(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Flavivirus genome recoding by codon optimisation confers genetically stable in vivo attenuation in both mice and mosquitoes [1]

['Wei-Xin Chin', 'Laboratory Of Molecular Rna Virology', 'Antiviral Strategies', 'Department Of Microbiology', 'Immunology', 'Infectious Diseases Translational Research Programme', 'Yong Loo Lin School Of Medicine', 'National University Health System', 'National University Of Singapore', 'Hao Yuin Kong']

Date: 2023-11

Virus genome recoding is an attenuation method that confers genetically stable attenuation by rewriting a virus genome with numerous silent mutations. Prior flavivirus genome recoding attempts utilised codon deoptimisation approaches. However, these codon deoptimisation approaches act in a species dependent manner and were unable to confer flavivirus attenuation in mosquito cells or in mosquito animal models. To overcome these limitations, we performed flavivirus genome recoding using the contrary approach of codon optimisation. The genomes of flaviviruses such as dengue virus type 2 (DENV2) and Zika virus (ZIKV) contain functional RNA elements that regulate viral replication. We hypothesised that flavivirus genome recoding by codon optimisation would introduce silent mutations that disrupt these RNA elements, leading to decreased replication efficiency and attenuation. We chose DENV2 and ZIKV as representative flaviviruses and recoded them by codon optimising their genomes for human expression. Our study confirms that this recoding approach of codon optimisation does translate into reduced replication efficiency in mammalian, human, and mosquito cells as well as in vivo attenuation in both mice and mosquitoes. In silico modelling and RNA SHAPE analysis confirmed that DENV2 recoding resulted in the extensive disruption of genomic structural elements. Serial passaging of recoded DENV2 resulted in the emergence of rescue or adaptation mutations, but no reversion mutations. These rescue mutations were unable to rescue the delayed replication kinetics and in vivo attenuation of recoded DENV2, demonstrating that recoding confers genetically stable attenuation. Therefore, our recoding approach is a reliable attenuation method with potential applications for developing flavivirus vaccines.

The mosquito-borne flaviviruses such as dengue virus (DENV) and Zika virus (ZIKV) have established themselves as major human pathogens. Live attenuated vaccines are seen as the most effective method for preventing flavivirus infection. Flavivirus genome recoding has emerged as a next-generation vaccine development method that acts by rewriting the flavivirus genome. Previous flavivirus genome recoding attempts were based on deoptimising the flavivirus genome. However, these deoptimised flaviviruses were found to be attenuated in a species dependent manner. For example, deoptimised DENV and ZIKV did not demonstrate attenuation in mosquito cells or mosquito animal models, which is undesirable because these mosquito-borne flaviviruses should be attenuated in their mosquito vector to prevent vaccine escape. To overcome these limitations, we adopted a flavivirus genome recoding approach based on the contrary approach of optimising the flavivirus genome and applied it to DENV2 and ZIKV. We found that this genome recoding approach of codon optimisation could confer attenuation in both mouse and mosquito animal models. This indicates that our flavivirus genome recoding approach may be used as a reliable method to construct attenuated vaccine backbones for the mosquito-borne-flaviviruses in general.

Funding: This work was supported by the MOE Tier 2 2017 grant (MOE 2017-T2-1-078) and MOE Tier 2 2021 grant (MOE T2EP3-02-2021) awarded to J.J.H.C. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Data Availability: All data generated or analysed during this study are included in this published article (and its supplementary information files). High resolution RNA structure are available in the supplementary data. The sequence of DENV2-rcCap-NS1, the most extensively recoded clone, is also available on Genbank (accession number OP909734). The sequence of DENV2-rcCap-NS1, the most extensively recoded clone, is also available on Genbank (accession number OP909734). We will make the sequence of OP909734 available to the public upon acceptance of the manuscript.

