In the genome of SeV-ΔP [1 Wiegand MA, Bossow S, Schlecht S, Neubert WJ. De novo synthesis of N and P proteins as a key step in Sendai virus gene expression J Virol 2007; 81(24): 13835-44.] the entire P ORF was deleted – leaving behind an “empty” transcription unit. In this way, none of the different P gene-encoded non-structural (NS) proteins (Fig. 1A) nor the P protein can be expressed. This design allowed insertion of a selected alternate P ORF encoding a N-terminally truncated P protein lacking the amino acids 2 to 77. These residues are thought to harbor the domain for formation of the P:N° complex which is essential for genome replication in vitro [4 Curran J, Pelet T, Kolakofsky D. An acidic activation-like domain of the Sendai virus P protein is required for RNA synthesis and encapsidation Virology 1994; 202(2): 875-4.] - thus presumably generating a replication-deficient SeV. For monitoring viral gene expression an EGFP ORF was inserted into an additional transcription unit in the genomes of all viral constructs used here (Fig. 1B).

All recombinant SeV could be rescued successfully from cDNA and amplified using a helper cell line (HCL) expressing the Sendai P protein constitutively [12 Willenbrink W, Neubert WJ. Long-term replication of Sendai virus defective interfering particle nucleocapsids in stable helper cell lines J Virol 1994; 68(12): 8413-7.]. RT-PCR and sequence analyses of recombinant SeV after passage 7 in cell culture confirmed both presence and integrity of the originally cloned variants demonstrating genomic stability during virus propagation.

In order to investigate any influence of the P gene deletions on viral replication, amplification of the SeV P mutants on helper cells was compared to SeV-wt. Besides the full length P protein several non-structural (NS) proteins. which are reported to have an impact on viral replication, are expressed from the P gene (Fig. 1A): a nested set of four C proteins (C’, C, Y1, Y2) is translated from a +1 frame relative to the P frame by using alternative start codons [13 Latorre P, Kolakofsky D, Curran J. Sendai virus Y proteins are initiated by a ribosomal shunt Mol Cell Biol 1998; 18(9): 5021-31.]. In addition, V and W proteins result from a co-transcriptional RNA editing process [14 Vidal S, Curran J, Kolakofsky D. A stuttering model for paramyxovirus P mRNA editing EMBO J 1990; 9(6): 2017-2.], leading to a frame-shift after codon 317. Here, amplification of SeV-P and SeV PΔ2-77 is performed in a HCL which expresses only the full length P protein but no NS proteins. Thus, we have three different settings in this comparative study: (i) SeV-ΔP with expression of P (in trans from HCL) but without any NS proteins; (ii) SeV-PΔ2-77 with expression of P and truncated VΔ2-77 and WΔ2-77 but no C proteins; (iii) SeV-wt with expression of P and all NS proteins.

After infection of cells (MOI = 3) virus-containing supernatants of five consecutive days were titrated (Fig. 2): SeV-wt infection produced the highest amount (41 x 10⁶ ciu/ml in total) but all cells were already dead after 3 days. SeV-ΔP and SeV-PΔ2-77 provoked a delayed cytopathic effect (CPE; data not shown) and produced titers which were two times lower than the SeV-wt over the period of 5 days. This clearly shows that amplification of the deletion mutants with P protein provided in trans from the HCL allows high titer virus production. In addition no trans-dominant negative effect develops during co-expression of the virus-encoded PΔ2-77 and the full-length P protein, provided by the HCL, as demonstrated by similar titers of SeV-ΔP and SeV-PΔ2-77. Furthermore, the NS proteins expressed in SeV-wt infections but absent in SeV-ΔP or partially absent and truncated in SeV-PΔ2-77 infections had, at best, a minor effect on viral amplification. However, the V and W proteins were reported to have an influence on viral replication and gene expression [15 Horikami SM, Smallwood S, Moyer SA. The Sendai virus V protein interacts with the NP protein to regulate viral genome RNA replication Virology 1996; 222(2): 383-90., 16 Yu D, Shioda T, Kato A, Hasan MK, Sakai Y, Nagai Y. Sendai virus-based expression of HIV-1 gp120: reinforcement by the V(-) version Genes Cells Mol Cell Mech 1997; 2(7): 457-66.]. Importantly, both proteins harbor the domain from aa 2 to 77 known to be responsible for binding of P to N (P:N°), essential for viral replication. Comparing propagation of SeV-ΔP and SeV-PΔ2-77 did not reveal any significant difference. This conforms to earlier studies when recombinant viruses with an interrupted expression of the V protein were shown to generally grow well in tissue cultures [17 Delenda C, Hausmann S, Garcin D, Kolakofsky D. Normal cellular replication of Sendai virus without the trans-frame, nonstructural V protein Virology 1997; 228(1): 55-62., 18 Kato A, Kiyotani K, Sakai Y, Yoshida T, Nagai Y. The paramyxovirus, Sendai virus, V protein encodes a luxury function required for viral pathogenesis EMBO J 1997; 16(3): 578-87.]. Obviously the truncated V and W proteins do not influence replication – possibly due to the deletion which prevents the proteins from interaction with N°.

