Insertion position as well as the inserted TRS and gene sequences differentially affect the retention of foreign gene expression by simian hemorrhagic fever virus (SHFV)

Abstract

Recombinant SHFV infectious cDNA clones expressing a foreign gene from an additional sg mRNA were constructed. Two 3′ genomic region sites, between ORF4′ and ORF2b and between ORF4 and ORF5, were utilized for insertion of the myxoma M013 gene with a C-terminal V5 tag followed by one of the three inserted transcription regulatory sequences (TRS), TRS2′, TRS4′ or TRS7. M013 insertion at the ORF4′/ORF2b site but not the ORF4/ORF5 site generated progeny virus but only recombinant viruses with an inserted TRS2′ retained the entire M013 gene through passage four. Insertion of an auto-fluorescent protein gene, iLOV, with an inserted TRS2′ at the ORF4′/ORF2b site generated viable progeny virus and iLOV expression was maintained through passage eight. Although regulation of SHFV subgenomic RNA synthesis is complex, the ORF4′/ORF2b site, which is located between the two sets of minor structural proteins, is able to tolerate foreign gene insertion.

INTRODUCTION

Simian hemorrhagic fever virus (SHFV) is a member of the family Arteriviridae. SHFV infection of African monkeys induces an asymptomatic acute or persistent infection whereas infection of macaque monkeys triggers an acute hemorrhagic fever disease characterized by an elevated inflammatory response that is typically fatal (Johnson et al., 2011; Vatter and Brinton, 2014; Vatter et al., 2015). The 5′ two-thirds of an arterivirus genome encodes two polyproteins (pp1a and pp1ab) that are processed by viral proteases to the mature nonstructural proteins, which function in viral replication and transcription (Brinton et al., 2015; Snijder et al., 2013). Among the arteriviruses, only the simian arterivirus genomes encode three instead of two papain-like protease (PLP) 1 domains at the 5′ end. All three of the SHFV PLP1s are active proteases (Vatter et al., 2014b). The 3′ one-third region of all arterivirus genomes encodes the viral structural proteins that are translated from a 3′−5′ co-terminal, nested set of sub-genomic (sg) mRNAs. The transcription of a minus strand template for each sg mRNA from the genome RNA is regulated by a transcription regulatory sequence (TRS) that consists of a six to nine nt core sequence flanked on each side by 15–20 nts (Godeny et al., 1998; Sola et al., 2005). A recent study showed that multiple TRSs contribute to regulating sg mRNA production for the majority of the structural proteins encoded by the SHFV genome (Di et al., 2017). In addition to the three major structural proteins, glycoprotein 5 (GP5), membrane protein (M) and nucleocapsid (N), the genomes of the arteriviruses, equine arteritis virus (EAV), porcine reproductive and respiratory syndrome virus (PRRSV) and lactate dehydrogenase-elevating virus (LDV), encode a single set of minor structural proteins, E, GP2, GP3, and GP4. The SHFV genome encodes an additional set of minor structural proteins, E′, GP2′, GP3′ and GP4′ (Smith et al., 1997; Snijder et al., 2013). Both sets of the SHFV minor structural proteins are functionally important for generating infectious progeny virus (Vatter et al., 2014a). The full-length cDNA of the SHFV strain LVR 42–0/6941 was successfully cloned into a pACYC184 vector and the SHFVic-WT virus generated from RNA copied from this infectious cDNA clone was shown to have growth characteristics similar to those of the parental virus (Vatter et al., 2014a).

The potential use of arteriviruses as viral vectors was previously investigated using EAV and PRRSV infectious cDNA clones. In an initial study, fusion of the influenza virus HA epitope to either the N or the C terminus of the PRRSV nucleocapsid protein decreased virus yields of the recombinant PRRSV produced and the HA epitope was not maintained stably in the recombinant viruses (Bramel-Verheije et al., 2000). In subsequent studies, viable recombinant virus was generated only when the green fluorescent protein (GFP) or the mCherry cDNA fused to the foot-and-mouth disease virus 2A self-cleaving oligopeptide cDNA were inserted into the EAV ORF1a region at the cleavage site between nsp1 and nsp2 (Mondal et al., 2015; van den Born et al., 2007). Deletions and insertions in the nsp2 gene hypervariable region were identified in natural strains of PRRSV and this site was also used for foreign gene insertion (Fang et al., 2006; Han and Yoo, 2014; Xu et al., 2012). An additional strategy tested was to express a foreign gene from an additional sg mRNA with the production of this extra sg mRNA regulated by the genomic TRS upstream of the inserted gene. A duplicated viral TRS was inserted after the foreign gene cDNA to regulate the production of minus strand sg RNA for the downstream viral gene because it no longer had a TRS (de Vries et al., 2001; Pei et al., 2009; Wang et al., 2013).

The SHFV genome is longer (15.7 kb) than those of the non-simian arteriviruses. Its ability to function as a viral expression vector has not previously been tested. Two small protein genes, M013 and iLOV, were chosen for insertion because the amount of additional sequence that could be added without negatively affecting efficient SHFV genome encapsidation was not known. M013 is a small (~14 kDa) myxoma virus protein that can antagonize proinflammatory cytokine production induced through both the inflammasome and NF-κB pathways (Rahman and McFadden, 2011; Rahman et al., 2009). SHVF infection induces proinflammatory cytokine production in macaque macrophages and macaques (Vatter and Brinton, 2014; Vatter et al., 2015). A previous study showed that expression of M013 from an adenovirus-associated viral vector relieved ocular inflammation in a mouse disease model (Ildefonso et al., 2015). Virus genomes expressing an auto-fluorescent tag provide a means for rapid detection of infected cells. A small flavin-based fluorescent protein (~12 kDa) derived from the light, oxygen or voltage-sensing (LOV) domain, iLOV, was previously used as an alternative to GFP for insertion into viral genomes with size constraints to allow rapid detection of infected cells (Buckley et al., 2015; Chapman et al., 2008).

In the present study, the ability of recombinant SHFV genomes to express a foreign gene from an additional sg mRNA was analyzed. Two 3ʹ insertion sites, between ORF4′ and ORF2b or between ORF4 and ORF5, which overlap by 1 nt or are separated by 2 nts, respectively, were tested. The M013 gene with a C-terminal V5 tag was inserted into the SHFV genome at either site followed by one of three inserted downstream TRSs, TRS2′, TRS4′ or TRS7. Only insertion at the ORF4′/ORF2b site generated viable progeny virus. Although all three of the duplicated TRSs initially generated viable recombinant viruses, only genomes with an inserted TRS2′ retained the entire M013 gene through passage. Insertion of the iLOV gene at the ORF4′/ORF2b site followed by an inserted TRS2′ generated progeny virus that retained iLOV expression through passage 7. Passage one (P1) recombinant SHFV expressing either M013 or iLOV and containing an inserted TRS2′ had growth kinetics similar to the wild type P1 SHFV infectious cDNA clone virus. The data indicated that insertion of a foreign gene between the two sets of SHFV minor structural protein genes was tolerated. However, the sequences of the inserted genes and the choice of the inserted TRS differentially affected the retention of the foreign gene.

