Viral and Antibody Kinetics, and Mosquito Infectivity of an Imported Case of Zika Fever Due to Asian Genotype (American Strain) in Singapore

Abstract

We report a case of a Singaporean who acquired Zika virus (ZIKV) during a visit to Cuba. The infection was confirmed using molecular and serological methods. This report highlights potential drawbacks of using IgG serology for diagnosis of flavivirus infections in endemic regions. The low viremia detected during the early phase of this case resulted in low mosquito infectivity rates, suggesting the possibility of ZIKV transmission prior to clinical onset. The report also emphasizes the challenges of public health interventions for Zika fever and the importance of sustaining a low vector population to reduce the risk of arbovirus transmission in vulnerable regions.

1. Introduction

Zika virus (ZIKV) caused a major outbreak in the Pacific Island of Yap in 2007 [1]. Global spread brought the virus to the Americas, where it has been associated with neurological, developmental, and ophthalmological abnormalities [2,3]. For the past four decades, Zika virus has been reported in South East Asia—including Malaysia in 1969 and Indonesia in 1977 [4,5]. However, it attracted little attention until the huge outbreaks in the Americas [6] and since then, several countries had reported autochthonous transmission of ZIKV [7]. This included Singapore where transmission, limited to a single lineage of the Asian genotype, has been documented in Singapore since August 2016 [8]—458 cases in 2016 and 67 cases in 2017 [9]. ZIKV is primarily transmitted by the Aedes species of mosquitoes, which also transmits the Dengue virus (DENV) [10]. Because the two closely related flaviviruses overlap greatly in geographical spread [11], and the serological tests of the two viruses inter-cross react [12], ZIKV diagnosis using serology remains a challenge in endemic settings [1,13].

2. The Study

In August 2017, a 51-year old female infected with the American strain of Asian genotype (NCBI accession No. MF988734) was reported in Singapore. The patient left Singapore for Havana, Cuba on 6 August 2017, and returned on 13 August, one day before the onset of symptoms. She initially presented with headache and neck/shoulder ache (day 0) and became febrile (37.8–38 °C) on the following day (Figure 1). Fever was associated with an itchy maculopapular rash and mild conjunctivitis (Figure 2A,B). The rash initiated from the arms and back of the torso, symmetrically disseminated to the face and whole body (including soles, palms and between fingers and toes) by day 3–4, and subsided on day 5. The rash reappeared on day 8 and lasted for two days. She developed bilateral numbness of thumb, index and middle fingers on day 4 and swelling of the wrist and hand from day 4 to 6. Despite uneventful recovery on day 9, the patient continued to experience skin tenderness when touched (tactile allodynia) for two days. Case had loose stool, once or twice a day from day 3 to 9. Fever and body aches were relieved by the antipyretics and analgesics.

Laboratory tests conducted on day 4 of illness showed leukopenia (3.1 × 10⁹/L) and reactive lymphocytosis (0.28 × 10⁹/L). A nerve conductance test on day 5 provided no evidence of diffuse peripheral neuropathy, with motor, sensory and late responses of the upper and lower limbs within normal ranges. Guillain-Barre Syndrome (GBS) was ruled out, and the numbness of the fingers was probably due to transient compression of the median nerve as a result of wrist joint swelling.

ZIKV infection was confirmed by RT-qPCR [1] on the urine sample collected on day 1 of illness. ZIKV RNA (4.59–6.69 log₁₀RNA copy/mL) remained detectable in all urine samples (n = 15) up to day 7 and RNA titres had no correlation with the hour of urine collection. Viral RNA was detectable up to day 6 in whole blood, but up to day 3 in sera. Day 1 sera yielded the highest viremia (7.49 log₁₀RNA copy/mL; 50 pfu/mL by plaque assay), which dropped on day 2 to 6.03 log₁₀RNA copy/mL and day 3 to 5.02 log₁₀RNA copy/mL. ZIKV RNA was also detectable in day 2–3 saliva samples. Virus was isolated from urine and sera samples using Aedes albopictus cell line (C6/36, ATCC© CRL-1660™) for at least three passages, and confirmed by immunofluorescence assay with flavivirus group antigen complex (ATCC© HB112™) and sequencing. No cytopathic effect was observed in the infected cell cultures.

IgM was first detected on day 3, and IgG on day 1. Patient had previously tested positive for Dengue virus (DENV)-1 and -2 neutralising antibodies by Plaque Reduction Neutralisation Test (PRNT₅₀) assay [15], and had vaccination against Yellow Fever and Japanese encephalitis viruses in 2008 and 2012 respectively. The early emergence of ZIKV-reactive IgG was likely due to the stimulation of cross-reactive antibodies from past flavivirus exposures—similar to the circumstances surrounding a secondary dengue infection. PRNT₅₀, adapted from Low et al. [15], using three strains of ZIKV (Uganda strain, MR 766, ATCC VR-84; Thailand strain, NCBI accession No. KF993678; Puerto Rico strain, PRVABC59, ATCC VR-1843) showed an increasing titre from <10 on day 1, 32–90 on day 4 to 180–450 on day 8. Case was negative for dengue (Non-structural 1 (NS1) antigen and IgM) and chikungunya (PCR and IgM).