Copyright: © 2023 Chin et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Therefore, aim of this study is to demonstrate a flavivirus genome recoding approach that can produce attenuated dengue virus type 2 (DENV2) and Zika virus (ZIKV) strains. This virus genome recoding approach is to codon optimise the flavivirus protein coding region, leading to the disruption of functional RNA elements that are essential for efficient virus replication [ 26 , 31 – 40 ].

We propose an alternative approach to virus genome recoding. The flavivirus genome contains many functional RNA elements that are essential for efficient virus replication [ 26 , 31 – 40 ]. These functional RNA elements may take the form of pseudoknots, RNA secondary structures, or long-range RNA interactions [ 26 , 31 – 40 ]. For example, the capsid coding region contains conserved RNA elements such as the capsid coding region hairpin element (cHP), the 5′ cyclisation sequence (5′CS), and the downstream of 5′ cyclization sequence pseudoknot (DCS-PK) [ 32 , 35 , 40 ]. Some RNA elements play a role in the cyclisation of the viral RNA genome, which is required for the virus genome to transition from a linear protein translation state to a circularised RNA replication state [ 32 , 35 – 40 ]. Silent mutations that target the cHP, 5′CS, or DCS-PK elements can inhibit genome cyclisation, which in turn reduces viral RNA replication efficiency [ 32 , 35 , 40 ]. Therefore, we hypothesised that a recoding approach that targets these RNA elements would lead to reduced replication efficiency and attenuation. Furthermore, because the function of these RNA elements is something inherent to flaviviruses, the attenuation mechanism should function regardless of cell type or animal species. For this purpose, we chose codon optimisation as a method of introducing a sufficient number of well-spaced silent mutations to disrupt the sequence, structure and function of flavivirus RNA elements.

An alternative rational design approach is synonymous virus genome recoding. Virus genome recoding is an attenuation method that typically involves deoptimising a virus genome by altering the frequencies of favourable codons, unfavourable codons, or CpG and UpA dinucleotides [ 8 , 12 – 25 ]. For flaviviruses, deoptimisation for a human host is usually performed by optimising the virus for an insect or mosquito host, but this leads to inconsistent results that are highly dependent on cell type or animal species [ 13 , 14 , 19 , 20 , 22 , 26 ]. For example, these deoptimisation approaches do not affect DENV replication in mosquito cells or mosquitoes [ 13 , 19 ]. This lack of effect in mosquitos is especially undesirable because vaccines for the mosquito-borne flaviviruses must lack transmissibility by their mosquito vectors [ 27 – 30 ].

The mosquito-borne flaviviruses have emerged as major threats to human health and quality of life [ 1 , 2 ]. There are several established live attenuated vaccines for flaviviruses, as well as live dengue virus (DENV) vaccines that have shown promise in phase II & III clinical trials [ 3 , 4 ]. Live attenuated flavivirus vaccines are considered safe and effective at protecting against flavivirus infection as they can confer life-long immunity with a single dose [ 3 , 4 ]. However, that there is no consistent or reliable method for generating a sufficiently attenuated flavivirus vaccine strain; methods that have proven successful in producing attenuated vaccine strains for some flaviviruses have failed to produce attenuated vaccine strains for other flaviviruses [ 4 – 11 ]. Even rationally designed live vaccines can suffer from the same problems of inconsistency and unpredictability [ 9 – 11 ].

Results

Recoded clones have higher protein expression efficiency but lower RNA replication efficiency Next, we investigated if DENV2 genome recoding affects viral protein expression or viral RNA replication. A translation reporter construct was used to investigate viral protein expression efficiency, while a subgenomic replicon was used to investigate RNA replication efficiency (Fig 1B) [41]. We cloned the rcE2-90, rcE2, and rcNS1 mutations into these constructs (Fig 1B) and investigated their effects in BHK-21 cells. We found that the recoded rcE2-90, rcE2, and rcNS1 translation reporter constructs did have higher firefly luciferase activities compared to the non-recoded control (Fig 1D), while the rcE2-90, rcE2, and rcNS1 replicons had lower replicon RNA levels compared to the non-recoded control (Fig 1E). This indicates that DENV2 recoding results in a simultaneous enhancement of viral protein translation efficiency and reduction of viral RNA replication efficiency. This is consistent with the disruption of an RNA element that regulates the transition of the DENV2 RNA genome from the linear protein translation state to the competing circularised RNA replication state [37–40].