RESULTS

Generation of recombinant SHFVic viruses expressing the M013 gene from a separate sg mRNA.

All of the 3′ ORFs in the SHFV genome overlap each other except for ORF4 and ORF5, which are separated by two nucleotides. However, ORF4′ and ORF2b overlap each other by only 1 nucleotide. These two ORF junction regions were chosen as insertion sites for the M013 gene to minimize disruption of the genome organization. Two sets of primers (Table 1) were designed to introduce two unique restriction sites (AflII and BglII) between ORF4 and ORF5 or between ORF4′ and ORF2b in the appropriate SHFV cDNA fragment clone. M013 inserted at the ORF4/ORF5 site would be expressed from all the sg mRNAs that are copied from the minus strand sg RNA templates generated from the eleven functional body TRSs for ORF5 (Di et al., 2017). The two body TRSs for ORF2b would be used for the expression of M013 when it is inserted at the ORF4ʹ/ORF2b site. Insertion of an additional body TRS after the foreign gene sequence is required to restore minus strand sg RNA synthesis for the downstream SHFV ORF. A previous study on a recombinant PRRSV showed that if the same TRS was inserted after the foreign gene as was located before it in the genome, copy choice recombination occurring within the homologous duplicated and natural TRS sequences resulted in excision of the inserted gene (Lai, 1992; Wang et al., 2014). Recombination between more distantly located duplicated TRSs was also detected but with a lower efficiency. Therefore, a different, single, body TRS than the one regulating the transcription of minus strand sg RNA for the inserted foreign gene was inserted to regulate the transcription of the downstream viral ORF. Three body TRSs, TRS2′, TRS4′ and TRS7, which were each shown to regulate sg mRNA abundance to different levels in SHFV infected cells (Di et al., 2017), were separately inserted as the duplicated body TRS for the downstream viral ORF.

Table 1.

Primer sequences

Primer name	Sequence (5′to 3′)
BamHI-M013-V5-XbaI-F	ATAAGGATCCACCATGGAGCACCGAG a
BamHI-M013-V5-XbaI-R	GGCGTCTAGATCACGTAGAATCGAGACCGAGGAGAGGGTTAGGGATAGGCT TACCAAATAACAATTTGCGAC b
Fragment IV-AflII+BglII-QC-F	GGCGATCTTTAGAGCATTCTATTCTTAACTTAAGAGATCTATGGGTTCTATACTC ACCC
Fragment IV-AflII+BglII-QC-R	GGGTGAGTATAGAACCCATAGATCTCTTAAGTTAAGAATAGAATGCTCTAAAGA TCGCC
Fragment V-AflII+BglII-QC-F	CCGCTCCTTAACTACCTAACTTAAGAGATCTATGTACTTATGTTTAGGG
Fragment V-AflII+BglII-QC-R	CCCTAAACATAAGTACATAGATCTCTTAAGTTAGGTAGTTAAGGAGCGG
AflII-M013-V5-BglII-F	GGGCTTAAGATGGAGCACCGAGGCGTCATTATAACCGTGTTGG
AflII-M013-V5-TRS2′-BglII-R	GCAGATCTcaggctgtaggtgtgaaacagaagttaaagaccttggctagTCACGTAGAATCGAGA CC c
AflII-M013-V5-TRS4′-BglII-R	TGAGATCTtgcggctgagaccaaagagttcagggtgaaaggttaaggcagctaTCACGTAGAATCG AGACC
AflII-M013-V5-TRS9-BglII-R	TGAGATCTtggtttgccagccacacttcctcaggttaacaagaccccttttggcggcaacccttTCACGTAG AATCGAGACC
AflII-iLOV-BglII-F	GGGCTTAAGATGATCGAGAAGAACTTCGTGATCACCG
AflII-iLOV-TRS2′-BglII-R	TTAGATCTcaggctgtaggtgtgaaacagaagttaaagaccttggctagTCACACGTGGTCGCTG CCG
Fragment IV-RTPCR-F	TATCGTCCTCTGCTGGAACGGG
Fragment IV-RTPCR-R	CAGGTCAAAGAACGAGACGAGG

^aThe introduced restriction sites are underlined

^bThe V5 tag sequence is in bold letters

^cLowercase letters indicate inserted TRS2′, TRS4′ or TRS7 core and flanking sequences

Six infectious cDNA clones, SHFVic-4/5-M013-TRS2′, SHFVic-4/5-M013-TRS4′, SHFVic-4/5-M013TRS7, SHFVic-4′/2b-M013-TRS2′, SHFVic-4′/2b-M013-TRS4′, and SHFVic-4′/2b-M013-TRS7 (Fig. 1), were constructed as described in Materials and Methods and used to transcribe genomic RNA in vitro. MA104 cells were either mock transfected or transfected with SHFVic-WT RNA or with one of the recombinant viral RNAs. The majority of the cells in all of the wells transfected with SHFVic-WT RNA were rounded by 5 days after transfection and this was also the case for the cells in all of the wells transfected with viral RNA transcribed from the three SHFVic-4′/2b-M013-TRS constructs. The culture fluids were collected and designated as P0 virus. Cell lysates were also harvested and analyzed by western blotting using anti-nsp1β antibody (Fig. 2). M013 expression was detected in the lysates from wells transfected with each of the three SHFVic-4′/2b-M013-TRS constructs. The data indicated that the initial cells transfected with either the WT viral RNA or the viral RNA from each of the three SHFVic-4′/2b-M013-TRS recombinants produced infectious progeny virus that spread and amplified the infection throughout the culture and that M013 was expressed in the infected cells.

Figure 1.

Diagram of the SHFV genome ORFs with the insertion sites and foreign gene plus TRS inserts indicated.

Figure 2. Analysis of protein expression in cells transfected with a recombinant SHFV genome RNA.

MA104 cells were either mock transfected (M), transfected with SHFVic-WT RNA (WT) or transfected with one of the three SHFVic-4′/2b-M013-TRS RNAs. At 5 days after transfection, cell lysates were harvested in RIPA buffer and subjected to Western blotting using anti-β actin, anti-nsp1β or anti-V5 antibody.

In contrast, for three SHFVic-4/5-M013-TRS recombinant viruses, rounded cells were not detected by 5 days after transfections and Western blot analyses detected no viral protein in the transfected cell lysates harvested at 5 days after transfection. The same results were obtained with three repeat experiments. For each of these repeats, SHFVic-WT RNA genomes synthesized in vitro at the same time and transfected under the same conditions produced infectious virus. Typically, the number of full length genomes produced in the in vitro transcription reactions is low and the full length RNA transfection efficiency is also low. Virions produced from the few successfully transfected cells infect additional cells allowing the spread and amplification of the infection. If infectious progeny virus is not produced from the initially transfected cells and additional cells are not infected, the amount of viral protein present in the cell extracts would be below the level needed for detection by western blotting.

