Biochemistry and Cell Biology

Impact of double-stranded RNA characteristics on the activation of human 2’-5’-oligoadenylate synthetase 2 (OAS2).

Journal: Biochemistry and Cell Biology

Manuscript ID bcb-2019-0060.R1

Manuscript Type: Article

Date Submitted by the 29-Mar-2019 Author:

Complete List of Authors: Koul, Amit; University of Manitoba Deo, Soumya; University of Lethbridge Booy, Evan; University of Manitoba Orriss, George;Draft University of Manitoba Genung, Matthew; University of Manitoba McKenna, Sean; University of Manitoba

Keyword: oligoadenylate synthetase, RNA, viral RNA, enzymology

Is the invited manuscript for consideration in a Special Ribowest RNA Issue? :

https://mc06.manuscriptcentral.com/bcb-pubs Page 1 of 39 Biochemistry and Cell Biology

Impact of double-stranded RNA characteristics on the activation of human 2’-5’- oligoadenylate synthetase 2 (OAS2).

Amit, Koul1, Soumya Deo3, Evan P. Booy1, George L. Orriss1, Matthew Genung4 and Sean A. McKenna1, 2

Department of Chemistry1, Department of Biochemistry and Medical Genetics2, University of Manitoba, 144 Dysart Road, Winnipeg, Manitoba, R3T2N2, Canada. Alberta RNA Research and Training Institute3, Department of Chemistry and Biochemistry, University of Lethbridge, 4401 University Drive, Lethbridge, Alberta T1K 3M4, Canada. Max Rady College of Medicine4, Rady Faculty of Health Sciences, University of Manitoba, 750 Bannatyne Avenue, Winnipeg, Manitoba, R3E 0W2, Canada.

Corresponding Author: Dr. Sean A. McKenna [email protected], +1-204-272-1562

Draft

ACKNOWLEDGEMENTS

This work was supported by Natural Sciences and Engineering Research Council of Canada

(NSERC). We would like to thank Emy Komatsu, Dr. Hélène Perreault, and Dr. John Sorensen for

help with mass spectrometry analysis of 2-5A. We also thank and acknowledge the financial support

provided by the University of Manitoba to AK through a University of Manitoba Graduate

Fellowship.

1 https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 2 of 39

ABSTRACT

Human 2’-5’ oligoadenylate synthetases (OAS) are a family of interferon inducible proteins which, upon activation by double-stranded RNA, polymerize ATP into 2’-5’ linked oligoadenylates. In this study, we probe the RNA cofactor specificity of the two smallest isozymes,

OAS1 and OAS2. First, we developed a strategy for the expression and purification of recombinant human OAS2 from eukaryotic cells and quantified the activity of the enzyme relative to OAS1 in vitro. We then confirmed that both OAS2 domains, as opposed to only the domain containing the canonical catalytic aspartic acid triad, are required for enzymatic activity. Enzyme kinetics of both

OAS1 and OAS2 in the presence of a variety of RNA binding partners enabled characterization of the maximum reaction velocity and apparent RNA-protein affinity of activating RNAs. While in this study OAS1 can be catalytically activatedDraft by dsRNA of any length greater than 19bp, OAS2 showed a marked increase in activity with increasing dsRNA length with a minimum requirement of 35bp. Interestingly, activation of OAS2 was also more efficient when the dsRNA contained 3’- overhangs despite no significant impact on binding affinity. Highly structured viral RNAs that are established OAS1 activators were not able to activate OAS2 enzymatic activity based on the lack of extended stretches of dsRNA of greater than 35 bp. Together these results may highlight distinct subsets of biological RNAs to which different human OAS isozymes respond.

Key words: Oligoadenylate synthetase, oligoadenylate, viral RNA, RNA.

2 https://mc06.manuscriptcentral.com/bcb-pubs Page 3 of 39 Biochemistry and Cell Biology

INTRODUCTION

2’ 5’-oligoadenylate synthetases (OAS) are a family of interferon-inducible enzymes that

act as pattern recognition receptors (PRR) to bind double-stranded RNA (dsRNA) during viral

infection (Hovanessian 2007). Upon interaction with dsRNA, activated OAS catalyzes the

polymerization of ATP into 2’ 5’- linked oligoadenylate (2-5A) chains, which in turn bind and

activate their only known target, (RNase L) (Zhou et al. 1993; Dong et al. 1997).

Activated RNase L degrades both viral and cellular RNA, thereby attenuating viral replication (Li

et al. 2016). The OAS-RNAse L system is therefore one of the main interferon-inducible antiviral

pathways that prevent viral propagation (Player et al. 1998; Han et al. 2012) The human OAS family

is composed of four , OAS1, OAS2, OAS3 and OASL, present on 12, of which

only the first three produce a catalyticallyDraft active enzyme (Hovnanian et al. 1998; Rebouillat et al.

1998). These genes encode proteins with one (OAS1), two (OAS2) and three (OAS3) OAS core

domains, and are hypothesized to form tetramer, dimer and monomer respectively in their catalytic

state (Chebath et al. 1987; Hovanessian et al. 1987). Through alternative splicing, these genes

encode 8 different catalytically active isoforms, including five OAS1 isoforms (p42, p44, p46, p48,

and p52), two OAS2 isoforms (p69 and p71), and a single OAS3 (p100) isoform, (Lin et al. 2009).

A core catalytic domain corresponding to the first 346 amino acids of human OAS1 is highly

conserved across different species; this conserved domain is also present in OAS2 and OAS3 in

two and three homologous copies (Hovanessian et al. 2007). OAS2 domain I and domain II share

41 and 53% identity with the core catalytic domain of OAS1, whereas OAS3 domains I, II and III

share 42, 47 and 60% amino acid sequence identity with the OAS1 core catalytic domain

(Hovanessian et al. 2007).

The core catalytic OAS domain is a polymerase beta (pol-β) nucleotidyl transferase domain

that belongs to the superfamily that includes poly-A polymerase (PAP1) (Hartmann et al. 2003) and

3 https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 4 of 39

CCA adding enzymes (Torralba et al. 2008). Human OAS1 consists of a single pol-β like domain, whereas OAS2 and OAS3 consist of two and three such domains respectively, likely acquired by duplication during evolution (Justesen et al. 2000). However, OAS enzymes are distinct in this superfamily as they require a cofactor (dsRNA) for catalytic activity, in contrast to CCA and

PAP1 that are constitutively active (Hovanessian et al. 2007). The core OAS1 domain adopts a bilobal structure with the catalytic aspartic acid triad located at the intersection of the N- and C- terminal lobes (Hartmann et al. 2003). On the opposite face from the active site, two dsRNA binding regions (separated by approximately 30 Å) enable recognition of two adjacent minor grooves encompassing approximately 17 base pairs (bp) (Donovan et al. 2013). The dsRNA-protein interaction is largely mediated through recognition of ribose 2’-OH in the minor groove and electrostatic interactions via basic aminoDraft acids in the dsRNA binding regions (Hartmann et al.

2003). Interaction with dsRNA requires reorientation of helix N3 in OAS1, leading to compression of active site aspartic acid residues (D75, D77, D148) into close proximity, enabling coordination of two magnesium ions required for substrate (ATP) binding during catalysis (Justesen et al. 2000;

Donovan et al. 2013).

The substrate specificities of the three OAS isozymes have begun to be investigated. OAS1, comprising a single domain, will bind dsRNA of any length provided that it exceeds 17 bp

(Donovan et al. 2013). OAS1 also shows enhanced affinity for dsRNA with 5’ di/triphosphate ends as opposed to 5’ -OH (Meng et al. 2012; Donovan et al. 2015). A 3’ single-stranded pyrimidine overhang has been shown to enhance the catalytic activity of both perfect dsRNA and highly structured RNA containing sufficient stretches of dsRNA (Vachon et al. 2015). dsRNA with a single 3’ uridine overhang appears to enable optimal OAS1 activation (Vachon et al. 2015). In

OAS3, the first two core domains (DI & DII) lack the aspartic acids required for catalytic activity but retain dsRNA binding, while the third domain (DIII) is catalytically active (Donovan et al.

4 https://mc06.manuscriptcentral.com/bcb-pubs Page 5 of 39 Biochemistry and Cell Biology

2015). The low-resolution solution structure of OAS3 suggests that the three domains are in an

extended linear arrangement (Ibsen et al. 2014). A strong preference of OAS3 for dsRNA longer

than 50 bp has been observed, and is suggested to require interaction with all three domains for

optimal catalytic activity (Donovan et al. 2015). This length dependence is logical as a single OAS

domain requires approximately 17 bp. Comparatively, little is known about OAS2, which contains

an N-terminal catalytically inactive domain (DI) lacking the aspartic acid triad and a C-terminal

catalytic domain (DII) (Sarkar et al. 1999b).

OAS enzymes are differentially expressed in specific cell types, and their presence in

different sub-cellular fractions such as mitochondria, nuclear, and rough/smooth microsomal

fractions suggest that these proteins may have distinct functional roles (Chebath et al. 1987;

Hovanessian et al. 1987). A comprehensiveDraft study of different OAS enzymes in dengue viral (DEN)

infection demonstrated that OAS1 and OAS3, but not OAS2, were able to show anti-DEN activity

(Lin et al. 2009). In contrast, two single nucleotide polymorphism (SNPs) in the OAS2 gene have

been suggested to increase the susceptibility against DEN infection and no such association was

found with SNPs of OAS1 and OAS3 (Thamizhmani et al. 2014). Murine cells expressing OAS2

isoforms show resistance against encephalomyocarditis virus (ECMV) infection but not towards

the vesicular stomatitis virus (Marié et al. 1999). Porcine OAS2 has been shown to provide

resistance against the porcine reproductive and respiratory syndrome virus (PRRSV) (Zhao et al.

