to contain infectious MAYV. Of the 39 mosquitoes with disseminated infections in legs and wings, 4 (10%) had infectious MAYV in saliva (Fig 5C). Of these four mosquitoes, one fed on RM 3, two fed on RM 4 and one fed on RM 12. The overall potential transmission rate for all 217 Ae. aegypti that fed on RM in this study was 2% (4/217), and 4% (4/104) of mosquitoes with infected bodies. These data show that while Ae. aegypti infected from RM may transmit MAYV, the probability of transmission appears to be low.

Individual rhesus macaques are variably infectious to Ae. aegypti

To assess MAYV infectivity in Ae. aegypti that ingested MAYV from viremic RM 2 dpi, we quantified infectious MAYV in bodies (representing infection) and legs & wings (representing disseminated infection) 10 dpf. The mean infection rates were not significantly different in cohorts that fed on RM inoculated with different MAYV doses or routes (Kruskal–Wallis, p > 0.05).(Fig 6A). Eleven of 12 RM infected at least one Ae. aegypti body. Infection and dissemination rates varied widely within cohorts that fed on different RM, ranging from 5% (1/20) to 90% (18/20) infected and 0% (0/1) to 100% (4/4) disseminated (Fig 6B).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 6. MAYV infects and disseminates Ae. aegypti that ingest bloodmeals from viremic rhesus macaques 2 dpi.

(A) Cohort infection rates of Ae. aegypti exposed to RM treated with the same dose and route of MAYV. Each dot represents the mean infection rate from a cohort that fed on a single RM, with error bars indicating standard deviations. Fractions on bars represent number of positive mosquito bodies divided by the number of bloodfed mosquitoes. (B) Rates of infection (bodies, solid bars) and dissemination (legs and wings, open bars) in cohorts of Ae. aegypti that ingested bloodmeals from each viremic RM. Fractions above bars show the number of positive tissue samples divided by the number tested. The colors and shape identifiers denote individual RM, which are also identified with a unique number. (C-E) Infectious MAYV levels in bodies and legs of individual Ae. aegypti that ingested bloodmeals from viremic RM and which became infected. Doses and routes of RM treatment are indicated by colored data points, and red numbers show the infectious plasma MAYV RM titers at the time of mosquito feeding, 2 dpi. Each dot represents a single mosquito tissue titration, with solid horizontal lines showing the mean values. Mosquito tissues with no detectable infectious MAYV above the LOD are not included in means; the numbers (N) of negatives are shown below the LOD line. The horizontal dotted line shows the assay limit of detection (LOD), 3.3 log₁₀ PFU/tissue. The colors and shapes of data points follow designations for individual RM, and data points with black outlines represent mosquitoes that fed on female RM. ns denotes not significant, p < 0.05, ANOVA for panel A and Kruskal-Wallis for panels C-E.

https://doi.org/10.1371/journal.pntd.0013061.g006

Ae. aegypti that feed on viremic rhesus macaques develop a range of MAYV levels

Mean Ae. aegypti body or leg and wing titers were not significantly different across cohorts that fed on RM inoculated with different MAYV doses or routes (one-way ANOVA, p > 0.05) (S2 Fig). Ae. aegypti developed a range of MAYV titers, with cohort body means ranging from 5.1 to 6.8 log₁₀ PFU. For legs and wings, mean titers ranged from 4.2 to 5.8 log₁₀ PFU. Mean body titers in cohorts that fed on RM in the same group were not different (Kruskal Wallis, p > 0.05, [mean leg and wing titers were not compared statistically due to small and uneven sample sizes]) (Fig 6C–6E). The results from these studies demonstrate individual RM are variably infectious to Ae. aegypti 2 dpi, with the majority of animals infecting mosquitoes. Individual mosquitoes that fed on the same RM showed variable rates of infection and dissemination and variable magnitudes of infection.

Rhesus macaque MAYV viremia at the time of mosquito feeding does not predict Ae. aegypti infection rates or magnitudes

The variable infectivity of RM for Ae. aegypti was accompanied by a 3 log₁₀ viremia range (2–5 log₁₀ PFU/ml blood, as indicated by the red values in Figs 5B and 6C–6E) at mosquito presentation. This prompted us to investigate whether RM viremia at the time of mosquito feeding correlates with mosquito infection rates. We predicted that, similar to conventional vector competence studies that use virus-spiked bloodmeals, Ae. aegypti infected by viremic RM would show a dose-dependent response to MAYV infection. To test this, we performed a Spearman’s correlation analysis comparing infection rates in Ae. aegypti that fed on different MAYV-infected RM (Fig 7A). The correlation between RM viremia and Ae. aegypti infection rate (rho = 0.43), showed a moderately positive relationship but was not statistically significant (p = 0.16). We also assessed whether the magnitude of viremia at mosquito presentation correlated with mean MAYV body titers in cohorts of infected mosquitoes that fed on each RM (Fig 7B). The rho value of 0.0046 (p = 0.99) showed no association between these two variables. These results indicate that the magnitude of MAYV RM viremia at the time of Ae. aegypti presentation does not predict Ae. aegypti infectiousness or infection levels in mosquitoes.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 7. Rhesus macaque MAYV viremia levels do not correlate with Ae. aegypti (A) infection rates or (B) mean body titers.

Lines show linear regression and the R² value represents the coefficient of determination, p shows p values, Spearman’s correlation analyses.

https://doi.org/10.1371/journal.pntd.0013061.g007

Mayaro virus is genetically stable during rhesus macaque infection and transmission from Ae. aegypti

To evaluate whether genetic changes in MAYV arising during infection of RM or Ae. aegypti may influence infectivity of RM to mosquitoes which could explain the lack of correlation between the magnitude of RM viremia and Ae. aegypti infection rates, we deep sequenced the MAYV genomes in plasma from all 12 RM at 2 dpi and compared it to the MAYV plasmid sequence that was used to generate the virus stock that was inoculated into RM. We also sequenced all 4 Vero passaged MAYV positive Ae. aegypti saliva samples. All 12 plasmas and 3 of the 4 Ae. aegypti saliva samples were identical across the genome at the consensus level (>50% of RNAs) to the sequence of MAYV used to produce the virus stock for RM inoculations. There was a single non-coding mutation (G5758A) in 1 Vero passaged mosquito saliva sample from a mosquito that fed on RM 3. These data do not support viral genetic differences as a contributor to the variable infectiousness of RM for Ae. aegypti.