Comparison of the retention of M013 expression during serial passage of recombinant SHFVic-4′/2b-M013-TRS viruses.

The P0 virus stocks generated from each of the three SHFVic-4′/2b-M013-TRS constructs were passaged onto fresh confluent MA104 monolayers as described in Materials and Methods. At 24 hpi, culture fluid was harvested and designated as P1 virus. Cell lysates were also harvested and M013 protein expression as well as nsp1β levels were assessed by western blotting. The SHFVic-4′/2b-M013-TRS2′ virus maintained M013 expression during passage 1 whereas neither the SHFVic-4′/2b-M013-TRS4′ nor the SHFVic-4′/2b-M013-TRS7 viruses produced detectable M013 expression during this passage (Fig. 3A). The detection of nsp1β indicated that all three recombinant P0 viruses produced infectious virus.

Figure 3. M013 expression and gene insert retention during passage of recombinant SHFVic-4′/2b-M013-TRS viruses.

MA104 cell cultures were separately transfected with in vitro transcribed viral genomic RNA for each SHFVic-4′/2b-M013-TRS recombinant. Culture fluid harvested 5 days after transfection was designated P0 virus. To generate P1 virus, 500 μl of P0 culture fluid was transferred to a fresh culture and culture fluid was harvested at 24 hpi and titered for infectivity by plaque assay. For each subsequent passage (P2 to P7), fresh cultures were infected at a MOI of 1 and culture fluids were harvested at 24 hpi. Culture fluids from mock transfected (M) and SHFVic-WT RNA (WT) transfected MA104 cell cultures were used as negative and positive controls, respectively. (A) At 24 hpi, cell lysates were harvested and analyzed by Western blotting using anti-nsp1β or anti-V5 antibodies. Beta actin was used as a loading control. (B) Passaged culture fluids were collected at 24 hpi, viral genomic RNA was extracted and the M013-TRS insert region was amplified by RT-PCR. The RT-PCR products were separated on a 1% agarose gel and imaged under UV light. L, DNA ladder.

To check whether the inserted M013-V5-TRS sequence was maintained, viral genomic RNA was extracted from the P0 and P1 stocks of each of the SHFVic-4′/2b-M013-TRS recombinant viruses and amplified by RT-PCR using primers flanking the genomic region where M013-V5-TRS had been inserted (Table 1). The PCR product amplified from the SHFVic-WT genomic RNA was used as a negative control. The RT-PCR products were separated on a 1% agarose gel and imaged (Fig. 3B). The full-length M013-V5-TRS insert was detected in all of the P0 SHFVic-4′/2b-M013-TRS viral RNA samples but was only detected in the P1 SHFVic-4′/2b-M013-TRS2′ viral RNA sample. In addition to the full length band, the P0 SHFVic-4′/2b-M013-TRS4′ and SHFVic-4′/2b-M013-TRS7 viral RNA samples also contained a predominant RT-PCR band migrating faster than the full length M013-V5-TRS insert. This was the only band detected in the P1 samples for these two recombinants. In contrast, a faster migrating RT-PCR band was not detected in the P0 SHFVic-4′/2b-M013-TRS2′ viral RNA sample and was a minor component in the P1 SHFVic-4′/2b-M013-TRS2′ viral RNA sample. The results indicated that partial deletion of the inserted M013 sequence from the recombinant genomes consistently correlated with the loss of M013 protein expression.

The P1 SHFVic-4′/2b-M013-TRS2′ virus was serially passaged on MA104 cells at an MOI of 1 through P7. Virus was harvested at 24 hpi at each passage. Western blot analysis of lysates harvested from MA104 cells infected with P2 through P7 SHFVic-4′/2b-M013-TRS2′ virus indicated that high levels of M013 expression were maintained through P3, the levels decreased markedly at P4 and were below the limit of detection at P5 (Fig. 3A). Viral genomic RNA was extracted from the P2 through P7 SHFVic-4′/2b-M013-TRS2′ virus samples and the M013 insertion region was amplified by RT-PCR. The full-length M013-V5-TRS insert was detected through P4 but was not detected in the P5 virus (Fig. 3B). A faster migrating band was detected as a minor component in the P1 virus sample, increased in abundance with subsequent passage and was the only band detected in the P5 virus sample. Although infectious progeny virus expressing M013 protein was produced from cells transfected with each of the three in vitro synthesized SHFVic-4′/2b-M013-TRS viral genomic RNAs, only the SHFVic-4′/2b-M013-TRS2′ genome maintained the full-length M013-V5-TRS insert during passage and continued to express M013 protein through P4.

Analysis of the kinetics of M013 gene expression in cells infected with P1 SHFVic-4′/2b-M013-TRS2′ virus.

SHFV nsp1β is cleaved from the viral nonstructural polyproteins pp1a and pp1ab that are translated from the 5′ end of the genomic RNA starting immediately after genome uncoating. However, viral minus strand sg RNA synthesis does not begin until pp1ab has been translated and the viral replication complex proteins have been processed. The minus strand sgRNAs are then used as templates for copying sg mRNAs from which structural protein are translated. To analyze the kinetics of M013 expression, MA104 cells were infected with P1 SHFVic-4′/2b-M013-TRS2′ virus at an MOI of 1. At different times after infection, cell lysates were harvested and subjected to western blot analysis using anti-V5 and anti-viral nsp1β antibodies. Nsp1β expression was detected by 2 hpi but M013 expression was not detected until 12 hpi (Fig. 4A). The expression of M013 peaked at 48 hpi and did not decrease by 72 hpi. The delayed detection of M013 were consistent with its expression from a sg mRNA.

Figure 4. Analysis of M013 expression kinetics in SHFVic-4′/2b-M013-TRS2′ virus infected cells.

(A) MA104 cells were infected with P1 SHFVic-4′/2b-M013-TRS2′ virus at a MOI of 1. At different times after infection, cell lysates were harvested in RIPA buffer and M013 expression was analyzed by Western blotting with anti-V5 antibody. Viral nsp1β and cellular β actin were also detected using protein-specific antibodies. (B) MA104 cells seeded on cover slips were infected with SHFVic-WT virus or P1 SHFVic-4′/2b-M013-TRS2′ virus at an MOI of 1. At 24 hpi, cells were fixed, permeabilized, blocked and processed for IFA using anti-V5 antibody (green), anti-dsRNA antibody (red) and Hoechst 33342 (blue). The cells were imaged with Zeiss Axio Observer 1 microscope using a 63× oil emersion objective. Scale bar, 25 μm.

