key: cord-0300768-kmp6l4f7
authors: Iida, Kei; Ajiro, Masahiko; Muramoto, Yukiko; Takenaga, Toru; Denawa, Masatsugu; Kurosawa, Ryo; Noda, Takeshi; Hagiwara, Masatoshi
title: Switching of OAS1 splicing isoforms mitigates SARS-CoV-2 infection
date: 2021-08-23
journal: bioRxiv
DOI: 10.1101/2021.08.23.457314
sha: 6f701e218d5c4357e7ceeb148a4a54794cb003e7
doc_id: 300768
cord_uid: kmp6l4f7

Background The rapidly accumulating disease susceptibility information collected from coronavirus disease (COVID-19) patient genomes must be urgently utilized to develop therapeutic interventions for SARS-CoV-2 infection. Chromosome 12q24.13, which encodes the 2’-5’-oligoadenylate synthetase (OAS) family of proteins that sense viral genomic RNAs and trigger an antiviral response, is identified as one of the genomic regions that contains SNPs associated with COVID-19 severity. A high-risk SNP identified at the splice acceptor site of OAS1 exon 6 is known to change the proportions of the various splicing isoforms and the activity of the enzyme. Methods We employed in-silico motif search and RNA pull-down assay to define a factor responsible for the OAS1 splicing. Next, we rationally selected a candidate for slicing modulator to modulate this splicing. Results We found that inhibition of CDC-like kinase with a small chemical compound induces switching of OAS1 splice isoforms in human lung cells. In this condition, increased resistance to SARS-CoV-2 infection, enhanced RNA degradation, and transcriptional activation of interferon β1, were also observed. Conclusions The results indicate the possibility of using chemical splicing modifiers aided by genome-based precision medicine to boost the innate immune response against SARS-CoV-2 infection.

incubator at 37 °C with 5 % CO2, and mycoplasma was confirmed negative in routine polymerase 109 chain reaction tests. VeroE6/TMPRSS2 cells were obtained from JCRB Cell Bank, and cultured in 110 Dulbecco's modified Eagle's medium (DMEM) (Sigma-Aldrich) supplemented with 10 % fetal 111 bovine serum, 100 U/mL penicillin, and 100 µg/mL streptomycin. The cells were maintained in an 112 incubator at 37 °C with 5 % CO2. 113 114 RNA pull-down assay and western blot 115

Calu-3 whole cell lysates were prepared using a lysis buffer containing 10 mM Tris-HCl (pH 7.4), 150 116 mM NaCl, 1 mM ethylenediaminetetraacetic acid, 1% Triton X-100, 0.1% sodium dodecyl sulphate, 117 0.25% sodium deoxycholate, and 10% glycerol with protease inhibitors (Nacalai Tesque) and 118 phosphatase inhibitors (Sigma-Aldrich, Munich, Germany), followed by sonication, treatment with 119 DNase I (Promega, Madison, WI, USA) at 37 °C for 5 min, and centrifugation (24,000 x g at 4 °C for 120 15 min). The supernatant was used as the soluble fraction for the RNA-pull down assay. 121

For the RNA pull-down assay, 5′-biotin, 3′-dTdT-attached RNA, designed for the sequence adjacent 122 to the OAS1 exon 5 splice donor (5′-CUGCUGGUGAGACCUCCUGCUUCC-3′ (oAM685), was 123 incubated with NeutrAvidin beads (Thermo Fisher Scientific, Waltham, MA, USA) for 2 h at 4 °C (no 124 bait RNA was used for the negative control), followed by wash with 1X TBS thrice. RNA-bound 125 were identified using the method described above. Differential alternative splicing (DAS) events were 146 analyzed with the method previously described (Sakuma et al., 2015) using the rMATS program 147 (http://rnaseq-mats.sourceforge.net/rmats4.1.1/). DAS was defined by the following criteria: FDR < 148 0.01, read counts ≥ 15, and delta Percent Spliced-In (PSI) ≥ 0.05. We compared the DAS events with 149 the gene annotation information, then classified the event types into events on the exons constituting 150 the productive mRNAs, events on additional exons/regions of the productive mRNAs, and others. 151

For characterizing gene set and transcriptome profiles, we used the Metascape website 152 (https://metascape.org/) and Gene Set Enrichment Analysis (GSEA, https://www.gsea-153 msigdb.org/gsea/). The same samples were also used for RT-PCR. 154

