Mutations in the non-structural protein coding region regulate gene expression from replicon RNAs derived from Venezuelan equine encephalitis virus

Self-replicating RNA (repRNA) derived from Venezuelan equine encephalitis (VEE) virus is a promising platform for gene therapy and confers prolonged gene expression due to its self-replicating capability, but repRNA suffers from a suboptimal transgene expression level due to its induction of intracellular innate response which may result in inhibition of translation. To improve transgene expression of repRNA, we introduced point mutations in the non-structural protein 1–4 (nsP1-4) coding region of VEE replicon vectors. As a proof of concept, inflammatory cytokines served as genes of interest and were cloned in their wild type and several mutant replicon vectors, followed by transfection in mammalian cells. Our data show that VEE replicons bearing nsP1GGAC-nsP2T or nsP1GGAC-nsP2AT mutations in the nsP1-4 coding region could significantly reduce the recognition by innate immunity as evidenced by the decreased production of type I interferon, and enhance transgene expression in host cells. Thus, the newly discovered mutant VEE replicon vectors could serve as promising gene expression platforms to advance VEE-derived repRNA-based gene therapies.


Introduction
The strategy of introducing one or more target genes into specific cells or tissues has been widely used in the prevention and treatment of a variety of diseases. The gene(s) of interest (GOI) need to be cloned into the gene expression vector for the intracellular delivery of nucleic acids such as DNA or RNA. Inside the cell, DNA molecules need to escape from endosomes and get access to the nucleus before they can be transcribed into mRNA to exert biological activity, and the exogenous DNA delivered into the nucleus may induce genomic instability and tumorigenesis, affecting the clinical application of DNA-based gene therapies. The messenger RNA (mRNA) exerts its biological activity within the cytoplasm of target cells and does not translocate into the nucleus. mRNA is only transiently active and completely degraded by metabolic pathways without toxic accumulation (Klausner et al. 1993). Therefore, mRNA therapeutics enhance the safety of gene therapy with no risk of genomic integration (Weissman and Karikó 2015;Zhong et al. 2018). The in vitro production of mRNA is relatively simple in a cell-free manner, avoiding complex manufacturing processes associated with other biologics such as recombinant proteins and viral vectors (Weng et al. 2020). Thus, the development of mRNA-based therapeutics has received extensive attention. However, conventional mRNA often transiently expresses the target gene for only 2 ~ 3 days. In order to maintain its therapeutic effect, multiple administrations or large doses are often required, which limits its clinical application.
Self-replicating RNA (repRNA) are genetically engineered replicons derived from self-replicating single-stranded RNA (ssRNA) viruses, in which the structural proteins are removed and replaced with exogenous genes of interest (Lundstrom 2018). Such replicon vectors can be derived from alphaviruses, flaviviruses and rhabdoviruses. RepRNAs are typically delivered within virus-like replicon particles (VRPs) that are produced by transfecting replicon RNA and two helper RNAs into permissive cells (Kamrud et al. 2010). Alphavirus VRPs have shown promising responses in the treatment of infectious disease and cancer (Davis et al. 2002;Hastie and Grdzelishvili 2012;Lundstrom 2020;Lücke et al. 2022). The most common alphaviruses that are engineered as expression vectors include Venezuelan equine encephalitis virus (VEE), sindbis virus (SIN) and semliki forest virus (SFV) (Rayner et al. 2002;Lundstrom 2021). The repRNA consists of untranslated region (UTR), non-structural proteins 1-4 (nsP1-4) and subgenomic promoter (SGP) of alphavirus, but the structural proteins encoding the genes of alphavirus have been replaced with exogenous genes of interest (Rayner et al. 2002). The non-structural proteins of alphavirus are essential for replicon activity, since they form the RNA-dependent RNA polymerase (RdRp) complex which is crucial for repRNA replication and amplification in the cytoplasm. Thus, one repRNA molecule can lead to the production of many copies of transcripts and proteins of interest in situ (Rupp et al. 2015;Carrasco et al. 2018). Due to this self-replicating activity, repRNAs can achieve long-term and high-level expressions of coding genes compared to other RNA molecules.