2017) whereas porcine OAS1 has been shown to inhibit Japanese encephalitis virus (JEV)

replication better than porcine OAS2 (Zheng et al. 2015). SNPs in OAS2 have also demonstrated

an association with a predisposition to chronic hepatitis C (HCV) infection in the Russian

population (Barkhash et al. 2014). In cases of severe chronic obstructive pulmonary disease,

exposure to cigarette smoke can lead to more severe outcomes from influenza infection (Zhou et

al. 2013). OAS2, but not OAS1, is produced in mice treated with both cigarette smoke and

5 https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 6 of 39

influenza virus (Zhou et al. 2013). Taken together, these results strongly suggest that OAS enzymes may mediate alternative and sometimes overlapping responses to different viral challenges.

While data is available concerning the dsRNA cofactor specificities of the smallest and largest human OAS isoforms, similar studies investigating OAS2 are lacking. In particular, dsRNA length requirement for OAS2 activation, specificity for viral RNAs, and function of the individual

OAS domains of OAS2 has yet to be explored. Here we report the expression and purification of recombinant and enzymatically active human OAS2 from eukaryotic cells to study the substrate specificity of OAS2. We synthesized dsRNA of varying lengths and used enzyme kinetics to compare the requirements for OAS2 activation relative to OAS1. We additionally examined the impact of 5’ end phosphorylation and 3’ single-stranded overhangs on dsRNA activation of OAS.

To relate substrate specificity to antiviralDraft activity, we studied a panel of established viral dsRNA activators of OAS1 for their ability to interact with and activate OAS2. Our data suggests that OAS1 and OAS2 may have distinct substrate specificity in terms of dsRNA length, which may explain their divergent roles in antiviral processes.

MATERIALS AND METHODS

Cell culture and reagents. The HEK293T cell line was a gift from Dr. Thomas Klonisch

(University of Manitoba). Plasmid pFastBac was purchased from Thermo-Fisher Scientific

(Ottawa, ON, Canada) for expression of the OAS2 gene transcript variant (tv2). Synthetic 19, 25,

30, 35 and 40 (bp) dsRNA were purchased from Integrated DNA technologies (Coralville,

IA, USA) (Fig. S1A). Plasmid PL4440 (Whyard et al. 2015) used for generation of dsRNA was a gift from Dr. Steve Whyard (University of Manitoba). Polyclonal rabbit OAS2 antibody was purchased from Proteintech (Rosemont, IL, USA) and mouse anti-FLAG M2 antibody was purchased from Sigma-Aldrich (Sigma-Aldrich, Oakville, ON, Canada). FLAG affinity resin was

6 https://mc06.manuscriptcentral.com/bcb-pubs Page 7 of 39 Biochemistry and Cell Biology

purchased from GenScript (Picastaway, NJ, USA). Polyethylenimine (PEI) transfection reagent

was purchased from Polysciences (Warrinton, PA, USA) SYBR Gold was purchased from Thermo-

Fisher Scientific (Ottawa, ON, Canada). All other standard laboratory chemicals and reagents were

purchased from Thermo-Fisher Scientific (Ottawa, ON, Canada).

RNA preparation. Adenovirus associated-RNA I (VAI), VAI lacking the terminal stem loop

(VAIΔTS), Epstein-Barr encoded RNA-1 (EBER-1), 5’/3’ terminal regions (TR) of West Nile Viral

RNA (5’TR, 3’TR WNV), human immune deficiency virus -Transactivation response element

(HIV-TAR) and single stranded (ssRNA) RNAs were in vitro transcribed from linearized plasmid

DNA templates using T7 RNA polymerase as described previously (McKenna et al. 2006, 2007).

RNA was purified using size exclusion chromatographyDraft (SEC) using an AKTA FPLC system with

a Superdex 200 26/60 size exclusion column (GE Healthcare) (Booy et al. 2013). 40, 50, 60, 70,

80, 90 and 120 bp dsRNA with an additional three bp at 3’ end (due to the presence of three G

nucleotides in the T7 reverse primer) was produced by in-vitro transcription of the PCR product

which was amplified from plasmid pL4440 (Whyard et al. 2015) using primers containing T7

promoters at each 5’ ends (Supplementary Table 1). The PCR product was first separated by 1%

agarose gel electrophoresis with extracted bands purified using the Genejet gel extraction kit

(Thermo-Fisher Scientific). Each strand of the PCR product contains a T7 promoter at the 5’ end

for in vitro transcription of both strands simultaneously. T7 RNA polymerase was used for in vitro

transcription as performed previously (Booy et al. 2013) with the following exceptions: PCR

product was used as template DNA, an extended incubation of 6 hours was performed, and

transcription products were treated with DNase I (Thermo-Fisher Scientific) to remove the DNA

template. The dsRNA produced had a 17 nt single stranded regions at both 3’ ends (referred to as

ss-dsRNA). Native Tris borate-EDTA (TBE) and denaturing tris borate–EDTA gels were used to

7 https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 8 of 39

verify the purity and homogeneity of RNA transcripts (Fig. S1B, S1C). Before loading RNA on the

gel, it was mixed with 2X denaturing RNA loading buffer (95% deionized formamide, 0.5mM

EDTA, 0.025% Xylene Cyanol, 0.025% Bromophenol Blue, 0.025% Sodium dodecyl sulfate),

heated for 5 minutes at 95°C and ran in 1X TBE buffer (89 mM Tris Base, 89 mM Boric acid, 2

mM EDTA, pH 8.0). For native RNA gels, RNA was first mixed with 10X Native RNA loading

buffer (0.025% Xylene Cyanol, 0.025% Bromophenol Blue, 50% Glycerol) and ran in 1X TBE

buffer. Final RNA concentration was determined by spectrophotometry using absorbance at 260

nm and the predicted extinction coefficient from IDT online tool OligoAnalyzer 3.1: 40bp (113100

mol-1 cm-1), 50bp (1326600 mol-1 cm-1), 60bp (1519600 mol-1 cm-1), 70bp (1716000 mol-1 cm-1),

80bp (1905600 mol-1 cm-1), 90bp (2092200 mol-1 cm-1), 120bp (2660400 mol-1 cm-1) (Owczarzy et al. 2008) . Draft

RNase A cleavage of 3’ ends. The in vitro transcribed 43-123bp ss-dsRNA contained 17 single stranded nucleotides at the 3’ ends. In order to remove these overhangs, we incubated 10 µg of purified RNA transcript with RNase A (10 µg/mL) in 50 mM Tris, 300 mM NaCl (pH 7.5) at 37°C for 30 minutes. RNase A was removed from dsRNA by phenol:chloroform extraction by phase lock gel (Quantabio) and dsRNA was subsequently concentrated by the RNA Cleanup and

Concentration Micro Kit (Thermo-Fisher Scientific). RNase A cleavage leaves a monouracil at the

3’ end of each RNA (referred to as U-dsRNA). The concentration of the RNA was determined by absorbance at 260nm and extinction coefficient obtained from IDT: 40bp (785200 mol-1 cm-1), 50bp

(980800 mol-1 cm-1), 60bp (1173800 mol-1 cm-1), 70bp (1370200 mol-1 cm-1), 80bp (1559800 mol-

1 cm-1), 90bp (1746400 mol-1 cm-1), 120bp (2314600 mol-1 cm-1). To ensure the purity of dsRNA post RNase A cleavage of single stranded ends, denaturing TBE gel was performed as described above.

8 https://mc06.manuscriptcentral.com/bcb-pubs Page 9 of 39 Biochemistry and Cell Biology

5’-end dephosphorylation of dsRNA. Under standard conditions, in vitro transcription generates a

5’ triphosphate (5’-PPP) due to the inclusion of nucleoside triphosphates in the transcription

reaction. Transcribed RNAs (7 µg) were treated with FAST AP thermosensitive alkaline

phosphatase (7 units, Thermo-Fisher Scientific) at 37 C for 20 minutes in a 200 µL final volume.

RNA was purified by phenol:chloroform extraction and further purified by RNA Cleanup and

Concentration Micro Kit (Thermo-Fisher Scientific). Denaturing RNA TBE gels were run to

confirm the RNA was intact and pure after the treatment. We used 10ug/mL of the ss-dsRNA, U-

dsRNA and 5’ end dephosphorylated RNAs to compare the OAS1 and OAS2 activity in our assays.

To confirm the effectiveness of phosphatase treatment we dephosphorylated VAI ∆TS in parallel

with ss-dsRNA and U-dsRNA and activatedDraft OAS1 (data not shown). Dephosphorylation of VAI

∆TS resulted in a decrease of OAS1 activity by approximately 50% as has been demonstrated

previously (Meng et al. 2012), however no change was observed in the OAS1 activity by either 5’-

PPP or dephosphorylated states of ss-dsRNA or U-dsRNA.