MA104 cells were infected with either P1 SHFVic-WT virus or P1 SHFVic-4′/2b-M013-TRS2′ virus at an MOI of 1. At 24 hpi, cells were processed for IFA using anti-dsRNA and anti-V5 antibodies. A dsRNA signal was detected in cells infected with either virus, but the M013-V5 signal was detected only in cells infected with the SHFVic-4′/2b-M013-TRS2′ virus (Fig. 4B). The intracellular localization of the M013-V5 protein in infected cells appeared to be mainly perinuclear, which is similar to what was observed in previous studies that analyzed overexpression of a tagged M013 protein in mammalian cells (Ildefonso et al., 2015; Rahman et al., 2009).

Generation of a SHFVic virus expressing the auto-fluorescent iLOV protein from a sg mRNA.

The cDNA for the auto-fluorescent protein iLOV (~12 kDa) was next inserted into the SHFV genome at the viable ORF4′/2b insertion site followed by an inserted TRS2′ (Buckley et al., 2015; Chapman et al., 2008). The SHFVic-4′/2b-iLOV-TRS2′ construct was generated as described in Materials and Methods and transfected into MA104 cells. At 5 days after transfection, the majority of the cells in the wells were rounded. The culture media was harvested (P0 virus) and used to infect MA104 monolayers to generate a P1 virus stock. MA104 cells seeded in 2-well chamber slides were infected with either P1 SHFVic-WT or P1 SHFVic-4′/2b-iLOV-TRS2′ virus at an MOI of 0.1. At 24, 48 and 72 hpi, Hoechst 33342 was added to the culture fluid to stain nuclei 25 minutes prior to live-cell imaging (Fig. 5A). No auto-fluorescence was detected in cells infected with the SHFVic-WT virus but green fluorescence was detected in cells infected with P1 SHFVic-4′/2b-iLOV-TRS2′ virus at each of the times analyzed with the percentage of cells containing green fluorescence increasing with time after infection. The data indicated that a recombinant SHFVic virus expressing the iLOV gene from a separate sg mRNA was successfully generated.

Figure 5. Expression of iLOV auto-fluorescence and of gene insert retention during serial passage of SHFVic-4′/2b-iLOV-TRS2′ virus.

(A) MA104 cells were seeded in 2-well chamber slides and infected with P1 SHFVic-4′/2b-iLOV-TRS2′or WT SHFVic virus at a MOI of 0.1. At different times after infection, cells were stained with Hoechst 33342 (2 μg/ml) (blue) for 25 min and then live cells were imaged with a Zeiss Axio Observer 1 microscope using a 10× objective. iLOV auto-fluorescence (green), nuclei (blue). Scale bars, 80 μm. (B) MA104 cells were seeded in 4-well chamber slides and infected with P1 through P10 SHFVic-4′/2b-iLOV-TRS2′ virus at a MOI of 6. At 24 hpi, cells were stained with Hoechst 33342 (blue) for 25 min and then live cells were imaged with a Zeiss Axio Observer 1 microscope using a 10× objective. iLOV auto-fluorescence (green), nuclei (blue). Scale bar, 80 μm. (C) SHFVic-4′/2b-iLOV-TRS2′ recombinant virus was serially passaged at a MOI of 1 on MA104 cell monolayers in 6-well plates. Viral genomic RNA was extracted from clarified culture fluid harvested 24 hpi at each passage. The iLOV-TRS2′ insert region was amplified by RT-PCR and the RT-PCR products were separated on a 1% agarose gel and imaged under UV light. L, DNA ladder.

Analysis of foreign gene retention in the SHFVic-4′/2b-iLOV-TRS2′ virus during serial passage.

To analyze retention of the foreign gene insert, SHFVic-4′/2b-iLOV-TRS2′ P1 virus was serially passaged nine times on MA104 cells using an MOI of 1 at each passage. The P1 through P10 virus stocks generated were then used to infect MA104 cells seeded in 4-well chamber slides at an MOI of 6. At 24 hpi, cells were stained with Hoechst 33342 and live cells were imaged (Fig. 5B). Green auto-fluorescence was detected in the majority of the cells after infection with the P1 through the P7 virus stocks. The number of fluorescent cells markedly decreased in cultures infected with the P8 virus stock and only a few fluorescent cells were detected in cultures infected with the P10 virus. Viral genomic RNA was extracted from culture fluid harvested from each passage of the SHFVic-4′/2b-iLOV-TRS2′ virus and the genomic region with the iLOV insert was amplified by RT-PCR. The RT-PCR products were separated on a 1% agarose gel and imaged (Fig. 5C). The full length iLOV-TRS2′ band was detected in the P1 through P8, but not in the P9 or P10 virus RNA samples. A faster migrating RT-PCR product was first detected as a minor component at P6. The relative amount of this PCR product increased with further passage, was the major band in the P8 sample and the only band detected in the P10 sample. The data indicate that even though the iLOV gene was inserted at the same genomic location as M013 and was followed by an inserted TRS2′, loss of the iLOV-TRS2′ insert was observed to occur five passages later than that of the M013-TRS2′ insert, indicating increased retention of the iLOV insert.

Growth kinetics of P1 SHFVic virus expressing either M013 or iLOV.

A previous study showed that insertion of a foreign gene (GFP) into the genome of a highly pathogenic strain of PRRSV decreased the efficiency of viral replication (Wang et al., 2014). To analyze the effect of foreign gene insertion on SHFV replication efficiency, MA104 cells were infected with either P1 SHFVic-WT, P1 SHFVic-4′/2b-M013-TRS2′ or P1 SHFVic-4′/2b-iLOV-TRS2′ at an MOI of 1. Virus infectivity titers in culture fluids collected at different times after infection were determined by plaque assay. The growth kinetics of the two viruses carrying a foreign gene were similar to those of the WT virus but the peak titers for both were slightly lower (Fig. 6). The finding that both SHFVic-4′/2b-M013-TRS2′ and SHFVic-4′/2b-iLOV-TRS2′ had similar growth kinetics to the SHFVic-WT virus suggested that the different stabilities of M013 and iLOV inserts were not due to differential effects of these two genes on virus replication efficiency.

Figure 6. Comparison of the growth kinetics of SHFVic-WT virus and SHFVic viruses expressing a foreign gene.

MA104 cells were infected with P1 SHFVic-WT virus or a P1 SHFVic virus expressing either M013 or iLOV or a P1 virus containing only a restriction site insert at a MOI of 1. At different times after infection, an aliquot of culture fluid was collected and viral infectivity was determined by plaque assay on MA104 cells. Average values are shown for triplicate biological samples that were each titered in duplicate. (A) Growth curves of P1 SHFVic-4′/2b-M013-TRS2′ and P1 SHFVic-WT virus. (B) Growth curves of P1 SHFVic-4′/2b-iLOV-TRS2′ and P1 SHFVic-WT virus. (C) MA104 cells were infected with P1 SFHVic-WT, or P1 SHFVic-4/5-AflII-BglII or P1 SHFVic-4′/2b-AflII-BglII virus at a MOI of 1. At 24 hpi, cell lysates were harvested and was analyzed nsp1β expression by Western blotting.