Calu-3 cells were pre-treated with 10 μM CaNDY or 0.1% DMSO for 24 h, infected with SARS-CoV- 

Total RNA extracted from cultured cells was used for reverse transcription using the PrimeScript 179 RTase (Takara Bio, Shiga, Japan) with random hexamers, and the products were then amplified with 180 ExTaq DNA polymerase (Takara Bio) with target-specific primer sets. Primers used in RT-PCR are 181 listed in Table S1 . Detection of RT-PCR products was conducted using the ChemiDoc MP Imaging 182 System (Bio-Rad), with subsequent analysis by Image Lab software (Bio-Rad). RT-PCR was 183 conducted in three technical replicates for a total RNA sample. infectious person as 1. Beta, a parameter for infection rate per day, per person, was set as 0.5 or 0.28 190 (i.e., 0.5 x 0.56), and Gamma, a parameter for the removal or the recovery rate per day, per person, 191 was set as 0.1, according to a report on the spread of SARS-CoV-2 in April 2020, in the USA (Adam, 192 2020 

The rs10774671 A/G SNP is known to control the production of OAS1 splice variants. The G allele of 203 rs10774671 (G-allele) creates the AG-dinucleotide that is essential for the recognition of the exon 6 204 splice acceptor site for p46 variant production. The A allele of rs10774671 (A-allele) leads to the 205 alternative splicing of OAS1 pre-mRNA to produce p42, p48, p44a, and p44b variants (Noguchi et al., 206 2013) (Fig. 2a) . OAS1 protein senses the double-stranded RNA structure, including RNA duplication 207 intermediates of SARS-CoV-2 (Sadler and Williams, 2008; Schlee and Hartmann, 2016), to 208 synthesize 2′-5′-oligoadenylates, which in turn trigger the activation of latent ribonuclease L (RNase 209 L) for viral RNA degradation (Sadler and Williams, 2008; Schlee and Hartmann, 2016) . Previous 210 studies revealed that the catalytic activity of OAS1 varies with its splice variants; p46, the major 211 isoform of OAS1 produced in the presence of the G-allele, presents optimal activity, while p42, the 212 major isoform produced in the presence of the A-allele, shows poor activity (Carey et al., 2019; Di et 213 al., 2020) . 214

In A-allele individuals, disruption of the exon 6 splice acceptor site induces OAS1 alternative splicing, 215 yielding p44a, p44b, and p48 through activation of acceptor sites downstream of the exon 5, and p42 216 as a consequence of the skipped recognition of exon 5 donor site (Fig. 2a) . Assuming that the splice-217 shift from the inactive p42 to other variants may improve the immune-response that is weakened due 218 to the presence of OAS1 A-allele, we first looked into the mechanism by which p42 variant is 219 dominantly produced. In searching for splice motifs surrounding the OAS1 exon 5 donor sites with the 220 ESEfinder tool (Cartegni et al., 2003) , we noticed a binding motif for serine/arginine-rich splicing 221 factor 6 (SRSF6) (5′-UGCUUC-3′) immediately downstream to the exon 5 donor site (Fig. 2b) . We 222 then speculated that SRSF6 interaction with this site could prevent U1snRNP from binding to the exon 223 expressed from OAS1 G-allele, while the p42 variant was expressed from the A-allele (Fig. 2d) . 239

Intriguingly, CLK inhibition induced splice-switching for OAS1 with the A-allele, by suppressing p42 240 splicing while promoting p44a and p44b variant production. This observation is consistent with the 241 model suggested by the RNA pull-down assay (Fig. 2c) , where the exon 5 donor site is made available 242 for U1snRNP binding upon CLK inhibition (Fig. 2b) . However, the p48 variant expression was under 243 the detection limit in both A-and G-alleles (Fig. 2d) . The splicing shift of OAS1 with A-allele upon 244 CLK inhibition was also evident in RNA-Seq analysis of Calu-3 cells, in which the p42 variant was 245 suppressed by 51%, and p44a and p44b were increased by 130% and 100%, respectively; the p44b 246 variant was found to be a minor form of OAS1, accounting for 1-4% of OAS1 transcripts (Fig. 2e, 2f) . 247