RepRNAs contain native alphavirus motifs and can mimic viral translation in situ, and several ssRNA and double-stranded RNA (dsRNA) species are formed during the amplification process of repRNAs, which could lead to enhanced immunogenicity through stimulation of pattern recognition receptors (PRRs). These PRRs activate innate immunity and induce the production of type I interferons (IFN-I) which could potentially suppress gene expressions by blocking host protein translational mechanisms (Sellins et al. 2005;Ivashkiv and Donlin 2014;Linares-Fernández et al. 2020). The non-structural proteins (nsP1-4) in repRNA sequence modulate repRNA replication and subgenome expression, and play an important role in the induction of innate immunity. Therefore, mutations in nsP1-4 may affect the genome replication by RdRp, the subgenomic repRNA transcription and its immunogenicity, thereby regulating transgene expression under control of the SGP.
In this study, we constructed TC-83 VEE-derived alphavirus replicon vectors, and introduced point mutations at multiple sites in their nsP1-4 region. As a proof of concept, we investigated effects of nsP1-4's mutations on the transgene expression levels of repRNAs encoding a number of inflammatory cytokines and chemokines (IL-2, IL-12, IL-15, IFNγ, GM-CSF), since these cytokines and chemokines are widely used for disease interventions (Santello et al. 2008;Berraondo et al. 2019). We identified several point mutations in the nsP1-4 region of replicon constructs that could dramatically enhance transgene expression of VEE-derived repRNAs.
Plasmid construction PCR-mediated site-directed mutagenesis method was used to generate different VEE mutants. In brief, primers were designed to provide two different PCR products with the same sequence region. The two fragments with overlapping regions were then fused in subsequent PCR amplification. VEE mutants were generated based on the modification of primers.
Codon optimized GM-CSF, IFN-γ, IL-2, IL-12, or IL-15 sequences were then cloned into VEE wild type (WT) plasmids and mutant plasmids by Gibson assembly, respectively. The primer sequences are indicated in Table 1.

In vitro transcription of repRNAs
The replicon RNAs were in vitro transcribed from the templates of linearized VEE constructs and capped by the addition of a 7-methyl guanosine (m7G) structure at the 5′ end following the manufacturer's instructions. Briefly, VEE constructs were linearized by enzymatic digestion using restriction   TCC TTT TCC  IL-2F  GTC TAG TCC GCC AAG TCT AGC ATA TGG CCA CCA TGG AGA CAG ACA CAC  IL-2R  GCG AGT TCT ATG TAA GCA GCT TGC CAA TTC TTA TTG AGG GCT TGT TGA GATG  IL-12F  GTC TAG TCC GCC AAG TCT AGC ATA TGG CCA CC  IL-12R  AAG CAG CTT GCC AAT TCC CGC GGT TAG GCG GAG CTC AGA TAG C  IL-15F  GTC TAG TCC GCC AAG TCT AGC ATA TGG CCA CCA TGG CCT CGC CGC AGC TCCG  IL-15R TAT GTA AGC AGC TTG CCA ATT CTT AAG AGG TAT TAA TAA ACA TCTG endonuclease MluI, followed by phenol-chloroform extraction and ethanol precipitation. Linearized DNA templates were transcribed into RNA using a MEGAscript T7 transcription kit. RNA transcripts were subjected to purification by lithium chloride precipitation before capping with vaccinia capping enzyme and 2′-O-methyltransferase. A second round of RNA purification was conducted by using a Monarch RNA cleanup kit. The RNA concentration was determined by NanoDrop One instrument (Thermo Fisher Scientific, Rockford, IL), and the RNA quality was assessed by gel electrophoresis.
Cell culture and repRNA transfection 293 T cells and RAW-Lucia ISG cells were cultured in Dulbecco's modified Eagle medium (DMEM) supplemented with 10% fetal bovine serum (FBS), penicillin (100 U/mL) and streptomycin (100 μg/mL) at 37 °C in an atmosphere containing 5% CO 2 . One day before transfection, cells were seeded in a 24-well plate. Cells were washed with phosphate buffered saline (PBS) and transfected 750 ng repRNA with Lipo2000 in Opti-MEM medium according to the manufacturer's instructions. Thirty-six hours after transfection, samples were collected for ELISA assay.

