key: cord-353731-7xn7m662
authors: Heaton, Brook E.; Trimarco, Joseph D.; Hamele, Cait E.; Harding, Alfred T.; Tata, Aleksandra; Zhu, Xinyu; Tata, Purushothama Rao; Smith, Clare M.; Heaton, Nicholas S.
title: SRSF protein kinases 1 and 2 are essential host factors for human coronaviruses including SARS-CoV-2
date: 2020-08-18
journal: bioRxiv
DOI: 10.1101/2020.08.14.251207
sha: 
doc_id: 353731
cord_uid: 7xn7m662

Antiviral therapeutics against SARS-CoV-2 are needed to treat the pandemic disease COVID-19. Pharmacological targeting of a host factor required for viral replication can suppress viral spread with a low probability of viral mutation leading to resistance. Here, we performed a genome-wide screen in human lung epithelial cells to identify potential host therapeutic targets. We report that the kinase SRPK1, together with the closely related SRPK2, are jointly essential for SARS-CoV-2 replication; inhibition of SRPK1/2 with small molecules led to a dramatic decrease (more than 100,000-fold) in SARS-CoV-2 virus production in immortalized and primary human lung cells. Subsequent biochemical studies revealed that SPRK1/2 phosphorylate the viral nucleocapsid (N) protein at sites highly conserved across human coronaviruses and, due to this conservation, even a distantly related coronavirus was highly sensitive to an SPRK1/2 inhibitor. Together, these data suggest that SRPK1/2-targeted therapies may be an efficacious strategy to prevent or treat COVID-19 and other coronavirus-mediated diseases.

In December 2019, a novel human coronavirus, now known as SARS-CoV-2, emerged 61 and began causing a human disease termed COVID-19 1,2 . Since then, a global pandemic 62 has infected millions and caused hundreds of thousands of deaths to date. Due to the 63 prevalence and severity of this disease, the development of therapeutic interventions is 64 of the highest importance. Much attention has been focused on targeting viral proteins 65

and their associated enzymatic activities. In particular, the virally encoded RNA-66 dependent RNA polymerase (RdRp) and the viral proteases, are attractive potential 67 targets. Remdesivir, the only FDA-approved antiviral for SARS-CoV-2, is a nucleoside 68 analogue which targets the viral RdRp and causes premature termination of 69 transcription 3 . Efficacy of this treatment however, unfortunately, appears limited 4 . 70

In addition to targeting viral proteins directly, other antiviral development strategies 72 attempt to target host factors that the virus requires to complete its lifecycle. Relative to 73 their hosts, viruses have dramatically less coding space in their genomes and therefore 74 utilize host proteins to supplement the activities of their own proteins. The major 75 advantages of inhibiting a virus indirectly via an essential host factor are two-fold: (1) 76 many viruses may utilize the same host protein, therefore host-directed therapeutics have 77 the potential to be broadly acting and (2) while direct targeting of the virus can rapidly 78 select for resistant viral mutants, it is thought to be much more difficult for a viral mutation 79 to overcome inhibition of a co-opted host protein. The development of host-directed COVID-19 therapeutics has been hindered however, 82 due to an incomplete understanding of the host factors required for SARS-CoV-2 83 replication; this is in no small part due to its recent emergence into the human population. sgRNA sequencing data indicated that the host gene with the highest probability of being 125 required for SARS-CoV-2 infection was the serine/arginine-rich protein kinase, SRPK1 126 ( Figure 1B) . We did not enrich for ACE2, presumably due to the numerous lentivirus-127 integrated copies of ACE2 causing inefficient targeting in our cell line. We chose to move 128 forward by interrogating potential roles for SRPK1 during viral infection as kinases are 129 generally thought to be good drug targets 9 , and identifying drug targets was the initial goal 130 of our screen. Since a number of SARS-CoV-2 proteins have been reported to be 131 phosphorylated, we took a bioinformatic approach to predict potential SRPK1 132 phosphorylation sites 10 across the entire SARS-CoV-2 genome ( Figure 1C , 133

Supplemental Table 1 ). The viral protein with by far the highest confidence predicted 134 SRPK1 sites was the viral nucleoprotein (N). Phosphoproteomic studies of SARS-CoV-135 2 proteins have reported that some of the predicted SRPK1 phosphorylation sites in the 136 N protein are indeed phosphorylated 11-13 (Fig 1C) . We therefore hypothesized that 137 SRPK1 was likely responsible for the phosphorylation of at least some sites in the N 138