Expression and purification of recombinant human OAS proteins. Expression of recombinant

human OAS1 p42 isoform (transcript variant 2) as a fusion protein in BL21 DE3 RIL cells

(Invitrogen, USA) and subsequent purification of the free protein from the cleaved N-terminal

affinity tag (GNSHT: GB1, Nus A, Streptavidin, 6xHis and TEV protease site) was performed as

described previously (Meng et al. 2012). The protein was dialyzed in storage buffer (50mM tris,

100mM NaCl, pH 7.4, 10% (v/v) glycerol) and stored at 4°C.

Recombinant human OAS2 p69 (transcript variant 2) was amplified by PCR from pFastBac

using DNA primers that were designed to encode an N-terminal DYKDDDDK (FLAG) tag. The

resulting PCR product was digested with XhoI and HindIII and ligated using T4 DNA ligase

9 https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 10 of 39

(Thermo-fisher scientific) into pCDNA3 vector (Thermo-Fisher Scientific) and confirmed by sequencing. pCDNA3 vector containing FLAG-OAS2 was transformed into NEB Turbo competent

E. coli (NEB, ON, Canada) using the manufacturer’s protocol. Maxi-preps (Thermo-Fisher

Scientific) were performed to obtain plasmid DNA for transfection of HEK293T cells. HEK293T cells were grown on ten 150mm plates (Thermo-Fisher Scientific) with Dulbecco’s Modified

Eagle’s Medium (DMEM) (Thermo-Fisher Scientific) supplemented with 10% fetal bovine serum

(FBS) such that the confluence of the cell is 80% at the time of transfection. On the day of transfection, media was replaced with 24 mL of fresh DMEM containing 10% FBS. Transfection mixture was made by adding 300 µg of FLAG-OAS2 vector into each of two 50 mL sterile conical tubes, each containing 20 mL serum free DMEM. Next 490 µL (1mg/mL) of polyethylenimine

(PEI) (Polysciences, PA, USA) was addedDraft to each tube and the final volume of each tube was brought to 30 mL with serum free DMEM. Transfection mixture was mixed thoroughly by pipetting the entire volume up and down several times and incubated at room temperature for 20 minutes. 6 mL of the transfection mixture was added dropwise to each plate and the plates were further incubated at 37 °C for 48 hr. for protein expression. OAS2 was purified by harvesting the cells from each 150mm plate using 10 mL PBS (Thermo-Fisher Scientific) and pelleting by centrifugation at

1000 xg for 5 minutes at 4°C. Samples were kept on ice throughout the remaining procedure.

Ten 150mm plates of cells were re-suspended in 2 mL of 1% lysis buffer- 50 mM Tris Base,

400 mM NaCl, 1% octylphenoxy poly(ethyleneoxy) ethanol (IGEPAL), 1 mM Dithiothreitol (DTT) supplemented with 1X halt protease cocktail inhibitor (Thermo-Fisher Scientific). Lysate was further diluted in 18 mL of lysis buffer (50 mM Tris Base, 400 mM NaCl, 1 mM DTT, pH 7.4) supplemented with 1X halt protease cocktail inhibitor (Thermo-Fisher Scientific). The lysate was vortexed for 10 seconds and then sonicated for 6 pulses of 10 seconds each at 35% output on ice

(Branson Sonic Dismembrator). The lysate was centrifuged at 20,000 xg for 20 minutes and filtered

10 https://mc06.manuscriptcentral.com/bcb-pubs Page 11 of 39 Biochemistry and Cell Biology

using a 0.2 µM syringe filter. The filtered lysate was applied to a gravity column filled with 2 mL

of anti-DYKDDDDK G1 affinity resin (Genscript) pre-equilibrated with 0.1% lysis buffer (50 mM

Tris Base pH 7.4, 400 mM NaCl, 0.1% IGPAL, 1 mM DTT). The column was washed with 15

column volumes of cold wash buffer-1 (50 mM Tris base pH 7.4, 400 mM NaCl, 0.1% IGPAL, 1

mM DTT) and 30 column volumes of wash buffer-2 (50 mM Tris Base pH 7.4, 100 mM NaCl,

0.1% IGEPAL, 1 mM DTT). The protein was eluted as 1 mL fractions by incubation for 15 minutes

at room temperature with elution buffer (50 mM Tris base pH 7.4, 100 mM NaCl, 1 mM DTT, 10%

Glycerol) supplemented with 0.1 mg/mL FLAG peptide (Genscript). Fractions collected were

analyzed for purity by SDS-PAGE. The pure fractions were combined and dialyzed overnight in to

storage buffer (50mM Tris Base pH 7.4, 100mM NaCl, 1mM DTT, 10% glycerol) to remove excess

FLAG peptide. Size exclusion chromatographyDraft was performed to remove any remaining protein

bound to nucleic acid from free protein. Protein concentration was determined

spectrophotometrically by absorbance at 280 nm using the extinction coefficient (129830 mol-1 cm-

1) calculated with the ProtParam tool on ExPASy servers (Wilkins et al. 1999; Gasteiger et al.

2005). Protein yield was approximately 200 µg/150 mm plate.

Size-exclusion chromatography-multi angle light scattering analysis (SEC-MALS). SEC-MALS

measurements were performed on previously purified OAS2 (350 µL at 1mg/mL) using a

Superdex200 size exclusion column, (GE Healthcare, Mississauga, ON, Canada) connected to a

DAWN HELEOS-II 18-angle light scattering detector and an Optilab T-rEX differential

refractometer. The ASTRA 6 package was used for data analysis (Wyatt Technology Corporation,

Santa Barbara, CA, USA).

11 https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 12 of 39

Expression/purification of individual OAS2 domains. Construction of vectors for FLAG-tagged

OAS2 domain I (DI1-331, 39.2 kDa) or domain II (DII332-687, 41.5 kDa) was performed as follows: for DI1-331, an N terminal FLAG sequence (DYKDDDDK) was added to the forward primer and a stop codon (TAA) was added to the reverse primer (after amino acid proline at position 331) and inserted by PCR. Similarly, for DII332-687 N terminal FLAG sequence was added before amino acid serine at position 332 and a stop codon (TAA) in the reverse primer. The PCR products were digested with XhoI and HindIII and ligated using T4 DNA ligase (Thermo-Fisher Scientific) to pre- digested pCDNA3. Individual domains were expressed and purified following the same protocol used for FLAG-OAS2. To visualize the purity of individual domains, SDS-PAGE gel was performed. We further confirmed the expression of individual OAS2 domains by a western blot using antibodies specific for the FLAG Draftpeptide (Fig. S1D).

OAS enzyme activity assay. OAS activity was measured using an established colorimetric assay that quantifies the amount of pyrophosphate (PPi) produced as a byproduct of 2-5A formation

(Meng et al. 2012). Kinetic experiments were performed using RNA in the range 5–800 nM. OAS

(400 nM) was incubated with RNA in the presence of 20 mM Tris-HCl pH 7.4, 5 mM MgCl2, 1 mM DTT, 2 mM ATP and 1% glycerol in 70 µL total volume at 37 °C. Aliquots of 20 µL were removed and immediately quenched by adding to 50 mM EDTA pre-loaded into the wells of a 96- well plate. At completion of the time course, 20 µL of molybdate reagent (2.5% ammonium molybdate in 2.5 M H2SO4), 20 µL β-mercaptoethanol (0.5M) and 8 µL of Eikonogen reagent

(0.125 g of 1-amino-2- naphthol-4-sulfonic acid, 0.125 g of sodium sulfite, and 7.325 g of sodium bisulfite to 100 mL of ddH2O) were added to each well and the final volume brought to 200 µL with water. Absorbance at 580 nm was measured using Epoch plate reader (BioTek, VT, USA) and readings converted to PPi produced by comparison with PPi standards subjected to identical

12 https://mc06.manuscriptcentral.com/bcb-pubs Page 13 of 39 Biochemistry and Cell Biology

process, as described previously (Meng et al. 2012). To account for minor variations between OAS

preparations we plotted data relative to poly (I:C) when comparing the effects of 3’ end cleavage.

For kinetic analysis of OAS1/OAS2 at different dsRNA concentrations, linear regression analysis

of early time points was used to obtain the initial rates (Vo). Initial rates were plotted against RNA

concentration and the apparent dissociation constant (Kapp) and the maximum reaction velocity

(Vmax) were determined by fitting the data to the following equation: V= (Vmax/(1+Kapp/[RNA]))

(Hartmann et al. 1998). Values were plotted and curve fit using Prism 7 (GraphPad).

2-5A formation confirmation by mass spectrometry. Activation assays (500 µL) were performed

by incubating OAS2 (300nM) in presence of 20 mM Tris-HCl pH 7.4, 5 mM MgCl2, 1 mM DTT,

2 mM ATP, 1% glycerol and 123 bp ss-dsRNADraft (20ug/mL) for 1hr at 37 °C. 40 µL of the reaction

(along with appropriate controls) was loaded on TBE denaturing RNA gel (10%) and ran at 200v

for 20 minutes, and stained with SYBR gold (Invitrogen) for 10 minutes. The remaining reaction

volume was passed through an Ultracel 10 kDa MW spin concentrator (Millipore) to remove the

RNA (123 bp ss-dsRNA) and OAS2. A Bruker compact QqTOF LC-MS/MS was then used to

detect 2-5A formation. 2µL of the sample was injected into a reverse phase column (Symmetry

shield TM RP18 5µm, 2.1×50mm). HPLC eluents were freshly prepared. Eluent-A was a mixture of

water and 0.1% formic acid. Eluent-B was a mixture of acetonitrile and 0.1% formic acid.