The envelope stem coding region contains a putative RNA element We wanted to determine if the phenotype of DENV2-EGFP-rcE2-90 requires the recoding of any specific region. We constructed three additional recoded clones, rcE2-60, rcE2-50, and rcE2-40, with the respective recoded regions narrowed down to 60, 50, and 40 codons respectively (S2A Fig and Table 1). We DNA-launched these recoded DENV2-EGFP clones in BHK-21 cells and used fluorescent microscopy to compare their replication efficiency. The rcE2-60 and rcE2-50 recoded clones retained the reduced replication efficiency of the rcE2-90 clone, while the rcE2-40 clone replicated faster (S2B Fig). The rcE2-50 clone differs from the rcE2-40 clone by a region corresponding to nucleotides 2197 to 2226 of the DENV2 genome and codons 421 to 430 of the envelope protein coding region. Codons 421 to 430 encode for the envelope protein stem region [42]. In wildtype DENV2, nucleotides 2197 to 2226 are predicted to be part of a RNA hairpin structure (S2C and S5 Figs), and the recoding mutations are predicted to disrupt this RNA hairpin. Therefore, this RNA hairpin may be a RNA element that contributes to efficient DENV2 replication. We named this putative RNA element the envelope stem RNA element (ESRE).

Degree of recoding is correlated with slower DENV2-EGFP replication Next, we investigated the effects of increasing the degree of genome recoding in the DEVN2-EGFP clones by constructing the rcCap-Env and rcCap-NS1 clones (Fig 1A and Table 1). We then DNA-launched wildtype (non-recoded) DENV2-EGFP as well as the rcCap-Env and rcCap-NS1 clones in BHK-21 cells to compare their replication efficiency. Compared to non-recoded DENV2-EGFP, the recoded rcCap-Env and rcCap-NS1 clones all demonstrated a great reduction in replication efficiency, as indicated by great reduction in EGFP positive cells (Fig 1F). This demonstrates that increasing the degree of genome recoding can lead to a further decrease in virus replication efficiency.