IFN-I activity assay
The IFN-I production after repRNA transfection was measured using RAW-Lucia ISG cells. This IFN reporter cell line is derived from a murine RAW 264.7 macrophage cell line with stable integration of luciferase reporter gene induced by interferon regulatory factor (IRF), and the induction of IFN-I can be determined by measuring the luciferase activity in the cell culture supernatant. RAW-Lucia ISG cells were transfected with WT and mutant replicons encoding GM-CSF, IFN-γ or IL-12. The cell culture supernatants at 12 h and 24 h were mixed with QUANTI-Luc 4 Lucia/gauss reagent according to the manufacturer's instructions, and the luminescence signals were detected immediately by a microplate reader (Tecan infinite 200Pro), as the indication of IFN-I activity.
ELISA assay ELISA assay was performed to detect cytokine secretion in cell culture supernatants and cytokine production in cells. Briefly, 96-well microplates were coated with capture antibody overnight at room temperature (RT). The coated plates were washed three times with wash buffer and blocked with reagent diluent at RT for 1 h. Plates were then washed three times with wash buffer and incubated with serial dilutions of samples and standards at RT for 2 h. After washing each well three times, plates were incubated with detection antibody at RT for 2 h. Plates were washed with wash buffer and then incubated with streptavidin-HRP. After incubation for another 20 min at RT, the reaction was developed by 3, 3′, 5, 5′-tetramethylbenzidine and the optical density of each well was determined at 450/540 nm using Tecan infinite 200Pro instrument. The sample concentration was calculated using the standard curve.

Statistical analysis
Data were expressed as means ± standard deviation of the mean. Student's t test was applied to compare the means of two groups, and ANOVA was used to compare the means of multiple groups. P-values below 0.5 were considered statistically significant.

Results and discussion
Mutation sites in the nsP1-4 region of repRNA It has been reported that mutations in the nsP1-4 region may affect many aspects of replicon biology such as expression levels of GOI and host immune responses against repRNA-encoded immunogens (Frolova et al. 2002;Kääriäinen and Ahola 2002;Bhalla et al. 2016;Li et al. 2019). To enhance the transgene expression mediated by TC-83 VEEderived replicon vectors, we introduced several point mutations at multiple sites in their nsP1-4 region that is the key component responsible for repRNA replication. We selected the mutation sites G3892C, A4714G and A3821T according to previous reports (Li et al. 2019;Bhalla et al. 2016), and other mutation sites were designed based on our experimental experience. We integrated these mutation sites into a number of new nsP1-4 mutants. As shown in Fig. 1, these mutation sites include G357C, G1569A, A1572C and C1575T in nsP1; A3821T, G3892C and T3922C in nsP2; A4714G in nsP3. Furthermore, we constructed the following repRNA constructs, each bearing multiple mutations in a single replicon vector: (1) the combination of G357C, G1569A, A1572C and C1575T in nsP1 (nsP1GGAC); (2) the combination of nsP1G-GAC and T3922C in nsP2 (nsP1GGAC-nsP2T); (3) the combination of nsP1GGAC-nsP2T, G3892C in nsP2 and A4714G in nsP3 (nsP1GGAC-nsP2GT-nsP3A); (4) the combination of G3892C in nsP2 and A4714G in nsP3 (nsP2G-nsP3A).

VEE nsP1GGAC-nsP2T promotes transgene expression
We first subcloned interleukin (IL)-15 or IL-12 sequences into T7-VEE plasmids bearing wild type (WT) nsP1-4 sequence and the above mutant nsP1-4 sequences. The resulting T7-VEE plasmids were linearized by restriction endonuclease digestion and the linearized DNA templates were in vitro transcribed into repRNAs encoding IL-15 or IL-12 (repRNA-IL-15 or repRNA-IL-12). We then evaluated IL-15 or IL-12 expression in cells and cell culture supernatants after transfecting 293 T cells with repRNA-IL-15 or repRNA-IL-12 for 36 h. Compared with WT repRNA-IL-15, repRNA-IL-15 bearing nsP1GGAC-nsP2T mutations significantly up-regulated IL-15 expression in cells and IL-15 secretion in cell culture supernatants after transfection using Lipofectamine 2000 (Fig. 2a). In contrast, repRNAs-IL-15 bearing other mutation(s) failed to up-regulate IL-15 expression in cells and decreased IL-15 secretion in cell culture supernatants (Fig. 2a). In addition, compared with WT repRNA-IL-12, transfection with repRNA-IL-12 bearing nsP1GGAC-nsP2T or nsP1GGAC mutations significantly increased IL-12 expression in cells and IL-12 secretion in cell culture supernatants (Fig. 2b). In contrast, repRNAs-IL-12 bearing other mutation(s) led to undetectable IL-12 expression in cells and little IL-12 secretion in cell culture supernatants (Fig. 2b). These data indicate that VEE-derived repRNAs bearing nsP1GGAC-nsP2T mutation could enhance transgene expression under control of the SGP. On the contrary, the introduction of A4714G mutation in nsP3 sequence could potentially compromise the expression of target genes downstream of SGP.