Separation was performed at a constant flow rate of 0.2mL/min with the following gradient: 20%

of eluent A for 5 minutes and 80-2% of eluent B for 30 minutes followed by 12% eluent B for 10

minutes and 80% of eluent B for 4 minutes. Mass spectra were obtained over the first 30 minutes

of separation in negative mode (electrospray ion source).

RESULTS

13 https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 14 of 39

Purification of active recombinant human OAS2 expressed in HEK293T cells

We expressed FLAG-tagged recombinant human OAS2 (69 kDa) in HEK293T cells and purified protein using affinity chromatography (Fig. 1A). Western blotting with OAS2-specific antibody confirmed the purified protein as OAS2 (data not shown). To remove OAS2 that was co-purifying with bound cellular nucleic acids, the protein was subsequently purified by size exclusion chromatography (Fig. 1B). OAS2 eluted as a single peak at approximately 170 mL with an

A280:A260 ratio consistent with free protein. Size-exclusion chromatography coupled with multi angle light scattering analysis (SEC-MALS) of purified OAS2 suggested a molecular weight of 135

± 9 kDa consistent with a dimer of OAS2 monomers (Fig. 1C). To initially confirm that the purified

OAS2 is catalytically active and can polymerize ATP into 2-5A chains, we mixed the enzyme in a buffered solution that contained ATP andDraft dsRNA (with appropriate negative controls), and detected a smear, potentially corresponding to 2-5A chain formation (<25 nt in length), on a denaturing gel that could be separated from activating RNA and OAS2 by collecting the flow through from spin concentration (Fig. 2A). To confirm that 2-5A chains were formed, we subjected the same flow through to LC-MS/MS. We observed 2-5A dimer (836.0 Da; retention time of 15.8-16.6 min), 2-

5A trimer (1165.1 Da; 17-18.4 min) and 2-5A tetramer (1494.1 Da; 20.5-22.4 min) (Fig. 2B, C, D) consistent with the published values (Morin et al. 2010). An analysis in the absence of RNA activator or absence OAS2 did not result in production of 2-5A chains by LC-MS/MS, as expected.

After confirming the production of 2-5A chains by OAS2, for any further activation assays we used an established colorimetric assay that detects inorganic pyrophosphate (PPi), a byproduct of 2-5A chain formation from ATP by the enzyme (Meng et al. 2012). Briefly, PPi (in the presence of concentrated sulfuric acid) reacts with ammonium molybdate to form a PPi-molybdate complex, which is then reduced by thiol reagent (β-mercaptoethanol) and detected by absorbance at 580 nm after addition of a developing reagent (Ekinogen) that produces a characteristic blue color (Putnins

14 https://mc06.manuscriptcentral.com/bcb-pubs Page 15 of 39 Biochemistry and Cell Biology

et al. 1975). We benchmarked enzymatic activity using poly (I:C), a long synthetic dsRNA. We

first performed complementary experiments where both OAS2 (50-500 nM) and poly (I:C) (0-60

µg/mL) concentrations were varied (Fig. 3A, 3B). The results clearly demonstrate OAS2 as active,

with a general increase in activity as either enzyme or dsRNA activator concentration is increased.

OAS2 concentrations between 300-400 nM and poly (I:C) concentrations between 20-40 µg/mL

were used for subsequent kinetic analyses based on these results. OAS2 activity was next compared

to OAS1 using these conditions (Fig. 3C). In the absence of poly (I:C), no detectable catalytic

activity is observed. OAS2 in presence of poly (I:C) promotes 2-5A chain formation to similar

levels as OAS1 at early time points, but OAS1 promotes more 2-5A chain formation at longer time

points. To quantitatively confirm these results, we determined the maximum reaction velocity

(Vmax) and the apparent binding affinity Draft(Kapp) for OAS enzymes in the presence of increasing poly

(I:C) concentrations (at fixed excess ATP) by determining the initial reaction velocity from the

linear phase of time courses (30 min) (Fig. 3D). OAS1 exhibited a Vmax of 7.7 + 0.3 µM/min and a

Kapp of 35 + 5 nM (Table 1). The kinetic parameters determined for OAS2 in the presence of poly

(I:C) were similar to OAS1 (Vmax=5.6 ± 0.3 M/min; Kapp=36 ± 6 nM).

Individual OAS2 domains are not sufficient for catalytic activity

As the C-terminal catalytic domain of OAS2 (DII) shares strong and

catalytic aspartic acids (D408, D410, D481) with OAS1 (Justesen et al. 2000), we next wanted to

determine whether DII alone was sufficient for 2-5(A) chain formation. Individual DI and DII were

expressed in HEK293T cells as N-terminal FLAG fusion proteins and purified in a similar manner

to full length OAS2 (Fig. 4A, S1D). While the yield of individual OAS2 domains was lower than

the full-length protein, soluble DI and DII were purified (Fig. 4B). Neither DI nor DII individually,

nor when added in trans, demonstrated the capacity to synthesize to 2-5A chains to a level above

15 https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 16 of 39

the negative control (no poly I:C), whereas enzymatic activity was observed for both full-length

OAS1 and OAS2 (Fig. 4C). Therefore, both OAS2 domains within the intact protein are required to achieve catalytic activity.

OAS2 requires longer dsRNA for activation than OAS1

Given that both domains from OAS2 were required for enzymatic activity, we hypothesized that only dsRNAs capable of interacting with both domains simultaneously would serve as activators. A similar observation has been made for the longer isozyme, OAS3, lending support to our hypothesis (Donovan et al. 2015). If each OAS2 domain recognized approximately 17 bp, the minimum length required for OAS1, then we would anticipate requiring close to 35 bp. To initially test this hypothesis, we used chemicallyDraft synthesized perfectly duplexed dsRNA (no overhangs) of different lengths (19, 25, 30, 35 and 40bp) that were purified by size exclusion chromatography

(Fig. S1A). In addition, single stranded RNA (ssRNA) were used as a negative control. Using the colorimetric assay to detect PPi formation (indicative of 2-5A chain formation), the ability of the synthetic dsRNA and poly (I:C) to activate both OAS1 and OAS2 was determined over a time course.

As expected, poly (I:C) and synthetic dsRNA 19bp or longer were able to activate OAS1

(Fig. 5A). Interestingly, while poly (I:C) was a potent OAS2 activator, only 35 and 40 bp synthetic dsRNA (and not 19, 25 or 30 bp) could activate OAS2 above baseline (Fig. 5B). ssRNA did not activate either OAS enzyme. We next determined the kinetic parameters for 25, 30, 35 and 40 bp dsRNA to quantify the differences observed (Fig. 5C, 5D). OAS1 shows similar maximum reaction velocity (Vmax) values for 25, 30, 35 and 40 bp dsRNA and demonstrates a modestly higher apparent affinity (Kapp) for 40 bp versus 25 bp dsRNA (Table 1). For OAS2, kinetic parameters could not be reliably obtained for 25, 30 bp synthetic dsRNA due to the lack of enzymatic activity. OAS1

16 https://mc06.manuscriptcentral.com/bcb-pubs Page 17 of 39 Biochemistry and Cell Biology

achieves a Vmax 8-fold less and a Kapp 4.5-fold less relative to OAS1 with 35bp dsRNA. 40bp dsRNA

achieves a Vmax 3-fold less for OAS2 relative to OAS1 with Kapp consistent with that observed for

OAS1 with 25bp dsRNA (Table 2).

OAS2 activation potential is dependent on dsRNA length

To characterize length requirements of dsRNA activation of OAS enzymes, we in vitro

transcribed a series of dsRNAs ranging in size from 43-123 bp. As a byproduct of transcription,

dsRNA were originally transcribed with a 17 bp single-stranded overhang at each 3’-end (ss-

dsRNA) (Fig. 6A). ss-dsRNA were tested for OAS1/OAS2 activation at a fixed concentration (10

µg/mL) for 60 minutes to compare enzymatic activity relative to poly (I:C) (positive control; 10

µg/mL). All ss-dsRNA were equally effectiveDraft at OAS1 activation regardless of length and were

within error of poly (I:C) (Fig. 6B). When ss-dsRNA were used for OAS2 activation, the activity

increased steadily from 43 to 123 bp, achieving 85% of poly (I:C) levels at 123 bp. (Fig. 6C).

To quantitatively confirm these observations over a range of ss-dsRNA concentration,

kinetic measurements of OAS enzyme activity in the presence of ss-dsRNA of variable length was

performed to determine Vmax and an apparent affinity (Kapp). Linear regression analysis of early time

points of OAS1/OAS2 activity with variable ss-dsRNA concentration was used to obtain the initial

rates (Vo) (Fig. 7A). Initial rates were plotted against RNA concentration and the apparent

dissociation constant (Kapp) and the maximum reaction velocity (Vmax) were determined (refer to

Materials and Methods). Although data were initially collected with an equal mass concentration

of each ss-dsRNA, results are instead presented in terms of the molar dsRNA concentration to

ensure that an equal number of ss-dsRNA molecules were used for comparison. For OAS1, Kapp

decreased with an increase in ss-dsRNA length (3.2-fold over the full range) and Vmax results fell

within a narrow range (6.6-9.0 µM/min) consistent with poly I:C (7.7 µM/min) (Fig. 7B, Table 1).