Recoded DENV2 clones have reduced replication efficiency We wanted to confirm that our results were not an experimental artifact of using an EGFP reporter virus. Therefore, we constructed three recoded DENV2 clones that are based on a wildtype DENV2-16681 backbone that does not carry any trans-genes. These clones were named rcCap-prM, rcCap-Env, and rcCap-NS1, which roughly corresponds to the regions in the genome that were targeted for recoding (Fig 2A and Table 1). The codon optimisation process introduced numerous silent mutations into the viral genome of these rcCap-prM, rcCap-Env, and rcCap-NS1 clones. S3 Fig shows an alignment that compares the recoded and non-recoded sequences from part of the prM and Env coding region. For any given region of recoding, approximately 57% to 61% of codons are mutated with silent mutations (S3 Fig and Table 1). The silent mutations alter the nucleotide sequence but not the encoded amino acid residue. These silent mutations may be a typical single nucleotide substitution at the third nucleotide of a codon. For example, the 89th codon of the prM coding region is mutated from GAA to GAG, both of which code for Glutamic acid (S3A Fig). The silent mutations may also take the form of double or even triple nucleotide substitutions. For example, the 92nd codon of the prM coding region is mutated from TCA to AGC, both of which encode for Serine (S3A Fig). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. Recoded DENV2 clones have small plaque phenotype and reduced replication efficiency. (a) Genomic maps showing regions of the DENV2 genome recoded with silent mutations. (b) BHK-21 hamster kidney cells, Huh-7 human hepatocarcinoma cells, HepG2 hepatoma cells, and C6/36 Aedes mosquito cells were infected with wildtype or recoded DENV2 clones (rcCap-prM, rcCap-Env, and rcCap-NS1) at an MOI of 0.02. Viral titres were measured using plaque assay. Limit of detection for our plaque assay is 10 PFU/ml. (c) Plaque formation assay of wildtype and recoded DENV2 clones. Serial passage of wildtype and recoded DENV2 clones (rcCap-prM, rcCap-Env, and rcCap-NS1) in BHK-21 cells results in the emergence of large plaque mutants. Large variations in plaque size are due to the resulting mixed plaque population. Plaque sizes were measured in ImageJ using ViralPlaque Fiji macro. Statistical analysis of recoded virus plaque sizes at passage 7 was performed using the wildtype virus at passage 7 as a control. ***: p-value of <0.001. https://doi.org/10.1371/journal.ppat.1011753.g002 The recoded DENV2 clones are derived from the same master sequence. For example, the recoded rcCap-NS1, rcCap-Env, and rcCap-prM clones share the exact same recoding mutations for the capsid and prM protein coding regions, an example of which is shown in S3A Fig. The recoded rcCap-NS1 and rcCap-Env clones also share the exact same recoding mutations for the Env protein coding region, an example of which can be seen in S3B Fig. Since the Env protein coding region of the rcCap-prM clone is not recoded, it shares the same Env protein coding sequence as wildtype DENV2 (S3B Fig). In other words, the rcCap-Env and rcCap-prM clones serve to narrow down the region of recoding seen in the rcCap-NS1 clone. We compared the viral replication kinetics of wildtype and recoded DENV2 clones in BHK-21 hamster kidney cells, Huh-7 human hepatocarcinoma cells, HepG2 hepatoma cells, and C6/36 Aedes mosquito cells. The cells were inoculated at an MOI of 0.02 and plaque assay was used to compare viral titres. When compared to wildtype DENV2, the recoded DENV2 clones demonstrated delayed growth kinetics. In BHK-21, Huh-7, and HepG2 cells, the titres for the recoded DENV2 clones peaked at later timepoints. In HepG2 and C6/36 cells, the recoded DENV2 clones also replicated to lower viral titres (Fig 2B and S1 Table). The reduced replication efficiency in C6/36 mosquito cells is a favourable marker of attenuation that is correlated with reduced transmissibility by mosquitoes [27–30]. The rcCap-Env clone had the lowest replication efficiency in BHK-21, Huh-7, and HepG2 cells, and the second lowest replication efficiency in C6/36 cells. These results confirm that virus genome recoding by codon optimisation can reduce virus replication efficiency regardless of cell type or animal species.

Predicted structure of DENV2 DCS-PK with recoding and rescue mutations The DENV2-rcCap-prM, rcCap-Env, and rcCap-NS1 clones all have the same recoded capsid coding sequence. This includes three nucleotide substitutions that can be mapped to the previously reported DCS-PK element: a177g, u196a, and c197g. The rescue mutations at nucleotide positions 158, 173, 181, and 192 can also be mapped to the DCS-PK. Therefore, we performed in silico modelling to investigate the effects of the recoding and rescue mutations on the DCS-PK [32,44–46]. The a177g mutation in stem 2 as well as the u196a and c197g mutations in stem 3 were found to disrupt the structure of the DCS-PK in recoded DENV2 (Fig 3A and 3B). The a158u and a192u rescue mutations are predicted to base pair with a192 and a158 respectively, and these interactions are predicted to restore stem 3 of the DCS-PK (Fig 3C and 3D). The a158u and a192u rescue mutations cannot base pair with each other, which would explain their negligible co-occurrence frequency (Table 3). The u173c and c181u are also predicted to create additional base pairings within the DCS-PK stem 2 (Fig 3C and 3D). The u173c mutation introduces a base pairing with a177g, which is one of the original recoding mutations, while the c181u mutation introduces a base pairing with a169. Therefore, the capsid rescue mutations are gain-of-function mutations that create additional base pairings within the DCS-PK. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. Predicted structure of DCS-PK RNA element in wildtype and recoded DENV2 clones. Capital letters indicate the original nucleotide sequence. Small letters indicate nucleotide substitutions. Black arrows indicate the positions of silent mutations introduced during codon optimisation: a177g in stem 2, as well as u196a and c197g in stem 3. (a) Predicted structure for Wildtype DENV2. (b) Predicted structure for recoded DENV2 clones (rcCap-prM, rcCap-Env, and rcCap-NS1). (c) & (d) Predicted structure of recoded DENV2 DCS-PK with rescue mutations acquired during serial passage. (c) Recoded rcCap-Env virus with the predominant a158u and u173c rescue mutations (indicated by the red arrows). (d) Recoded rcCap-NS1 virus with the predominant a158u and c181u rescue mutations (indicated by the red arrows). https://doi.org/10.1371/journal.ppat.1011753.g003