(D) Growth curves of P1 SHFVic-WT and P1 SHFVic-4′/2b-AflII-BglII virus.

The effect of insertion of only the short restriction site fragment at the ORF 4′/ORF2b site on viral peak titers was also analyzed. The sequences of the two infectious clone fragments with the restriction site inserts were confirmed prior to their use in the assembly of the full length infectious clone. MA104 cells were transfected with RNA transcribed from the SHFVic-4/5-AflII-BglII and SHFVic-4′/2b-AflII-BglII clones and cell extracts harvested at 5 days after transfection or 24 hpi with infection with P0 supernatant were analyzed by western blotting (Fig. 6C). As observed with the SHFVic-4/5-M013-TRS2′, SHFVic4/5-M013-TRS4′, SHFVic-4/5-M013-TRS7 RNAs, nsp1β was not detected after SHFVic-4/5-AflII-BglII RNA transfection. This result indicated that insertion of even a short sequence at the ORF4/ORF5 site in the SHFV genome severely compromised infectious virion production and spread. Nsp1β was detected in cells transfected with SHFVic-WT and SHFVic-4′/2b-AflII-BglII RNA and also in cells infected with both of these P1 viruses. The growth kinetics of these two viruses were similar but the P1 SHFVic-4′/2b-AflII-BglII titers were slightly lower (Fig. 6D).

Analysis of the boundaries of the spontaneous deletions in the M013 and iLOV genes in passaged recombinant viruses.

RT-PCR amplification of the inserted regions in the genomes of passaged SHFVic-4′/2b-M013-TRS and SHFVic-4′/2b-iLOV-TRS viruses detected bands of a smaller size than the full-length insert, suggesting partial deletion of the insert had occurred (Figs. 3B and 6B). To determine the boundaries of the deleted regions in the passaged viruses, the smaller RT-PCR products generated from the P1 SHFVic-4′/2b-M013-TRS4′, P1 SHFVic-4′/2b-M013-TRS7, P5 SHFVic-4′/2b-M013-TRS2′ and P8 SHFVic-4′/2b-iLOV-TRS2′ genome RNAs were separately extracted from the gel and cloned into a pCR4-TOPO vector. Plasmid DNA from selected clones was digested with EcoRI and the insert fragments were separated on a 1% agarose gel. Each of the different smaller sized PCR products detected was sequenced and the sequences were aligned to the respective parental SHFVic-4′/2b-M013-TRS or SHFVic-4′/2b-iLOVTRS2′ sequence. The three different SHFVic-4′/2b-M013-TRS viruses differed in both the length of and extent of variation of the internal deletions detected but all of the deletions occurred within the inserted sequences (Fig. 7A, B and C). In contrast, some of the SHFVic-4′/2b-iLOV-TRS2′ genomes contained a deletion that included not only the entire iLOV-TRS2′ insert, but also the introduced AflII restriction site and up to 5 nts at the 3′ end of the ORF4′ gene, which included the stop codon (Fig. 7D). Interestingly, the fusion of the two viral genome regions flanking the deletion in these genomes created a new stop codon for ORF4′.

Figure 7. The boundaries of spontaneous internal deletions detected in the genomes of passaged recombinant SHFVic viruses expressing a foreign gene.

(A-C) Internal deletions found in the genomes of SHFVic-4′/2b-M013-TRS4′, SHFVic-4′/2b-M013-TRS7 and SHFVic-4′/2b-M013-TRS2′ after one or more passages. (D) Internal deletions found in the genome of SHFVic-4′/2b-iLOV-TRS2′ after passage.

The ORF2b, M013 and iLOV coding sequences are indicated by blue boxes. The V5 tag is indicated by a yellow box, the inserted body TRS sequence is indicated by a green box. Red and orange boxes indicate the introduced AflII and BglII restriction sites, respectively. The numbers above the boxes indicate the nt boundaries of the inserts and internal deletions. Two nt numbers separated by a slash, indicate that clones with a deletion nt boundary at each of these nt positions were detected. Dashed lines indicate deleted regions.

DISCUSSION

Recombinant EAV expressing mCherry and PRRSV expressing GFP or luciferase from an additional sg mRNA were previously generated (Gao et al., 2016; Mondal et al., 2015; Pei et al., 2009; van den Born et al., 2007; Wang et al., 2013). The site of foreign gene insertion utilized in these EAV and PRRSV infectious cDNA clones was between ORF1b and ORF2a, the junction between the nonstructural and structural gene coding regions. These two ORFs do not overlap in the PRRSV genome and minimally overlap in the EAV genome.(de Vries et al., 2001; Han and Yoo, 2014; Pei et al., 2009; van den Born et al., 2007) In the SHFV genome, ORF1b and ORF2a′ overlap by 43 nts and this site was not used due to the long overlap. A foreign gene was not inserted between SHFV ORF7 and the 3′ untranslated region (3′UTR) because it was previously shown that the interaction between a stem-loop structure located in the 3′-UTR and a hairpin structure in ORF7 is required for arterivirus RNA synthesis (Beerens and Snijder, 2007; Verheije et al., 2002). However, two subsequent publications reported successful generation of recombinant PRRSV expressing a foreign gene inserted between ORF7 and the 3′-UTR (Wang et al., 2013; Zhang et al., 2015), suggesting this as a possible future alternative site for foreign gene insertion in the SHFV genome.