17 https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 18 of 39

For OAS2, Kapp decreased with an increase in ss-dsRNA length; 4-fold over the full range.

Interestingly, the affinity for OAS2 only approached poly (I:C) levels at the largest RNA length

(123 bp) (Fig. 7C), whereas for OAS1 significantly shorter RNA was required. A 4.5-fold increase in Vmax was observed when comparing OAS2 activation by 43 bp to 123 bp ss-dsRNA, achieving a value similar to poly (I:C) by 73-83 bp (Table 2). Taken together, these results suggest that both binding affinity and catalytic rate are dependent upon dsRNA length beyond 35 to 40 bp for OAS2.

OAS2 activity is enhanced by the single stranded 3’ overhangs on dsRNA.

Single-stranded 3’ overhangs have been shown to play an important role in OAS activation; for example single stranded pyrimidine rich (CUUU) at 3’ ends of an adenoviral dsRNA has been shown to enhance OAS1 activation (VachonDraft et al. 2015). To test the importance of 3’-overhangs, we RNAase A-digested each of the ss-dsRNAs to leave 43-123 bp dsRNA with a monouracil at each 3’-end (U-dsRNA) (Fig. 6A, arrows). To ensure that RNAse A treatment did not digest the dsRNA segment, the length of cleaved RNA was confirmed by denaturing gel electrophoresis (Fig.

7D). To compare the effects of a 17 nt overhang versus a monouracil, dsRNA were tested for

OAS1/OAS2 activation at a fixed concentration (10 µg/mL) for 60 minutes to compare enzymatic

activity relative to poly (I:C) (positive control; 10 µg/mL). No significant difference between ss-

dsRNA and U-dsRNA for OAS1 activation over the range of dsRNA lengths (Fig. 6B, red vs.

purple) were observed. Conversely, a 2-fold reduction in OAS2 activation was observed when the

longer 3’ overhang was reduced to a monouracil for all dsRNA lengths examined (Fig. 6C). The

dependence on a longer single-stranded overhang for OAS2 was then quantitatively confirmed by

kinetic analysis over a range of U-dsRNA concentrations (Fig. 7E). A 17 nt 3’-overhang (ss-

dsRNA) leads to a Vmax approximately 2-fold higher relative to the monouracil counterpart (U- dsRNA) at any dsRNA length examined (Table 2). This decrease in catalytic activity did not result

18 https://mc06.manuscriptcentral.com/bcb-pubs Page 19 of 39 Biochemistry and Cell Biology

from a change in binding affinity, as similar Kapp values were obtained for ss-dsRNA and U-dsRNA

of the same length (Table 2).

Phosphorylation state of the 5’ end of dsRNA activators does not affect OAS2 activation.

The 5’ phosphorylation state of foreign dsRNA has been proposed to be used by the innate

immune system to discern self RNAs from foreign RNAs (Nallagatla et al. 2008). The presence of

a 5’-PPP, as opposed to monophosphate or 5’-OH, has been demonstrated to enhance the activation

of OAS1 by dsRNA (Meng et al. 2012). We therefore sought to determine the impact of the 5’

phosphorylation state of RNA on OAS2 activation. In vitro transcribed RNA has a 5’-PPP owing

to the use of 5’ nucleotide triphosphates as their building blocks. Alkaline phosphatase was used to

remove the 5’-PPP from both ss-dsRNADraft and U-dsRNA of varying length, leaving a 5’-OH. Neither

dephosphorylated ss-dsRNA nor U-dsRNA (of any length examined) has a significant impact

OAS1 or OAS2 relative to the same RNA with a 5’-PPP (Fig. 6B, 6C). We therefore conclude that

OAS2 5’-phosphorylation state is independent for dsRNA.

Different dsRNA length requirement of OAS2 suggests non-overlapping biological functions.

To explore the interaction between transcribed viral RNAs and OAS2, we investigated a

panel of 6 well-characterized viral dsRNAs that have known secondary structures (see Figure S2

for details) and whose efficiency of OAS1 activation in vitro has been previously established. The

5’-terminal region (5’-TR; nt 1 to 146) and 3’-terminal region (3’-TR; nt -1 to -111) of the West

Nile Virus (WNV) RNA genome are rich in conserved secondary structures amongst members of

Flaviviridae (Ng et al. 2017), and are both established activators of OAS1 (Deo et al. 2014).

Adenoviral RNA (VAI; 159 nt) synthesized by host cellular RNA polymerase III accumulates in

cells during the late stages of infection, and is a modest OAS1 activator in vitro (Thimmappaya et

19 https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 20 of 39

al. 1982; Mathews et al. 1991; Meng et al. 2012). The Dicer-processed VAI lacking the terminal

stem (VAIΔTS) shows binding and inhibition of OAS1 activation (Meng et al. 2012). Epstein-Barr

RNA (EBER-1) is a similarly transcribed non-coding RNA with a single report suggesting OAS1

binding and activation (Sharp et al. 1999). HIV transactivation response element (HIV-TAR; 57 nt)

is an RNA stem loop present in the 5’ terminal region of HIV mRNA and is essential for viral RNA

to interact with transacting protein Tat (Muesing et al. 1987; Dingwall et al. 1989).

We in vitro transcribed WNV 5’ TR, WNV 3’ TR, VAI, VAIΔTS, HIV-TAR and EBER-

1 from linearized plasmids using T7 polymerase. RNA homogeneity was confirmed by denaturing gel electrophoresis (Fig. 8A). Time courses measuring PPi production for OAS1 and OAS2 (Fig.

8B) were performed in the presence of each of the viral RNAs and poly (I:C) (positive control). As none of the viral dsRNAs studied have Draftstretches approaching 35 consecutive bp, we hypothesized that they would not serve as activators of OAS2. With the exceptions of HIV-TAR and EBER-1, all viral RNAs were able to activate OAS1 to some extent (Fig. 8B). Whereas 5’TR and 3’TR achieved potent activation, VAI and VAI∆TS were only modestly above baseline. Interestingly, none of the viral RNAs activated the catalytic activity of OAS2 to levels similar to OAS1 (Fig.8C).

Therefore, established in vitro activators of OAS1, particularly the 5’-TR and 3’-TR of WNV, were unable to affect the same catalytic enhancement for OAS2, reinforcing the dsRNA length dependencies of the enzymes.

DISCUSSION

While three highly related catalytically active OAS enzymes exist as part of the overall innate immune response to viral infection, understanding their specific contributions as pattern recognition receptors for viral dsRNA has remained largely elusive. While the in vitro RNA binding properties of the smallest member, OAS1, and the largest, OAS3, have been studied, the

20 https://mc06.manuscriptcentral.com/bcb-pubs Page 21 of 39 Biochemistry and Cell Biology

dsRNA cofactor specificity of human OAS2 has not been extensively characterized. As the

expression of OAS2 in bacterial cells (which lack most post-translational modifications) has been

shown to produce inactive protein (Sarkar et al. 1999a), we developed an expression/purification

protocol of human OAS2 in mammalian cells that produced 200 µg of pure recombinant human

OAS2 per 150mm plate. Using this approach, we produced OAS2 that had comparable, albeit

moderately lower, catalytic activity when compared to human OAS1 with poly (I:C) (Fig. 3C).

Once the minimum dsRNA length of 17 bp is satisfied, OAS1 has been shown to be

indiscriminately activated to a similar level by dsRNA or poly (I:C) of different lengths (Desai et

al. 1997; Donovan et al. 2015) Our results confirm these observations, as dsRNA of 43-123bp have

a similar impact on OAS1 activity (Vmax). These results support high-resolution structural studies

that demonstrate that two successive minorDraft grooves of an A-RNA helix are required for OAS1

interaction, consistent with 17bp (Donovan et al. 2013). OAS2 consists of a catalytically inactive

OAS domain (DI) and catalytically active OAS domain (DII) based on the absence or presence of

the catalytic aspartic acid triad. Notably, both domains maintain conservation of the basic dsRNA

binding region. Our data support a model where a single dsRNA of approximately 35 to 40 bp

interacts with both DI and DII simultaneously, spanning across 4 minor grooves. In the absence of

high resolution structure, we interpret this result as both dsRNA binding regions from DI and DII

being engaged with dsRNA to enable catalysis. We purified OAS2 as a stable dimer, and we are

currently investigating the structural and functional role that dimerization may play. For full

disclosure, our results contrast those of Sarkar et. al. (Sarkar et al. 1999a), which reported that a 25

bp dsRNA can activate OAS2 purified from insect cells. At this stage we cannot definitively

rationalize this difference, but the sequence of the dsRNA we used were different as were the cells

used for expression of OAS2. The current study did not comprehensively examine the impact of

dsRNA cofactor sequence on OAS2 activation, but should be considered for future studies as this

21 https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 22 of 39

effect has been previously highlighted (Kodym et al. 2009). OAS3 has three OAS domains and while the catalytically active DIII by itself is sufficient to catalyze 2’-5’(A), optimal activity requires all three domains and more than 50 bp dsRNA (Donovan et al. 2015). We therefore suggest that OAS1, OAS2, and OAS3 enable discrimination between increasingly longer viral dsRNA, with domain duplication representing an adaptation to different viral challenges and the highly structured

RNAs that accompany them. This hypothesis is bolstered by our observation that OAS1, but not

OAS2, can stimulate the production of 2-5A in the presence of the highly structured 5’-TR and 3’-

TR of the WNV genome. WNV belongs to the viral Flaviviridae family, which have positive single- stranded RNAs genomes, and include pathogenic viruses like Dengue and JEV (Chambers et al.