Constructing recapitulatory rescue clones of recoded DENV2 We cloned various combinations of the rescue mutations into the rcCap-Env infectious clone (S4 Fig). We chose the rcCap-Env clone for our downstream experiments because it showed the lowest replication efficiency in human and mammalian cells. The rcCap-Env+rsCE rescue clone contains the a158u (Cap-N21I), u173c (V26A), and a1522g (Env-M196V) mutations and recapitulates the dominant species of the rcCap-Env virus population at passage 10. We also constructed two additional rescue clones to deconvolute the contributions of the capsid and envelope rescue mutations: rcCap-Env+rsCap contains the a158u (Cap-N21I) and u173c (Cap-V26A) mutation while rcCap-Env+rsEnv contains only the a1522g (Env-M196V) mutation. As a control, we cloned the a1522g (Env-M196V) mutation into wildtype DENV2 to construct WT+rsEnv. WT+rsEnv recapitulates the genotype of one of the dominant species of the wildtype virus population at passage 10. All the rescue clones were viable as they were able to form plaques (S5 Fig). Amongst all the wildtype, recoded, and rescue clones, the WT+rsEnv rescue clone that possesses the Env-M196V mutation forms the largest plaques (S5 Fig). In contrast, the rcCap-Env+rsEnv rescue clone that also possesses the Env-M196V forms the smallest plaques.