We next wanted to define the degree to which inhibition of SRPK1 mediated N 141 phosphorylation would affect viral replication, especially since other non-SRPK1 kinases 142 have been predicted to be responsible for SARS-CoV-2 protein phosphorylation 12,13 . We 143 selected the SRPK1 inhibitor SPRIN340 ( Figure 1D ) as it has been reported to inhibit 144 SRPK1 (and SRPK2 with an ~10x higher IC50) without affecting other kinases including 145 the related serine/arginine-rich protein kinase family members (SRPKs) 14 . Treatment of 146 A549-ACE2 cells with increasing concentrations of SRPIN340 revealed significant toxicity 147 at a concentration of 160µM; lower concentrations however, were well tolerated ( Figure  148 1E). We next infected cells treated with non-toxic concentrations of SRPIN340 with 149 SARS-CoV-2 and allowed multicycle viral growth for 24 hours. We observed a dramatic 150 SRPIN340-mediated decrease in viral RNA and infectious viral titer (up to 100,000-fold) 151 at the higher drug concentrations (Figure 1F,G) . 152

The stronger viral inhibitory phenotype at higher concentrations of SRPIN340 suggested 154 that SRPK1 and SRPK2 (which are known to have similar substrate specificities 14,15 ) likely 155 need to be inhibited simultaneously for full suppression of viral replication. To further test 156 that hypothesis, we made use of another small molecule (SPHINX31) that targets SRPK1 157 without appreciably affecting SRPK2 16 . Consistent with our previous data, while we could 158 significantly inhibit SARS-CoV-2 replication by targeting SRPK1 alone, the magnitude of 159 the effect was smaller than with SRPIN340 (Supplemental Figure 2) . Finally, in order 160 to provide direct evidence that both SRPK1 and SRPK2 could indeed phosphorylate the 161 SARS-CoV-2 N protein, we performed in vitro phosphorylation assays with purified N, 162 SRPK1 or SRPK2, and 32-P ATP. In a dose dependent manner, increasing 163 concentrations of the kinases led to increased N protein phosphorylation ( Figure 1H,I) . 164

Together, these data indicate that while SRPK1 is the dominant kinase (and thus most 165 Table 2 ). We 181 therefore wanted to determine if the viral requirement for SRPK1/2 activity might be 182 broadly conserved as well.

SRPIN340 treatment of cells infected with the 183 alphacoronavirus 229E (which is only distantly related to betacoronavirus SARS-CoV-2) 184 inhibited the virus by more than 1,000-fold at non-toxic concentrations of the drug ( Figure  185 2G-H). These data indicate that the requirement for SRPK1/2 activity is not restricted to 186 Further, SRPK1 is upregulated in multiple cancers including prostate, breast, lung, and 206 glioma 22 . In particular, it is thought that its regulation of vascular endothelial growth factor 207 (VEGF) is central to its importance for a number of different diseases 23 . As such, a 208 number of different inhibitors of SRPK1 have been developed and used in various pre-209 clinical disease models, however none of these compounds have yet been FDA 210

approved. In addition to its normal physiological roles and potential roles in cancer, 211 SRPK1/2 kinase activity has been previously reported as important for the replication of have been reported to suppressed after treatment with GSK-3 inhibitors, presumably by 230 affecting N protein phosphorylation 34 . Further studies will be required to understand how 231 viral phosphorylation of SARS-CoV-2 protein affects their functions, as well as 232 experimentally define which kinases phosphorylate which viral proteins at which specific 233

locations. It will also be important to define potential kinase redundancy and the effects 234 on viral replication. At least for SARS-CoV-2, it appears that SRPK1/2 are central to the 235 regulation of viral infection, but it is worth noting that although our study has focused on 236 the phosphorylation of the N protein, we cannot rule out that the viral inhibition phenotype 237 observed is also or instead due to altered phosphorylation of other viral, or indeed host, (Thermo Fisher Cat #15596018) and total RNA was isolated by either phase separation 299 with chloroform and isopropanol or using the Zymo Direct-zol RNA Miniprep kit (Zymo 300

Cat #R2050). One-step qRT-PCR was performed using the Invitrogen EXPRESS One-301

Step Superscript qRT-PCR kit (Thermo Fisher Cat #11781200) and the commercial 302 Milipore, cat# AB3786) and SARS-CoV-2 (1:500, Genetex cat# GTX632604) in blocking 435 buffer at 4°C overnight. Membranes were then washed 3 times in PBST, incubated with 436 secondary antibodies in blocking buffer for 1h at room temperature followed be three 437 washes with PBST and mounted using Fluor G reagent with DAPI. All confocal images 438 were collected using Olympus Confocal Microscope FV3000 using a 40X objectives. 

COVID-19 pathophysiology: A review

Early Transmission Dynamics in Wuhan, China

Remdesivir is a direct-acting antiviral that inhibits RNA-464 dependent RNA polymerase from severe acute respiratory syndrome coronavirus 465 2 with high potency

Remdesivir for the Treatment of Covid-19 -Preliminary Report. 468

Genome-wide CRISPR screen reveals host genes that regulate 470 SARS-CoV-2 infection. bioRxiv

Severe Acute Respiratory Syndrome Coronavirus 2 from 472 Patient with Coronavirus Disease, United States

Improved vectors and genome-wide 475 libraries for CRISPR screening

MAGeCK enables robust identification of essential genes from 478 genome-scale CRISPR/Cas9 knockout screens

Protein kinases--the major drug targets of the twenty-first century?