1990). Flaviviridae family members share high secondary structural conservation within their 5’-

TR and 3’-TR regions owing to requisiteDraft long-range interactions between the regions prior to transcription (Friebe et al. 2010; Ng et al. 2017). Conserved TR secondary structures are, however, not perfectly double-stranded which results in stem loops capable of activating OAS1 (Meng et al.

2012; Deo et al. 2014), but likely insufficient to interact with both domains of OAS2 (Fig. 8B, 8C).

While a lower boundary for dsRNA cofactor length of approximately 35 bp is required for

OAS2 activity, we demonstrated that at the same molar concentration an increased OAS2-dsRNA apparent binding affinity (Kapp) and OAS2 catalytic activity (Vmax) was observed as dsRNA length was increased from 43-123 bp (Table 2). OAS1, in contrast, demonstrated almost no impact on catalytic activity for the same dsRNAs. The most plausible explanation for the length dependence observed for OAS2 is that two individual protein region binding events have to occur on the same dsRNA cofactor; increased dsRNA length increases the likelihood that once the first OAS2 domain interacts that there remains sufficient room on the dsRNA for the second domain to also bind. By

123 bp, Vmax appears to have reached close to its optimal value for the experimental conditions examined, again reinforcing that OAS2 may be fine-tuned to optimally sense viral dsRNA of

22 https://mc06.manuscriptcentral.com/bcb-pubs Page 23 of 39 Biochemistry and Cell Biology

intermediate length of approximately 2-10 turns of an A-RNA helix. Similar observations have

been made on another interferon inducible enzyme, PKR, for which it has been shown that 30 bp

dsRNA was able to activate PKR and that the activation potential increases with longer dsRNA,

eventually saturating at a length of 85 bp (Manche et al. 1992).

Our results also demonstrate that OAS2, but not OAS1, is sensitive to the presence of 3’

single stranded ends under the specific experimental conditions examined. Direct comparison of

ss-dsRNA (17 nt overhang) versus U-dsRNA (monouracil overhang) show that while binding

affinity (Kapp) is not impacted, the longer overhangs promote higher Vmax (~ 2-fold increase in Vmax)

(Table 2). While further studies are warranted to determine the importance of 3’ overhangs in a

cellular context, a 3’ overhang may stabilize the active site structure of the OAS enzyme as has

been suggested previously, or promoteDraft the ss-dsRNA-driven conformational change of OAS

enzymes leading to faster catalysis of 2-5A chains (Vachon et al. 2015). Since our colorimetric

assay only detects the production of PPi, we cannot comment on whether the increase in the OAS2

activity favors production of 2-5A of a specific or longer chain length. In a cellular context, 2-5A

chains are necessary for the activation of RNase L. Once bound to 2-5A chains, RNase L degrades

single-stranded RNAs and single-stranded regions present within loops and bulges in structured

viral RNAs preferentially at UA or UU dinucleotide sequences (Wreschner et al. 1981). RNase L

action may therefore increase the presence of dsRNA with a 3’ overhang, and thereby sensitize the

cell to viral invaders through rapid amplification of OAS activity and RNase L-mediated RNA

degradation (Vachon et al. 2015). This RNase L-dependent amplification has also been observed

in the context of short self-RNA cleavage (Malathi et al. 2007). While our work on dsRNA greater

that 43 bp showed no significant impact of a 3’ overhang on OAS1 activity, a minimal 18bp dsRNA

with 3’-overhang has been shown to increase OAS1 activity and 2-5A chain length (Vachon et al.

2015).

23 https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 24 of 39

Most processed host RNAs do not have 5’-PPP at the end and are typically processed as 5’

monophosphates for rRNA transcripts by Pol I or 7-methyl guanosine caps for mRNA transcripts

by pol II (Nallagatla et al. 2008). In contrast, many viral RNAs contain terminal 5’-PPP (Nallagatla

et al. 2008). There are contradicting reports about enhanced OAS1 activation in the presence of 5’-

PPP. We have previously shown that the presence of 5’-PPP in VAI RNA enhances OAS1

activation, while other literature suggests that a different 18bp dsRNA with 5’-PPP has no effect

on OAS1 activity (Meng et al. 2012; Vachon et al. 2015). Removal of 5’-PPP from model ss-dsRNA

cofactors has been shown to abrogate PKR activation, an interferon stimulated innate immunity

viral RNA binding protein, however perfectly double stranded RNA was shown not to have 5’-PPP

dependence (Nallagatla et al. 2008). We saw no impact of 5’-phosphorylation state on OAS1 or

OAS2 activity by comparing in vitroDraft transcribed dsRNAs (5’-PPP) versus enzymatically dephosphorylated dsRNA. The effects of 5’-PPP on OAS enzymes, specifically on OAS2 needs to be further evaluated with a variety of dsRNAs including viral RNAs. Considering our previous observation of OAS1 activation by VAI RNA with 5’-PPP, we conclude that the effects of 5’-PPP on activation of OAS enzymes (OAS1, OAS2) are RNA dependent.

In conclusion, we have begun to define the features of dsRNA required for efficient catalytic activity for OAS2; length and end modifications play a central role. Established viral RNA activators of OAS1 could not activate OAS2 to similar levels. Future work that broaden to evaluate different viral RNAs with end modifications in a cellular context will provide a better understanding of OAS2 role in viral infection.

24 https://mc06.manuscriptcentral.com/bcb-pubs Page 25 of 39 Biochemistry and Cell Biology

REFERENCES

Barkhash, A. V., Kochneva, G. V., Chub, E. V., Mikhailova, S. V., and Romaschenko, A.G. 2014. Association between polymorphisms in OAS2 and CD209 genes and predisposition to chronic hepatitis C in Russian population. Microbes Infect. 16(5): 445–449. doi:10.1016/j.micinf.2014.02.004. Booy, E.P., Meng, H., and McKenna, S.A. 2013. Native RNA purification by gel filtration chromatography. Methods Mol. Biol. 965(1): 1–13. doi:10.1007/978-1-62703-239-1_1. Chambers, T.J., Hahn, C.S., Galler, R., and Rice, C.M. 1990. Flavivirus genome organization, expression, and replication. Annu. Rev. Microbiol. 44: 649–688. doi:10.1146/annurev.mi.44.100190.003245. Chebath, J., Benech, P., Hovanessian, A., Galabru, J., and Revel, M. 1987. Four different forms of interferon-induced 2’,5’-oligo(A) synthetase identified by immunoblotting in human cells. J. Biol. Chem. 262(8): 3852–3857. Deo, S., Patel, T.R., Dzananovic, E., Booy, E.P., Zeid, K., et al. 2014. Activation of 2’ 5’- oligoadenylate synthetase by stem loops at the 5’-end of the west nile virus genome. PLoS One 9(3): e92545. Public Library of Science. doi:10.1371/journal.pone.0092545. Desai, S.Y., and Sen, G.C. 1997. Effects of Varying Lengths of Double-Stranded RNA on Binding and Activation of 2′-5′-Oligoadenylate Synthetase. J. Interf. Cytokine Res. 17(9): 531–536. Mary Ann Liebert, Inc., publishers.Draft doi:10.1089/jir.1997.17.531. Dingwall, C., Ernberg, I., J. Gait, M., Green, S.M., Heaphy, S., et al. 1989. Human immunodeficiency virus 1 tat protein binds trans-activation-responsive region (TAR) RNA in vitro. Proc. Natl. Acad. Sci. U. S. A. 86(18): 6925–9. doi:10.1073/pnas.86.18.6925. Dong, B., and Silverman, R.H. 1997. A bipartite model of 2-5A-dependent RNase L. J. Biol. Chem. 272(35): 22236–22242. doi:10.1074/jbc.272.35.22236. Donovan, J., Dufner, M., and Korennykh, A. 2013. Structural basis for cytosolic double-stranded RNA surveillance by human oligoadenylate synthetase 1. Proc. Natl. Acad. Sci. U. S. A. 110(5): 1652–7. doi:10.1073/pnas.1218528110/- /DCSupplemental.www.pnas.org/cgi/doi/10.1073/pnas.1218528110. Donovan, J., Whitney, G., Rath, S., and Korennykh, A. 2015. Structural mechanism of sensing long dsRNA via a noncatalytic domain in human oligoadenylate synthetase 3. Proc. Natl. Acad. Sci. 112(13): 201419409. doi:10.1073/pnas.1419409112. Friebe, P., and Harris, E. 2010. Interplay of RNA elements in the dengue virus 5’ and 3’ ends required for viral RNA replication. J. Virol. 84(12): 6103–18. doi:10.1128/JVI.02042-09. Gasteiger, E., Hoogland, C., Gattiker, A., Duvaud, S., Wilkins, M.R., et al. 2005. Protein Identification and Analysis Tools on the ExPASy Server. In The Proteomics Protocols Handbook. doi:10.1385/1-59259-890-0:571. Han, Y., Whitney, G., Donovan, J., and Korennykh, A. 2012. Innate Immune Messenger 2-5A Tethers Human RNase L into Active High-Order Complexes. Cell Rep. 2(4): 902–913. The Authors. doi:10.1016/j.celrep.2012.09.004. Hartmann, R., Justesen, J., Sarkar, S.N., Sen, G.C., and Yee, V.C. 2003. Crystal Structure of the 2 -Specific and Antiviral Protein 2 -5 -Oligoadenylate Synthetase. Structure 19: 1173–1185. Hartmann, R., Nørby, P.L., Martensen, P.M., Jørgensen, P., James, M.C., et al. 1998. Activation of 2’-5’- Oligoadenylate Synthetase by Single-stranded and Double-stranded RNA Aptamers. Biochemistry 273(6): 3236–3246. doi:10.1074/jbc.273.6.3236. Hovanessian, A. 2007. On the discovery of interferon-inducible, double-stranded RNA activated enzymes: The 2’-5’ oligoadenylate synthetases and the protein kinase PKR. Cytokine 25 https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 26 of 39