Separation of two adjacent 3′ EAV major structural protein ORFs (ORF5 and ORF6) that overlap by 12 nt by inserting a repeat of the overlapping sequence and an AflII restriction site did not negatively affect virus viability or yield (de Vries et al., 2000). Among the SHFV 3′ ORFs, ORF4 and ORF5 are separated by 2 nts and ORF4′ and ORF2b overlap but by only 1 nt. These two sites were separately used for foreign gene insertion. Genomes with M013 inserted at the SHFV ORF4/ORF5 site did not produce infectious recombinant virus. The ORF4/ORF5 site is the junction between the minor and major structural protein coding regions and insertion at this site of even a short additional sequence may have disrupted important RNA structures that function as regulatory elements for the differential control of the expression of the more abundant major and less abundant minor proteins. A recent NGS study of the SHFV transcriptome identified 96 body TRSs with a total of eleven TRSs regulating the expression of ORF5 (Di et al., 2017). Each of these TRSs is functional but they differ in their efficiency of ORF5 sg mRNA generation. Together these eleven ORF5 sg mRNAs produced 9.5% of the total SHFV sg mRNA at 18 hpi in MA104 cells (Di et al., 2017). A foreign gene inserted at this site would be expected to be expressed from the eleven sg mRNAs normally used for ORF5 expression and to be highly expressed. In contrast, the ORF5 minus strand sg RNA abundance generated from this recombinant genome would be regulated by a single inserted TRS. In recombinant EAV genomes, an inserted TRS7 was previously shown to rescue the expression of ORF6 when a GFP gene was inserted at the ORF5/ORF6 junction (de Vries et al., 2001). In recombinant PRRSV genomes, an inserted TRS6 was previously shown to successfully rescue the expression of ORF2a when a luciferase gene was inserted at the ORF1b/ORF2a junction (Gao et al., 2016). Three different SHFV body TRSs, TRS2′, TRS4′ and TRS7, known to regulate sg mRNA transcription at different efficiencies in their normal locations were tested for their ability to rescue ORF5 expression. The SHFV TRS2′ was previously shown to generate 4.382%, TRS4′ to generate 0.573% and TRS7 to generate ~50.554% of the total SHFV sg mRNA produced at 18 hpi in MA104 cells (Di et al., 2017). With insertion of TRS2′ or TRS4′, the amount of ORF5 expression would be expected to be much lower than normal and with insertion of TRS7, much higher than normal. A previous study with an nsp1 mutant EAV showed that when the relative abundance of sg mRNA5 and sg mRNA6 were decreased, production of infectious progeny virus was severely impaired (Nedialkova et al., 2010). The lack of infectious virus production by genomes with an insertion at this site may be due to inappropriate lower or higher levels of ORF5 sg mRNA regulated by the single inserted TRS. Alternatively, the inserted TRSs may not be functional in the structural context of this insertion site. Also, this insertion site was located immediately 5′ of the single core TRS for ORF5a. The ORF5a sg mRNA was the least abundant viral sg mRNA and insertion at the ORF4/ORF5 junction could further reduce the amount of ORF5a sg mRNA produced (Di et al., 2017).

Insertion of M013 with each of the different inserted TRSs at the ORF4′/ORF2b junction site located between the two sets of SHFV minor structural protein ORFs generated viable P0 virus. The expression of a foreign gene at this site would be regulated by the one major TRS (TRS2) and one alternative TRS that normally regulate ORF2b expression. These two ORF2b TRSs together generate 4.137% of the total SHFV sg mRNA produced at 18 hpi in MA104 cells (Di et al., 2017). Among the three inserted TRSs tested, TRS2ʹ is the most similar in the abundance of sg mRNA it generates (4.382% at 18 hpi in MA104 cells) to that of the normal abundance of the two ORF2b sg mRNAs (Di et al., 2017). Consistent with this, only recombinant virus with an inserted TRS2′ generated virus that maintained M013 expression during serial viral passage. It has been suggested that the core and flanking sequences of body TRSs may form an RNA secondary structure that contributes to determining the abundance of the sg mRNA generated (Pasternak et al., 2000; Pasternak et al., 2003; Sola et al., 2005). Insertion of a TRS sequence at a new location in the genome might result in it assuming a different folded RNA structure with an altered minus strand sg RNA regulatory efficiency.

Viable genomes with internal deletions in the M013 gene were detected in the P0 SHFVic-4ʹ/2b-M013-TRS4′ and P0 SHFVic-4ʹ/2b-M013-TRS7 recombinant virus RNA samples. Although the SHFVic-4ʹ/2b-M013-TRS2′ virus maintained M013 expression through passage 3, the M013 expression level decreased at passage 4 and was undetectable by western blotting at passage 5 due to deletions occurring within the M013 gene. The boundaries of the deletions in recombinant genomes for all three of these SHFVic-4ʹ/2b-M013-TRS viruses were similar and did not include deletion of the inserted TRS sequence. However, the different TRS sequences differentially affected the retention of the M013 sequence possibly due to their differential effects on the local RNA structural context. In a previous study with a recombinant EAV that had the GFP cDNA followed by a duplicated TRS7 inserted between ORF5 and ORF6, genomes with deletions in the GFP gene emerged at P2 and became dominant at P3 (de Vries et al., 2001). Another study with a recombinant PRRSV containing GFP cDNA followed by either a duplicated TRS2 or TRS6 inserted between ORF1b and ORF2a also showed that deletion mutants emerged during early passages (Wang et al., 2014). Loss of M013 sequence could be due to recombination occurring at homologous sequences or to polymerase jumping of a secondary RNA structure in the recombinant viral genomes.

Interestingly, the insertion of iLOV at the ORF4′/ORF2b junction site with an inserted TRS2′ produced a recombinant virus that efficiently expressed iLOV through passage 7. The smaller size of the iLOV gene (333 nts) compared to the M013-V5 gene (423 nts) could contribute to the prolonged retention of the iLOV insert in recombinant SHFV. However, it is more likely that the prolonged retention is due to the iLOV sequence either not containing homologous sequences that promote recombination or not causing a major disturbance in the folded structure of the recombinant viral genome. Consistent with these hypotheses, no internal iLOV deletions were detected. The iLOV deletions included some or all of the downstream inserted TRS2′ and in some cases, also the upstream inserted restriction site. In contrast, only internal M013 deletions were detected that sometimes also included V5 sequence.

SHFV infection of African monkeys such as baboons can result in long-term, asymptomatic persistent infections, while SHFV infection of Asian macaques causes a fatal hemorrhagic fever disease (Vatter et al., 2015). SHFV can infect primary baboon and macaque macrophages. The green monkey fetal kidney MA104 cell line and its derivatives are the only known cell lines that can be infected by SHFV. Transfection of cell lines from other host species, such as BHK (hamster) or HEK 293 (human), with in vitro synthesized viral genome RNA results in the production of infectious virions from the transfected cells but they cannot infect additional cells in the culture due to their inability to utilize the receptors on these cells. The transfected cultures must be overlaid with MA104 cells to amplify the virus. The ability of SHFV genome RNA to replicate in cells from other species including humans suggests that it has the potential to cross the species barrier. The use of baboon organs for human transplantation could facilitate the selection of mutants that could infect humans (Levy, 2000; Vine and Kier, 1993). Over expression of Myxoma virus M013 was previously shown to antagonize proinflammatory cytokine production induced through both the inflammasome and the NF-κB pathway (Rahman and McFadden, 2011; Rahman et al., 2009). NF-κB pathway activation was reported to facilitate efficient replication of EAV (Mottahedin et al., 2013). A recombinant SHFV expressing M013 might be attenuated due to less efficient virus replication and reduced induction of proinflammatory cytokine production in macaque macrophages and cause an asymptomatic acute or persistent infection in macaques. However, the delay in M013 expression observed when it was expressed from a sg mRNA suggests that expression of M013 from pp1a would have a greater attenuating effect.