Additionally, through the RNA-Seq analysis of Calu-3 cells with or without CaNDY treatment, we 248 identified differential expression of 98 genes, and splice alterations yielding protein-coding variants 249 for 63 genes; however, none of these events were associated with viral infection except for the OAS1 250 splice alterations (Tables S2, S3 , and Fig. S2 ). 251 foreign RNA molecules, but also the endogenous ribosomal RNA within the host cells (Lin et al., 265 2009 ). Thus, degradation of rRNA was investigated as a measure of RNase L activity. We confirmed 266 the appearance of a band that is expected to be a degradation product of rRNA in cells infected with 267 SARS-CoV-2 (Fig. 3c) . We also confirmed that the concentration of decomposition products, sized 268 2,000-3,000 nucleotides, increased slightly (fold change = 1.35), with a decrease in the molecular 269 weight and concentration of the 28S rRNA peak (fold change = 0.56), upon CaNDY treatment (Fig. These results indicate that CaNDY treatment enhances the infection-dependent RNase L pathway, and 276

The G-allele of SNP rs10774671 produces the p46 splice variant, which possesses 56 amino acids at 290 the C-terminus, compared to the 18 completely different amino acids present in the p42 splice variant 291 produced by the A-allele. This changes the localization of the OAS1 protein and its interacting 292 partners (Kjaer et al., 2014) , which may be linked to the differential 2'-5'-oligoadenylate synthesis 293 activities, RNase L activation, and interferon pathway activation. Several clinical trials that investigate 294 the efficacy of an intervention of the interferon β pathway to affect the progression of COVID-19 have 295 been conducted successfully (Hung et al., 2020; Monk et al., 2021; Shalhoub, 2020) . On the other 296 hand, inhibition of the interferon pathway is considered important to prevent cytokine storms in 297 patients with severe COVID-19 (Nile et al., 2020) . Since the OAS1 IVS5-1 SNP can alter the response 298 intensity of the interferon β pathway, stratification of COVID-19 patients depending on this SNP may 299 assist in choosing the best course of interferon treatment. 300

It is reported that protective SNPs at OAS1/OAS3/OAS2 loci originated in the Neanderthals (Zeberg 301 and Pääbo, 2021; Zhou et al., 2021) . Interestingly, the OAS1 IVS5-1 A-allele, which was confirmed to 302 be associated with the aggravation of COVID-19 in this study, was not found in the ancient human 303 genome, the sequenced Neanderthal genome, or the Denisovan genome (Mendez et al., 2013) . This 304 suggests that the G to A mutation at the OAS1 IVS5-1 position occurred at a relatively modern age in 305 human history. The OAS1 IVS5-1 A-allele has expanded to become a major allele in the population, 306 especially in the Asian region (Fig. S4 ) (Marcus and Novembre, 2017) . This suggests the possibility 307 that this variant has some positive effects on human survival, trading off the effect of weakening 308 resistance against viruses. 309

The results obtained in this study indicate that the OAS1 IVS5-1 SNP is involved in the exacerbation 310 of COVID-19 by eliciting changes in the OAS1 splicing isoform balance. Individuals with the A-allele 311 may have a higher risk for SARS-CoV-2 infection and its aggravation than those with the G-allele. 312

However, we successfully demonstrate a way to overcome these genetic risks by modulating the 313 splicing phenomenon. Splicing modulation decreased the virus infection rate 0.56 times in cell-based 314 assays (Fig. 3b) . This theoretically implies that the peak number of SARS-CoV-2-infected individuals 315 will be reduced by 42%, if we can alter the OAS1 splicing patterns for all the individuals carrying the 316 OAS1 IVS5-1 A-allele and boost the immunity of the entire population (simulation using the 317 Susceptible, Infected and Removed (SIR) model, see Methods for details). We hope that this approach 318 based on genome-based precision medicine with a splicing modulator will contribute to the 319 management of the COVID-19 pandemic. 320 321 doi:10.1038/d41586-020-02009-w nucleic acid polymorphisms of OAS1 and MxA genes: a case-control study. BMC Infect Dis 6:106. Original gel electrophoresis files for Figure 3D . 486 

Figures S1 to S4 Table S1  Legends for Tables S2, S3  Table S4 SI References

Tables S2 to S3 

Numbers of splice-junction reads for each OAS1 splice variants. Dots, read number of repeats

Genotyping of rs10774671 SNP. Results of the SNP genotyping for Calu-3 cells (a) and Daudi cells (b)

Visualizing the geography of genetic variants

Meta-and Orthogonal Integration of Influenza "OMICs" Data Defines a Role for UBR4 in Virus Budding