RNA SHAPE analysis shows loss of RNA structures in Recoded DENV2 We wanted to confirm if DENV2 genome recoding could disrupt genomic RNA structures, and if the disruption remained stable after serial passaging. Therefore, we performed RNA SHAPE-MaP analysis on wildtype DENV2 and the rcCap-Env+rsCap rescue clone [47]. Using the results of our SHAPE analysis as a constraint, we also performed in silico modelling of wildtype and recoded virus RNA structural elements [48,49]. We found that almost all the RNA structural elements that are found in the structural protein coding region of wildtype DENV2 are disrupted and no longer found in recoded DENV2 (Fig 4A, 4B, see S6 and S7 for high resolution figures). Furthermore, we found that the a158u (Cap-N21I) and u173c (Cap-V26A) rescue mutations did not play any meaningful role in reversing the large-scale disruptions to RNA structural elements. This confirms that DENV2 genome recoding by codon optimisation results in the disruption of potential RNA elements in the virus genome. It also indicates that these disruptions are genetically stable. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. Recoded DENV2 has disrupted RNA structural elements and genetically stable attenuation. (a) RNA SHAPE-Map analysis of DENV2 genomic RNA structures. Analysis was performed on BHK-21 cells infected with either wildtype DENV2 or the rcCap-Env+rsCap recoded rescue clone. After infection, NAI treatment was performed to modify single-stranded nucleotides, after which total RNA extraction, cDNA library preparation and Illumina sequencing was performed according to the SHAPE-Map protocol. Differences in SHAPE were assessed using ΔSHAPE. Inverted red triangle indicates location of mutations found in the rcCap-Env+rsCE clone. Positive and negative changes to reactivity compared to wildtype DENV2 are indicated in blue and purple respectively. Positive changes represent reduced reactivity, indicating increased base pairing by a particular nucleotide. (b) Predicted RNA secondary structures in the 5′UTR and structural protein coding region of wildtype DENV2-16681 and rcCap-Env+rsCap rescue clone, corresponding to nucleotides 1 to 2421 of the DENV2 genome. In silico modelling of DENV2 genomic RNA secondary structures was performed using the Superfold pipeline with RNAstructure v6.3 as the backend, with the results of our SHAPE analysis incorporated as a constraint. RNA structures were then visualized using VARNA 3.93 and a custom script to map SHAPE reactivity data onto the resulting figure. Black arrows indicate nucleotide positions 500, 1000, 1500, and 2000 of the respective DENV2 genomes. (c) Rescue clones of recoded DENV2-rcCap-Env retain attenuated growth kinetics. BHK-21 hamster kidney cells and Huh-7 human hepatocarcinoma cells were infected with wildtype DENV2, wildtype DENV2 with Env-M196V cell line adaptation mutation (WT+rsEnv), recoded DENV2 (rcCap-Env), and rescue mutants of DENV2-rcCap-Env (rcCap-Env+rsCE, +rsCap, and +rsEnv) at an MOI of 0.1. Viral titres were measured using plaque assay. Limit of detection for our plaque assay is 10 PFU/ml. (d) Recoded and rescue clones of DENV2 demonstrate attenuation of neurovirulence in suckling mice. Newborn outbred white ICR mice that were less than 24 hours old were challenged by intracranial inoculation with wildtype DENV2, WT+rsEnv, rcCap-Env, or rc+rsCE clones at a dose of 10^2 PFU per mouse. The mice were kept for four weeks and observed daily for clinical symptoms. Mice that reached a humane endpoint were euthanized. Group sizes: PBS control, n = 10; wildtype DENV2, n = 11; WT+rsEnv, n = 12; rcCap-Env, n = 9; rCap-Env+rsCE clone, n = 11; rcCap-Env+rsCap, n = 10; rcCap-Env+rsEnv, n = 9. **: p-value of <0.01. ***: p-value of <0.001. (e) Recoded DENV2 demonstrates attenuation in its Aedes albopictus mosquito vector. Aedes albopictus mosquitoes were fed an infectious blood meal containing 2.5 x 107 PFU/ml of either wildtype DENV2 or DENV2-rcCap-Env+rsCE. At 11 days post challenge, the mosquitoes were harvested and their infection status and viral loads were determined using plaque assay. Group sizes for both were n = 30. Limit of detection is 100 PFU/ml. ***: p-value of <0.001. https://doi.org/10.1371/journal.ppat.1011753.g004

Recapitulatory rescue clones of recoded DENV2 retain delayed viral growth kinetics Next, we compared the viral replication kinetics of rcCap-Env and the derivative rescue clones in BHK-21 cells and Huh-7 cells. The cells were inoculated at an MOI of 0.1 and plaque assay was used to compare viral titres. We found that the rescue mutations affect virus particle production, but do not rescue the delayed replication kinetics of recoded DENV2. In both BHK-21 and Huh-7 cells, the rsCap-Env+rsCE, +rsCap, and +rsEnv rescue clones all retained the delayed peak titre of parental rsCap-Env (Fig 4C and S2 Table). The addition of the a158u (Cap-N21I) and u173c (Cap-V26A) rescue mutations to the rcCap-Env backbone allows the resulting rcCap-Env+rsCap rescue clone to achieve higher peak titres in both BHK-21 and Huh-7 cells (Fig 4C and S2 Table). The a1522g (Env-M196V) rescue mutation acts as a more specific BHK-21 cell line adaptation; the addition of the a1522g (Env-M196V) mutation to the wildtype DENV2, rcCap-Env, and rcCap-Env+rsCap clones confers higher peak titres on the resulting WT+rsEnv, rcCap-Env+rsEnv, and rcCap-Env+rsCE clones respectively, but only in BHK-21 cells (Fig 4C and S2 Table). The opposite was true in Huh-7 cells, with the Env-M196V conferring lower peak titres (Fig 4C and S2 Table). Curiously enough, the addition of the a1522g (Env-M196V) mutation to the rcCap-Env backbone confers a further delay in replication kinetics on the resulting rcCap-Env+rsEnv rescue clone (Fig 4C and S2 Table). This effect was not observed in the rcCap-Env+rsCE rescue clone with the additional capsid rescue mutations or in the wildtype backbone. Therefore, the effect of the a1522g (Env-M196V) mutation depends on whether it is cloned into a wildtype backbone or into some specific recoded backbone. This makes sense when we consider that the underlying a1522g RNA mutation can have its own effects at the functional RNA level.