The expression of an auto-fluorescent protein from a virus genome provides a direct means of detecting infected cells during in vitro and in vivo studies for documentation of virus infection efficiency and spread. The iLOV protein is half the size of GFP. The smaller size of the iLOV gene has provided an advantage for the generation of auto-fluorescent recombinant viruses for those that have constraints on viral genome size due to capsid packaging limitations (Seago et al., 2013). Additional advantages of iLOV are that it is oxygen-independent and pH-stable and can recover fluorescence spontaneously after photobleaching. iLOV has outperformed GFP in multiple virus studies (Chapman et al., 2008; Mukherjee et al., 2013; Wang et al., 2017; Wingen et al., 2014). Also, the iLOV protein has been successfully modified either to improve its photostability or to shift its absorption and emission optical bands to longer or shorter wavelengths (Christie et al., 2012; Davari et al., 2016; Khrenova et al., 2017; Khrenova et al., 2015). The iLOV protein has also been modified into a highly sensitive sensor for intracellular arsenic ions (Ravikumar et al., 2017). These modifications make iLOV a versatile tool for enhancing future studies to characterize SHFV infections in vitro and in vivo.

MATERIALS AND METHODS

Cells and virus

The embryonic African green monkey kidney MA104 cell line was provided by Omana Nianan, Centers for Disease Control and Prevention (Atlanta, GA). These cells were cultured at 37⁰C in a 5% CO_2.atmosphere using Minimal Essential Medium supplemented with 1% L-glutamine, 10% fetal bovine serum and 0.1% gentamicin. Wild type SHFV LVR 42–0/6941 infectious cDNA clone virus (SHFVic-WT) was generated as previously described (Vatter et al., 2014a) and the titer of the P1 virus stock was determined to be 1.3 × 10⁷ PFU/ml by plaque assay on MA104 cells.

Plaque assay

MA104 cells were seeded in 6-well plates, grown to confluence, and infected with 100 μl of a serial 10fold dilution of a virus sample for 1 h at 37⁰C. The inoculum was removed and 2 ml of overlay media (2X MEM containing 2.5% FCS mixed 1:1 with 1% SeaKem ME agarose) was added. After incubation at 37⁰C for 72 h, the overlay media was removed and the cells were stained with 0.05% crystal violet in 10% ethanol prior to plaque counting. Three biological repeats of each sample were assayed in duplicate wells for each dilution.

Construction of an M013 mammalian expression plasmid pEF6-M013-V5

The pcDNA-DEST40-M013 construct was provided by Grant McFadden, University of Florida. A set of primers was designed to amplify the M013 gene from this construct and to add a V5 tag (GKPIPNPLLGLDST) fused at the C-terminus (Table 1). The M013 gene was amplified by PCR using AccuPrime Tag DNA polymerase (Thermo Fisher Scientific) and the following parameters: 35 cycles at 95^oC for 15 sec, 60^oC for 30 sec and 68^oC for 35 sec. The amplified PCR product was separated on a 1% agarose gel, digested with restriction enzymes and ligated into the mammalian expression vector pEF6-V5/His TOPO. After transformation into TOP10 competent cells, pEF6-M013-V5 plasmid DNA was extracted from selected bacterial clones and sequenced.

Insertion of the M013 or iLOV gene into the SHFVic genome.

The SHFV LVR 42–0/6941 full length cDNA was divided into five fragments and each fragment is maintained in a pCR-XL-TOPO vector as described previously (Vatter et al., 2014a). One set of primers was designed to separate the 1 nt overlapping ORF4′ from ORF2b and insert AflII and BglII restriction sites between these two ORFs in fragment IV (Table 1). Another set of primers was designed to insert AflII and BglII restriction sites in the region between the non-overlapping ORF4 and ORF5 in fragment V. The restriction sites were inserted into the Fragment IV and V plasmids using the respective primer set (Table 1) and a QuikChange lighting kit following the manufacturer’s protocol (Agilent Genomics). The mutated fragment IV and V plasmid DNAs were transformed into XL10-Gold Ultracompetent cells (Agilent Genomics) and DNA was purified from selected clones and sequenced.

One forward primer and three alternative reverse primers were designed to the ends of the M013-V5 sequence. Each reverse primer introduced the core and flanking sequences of a SHFV body TRS, TRS2′, TRS4′ or TRS7, on the downstream side of the exogenous gene sequence (Table 1). Each primer pair was used to amplify the M013-V5 gene from the pEF6-M013-V5 plasmid with AccuPrime Tag DNA polymerase (Thermo Fisher Scientific). The three PCR products, M013-V5-TRS2′, M013-V5-TRS4′ and M013-V5-TRS7 were separated on a 1% agarose gel, excised and the DNAs were extracted. Each PCR product was digested with AflII and BglII and separately ligated into a mutated fragment IV or fragment V DNA that had been digested with AflII and BglII. Each of the mutant fragment constructs was transformed into TOP10 competent cells and the DNA was purified from selected clones and verified by sequencing. To construct the full-length mutant SHFV genomes, each of the M013-containing fragment IV or fragment V constructs was digested with PflMI and ligated simultaneously with the other four PflMI-digested, wild type SHFV genome fragments. Each ligated full-length mutant SHFV sequence was then cloned into a pACYC184 vector to generate the SHFVic-4/5-M013-TRS2′, SHFVic-4/5-M013-TRS4′, SHFVic-4/5-M013-TRS7, SHFVic-4′/2b-M013-TRS2′, SHFVic-4′/2b-M013-TRS4′, and SHFVic4′/2b-M013-TRS7 constructs. Full-length mutant SHFV cDNA constructs containing only the AflII and BglII restriction sites at the ORF4′/2b or the ORF4/5 junction sites were also constructed and designated as SHFVic-4′/2b-AflII-BglII and SHFVic-4/5-AflII-BglII.

A set of primers that introduced the core and flanking sequences of SHFV body TRS2′ downstream of the exogenous gene coding sequence was designed (Table 1) and used to amplify the iLOV gene from the piLOV-C1 plasmid purchased from Addgene using the same protocol described in the previous section for M013. The PCR product, iLOV-TRS2′, was separated on a 1% agarose gel, excised and the DNA extracted. The extracted iLOV PCR product and the mutated SHFV fragment IV were each digested with AflII and BglII and ligated together to produce the fragment IV-iLOV-TRS2′ construct, which was transformed into TOP10 competent cells and then purified from a selected clone and validated by sequencing. To construct the full-length SHFVic-iLOV, the fragment IV-iLOV-TRS2′ construct was digested with PflMI and ligated simultaneously with the other four PflMI-digested, wild type SHFV fragments and cloned into a pACYC184 vector to generate the SHFVic-4′/2b-iLOV-TRS2′ construct.