In order to validate our experimental set-up which should particularly reflect the situation at the beginning of an infection when only one initial NC is present in the cell, we examined the potential replication deficiency of our SeV P mutants. Checking viral amplification in a cell line not expressing the viral P protein is the optimal set-up. Therefore, Vero cells infected with SeV-ΔP or SeV-PΔ2-77 (MOI = 1) were monitored for a period of 10 days for any viral spread – detectable via GFP expression: in SeV-ΔP infection no EGFP expression could be observed, in SeV-PΔ2-77 infections only single green cells (comparable to Fig. 5). No spread of green fluorescence to neighboring cells could be found. In contrast, SeV-wt infections spread rapidly and destroyed all cells until day 3 post infection (p.i.). As further and more sensitive control, aliquots of the supernatants of infected Vero cells were taken at day 3 and day 10 and transferred onto fresh Vero cells and HCLs. Over a period of 7 days only SeV-wt progeny but no progeny of the SeV-P mutants could be detected.

In addition to this sensitive biological test setting above, replication-deficiency of SeV-ΔP and SeV-PΔ2-77 was verified directly via RT-PCR. Total cellular RNA was isolated from infected Vero cells (MOI = 1) at 24 h and 48 h p.i. A 3 h incubation time served as reference (initial amount of 5 x 10⁵ input genomes), ensuring that virus adsorption and internalization is completed [19 Yonemitsu Y, Kitson C, Ferrari S, et al. Efficient gene transfer to airway epithelium using recombinant Sendai virus Nat Biotechnol 2000; 18(9): 970-3.] and primary transcription is initialized [20 Plumet S, Duprex WP, Gerlier D. Dynamics of viral RNA synthesis during measles virus infection J Virol 2005; 79(11): 6900-8.]. Equal amounts of RNA were subjected to selective reverse transcription of only the negative strand genomic viral RNA by employment of a SeV leader-specific positive-sense primer. Identical aliquots of the RT samples of each virus were used for a semi-quantitative endpoint PCR (Fig. 3A) and a quantitative PCR analysis (Fig. 3B). Both attempts reveal a constant (SeV-PΔ2-77) or marginally decreasing (SeV-ΔP) template number during the course of infection, in contrast to SeV-wt infections with a 366-fold (24 h) and 704-fold increase (48 h), respectively. In summary, our data demonstrate the incapability of both mutants SeV-ΔP and SeV-PΔ2-77 to replicate genomes in cells not providing the viral P protein. Thus, deletion the N-terminal 76 aa suffices to create a replication-deficient phenotype.

After validation of our test system which guarantees a constant minimal template number during infection experiments, we could next evaluate the influence of P gene products solely on viral transcription. First we looked at the immediate products of transcription, i.e. mRNA. Total RNA from infected Vero cells was isolated at 3 h (reference), 24 h and 48 h p.i. with SeV ΔP, PΔ2-77 and wt. After use of an oligo-dT primer for RT in order to transcribe only mRNA, a quantitative PCR analysis was performed amplifying a part of the N ORF (Fig. 4A). SeV-ΔP infected cells show no detectable increase from 3 h to 48 h p.i. Interestingly, SeV PΔ2-77 very clearly is able to transcribe in an enhanced mode – without being able to replicate viral genomes - showing a 5.7-fold (24 h) to 8.2-fold (48 h) increase in mRNA. Compared to SeV-wt (increase 596-fold at 24 h and 806-fold at 48 h p.i.), the rise in mRNA levels shows a parallel progression but in total at a 100-fold reduced amount. The 8.2-fold increase in 1° transcription during SeV-PΔ2-77 infections resulted statistically from only one template NC per infected cell. SeV-wt, as expected, transcribes at high rates, reaching the 800-fold amount of mRNA 48 h p.i. This high number is probably mainly based on 2° transcription after replication with 704-fold more genomic templates compared to 3 h (Fig. 3B).

Based on these surprising data that SeV-PΔ2-77 is able to efficiently support viral transcription at constant template number – even reaching SeV-wt performance when the effects from 2° transcription are deducted – we wanted to analyze next which consequences at the protein level could be detected. Here, we even have a special situation, because SeV-PΔ2-77 is unable to produce progeny virus in cells not trans-complementing P. This means synthesized proteins cannot be drained off the cell in form of released viral particles. Therefore, we compared protein expression levels from infected HCL and Vero cells. EGFP expression was monitored by fluorescence microscopy for a rough estimation (Fig. 4B) and by FACS analysis to determine the number and the intensity of fluorescent cells (Fig. 4C). In HCL, providing the full-length P protein in trans, both mutants induced a strong EGFP fluorescence at 48 h p.i. comparable to infections with SeV-wt (Fig. 4B, upper panel). In Vero cells, however, SeV-ΔP could not produce any visible fluorescence, while SeV-PΔ2-77 infected cells still showed a strong but slightly reduced expression compared to SeV-wt 48 h p.i. (Fig. 4B, lower panel).