The A3821 site of nsP2 is important for subgenome expression Furthermore, we mutated A to T at position 3821 of nsP2 in nsP1 GGAC-nsP2T mutant to generate a new nsP1GGAC-nsP2AT mutant (Fig. 1). We investigated if the introduction of A3821T mutation in nsP2 (nsP2 A3821T) could affect the transgene expression mediated by replicons bearing nsP1GGAC-nsP2AT mutation. As shown in Fig. 3a, compared with repRNA-IL-15 bearing nsP1GGAC-nsP2T mutations, repRNA-IL-15 bearing nsP1GGAC-nsP2AT mutations further increased IL-15 expression in 293 T cells and IL-15 secretion in cell culture supernatants at 36 h after transfection. Similar gene expression profile was observed when 293 T cells were transfected with repRNA-IL-12. Compared with repRNA-IL-12 bearing nsP1GGAC-nsP2T mutations, repRNA-IL-12 bearing nsP1GGAC-nsP2AT mutations further increased IL-12 expression in cells and IL-12 secretion in cell culture supernatants at 36 h after transfection (Fig. 3b). These results suggest that In addition, we synthesized WT replicons and the above mutants (replicons bearing nsP1GGAC-nsP2T or nsP1GGAC-nsP2AT mutations) encoding GM-CSF, IFN-γ or IL-2, and evaluated their transfection efficiency. Compared with WT repRNA-GM-CSF, repRNA-GM-CSF bearing nsP1GGAC-nsP2T or nsP1GGAC-nsP2AT mutations enhanced GM-CSF expression in cells and GM-CSF secretion in cell culture supernatants at 36 h after transfection (Fig. 4a). In addition, compared with WT repRNA-IFN-γ, repRNA-IFN-γ bearing nsP1GGAC-nsP2AT mutations enhanced IFN-γ expression in cells and IFN-γ secretion in cell culture supernatants; and repRNA-IFN-γ bearing nsP1GGAC-nsP2T mutations led to enhanced IFN-γ expression in cells and similar IFN-γ secretion in cell culture supernatants (Fig. 4b). RepRNA-IFN-γ bearing nsP1GGAC-nsP2AT mutations showed improved IFN-γ secretion in cell culture supernatants and similar IFN-γ expression in cells compared to repRNA-IFN-γ bearing nsP1GGAC-nsP2T mutations (Fig. 4b). We also analyzed the effects of mutation(s) in the nsP1-4 region on the expression of repRNA encoding IL-2. Compared with WT repRNA-IL-2, repRNAs-IL-2 bearing nsP1GGAC-nsP2T, nsP1GGAC-nsP2AT and nsP2G-nsP3A mutations enhanced IL-2 secretion by ~ 1.5fold, ~ 17.0-fold and ~ 3.0-fold in cell culture supernatants respectively, and increased IL-2 expression by ~ 1.5-fold, ~ 28.0-fold and ~ 3.0-fold in cells respectively (Fig. 4c). Above all, VEE-derived replicons bearing nsP1GGAC-nsP2AT mutations lead to optimal expression of target genes such as IL-2, IL-12, IL-15, GM-CSF and IFN-γ. VEE nsP1GGAC-nsP2T or nsP1GGAC-nsP2AT reduce interferon production RepRNAs in host cells could stimulate innate immunity and activate the IFN-I signaling pathway, thereby preventing the repRNAs from replication and To analyze whether VEE mutation-induced alterations in transgene expression are associated with innate immune signaling in host cells, we incubated WT repRNAs encoding GM-CSF, IFN-γ or IL-12 and their corresponding mutant repRNAs with RAW-Lucia ISG cells that serve as an IFN-I reporter cell line. As shown in Fig. 5, compared with WT repR-NAs encoding GM-CSF (Fig. 5a), IFN-γ (Fig. 5b) or IL-12 (Fig. 5c), their corresponding mutant repRNAs bearing nsP1GGAC-nsP2T or nsP1GGAC-nsP2AT mutations significantly down-regulated IFN-I production at 12 h and 24 h after transfection; while repR-NAs bearing other mutations failed to reduce IFN-I levels. These data suggest that VEE-derived repRNAs bearing nsP1GGAC-nsP2T or nsP1GGAC-nsP2AT mutations decreased the production of IFN-I in host cells and resulted in the evasion of interferon-mediated immune response, leading to enhanced transgene expression of mutant repRNAs in host cells. RdRp regulates the replication and amplification of replicon genomic RNA and subgenomic RNA through a complex and orderly mechanism. The nsP1 protein is responsible for mRNA capping and has both guanylyltransferase (GTase) and guanine-7-methyltransferase (MTase) activities, thereby guiding the capping and methylation of viral genomic and subgenomic RNAs (Cross. 