Growth Factor Rev. 18(5–6): 351–361. doi:10.1016/j.cytogfr.2007.06.003. Hovanessian, A., and Justesen, J. 2007. The human 2′-5′oligoadenylate synthetase family: Unique interferon-inducible enzymes catalyzing 2′-5′ instead of 3′-5′ phosphodiester bond formation. Biochimie 89(6–7): 779–788. doi:10.1016/j.biochi.2007.02.003. Hovanessian, A.G., Laurent, A.G., Chebath, J., Galabru, J., Robert, N., et al. 1987. Identification of 69-kd and 100-kd forms of 2-5A synthetase in interferon-treated human cells by specific monoclonal antibodies. EMBO J. 6(5): 1273–80. Available from http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=553929&tool=pmcentrez&rende rtype=abstract. Hovnanian, A., Rebouillat, D., Mattei, M.G., Levy, E.R., Marié, I., et al. 1998. The human 2’,5’- oligoadenylate synthetase locus is composed of three distinct genes clustered on chromosome 12q24.2 Encoding the 100-, 69-, and 40-kDa forms. Genomics 52(3): 267–277. doi:10.1006/geno.1998.5443. Ibsen, M.S., Gad, H.H., Thavachelvam, K., Boesen, T., Despres, P., et al. 2014. The 2’-5’- Oligoadenylate Synthetase 3 Enzyme Potently Synthesizes the 2’-5’-Oligoadenylates Required for RNase L Activation. J. Virol. 88(24): 14222–14231. doi:10.1128/JVI.01763- 14. Justesen, J., Hartmann, R., and Kjeldgaard, N.O. 2000. Gene structure and function of the 2’-5’- oligoadenylate synthetase family. Cell. Mol. Life Sci. 57(11): 1593–1612. doi:10.1007/PL00000644. Kodym, R., Kodym, E., and Story, M.D.Draft 2009. 2’-5’-Oligoadenylate synthetase is activated by a specific RNA sequence motif. Biochem. Biophys. Res. Commun. 388(2): 317–322. Elsevier Inc. doi:10.1016/j.bbrc.2009.07.167. Li, Y., Banerjee, S., Wang, Y., Goldstein, S.A., Dong, B., et al. 2016. Activation of RNase L is dependent on OAS3 expression during infection with diverse human viruses. : 3–8. doi:10.1073/pnas.1519657113. Lin, R.-J., Yu, H.-P., Chang, B.-L., Tang, W.-C., Liao, C.-L., et al. 2009. Distinct antiviral roles for human 2’,5’-oligoadenylate synthetase family members against dengue virus infection. J. Immunol. 183(12): 8035–8043. doi:10.4049/jimmunol.0902728. Malathi, K., Dong, B., Gale, M., and Silverman, R.H. 2007. Small self-RNA generated by RNase L amplifies antiviral innate immunity. Nature 448(7155): 816–819. doi:10.1038/nature06042. Manche, L., Green, S.R., Schmedt, C., and Mathews, M.B. 1992. Interactions between double- stranded RNA regulators and the protein kinase DAI. Mol. Cell. Biol. 12(11): 5238–48. doi:10.1128/MCB.12.11.5238. Marié, I., Rebouillat, D., and Hovanessian, A.G. 1999. The expression of both domains of the 69/71 kDa 2’,5’ oligoadenylate synthetase generates a catalytically active enzyme and mediates an anti- viral response. Eur. J. Biochem. 262(1): 155–165. doi:10.1046/j.1432- 1327.1999.00361.x. Mathews, M.B., and Shenk, T. 1991. Adenovirus virus-associated RNA and translation control. J. Virol. 65(11): 5657–5662. McKenna, S.A., Kim, I., Liu, C.W., and Puglisi, J.D. 2006. Uncoupling of RNA Binding and PKR Kinase Activation by Viral Inhibitor RNAs. J. Mol. Biol. 358(5): 1270–1285. doi:10.1016/j.jmb.2006.03.003. McKenna, S.A., Kim, I., Puglisi, E.V., Lindhout, D.A., Aitken, C.E., et al. 2007. Purification and characterization of transcribed rnas using gel filtration chromatography. Nat. Protoc. 2(12): 3270–3277. doi:10.1038/nprot.2007.480. Meng, H., Deo, S., Xiong, S., Dzananovic, E., Donald, L.J., et al. 2012. Regulation of the 26 https://mc06.manuscriptcentral.com/bcb-pubs Page 27 of 39 Biochemistry and Cell Biology

interferon-inducible 2’-5’-oligoadenylate synthetases by adenovirus VA(I) RNA. J. Mol. Biol. 422(5): 635–49. doi:10.1016/j.jmb.2012.06.017. Morin, B., Rabah, N., Boretto-Soler, J., Tolou, H., Alvarez, K., et al. 2010. High yield synthesis, purification and characterisation of the RNase L activators 5’-triphosphate 2’-5’- oligoadenylates. Antiviral Res. 87(3): 345–352. Elsevier B.V. doi:10.1016/j.antiviral.2010.06.003. Muesing, M.A., Smith, D.H., and Capon, D.J. 1987. Regulation of mRNA accumulation by a human immunodeficiency virus trans-activator protein. Cell 48(4): 691–701. doi:10.1016/0092-8674(87)90247-9. Nallagatla, S.R., Toroney, R., and Bevilacqua, P.C. 2008. A brilliant disguise for self RNA: 5’- end and internal modifications of primary transcripts suppress elements of innate immunity. RNA Biol. 5(3): 140–4. doi:10.4161/rna.5.3.6839. Ng, W.C., Soto-Acosta, R., Bradrick, S.S., Garcia-Blanco, M.A., and Ooi, E.E. 2017. The 5ʹ and 3ʹ untranslated regions of the flaviviral genome. Viruses 9(6): 1–14. doi:10.3390/v9060137. Owczarzy, R., Tataurov, A. V, Wu, Y., Manthey, J.A., McQuisten, K.A., et al. 2008. IDT SciTools: a suite for analysis and design of nucleic acid oligomers. Nucleic Acids Res. 36(Web Server issue): 163–169. doi:10.1093/nar/gkn198. Player, M.R., and Torrence, P.F. 1998. The 2-5A system: Modulation of viral and cellular processes through acceleration of RNA degradation. doi:10.1016/S0163-7258(97)00167-8. Putnins, R.F., and Yamada, E.W. 1975. Colorimetric determination of inorganic pyrophosphate by a manual or automated method.Draft Anal. Biochem. 68(1): 185–195. doi:10.1016/0003- 2697(75)90693-4. Rebouillat, D., Marié, I., and Hovanessian, A.G. 1998. Molecular cloning and characterization of two related and interferon- induced 56-kDa and 30-kDa proteins highly similar to 2’-5’ oligoadenylate synthetase. Eur. J. Biochem. 257(2): 319–330. doi:10.1046/j.1432- 1327.1998.2570319.x. Sarkar, S.N., Bandyopadhyay, S., Ghosh, A., and Sen, G.C. 1999a. Enzymatic characteristics of recombinant medium isozyme of 2’-5’ oligoadenylate synthetase. J. Biol. Chem. 274(3): 1848–1855. doi:10.1074/jbc.274.3.1848. Sarkar, S.N., Ghosh, A., Wang, H.W., Sung, S.S., and Sen, G.C. 1999b. The Nature of the Catalytic Domain of 2’-5’-Oligoadenylate Synthetase. 274(36): 25535–25542. Sharp, T. V., Raine, D.A., Gewert, D.R., Joshi, B., Jagus, R., et al. 1999. Activation of the interferon-inducible (2’-5’) oligoadenylate synthetase by the Epstein-Barr virus RNA, EBER-1. Virology 257(2): 303–313. doi:10.1006/viro.1999.9689. Thamizhmani, R., and Vijayachari, P. 2014. Association of dengue virus infection susceptibility with polymorphisms of 2’-5’-oligoadenylate synthetase genes: a case-control study. Braz. J. Infect. Dis. 18(x x): 43–45. Elsevier Editora Ltda. doi:10.1016/j.bjid.2014.03.004. Thimmappaya, B., Weinberger, C., Schneider, R.J., and Shenk, T. 1982. Adenovirus VAI RNA is required for efficient translation of viral mRNAs at late times after infection. Cell 31(3 PART 2): 543–551. doi:10.1016/0092-8674(82)90310-5. Torralba, S., Sojat, J., and Hartmann, R. 2008. 2′-5′ Oligoadenylate synthetase shares active site architecture with the archaeal CCA-adding enzyme. Cell. Mol. Life Sci. 65(16): 2613–2620. doi:10.1007/s00018-008-8164-5. Vachon, V.K., Calderon, B.M., and Conn, G.L. 2015. A novel RNA molecular signature for activation of 2’-5’ oligoadenylate synthetase-1. Nucleic Acids Res. 43(1): 544–552. doi:10.1093/nar/gku1289. Whyard, S., Erdelyan, C.N.G., Partridge, A.L., Singh, A.D., Beebe, N.W., et al. 2015. Silencing the buzz : a new approach to population suppression of mosquitoes by feeding larvae double- 27 https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 28 of 39

