Reverse genetics rescue of sylvatic dengue viruses

Abstract

There are an unknown number of sylvatic ("sylvan" = "of the forest") dengue viruses currently being sustained by nonhuman primates and mosquitoes in the forests of Africa and Asia. Humans are regularly infected with these viruses, occasionally resulting in small epidemics. One barrier to understanding sylvatic dengue virus biology is the scarcity of stocks available for study. While the full-length genome sequences of 28 sylvatic viruses exist on GenBank, accessible infectious stocks exist for only a handful of these. One way to overcome this obstacle is to rescue infectious virus stocks using reverse genetics. However, reverse genetic engineering of dengue viruses is notoriously difficult. Here, we optimize a reverse genetics method specifically for the rescue of sylvatic dengue virus stocks from sequence data. The key to our approach was the observation that mosquito cells, rather than mammalian cells, must be used to launch sylvatic dengue virus replication from assembled genomes. We demonstrate the success of this technique by rescuing seven sylvatic dengue viruses. With this unique collection, we then characterized the relative fitness of each virus strain on human, monkey, and mosquito cells. While mosquito cells are universally permissive for the growth of sylvatic dengue viruses, some sylvatic dengue virus strains showed significantly better replication in human and monkey cells than others. These sylvatic dengue virus strains may have a greater potential for human adaptation.

Importance: Given the enormous burden of the four human dengue viruses, which emerged from the sylvatic dengue virus reservoir, it is important that we consider the possibility of a new dengue virus emerging into the human population. Nonhuman primate species in Asia and Africa are suspected to be the natural reservoir hosts for sylvatic dengue viruses. Occasionally, these sylvatic dengue viruses infect humans, although there are few stocks of these viruses available for study in the lab. Here, we optimize a reverse genetics technique for sylvatic dengue viruses, and we rescue stocks of seven strains. With this method, theoretically, any sylvatic dengue virus sequence deposited on GenBank can be transformed into a high-titer infectious virus stock.

Keywords: dengue virus; reverse genetics.

Fig 1

The history of human infections with sylvatic dengue viruses. This figure summarizes the history of sylvatic dengue virus infections in humans, as reported in the literature from the 1960s to today. In 1966, three closely related strains of sylvatic dengue virus (strains IBH11234, IBH11208, and IBH11664) were isolated from humans in Nigeria (12, 18). In 1970 and 1990, strains DakHD10674 and DakHD76395 were isolated from humans in Senegal (12, 19). In 2005, strain P72-1244 was isolated from a patient with dengue fever in Malaysia, a strain that was first isolated from a sentinel monkey in 1972 (12, 20). In 2007 and 2008, two additional strains, DKE-121 and DKD811, were isolated from patients in Malaysia (21). In 2009, strain EEB-17 was isolated from a traveler returning from Guinea Bissau (22). In 2014 and 2015, strains Brun2014 and DSab2015 were isolated from travelers returning from Brunei and Malaysia, respectively (23, 24). In 2020, a small outbreak of sylvatic dengue virus occurred in Senegal, with 59 infections being documented and three strains sequenced (strains SH356683, SH356692, and SH356702) (17). A year later, an additional sylvatic dengue virus infection was confirmed in a man in Senegal (25).

Fig 2

In vitro assembly of a sylvatic dengue virus genome (strain DakArA1247). (A) The sequence of sylvatic dengue strain DakArA1247 was obtained from GenBank. From this, three overlapping DNA fragments were designed and synthesized by a commercial vendor. Diagonal lines indicate overhangs added by primers during fragment amplification. (B) In addition, two alternate “UTR-linker” sequences (fragment 4) were synthesized, differing only by the promoter they contain. Diagonal lines indicate overhangs added by primers during fragment amplification. The CPER assembles all four products into circularized form. (C) Upon receipt of synthetic DNA fragments 1–4, each fragment was amplified by PCR. Fragment 1 was amplified with a forward primer that added an overhang sequence complementary to the end of the promoter region of fragment 4; thus, two promoters are indicated in the fragment 1 lane. Also, the two versions of fragment 4 are shown. (D) PCR products of the four fragments were combined at equimolar ratios and cycled in a thermocycler to perform the CPER. Two bands are evident in the uncycled control: fragments 2 (4.1 kb) and 3 (3.8 kb) are similar sizes and presumably run together as the upper band, and the lower band is fragment 1 (2.8 kb). The linker fragment (~800 or 1,000 bp depending on the promoter) is too dilute to see. (E) The product of the CPER is then directly transfected into permissive cells to launch virus replication.