The complete genome sequence of the isolate shared the highest nucleotide (99.8%) and amino acid (100%) similarity with a Cuban strain (NCBI accession No. MF159531) detected in the United States of America in April 2017. The virus was distinct from autochthonous strains circulating in Singapore since August 2016, though both belonged to Asian genotype (Figure 3). Sequence analysis of the virus strain, along with the patient’s travel history, suggests that the infection occurred in Cuba. ZIKV whole genome was sequenced as per the protocol described in Supplementary File S1.

Day 1 to 4 residual ethylenediaminetetraacetic acid (EDTA) blood samples were fed to local Aedes aegypti mosquitoes (n = 100) [16], which were then sacrificed to test for ZIKV (

Table 1. Infection of Aedes aegypti (F2 generation).

Days Post Infection	Fever Days (Zika virus log10RNA Copy/mL)/Infection Rate (n)
	Day 1 (7.49)			Day 2 (6.03)			Day 3 (5.02)			Day 4 (Undetectable)
	MG	SG	Head	MG	SG	Head	MG	SG	Head	MG	SG	Head
5	10% * (10)	0 (10)	10% (10)	0 (10)	0 (10)	20% (10)	ND			ND
7	0 (10)	0 (10)	10% (10)	0 (10)	0 (10)	0 (10)	0 (10)	0 (10)	10% (10)	0 (10)	0 (10)	10% (10)
14	0 (10)	0 (10)	10% (10)	0 (10)	0 (10)	20% (10)	0 (10)	0 (10)	10% (10)	0 (10)	0 (10)	0 (10)

Female mosquitoes were fed with residual EDTA blood samples (drawn within 2 h). Fully engorged mosquitoes were incubated at 28 °C ± 1 °C with 80% ± 10% relative humidity. Midgut (MG) and salivary glands (SG) were dissected at 5, 7 and 14 days post infection and viral titres were determined using 50% tissue culture infectious dose assay (log₁₀ TCID₅₀/mL) [18]. Mosquito heads were analyzed for ZIKV RNA using RT-qPCR as previously described [1]. The discrepancy in the results between the cell-based assay and the RT-qPCR assay could be due to the higher sensitivity of the latter assay. The RT-qPCR assay measures not only the infectious virus but also non-infectious, immature and defective virions, which are not capable of further infection and replication. * 4.95 log₁₀TCID₅₀/mL midgut viral titre; ND = not detected.

). Viable ZIKV was recovered from only one midgut of 30 mosquitoes fed with day 1 blood sample. None of the mosquitoes yielded viable ZIKV in the salivary gland, suggesting non-infectivity. From each feed, ZIKV RNA was detected only in 10–20% of mosquito heads, showing low level of viral dissemination. The low infectivity of ZIKV was in contrast to DENV, in which the viremia of ~7 log₁₀RNA copy/mL resulted in 90–100% of midgut and 25–50% of salivary gland infection rates [16]. To the best of our knowledge, this is the first report using a patient’s venous EDTA blood to infect mosquitoes. Unfortunately, the ZIKV titre in the EDTA blood was low and could have resulted in the low mosquito infection rate. The low infectivity also greatly differed from the high infection rate (as high as 100%) reported earlier based on laboratory studies that used high ZIKV titres to feed mosquitoes [17,18]. However, it has been shown that the susceptibility of geographically different populations of Ae. aegypti mosquitoes vary among different ZIKV strains [19,20,21]. Pompon et al. [17] has also reported the absence of infection among local Ae. aegypti mosquitoes fed with low ZIKV titres (10² pfu/mL).

3. Conclusions

The case exhibited classical symptoms of Zika infection [22], with no evidence of GBS. Distinct virus lineages circulating in Asia and the Americas suggest that outbreaks on the two sides of the globe are independent, and possibly driven by environmental factors such as the El Nino of 2015–2016 [23]. This report highlights the limit of IgG serology against the backdrop of flavivirus endemicity. It also demonstrates the risk of cross border exchanges of emerging viruses. The low viremia and low mosquito infectivity rate, as early as day 2 of disease, and their subsequent plunge suggest that the case was probably more infective before the clinical onset. Together with the high rate of asymptomatic cases, these factors contribute to the challenges of public health interventions for Zika fever and highlight the importance of sustaining a low vector population to reduce the risk of arbovirus transmission in vulnerable regions.