Synergistic Internal Ribosome Entry Site/MicroRNA-Based Approach for Flavivirus Attenuation and Live Vaccine Development

Abstract

The recent emergence of Zika virus underscores the need for new strategies for a rapid development of safe flavivirus vaccines. Using another flavivirus (Langat virus [LGTV]) that belongs to the group of tick-borne flaviviruses as a model, we describe a dual strategy for virus attenuation which synergistically accesses the specificity of microRNA (miRNA) genome targeting and the effectiveness of internal ribosome entry site (IRES) insertion. To increase the stability and immunogenicity of bicistronic LGTVs, we developed a novel approach in which the capsid (C) protein gene was relocated into the 3' noncoding region (NCR) and expressed under translational control from an IRES. Engineered bicistronic LGTVs carrying multiple target sequences for brain-specific miRNAs were stable in Vero cells and induced adaptive immunity in mice. Importantly, miRNA-targeted bicistronic LGTVs were not pathogenic for either newborn mice after intracranial inoculation or adult immunocompromised mice (SCID or type I interferon receptor knockout) after intraperitoneal injection. Moreover, bicistronic LGTVs were restricted for replication in tick-derived cells, suggesting an interruption of viral transmission in nature by arthropod vectors. This approach is suitable for reliable attenuation of many flaviviruses and may enable development of live attenuated flavivirus vaccines.IMPORTANCE The recent emergence of Zika virus underscores the need for new strategies for a rapid development of safe flavivirus vaccines. Allied separately attenuating approaches based on (i) microRNA genome targeting and (ii) internal ribosome entry site insertion are not sufficient for relievable attenuation of neurotropic flavivirus pathogenesis. Here, we describe a novel dual strategy that combines the specificity of miRNA-based and the effectiveness of IRES-based attenuating approaches, allowing us to overcome these critical limitations. This developed approach provides a robust platform for reliable attenuation of many flaviviruses and may enable development of live flavivirus vaccines.

Keywords: attenuation; flavivirus; internal ribosome entry site; microRNA; vaccine.

FIG 1

Development of bicistronic LGTV. (A) Schematic representation of the viral genomes constructed for the study. “C trn (48AA)” denotes the replication promoter region of the C gene. The ORF-shifting insertion (red asterisk, nt position 151 of the LGTV genome) of a single A nucleotide (Fr Sh +1) and an ORF restoration (−1) are indicated. Yellow and red boxes denote the 2A protease gene of FMDV and mir-124 target (T) sequences, respectively. Gray boxes denote codon-optimized sequences of the C gene. (B and C) Growth kinetics of recovered viruses in Vero cells. Individual samples for each time point were titrated in Vero cells in duplicate. Results are presented as an average ± standard deviation (shown as error bar). The dashed line indicates the limit of virus detection (0.7 log₁₀ PFU/ml).

FIG 2

Synergistic effect between IRES- and miRNA-based strategies attenuates neuropathogenesis and neuroinvasiveness of bicistronic LGTVs in mice. (A) Schematic representation of recombinant LGTV (rLGTV) genomes used in this study. Red and blue boxes denote mir-124(T) and mir-9(T) sequences, respectively. Striped boxes indicate scrambled (synonymous) sequence for mir-124(T) and mir-9(T). (B) Growth kinetics of rLGTVs in the brains of newborn Swiss Webster (SW) mice, after i.c. inoculation with 10² PFU. Viral load in the brain (n = 3) ± standard deviation (SD) was determined by titration in Vero cells. (C) Survival of newborn SW mice (n = 10) inoculated i.c. with 10⁴ PFU/mouse of rLGTVs. (D) Survival of SCID mice (n = 4 or 5) inoculated i.p. with 10⁵ PFU/mouse of rLGTVs. (E and F) Mean titer in the serum (E) and brain (F) of SCID mice after i.p. infection with 10⁵ PFU/mouse. (E) Due to the death or paralysis of the animals, mouse serum samples from IRES-124/9(4m)*- and cap-124/9-infected groups were not collected after days 7 and 23, respectively. (F) Brains from healthy mice infected with IRES-124(4m) and IRES-124/9(4m) were collected on day 52. Brains from paralyzed/dead mice were isolated on day 23 [IRES-124/9(4m)* group] or from day 26 to 31 (cap-124/9 group). The dashed line indicates the limit of virus detection (1.7 log₁₀ PFU/ml of serum [F] or 1.7 log₁₀ PFU/g of brain [F]).

FIG 3

Adult B6 IFNRI^−/− mice immunized with bicistronic LGTVs are protected against lethal challenge with wt LGTV. (A) Survival of B6 IFNRI^−/− mice (n = 5) after immunization. Mice were inoculated i.p. with 10⁵ PFU/mouse of rLGTVs or with diluent alone (mock). (B) Survival of vaccinated mice after i.p. challenge with 10² PFU of wt LGTV. At 31 days postvaccination, animals were challenged with wt LGTV and monitored for neurological signs of disease. (C) Neutralizing antibody titer in the serum of B6 IFNRI^−/− mice at 28 and 56 days postinoculation with bicistronic LGTVs. Neutralizing antibody titer was determined using the 50% plaque reduction neutralization assay (PRNT₅₀) against wt LGTV. (D) Titer of wt LGTV in the serum of mice on days 1 and 4 postinfection. B6 IFNRI^−/− mice immunized with bicistronic LGTVs or mock-infected animals were challenged on day 32 with 10² PFU of wt LGTV. Virus titer in the serum was determined by titration in LLC-MK2 cells. The dashed line indicates the limit of virus detection (1.7 log₁₀ PFU/ml of serum).

FIG 4

Bicistronic LGTVs are growth restricted in tick-derived cells. Tick-derived ISE6 cells (A) or Vero cells (B) were infected with mono- or bicistronic LGTVs at an MOI of 0.1 in duplicate wells of 6-well plates. Virus titer in cell supernatant was determined on day 5 p.i. Mean virus titers ± SD are shown. The dashed line indicates the limit of virus detection (0.7 log₁₀ PFU/ml).

FIG 5

Relative efficiency of IRES-dependent and cap-dependent mechanisms of gene expression in bicistronic LGTV. (A) Schematic organization of subgenomic LGTV replicons expressing nLuc gene via IRES- or cap-dependent mechanisms. The replicons were created by modifying IRES-124(4m) virus; nLuc sequence is indicated as a black box, partial GFP sequence (ΔGFP) is represented by a green box, and 2A protease is shown as a yellow box. The target sequence for mir-124 is shown in red. Frameshift and frame restoration in Ctrn and Ctrn* are represented by +1 and −1, respectively. (B) Relative luminescence kinetics for the LGTV replicons transfected into Vero cells. Luminescence values are calculated as mean ± SD (shown as error bar) from 3 biological replicates of Vero cell extracts transfected with replicon RNAs.