GPS 5.0: An Update on the Prediction of Kinase-specific 483

Phosphorylation Sites in Proteins

Characterisation of the transcriptome and proteome of 486 SARS-CoV-2 reveals a cell passage induced in-frame deletion of the furin-like 487 cleavage site from the spike glycoprotein

The Global Phosphorylation Landscape of SARS-CoV-2

Growth factor receptor signaling inhibition prevents SARSCoV-2 492 replication. bioRxiv

Utilization of host SR protein kinases and RNA-splicing 494 machinery during viral replication

SRPK2: a differentially expressed SR protein-specific kinase 497 involved in mediating the interaction and localization of pre-mRNA splicing 498 factors in mammalian cells

Development of Potent, Selective SRPK1 Inhibitors as Potential 501

Topical Therapeutics for Neovascular Eye Disease

Structure of the SARS coronavirus nucleocapsid protein RNA-504 binding dimerization domain suggests a mechanism for helical packaging of viral 505 RNA

The Coronavirus Nucleocapsid Is a 507 Multifunctional Protein

Pathogenesis of COVID-19 from a cell biology perspective

Splicing kinase 512 SRPK1 conforms to the landscape of its SR protein substrate

Regulation of SR 515 protein phosphorylation and alternative splicing by modulating kinetic interactions 516 of SRPK1 with molecular chaperones

The many faces of SRPK1

SRPK1 inhibition in vivo: modulation of VEGF splicing and 521 potential treatment for multiple diseases

Serine-Arginine Protein Kinase 1 Regulates Ebola Virus 524 Transcription

Inhibition of hepatitis C virus replication by a specific inhibitor 526 of serine-arginine-rich protein kinase

529 Regulation of the subcellular distribution of key cellular RNA-processing factors 530 during permissive human cytomegalovirus infection

Human papillomavirus type 1 E1^E4 protein is a potent 533 inhibitor of the serine-arginine (SR) protein kinase SRPK1 and inhibits 534 phosphorylation of host SR proteins and of the viral transcription and replication 535 regulator E2

Molecular Mechanism of SR Protein Kinase 1 Inhibition by 537 the Herpes Virus Protein ICP27

Suppression of hepatitis B virus replication 539 by SRPK1 and SRPK2 via a pathway independent of the phosphorylation of the 540 viral core protein

Hepatitis 542 B virus core protein phosphorylation: Identification of the SRPK1 target sites and 543 impact of their occupancy on RNA binding and capsid structure

Identification of SRPK1 and SRPK2 as the major cellular protein 547 kinases phosphorylating hepatitis B virus core protein

Phosphorylation of the arginine/serine 550 dipeptide-rich motif of the severe acute respiratory syndrome coronavirus 551 nucleocapsid protein modulates its multimerization, translation inhibitory activity 552 and cellular localization

The severe acute respiratory syndrome coronavirus nucleocapsid 555 protein is phosphorylated and localizes in the cytoplasm by 14-3-3-mediated 556 translocation

Glycogen synthase kinase-3 regulates the phosphorylation of 559 severe acute respiratory syndrome coronavirus nucleocapsid protein and viral 560 replication

Serine-arginine protein kinase 1 (SRPK1) inhibition as a 562 potential novel targeted therapeutic strategy in prostate cancer

An orally bioavailable broad-spectrum antiviral inhibits 565 SARS-CoV-2 in human airway epithelial cell cultures and multiple coronaviruses 566 in mice

SARS-CoV-2 mRNA was quantified by qRT-PCR n=4. (e) Calu-3 cells were treated with 611

Infectious 612 virus was quantified by plaque assay n=4. (f) Diagram of the N protein of multiple

Red lines within protein indicate predicted SRPK1 phosphorylation sites. 614 (g) Huh7 cells were treated with different concentrations of SRPIN340 and cytotoxicity 615 was assayed with the CellTiter-Glo assay n=4. (h) Huh7 cells were treated with DMSO 616 or 17 µM SRPIN340 then infected with 229E MOI 0.01. Viral RNA

Diagram of primary type II pneumocyte cell culturing and infections. 618 (j) Primary type II pneumocytes were isolated from different donors, cultured on transwell 619 membranes, treated with DMSO or 53 µM SRPIN340 and then infected at MOI=1. After 620 24 hrs cells were fixed and stained for type

Scale bar= 100 µm. (k) Microscopy images from (j) were quantified 622 and % positive cells from four independent donors were plotted, n>3. For all panels, error 623 bars represent SEM and *p<0.05 **p<0.001 by an unpaired