Recombinant SHFVic virus cell passage protocol

Each recombinant SHFVic plasmid DNA was linearized at the introduced PvuI restriction site located downstream of the SHFV genome in the pACYC184 vector (Vatter et al., 2014a). Viral genomic RNA was in vitro transcribed from the linearized plasmid DNA and purified using a mMESSAGE mMACHINE SP6 Transcription kit according to the manufacturer’s protocol (Thermo Fisher Scientific). MA104 cells were grown in a 6-well plate to ~40–50% confluence and then each well was transfected with ~100 ng of purified recombinant SHFV genome RNA in DMRIE-C transfection reagent (Thermo Fisher Scientific) following the manufacturer’s protocol. Five days after transfection, culture fluid (2 ml) was collected from the wells, centrifuged at 4⁰C for 5 min at 100 × g, aliquoted and designated as passage 0 (P0) virus. P0 virus (500 μl) was then used to infect an MA104 monolayer in a 10 cm dish. Culture fluid (10 ml) was collected from the dish at 24 hpi, centrifuged, aliquoted and designated as passage 1 (P1) virus. The infectivity titer of the P1 virus stock was determined by plaque assay on MA104 cells. MA104 monolayers in 6-well plates were infected with P1 virus at an MOI of 1. At 24 hpi, culture fluid (2 ml) was collected, centrifuged, aliquoted, titered and designated as passage 2 virus (P2). The same protocol was repeated to generate each additional subsequent passage of recombinant SHFVic virus.

Reverse transcription and amplification of passaged recombinant SHFVic genome RNA

Viral genomic RNA was extracted from 250 μl of culture fluid harvested at each passage using TRI-LS reagent (Molecular Research Center, Inc.) according to the manufacturer’s protocol. Viral RNA (~50–100 ng) was amplified by RT-PCR using a SuperScript III one-step RT-PCR kit (Invitrogen) according to the manufacturer’s protocol with primers designed to the regions flanking the inserted M013 or iLOV gene (Table 1). The RT-PCR products (25 μl) were separated on a 1% agarose gel, stained with ethidium bromide and visualized under long-wavelength UV light. The bands of the appropriate sizes were excised from the gel and the extracted DNAs were either sequenced directly or subjected to TA cloning (see next section).

TA-cloning of RT-PCR products

The extracted RT-PCR product DNAs were ligated into a pCR4-TOPO cloning vector and the vector DNA was then transformed into TOP10 chemically competent cells (Thermo Fisher Scientific) using the manufacturer’s protocol. Eight to ten colonies were picked randomly from the plates and grown overnight in liquid culture (LB media plus 50 μg/ml kanamycin). Plasmid DNA was extracted from the liquid cultures and digested with FastDigest EcoRI (Thermo Fisher Scientific). The digested plasmid DNAs were then separated on a 1% agarose gel, stained with ethidium bromide and imaged with a BioDoc-it imaging system (UVP, LLC). Representative plasmids with different sized inserts were sequenced.

Western blot assay

Transfected or infected MA104 cell lysates were harvested in 1 × RIPA buffer (1X phosphate-buffered saline, 1% Nonidet P-40, 0.5% sodium deoxycholate, and 0.1% SDS) containing 1X Halt protease inhibitor cocktail (Thermo Fisher Scientific). Total protein in cell lysates was measured by a BCA assay (Pierce) according to the manufacturer’s protocol. The same amount of total protein was added to each well of a 15% SDS-PAGE gel and the separated proteins were transferred to a nitrocellulose membrane at 100 V for 1 h. The membrane was incubated in blocking buffer (1X Tris-buffered saline containing 5% non-fat dry milk and 0.1% Tween 20) at 4⁰C overnight and cut into strips. The strips were incubated at 4⁰C overnight with a primary anti-SHFV nsp1β peptide (FAQKVITAFPEGVLC) antibody (Abgent custom rabbit antibody) or an anti-V5 tag antibody (R960–25, Thermo Fisher Scientific) or an anti-β-actin antibody (C-11; Santa Cruz Biotechnology). The membrane strips were then washed three times for 10 min with 1X Tris-buffered saline containing 0.1% Tween 20, followed by incubation with a secondary antibody (horseradish peroxidase-conjugated anti-rabbit or anti-mouse antibody)(Santa Cruz Biotechnology) for 1 h at room temperature. After washing, the membrane strips were developed with a Super-Signal West Pico detection kit (Pierce) following the manufacturer’s protocol.

Immunofluorescence assay (IFA)

Cells grown on coverslips in a 24 well plate were fixed with 4% paraformaldehyde (PFA) for 10 min, followed by permeabilization for 10 min with 0.1% Triton-X and then blocking with 5% horse serum at room temperature for 1 h. After incubation with rabbit anti-V5 antibody (Abcam, 1:500) and mouse anti ds-RNA antibody (English & Scientific Consulting Kft., 1:1000) overnight at 4⁰C, cells were incubated with Alexa Fluor 488-donkey anti-rabbit antibody (Thermo Fisher Scientific, 1:400), Alexa Fluor 594 donkey anti-mouse antibody (Thermo Fisher Scientific, 1:400) and Hoechst 33342 (Thermo Fisher Scientific, 0.05%) for 1 h at room temperature. The cover slips were mounted on a glass slide with Prolong Gold Antifade reagent (Invitrogen) and the cells were imaged with a Zeiss Axio Observer 1 microscope using a 40X or 63X oil emersion objective.

To determine whether M013 overexpression affects SHFV replication, MA104 cells were seeded on cover slips in the wells of a 24-well plate and grown to ~70–80% confluency. Cells were transfected with 0.5 μg of pEF6-M013-V5 plasmid DNA using lipofectamine LTX/PLUS reagent (Thermo Fisher Scientific) following the manufacturer’s protocol for 24 h prior to infection with wild type SHFVic virus (SHFVic-WT, MOI of 1). Cells were processed for IFA at 24 hpi. To detect the expression of foreign genes from recombinant SHFVic genomes, MA104 monolayers were infected with either wild type or a recombinant SHFVic virus at an MOI of 1 and processed for IFA at 24 hpi.

For live cell imaging, MA104 cells were seeded in either 2-well or 4-well chamber slides (Thermo Fisher Scientific), grown to 80% confluency and then infected with either SHFVic-WT virus or a passage stock of the SHFVic-4′/2b-iLOV-TRS2′ virus at an MOI of 0.1 (for 2-well chamber slides) or MOI of 6 (for 4-well chamber slides). At different hpi, Hoechst 33342 (2 μg/ml) was added to the culture media and the cells were incubated at 37⁰C in a CO₂ incubator for 25 min before live imaging using a 10X objective on a Zeiss Axio Observer 1 microscope.

HIGHLIGHTS

• The length of the SHFV genome is at the upper limit of known arterivirus genomes.

• SHFV genomes expressing a foreign gene from an additional subgenomic mRNA were constructed.

• Insertion at the ORF4′/ORF2b but not the ORF4/ORF5 site gave sustained foreign gene expression.

• Inserted TRSs and foreign gene sequences differentially affected foreign gene retention.

• Viable gene insertion sites are not conserved between the genomes of different arteriviruses.