In the FACS analyses of the same setting quantification of gene expression revealed some interesting numbers. In the HCL infected with SeV-ΔP or SeV-PΔ2-77 a continuous distribution with fluorescence intensities comparable to wt infections could be observed (Fig. 4C, upper panel). In Vero cells, however, EGFP expression in SeV-ΔP infections was only weak, whereas SeV-PΔ2-77 induced a significant signal of medium intensity (peak at 300 FL-1H units) (Fig. 4C, lower panel). Strikingly, the EGFP level was only 7-10 times reduced compared to SeV-wt infections (peak at 2000 FL-1H units). This is a surprising result because it shows that gene expression from only one template on average can lead to a significant amount of protein, especially when taking the drastically differing levels of expressed mRNA into consideration (Fig. 4A).

Taking all results of the different experiments together, we could clearly show that mutant SeV-PΔ2-77, but not SeV-ΔP, is able to actively transcribe and to express proteins to significant levels. In conclusion, domain aa 2 to 77 of the Sendai P protein seems to be dispensable for transition from early 1° transcription to late 1° transcription at the beginning of an infection.

Besides mRNA and protein expression levels during 1° transcription we wanted to evaluate NC stability as another critical parameter at the beginning of an infection. The longer stability and functionality of the infiltrated NC lasts, the higher the likelihood of viral survival and the establishment of a productive infection. We applied our replication-deficient SeV-PΔ2-77 and simulated infections of Vero cells (Fig. 5) with statistically calculated only one NC per cell. Only from day 6 p.i. on a fading EGFP signal could be observed. To exclude that this fading fluorescence was the result of a reduced viability and arrested cell division, we split infected Vero cells twice (at day 7 and day 16 p.i.). However, the same observation was made: EGFP expression decreased from day 6 p.i. on. Interestingly, this decline progressed slowly. As long as until day 30 p.i. fluorescent cells were still detectable. SeV-ΔP infected cells exhibited also a constantly moderate fluorescence for a similar period of time (data not shown) when infected with a MOI of 15. A presumed CPE resulting from SeV-PΔ2-77 infection was not visible - indicating that the decline in EGFP expression was rather due to inactivation of the NC inside the cell than the consequence of cellular break-down due to the viral infection. In conclusion, with respect to the half-life of EGFP of 26 h [21 Corish P, Tyler-Smith C. Attenuation of green fluorescent protein half-life in mammalian cells Protein Eng 1999; 12(12): 1035-40.], we discovered that NCs remain functional during an infection for at least 5 days but maybe even much longer (> 15 days). This clearly extends existing knowledge that describes NC activity of SeV to last only 24 h. The documented period of at least 5 days seems to be a comfortable time frame for the virus to establish a productive infection.

Vero cells were purchased from ATCC (CCL-81). The helper cell line (HCL; derived from HEK 293 cells) expresses constitutively SeV P protein [12 Willenbrink W, Neubert WJ. Long-term replication of Sendai virus defective interfering particle nucleocapsids in stable helper cell lines J Virol 1994; 68(12): 8413-7.]. BSR-T7 - BHK cells expressing T7 RNA polymerase [25 Buchholz UJ, Finke S, Conzelmann KK. Generation of bovine respiratory syncytial virus (BRSV) from cDNA: BRSV NS2 is not essential for virus replication in tissue culture, and the human RSV leader region acts as a functional BRSV genome promoter J Virol 1999; 73(1): 251-9.] - were a kind gift from K.-K. Conzelmann (Munich). Cell lines were maintained in DMEM (Invitrogen) supplemented with 10% heat-inactivated FCS (Invitrogen). Recombinant SeV derived from strain Fushimi (murine PIV1, ATCC VR-105). All incubation steps of infected cells took place at 33 °C.