Recoded DENV2 demonstrates genetically stable attenuation in suckling mice We investigated whether recoded DENV2 demonstrates genetically stable in vivo attenuation. We investigated in vivo attenuation using a suckling mouse model of neurovirulence as it is a well-established model for studying flavivirus attenuation [27–29]. Newborn outbred white ICR mice that were less than 24 hours old were inoculated intracranially at a dose of 10^5 PFU/ml with wildtype, recoded, or rescue clones. The mice were kept for four weeks post inoculation and observed daily for clinical symptoms and euthanised when they reached a humane endpoint [27,28]. Wildtype DENV2 and WT+rsEnv had similar lethality rates of 91% and 92% (n = 10/11 and 11/12) respectively (Fig 4D). In contrast, recoded rcCap-Env and its derivative rescue clones all demonstrated in vivo attenuation; the lethality rate of the rcCap-Env, +rsCE, +rsCap, and +rsEnv was significantly lower at 11%, 9%, 30%, and 0% respectively (n = 1/9, 1/11, 3/10, and 0/9) (Fig 4D). This demonstrates that rcCap-Env possesses genetically stable attenuation, as serial passaging does not result in mutations that can restore virulence.

Recoded DENV2 demonstrates attenuation in Aedes albopictus mosquitoes Next, we investigated if recoded DENV2 is attenuated in its Aedes mosquito vector [30]. This is a blood meal challenge model, where mosquitoes are fed an infectious blood meal containing DENV2. The mosquitoes are then kept for 11 days after oral infection, after which we use plaque assay to determine their infection status and viral load. We typically challenge our mosquitoes with a blood meal containing DENV2 at a concentration of 5 x 106 PFU/ml, as this concentration is sufficient to infect mosquitoes with wildtype DENV2. However, when we attempted to challenge mosquitoes with recoded virus at the same virus concentration of 5 x 106 PFU/ml, none of the recoded viruses could establish an infection. Therefore, we repeated the mosquito challenge with the virus at a 5-fold higher concentration of 2.5 x 107 PFU/ml. While most of the recoded clones are unable to replicate to such a high titre, the recoded rcCap-Env+rsCE clone replicates well enough in BHK-21 cells to reach this titre. Therefore, we were able to challenge Aedes albopictus mosquitoes with an infectious blood meal containing 2.5 x 107 PFU/ml of either wildtype DENV2 or recoded DENV2-rcCap-Env+rsCE. The mosquitoes were kept for 11 days after oral infection, after which we used plaque assay to determine their infection status and viral load. Compared to wildtype DENV2, recoded DENV2-rcCap-Env+rsCE was attenuated in Aedes albopictus mosquitoes. Wildtype virus was able to establish infection in 29/30 mosquitoes, whereas recoded DENV2-rcCap-Env+rsCE was only able to only infected 24/30 mosquitoes (Fig 4E). Furthermore, mosquitoes that were infected with recoded DENV2-rcCap-Env+rsCE were also found to carry a lower viral load (Fig 4E). This demonstrates that recoded DENV2 has in vivo attenuation in both mice and mosquitoes.

[END]
---
[1] Url: https://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1011753

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.