1983;Ahola and Kääriäinen 1995). The nsP2 protein plays a variety of functions during alphavirus infection. The N-terminal of nsP2 protein contains a helicase structure which performs the first viral RNA capping reaction as an RNA triphosphatase and promotes RNA helicase activity as a nucleotide triphosphatase (NTPase) (Vasiljeva et al. 2000;Karpe et al. 2011). The C-terminal of nsP2 protein is identified as a papain-like cysteine protease similar to known cathepsins, which processes viral non-structural polyproteins (Russo et al. 2006). The Data are represented as mean ± SD, *p < 0.05, **p < 0.01, ***p < 0.001 nsP3 protein is necessary for repRNA synthesis and replication, but its specific role is not clear. The nsP3 protein contains the N-terminal macrodomain with nucleic acid binding and phosphatase capabilities, the alphavirus unique domain with a strong homology in alphavirus, and the C-terminal hypervariable region (Malet et al. 2009;Aaskov et al. 2011;Shin et al. 2012). Since the nsP4 protein, the most highly conserved region in alphaviruses, contains the core RdRp domain and motifs, it is directly responsible for RNA synthesis of the viral replicase complex (Rubach et al. 2009;Lello et al. 2021). The nsP1GGAC-nsP2T or nsP1GGAC-nsP2AT mutant VEE replicons could significantly enhance the expression of transgenes encoding inflammatory cytokines and chemokines, which may be related to the capping activity of nsP1 and the activity of helicase, triphosphatase or protease of nsP2. In this study, we discovered that VEE replicons bearing nsP1GGAC-nsP2T or nsP1GGAC-nsP2AT mutations in the nsP1-4 region could significantly enhance the transgene expression of repRNAs. Replicon RNAs can stimulate intracellular innate immune receptors and induce IFN-I production, followed by the inhibition of their replication and translation (Alsharifi et al. 2008;Fros and Pijlman 2016;Cagigi and Loré 2021). The replicon RNA produces single-stranded RNA (ssRNA) and double-stranded RNA (dsRNA) species in the process of replication. They can be recognized by a number of pattern recognition receptors (PRRs) such as toll-like receptors (TLRs). The activation of PRRs triggers the expression and secretion of IFN-I, which leads to the initiation of IFN-I signaling by binding of IFN-I to its receptor, followed by activating the janus kinase signal transducer and transcription activator (JAK-STAT) pathway. STAT1/2 dimer are phosphorylated and then transferred to the nucleus, resulting in the upregulation of IFN-stimulated genes (ISGs). The innate immune signaling activation inhibits repRNA replication and subsequent protein translation (Schilte et al. 2010;Hervas-Stubbs et al. 2011;tenOever. 2016). In this study, we found that VEE replicons bearing nsP1GGAC-nsP2T or nsP1GGAC-nsP2AT mutations in the nsP1-4 region could reduce the IFN-I production in repRNA-transfected cells, thereby affecting the immunogenicity of subgenomic repRNA and resulting in the evasion of interferonmediated immune response, leading to the up-regulation of repRNA transgene expression.