Numbering of positions given below refers to GenBank submission M30202, sequences are denoted in the positive antigenomic direction. The plasmid pRS3Gg, containing the SeV wt genome [26 Leyrer S, Neubert WJ, Sedlmeier R. Rapid and efficient recovery of Sendai virus from cDNA: factors influencing recombinant virus rescue J Virol Methods 1998; 75(1): 47-58.], was used as a template to create fragments harboring an alternate P gene product by PCR (“Expand High Fidelity PCR system”, Roche). The fragment encoding the N-terminally truncated protein PΔ2-77 was amplified with the primer pair PΔ2-77 Xho for (5’-CCCCCTTTTTCTCGAGATGTCGACCCAAGATAATCGATCAGGTGAGG-3’) and PΔ2-77 Xho rev (5’-TTTTT CCCCCCTCGAGTTACTAGTTGGTCAGTGACTCTATGTCCTC-3’), fusing the P start codon directly to the codon for amino acid 78 Ser and introducing XhoI restriction sites (underlined) flanking the ORF. Construction of the genomic plasmids containing the SeV wt genome or the mutant SeV ΔP lacking the entire P ORF (corresponding to original position 1844-3550) was described previously [1 Wiegand MA, Bossow S, Schlecht S, Neubert WJ. De novo synthesis of N and P proteins as a key step in Sendai virus gene expression J Virol 2007; 81(24): 13835-44.]. The plasmid of SeV PΔ2-77 was obtained by inserting the PCR product into the “empty” transcription cassette of SeV ΔP via XhoI. All genomic constructs contain an additional transcription unit with the EGFP ORF between the leader and the N gene [27 Bernloehr C, Bossow S, Ungerechts G, et al. Efficient propagation of single gene deleted recombinant Sendai virus vectors Virus Res 2004; 99(2): 193-7.] as well as the oligobasic F cleavage site F_mut and were confirmed by sequencing.

Recombinant viruses were recovered from transfected BSR-T7 cells as described in Wiegand et al. [1 Wiegand MA, Bossow S, Schlecht S, Neubert WJ. De novo synthesis of N and P proteins as a key step in Sendai virus gene expression J Virol 2007; 81(24): 13835-44.] with a modification of the rescue procedure for the P mutants. After three days, the cells were transfected a second time with 1.0 µg pTM-P/C^- and incubated for 2-3 more days. The recovered viruses were propagated on the P-expressing helper cell line (HCL) several times using DMEM + 5% FCS. Virus stocks were prepared from the 5^th passage and used for all following experiments.

Confluent HCL monolayers in 12-well plates (5 × 10⁵ per well) were inoculated with virus (MOI = 3). After adsorption (1 h at 33 °C), the inoculum was removed and cells were washed twice before adding 1 ml medium supplemented with 5% FCS. Supernatants were isolated at the indicated time points and replaced by 1 ml fresh medium. Released progeny viruses were titrated in duplicate on HCL in a dilution series, counting EGFP-positive cells. Cells were monitored by live cell fluorescence microscopy (488 nm excitation) using a Leica DM IL microscope (Leica Microsystems), a F-View II CCD camera and AnalySIS software (Soft Imaging Systems).

Total cellular RNA was isolated (RNeasy; Qiagen) from infected cells at the indicated time points. Applying 1 µg RNA each, viral (-)genomes were selectively reverse transcribed (Superscript III; Invitrogen) using the primer “ld-25” 5’-TGGAATATATAATGAAGTCAGA-3’ (nt 25 to 46). An aliquot of the resulting viral genome-specific cDNA (1/10^th) was employed in a standard PCR with the oligonucleotides “N-1280” 5’-GGATACAGCCAAGGAGA GGC-3’ (nt 1280 to 1299) and “M-4749+” 5’-GGGGCATT GTCGCAGGTC-3’ (nt 4749 to 4732). To amplify viral mRNAs, 1 µg of each total RNA was reverse transcribed with an oligo(dT)₂₀ primer. The conditions for the RT and PCR were as in Wiegand et al. [1 Wiegand MA, Bossow S, Schlecht S, Neubert WJ. De novo synthesis of N and P proteins as a key step in Sendai virus gene expression J Virol 2007; 81(24): 13835-44.].

For real-time quantitative PCR, different sets of primers were designed using the ABI Software Primer Express (Applied Biosystems). The primer pair “SendaiNF” 5’-GCCAG AGGAGCACAGTCTCAGT-3’/“SendaiNR” 5’-GTCCAAT GAGTGAGCTAGGAAGG-3’ was applied for amplification of a 101 nt long region within the N gene. Aliquots (1/20^th) of viral (-)genome-specific and mRNA-specific RT samples were amplified using “SYBR Green Taqman Universal Mastermix” (25 µl volume) with activation (50 °C, 2 min and 95 °C, 10 min), followed by 40 cycles of 95 °C, 15 sec and 60 °C, 1 min. The reactions were performed and detected in MicroAmp Optical 96-well Reaction Plates on an ABI PRISM 7900 Sequence Detection System (9600 emulation mode). At completion, the threshold cycles were determined and the data were analyzed via the Sequence Detection Software 2.1 (Applied Biosystems).

Infected cells were trypsinized, washed and resuspended in 200 µl PBS (approximately 5 × 10⁵ cells). The number of EGFP-expressing cells was immediately analyzed by flow cytometry (FACScalibur; Becton Dickinson) using CellQuest software.