Fig 3

Identification of cell lines that can launch sylvatic dengue virus replication from in vitro assembled genomes. (A–C) Dengue virus genomes (legend) were assembled by CPER, with either (A and B) the mammalian CMV promoter or (C) the insect OpIE2 promoter. Assembly products were transfected into either (A) Vero E6, (B) HEK 293T, or (C) C6/36 cells. Supernatants were collected at the indicated days post-transfection. The infectious virus in the supernatant at each time was determined by a focus forming unit (FFU) assay. Each line represents one of the three biological replicates. The dotted line represents the limit of detection (LD); points underneath the LD were arbitrarily assigned a value for visualization purposes. (D–F) CPER-assembled dengue virus genomes (top of graphs) were transfected into eight different cell lines. Supernatants were collected 7 days post-transfection, and virus titers were obtained by FFU assay. This entire process was repeated 6–9 times for each cell line, and each dot represents an independent experiment. The dotted line represents the limit of detection; points underneath were plotted at ND (not detected). The frequency of success for launching a virus in each cell type (bottom of graphs) was calculated by dividing the number of experiments that yielded at least one infectious virus (by FFU assay) by the total number of individual experiments attempted.

Fig 4

A sylvatic dengue virus (DakArA1247) rescued by this reverse genetics method exhibits similar properties to an existing stock obtained from BEI Resources. (A) C6/36 cells were exposed to stocks of DakArA1247 that were either obtained from BEI Resources or rescued via the method herein. The multiplicity of infection was 0.1, and cell supernatants were collected at 0, 1, 2, 3, 5, and 7 days post-exposure, from which viral RNA was extracted and analyzed by RT-qPCR. Synthetic dengue genomic RNA was used to generate a standard curve of genome copies per milliliter, from which Y-axis values were derived. The dotted line represents the limit of quantification. (B) Supernatant was collected 5 days post-exposure, and the concentration of infectious virus was determined by FFU assay. The dotted line represents the limit of detection. (C) Representative images of foci obtained by testing supernatant 5 days post-exposure by FFU assay. (D) Representative images of cytopathic effects observed on C6/36 cells 5 days post-exposure. (E) cDNA was generated from extracted viral RNA from each virus. PCR amplicons spanning bases 2,037–4,583, 4,118–6,695, and 6,031–8,515 of the DakArA1247 genome were amplified from cDNA in two independent reactions and sequenced. Sequencing coverage and identified mutations are listed.

Fig 5

Only some sylvatic dengue virus strains replicate in human cells. (A–C) The indicated human cell lines (top of graphs) were exposed to eight launched stocks of dengue virus (virus strains denoted on X axis) at a multiplicity of infection of 0.1. All are sylvatic strains except 16681, which is a human dengue virus control. Supernatants were collected 3 days post-exposure, and virus titers were determined by FFU assay (Y axis). Each dot represents a technical replicate, with the top of the bar representing the mean of all three technical replicates. The dotted line represents the limit of detection of the assay. Values under the limit of detection were assigned a value of 1 FFU/mL for the purpose of calculating the mean and are plotted at “not-detected” (N.D.) on the Y axis.

Fig 6

Only some sylvatic dengue virus strains replicate in monkey cells. (A) African green monkey, Chlorocebus sabaeus, photographed in The Gambia by Charles Sharp and used under CC BY-SA 4.0. (B and C) The indicated cell lines (top of graphs) were exposed to one of eight launched stocks of dengue virus (virus strains denoted on X axis) at a multiplicity of infection of 0.1. All are sylvatic strains except 16681, which is a human dengue virus control. Supernatants were collected 3 days post-exposure, and virus titers were determined by FFU assay (Y axis). Each dot represents a technical replicate, with the top of the bar representing the mean of all three technical replicates. The dotted line represents the limit of detection of the assay. Values under the limit of detection were assigned a value of 1 FFU/mL for the purpose of calculating the mean and are plotted at “not-detected” (N.D.) on the Y axis.

Fig 7

All tested sylvatic dengue virus strains replicate in mosquito cells. (A, B) The indicated cell lines (top of graphs) were exposed to one of eight launched stocks of dengue virus (X axis) at a multiplicity of infection of 0.1. All are sylvatic strains except 16681, which is a human dengue virus control. Supernatants were collected 3 days post-exposure, and virus titers were determined by FFU assay. The Y-axis represents FFUs per milliliter. Each dot represents a technical replicate, with the top of the bar representing the mean of all three technical replicates. The dotted line represents the limit of detection of the assay.