stranded RNAs. Parasites and Vectors 8(96): 1–11. doi:10.1186/s13071-015-0716-6. Wilkins, M.R., Gasteiger, E., Bairoch, A., Sanchez, J.C., Williams, K.L., et al. 1999. Protein identification and analysis tools in the ExPASy server. Methods Mol. Biol. Wreschner, D.H., McCauley, J.W., Skehel, J.J., and Kerr, I.M. 1981. Interferon action--sequence specificity of the ppp(A2’p)nA-dependent ribonuclease. Nature 289(5796): 414–417. doi:10.1038/289414a0. Zhao, M., Wan, B., Li, H., He, J., Chen, X., et al. 2017. Porcine 2′, 5′-oligoadenylate synthetase 2 inhibits porcine reproductive and respiratory syndrome virus replication in vitro. Microb. Pathog. 111: 14–21. Elsevier Ltd. doi:10.1016/j.micpath.2017.08.011. Zheng, S., Zhu, D., Lian, X., Liu, W., Cao, R., et al. 2015. Porcine 2′, 5′-oligoadenylate synthetases inhibit Japanese encephalitis virus replication in vitro. J. Med. Virol. (March 2015Zheng, Sheng): n/a-n/a. doi:10.1002/jmv.24397. Zhou, A., Hassel, B.A., and Silverman, R.H. 1993. Expression cloning of 2-5A-dependent RNAase: A uniquely regulated mediator of interferon action. Cell 72(5): 753–765. doi:10.1016/0092-8674(93)90403-D. Zhou, Y., Kang, M.-J., Jha, B.K., Silverman, R.H., Lee, C.G., et al. 2013. Role of Ribonuclease L in Viral Pathogen-Associated Molecular Pattern/Influenza Virus and Cigarette Smoke- Induced Inflammation and Remodeling. J. Immunol. 191(5): 2637–2646. doi:10.4049/jimmunol.1300082. Draft

28 https://mc06.manuscriptcentral.com/bcb-pubs Page 29 of 39 Biochemistry and Cell Biology

Table 1. Kinetic parameters determined for OAS1 activation with ss-dsRNA. OAS1 + ss-dsRNA

RNA Kapp Vmax (nM) (µM/min) 43bp dsRNA 87 ±13 6.6 ±0.3 73bp dsRNA 72 ±11 6.6 ±0.5 83bp dsRNA 42 ±7 7.0 ±0.5 93bp dsRNA 48 ±10 9.0 ±0.7 123bp dsRNA 27 ±3 8.0 ±0.2 40bp dsRNA (S) 362 ±67 5.0 ±0.4 35bp dsRNA (S) 258 ±48 4.0 ±0.3 30bp dsRNA (S) 415 ±101 4.0 ±0.3 25bp dsRNA (S) 613 ±156 5.0 ±0.6 Poly (I:C) * 35 ±5 7.7 ±0.3

Table 2. Kinetic parameters determined for OAS2 activation with ss-dsRNA and U-dsRNA. OAS2 +Draft ss-dsRNA OAS2 + U-dsRNA RNA Kapp Vmax Kapp Vmax (nM) (µM/min) (nM) (µM/min) 43bp dsRNA 217 ±62 1.4 ±0.2 202 ±78 0.6 ±0.1 53bp dsRNA 161 ±17 2 ±0.1 127 ±14 1.0 ±0.1 63bp dsRNA 147 ±43 3.3 ±0.5 120 ±32 2.4 ±0.3 73bp dsRNA 118 ±22 4.6 ±0.5 107 ±26 2.6 ±0.3 83bp dsRNA 108 ±22 5 ±0.5 100 ±25 3.3 ±0.4 93bp dsRNA 104 ±23 6 ±0.8 70 ±13 3.7 ±0.3 123bp dsRNA 53 ±11 6.3 ±0.6 50 ±12 3.7 ±0.4 40bp dsRNA (S) 613 ±181 1.7 ±0.3 N/A N/A 35bp dsRNA (S) 1211 ±388 0.5 ±0.1 N/A N/A 30bp dsRNA (S) N/A N/A N/A N/A 25bp dsRNA (S) N/A N/A N/A N/A Poly (I:C) * 36 ±6 5.6 ±0.3 N/A N/A

29 https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 30 of 39

FIGURE LEGENDS

Figure 1. Purification of recombinant human oligoadenylate synthetase 2 (OAS2). (A). Denaturing SDS-PAGE (10%) analysis of human OAS2 purification from HEK293T cells. Lanes (left to right) molecular weight marker, lysate, FLAG affinity column wash, elution from FLAG affinity column and purified protein after size exclusion chromatography are indicated. (B). Elution profile of OAS2 from SEC showing protein purity. A260 (dotted line) and A280 (solid line) are shown. (C). SEC-MALS showing molecular mass from light scattering (solid line) as a function of elution volume for OAS2. Refractive index (dashed line) and UV (dotted line) were used as concentration detectors.

Figure 2. Detection of 2-5A by mass spectrometry. (A). Denaturing gel electrophoresis (10%) was run at 200V for 20 minutes, stained with SYBR gold for 10 minutes to detect RNA-containing species. To generate 2-5A for mass spectrometry, OAS2 (300nM) and dsRNA (123 bp, 20 µg/mL) were incubated in the presence of ATP and MgCl2 at 37 °C, quenched, and passed through a spin concentrator (10 kDa cutoff, rightmost lane). The lanes from left to right correspond to: 25bp marker, 40bp marker, activating RNA alone (123 bp) alone, ATP alone, activation reaction lacking OAS2, activation reaction lacking RNA, activation reaction prior to spin concentration, activation reaction lacking ATP, and the sample after spin concentration for mass spectrometry. (B, C, D) LC- MS/MS spectra showing 2-5A. 2 μL of theDraft sample passed through the spin concentrator in (A) was injected onto a reverse phase column, and mass spectra in the negative mode were collected by compact QqTOF LC-MS/MS. (B) 2-5A dimer was detected at 836 Da [(417.0318 Da ×2)+2]; (C) 2-5A trimer was detected at 1165.1 Da [(581.5573 Da ×2)+2]; (D) 2-5A tetramer was detected at 1494.1 Da [(497.0561 Da ×3)+3]. A control reaction minus OAS2 was simultaneously performed and no 2-5A was detected.

Figure 3. Comparative kinetic analysis of OAS2 and OAS1 activity. (A). Enzyme activity (Vo) after 20 min reaction shown as a function of poly (I:C) concentration at increasing OAS2 concentration (shown in nM at endpoint of curves) or (B) OAS2 concentration at increasing poly (I:C) concentration (shown in µg/mL at endpoint of curves). Curves were fit using nonlinear regression. Error bars represent the standard deviation of three replicates. (C) Purified OAS1 or OAS2 (300 nM) in presence and absence of poly (I:C) (20 µg/mL) were incubated with ATP (2 mM) and MgCl2 (5 mM) at 37°C and quenched at time points (0-180) min in order to assay for the production of 2’-5’ (A) chain formation. Error bars represent the standard deviation from three replicates. (D) Kinetic analysis of OAS2 activity as a function of poly (I:C) concentration. OAS2 activity was determined by calculating initial rates in triplicate at the specified poly (I:C) concentrations (see Materials & Methods for details).

Figure 4. Purification of OAS2 domain I and II from HEK293T cells. (A) Schematic of OAS2 DI and DII adjacent to the full length OAS2 protein. Active site aspartic acids are indicated in OAS1 and OAS2 DII. (B) Denaturing SDS-PAGE gel (10%) of full length purified OAS1, OAS2 enzymes and OAS2 domains stained with Coomassie Blue. Protein was loaded on the SDS-PAGE gel and run at 200 V for 40 min. (C). Catalytic activity of OAS proteins (300 nM) incubated with poly (I:C) (20 µg/mL) in a buffered solution with ATP and MgCl2 at 37 °C. Reactions were quenched after 1hr. and assayed for the production of PPi. Error bars represent the standard deviation from triplicate experiments.

30 https://mc06.manuscriptcentral.com/bcb-pubs Page 31 of 39 Biochemistry and Cell Biology

Figure 5. Enzyme activity of OAS2 and OAS1 with perfectly duplexed synthetic dsRNA. (A) OAS1 (300nM), synthetic 19bp, 25bp, 30bp, 35bp, 40bp dsRNA (300nM) and poly (I:C) (20ug/mL) were incubated in presence of ATP and MgCl2 at 37 °C and quenched at time points. Error bars represent the standard deviation from three replicates of PPi quantification. (B) Identical experiment as in (A) were performed for OAS2. Kinetic analysis of OAS1 (C) and OAS2 (D) activity as a function of synthetic dsRNA concentration. Activity was determined by calculating initial rates in triplicate at the specified dsRNA concentrations (see Materials & Methods for details).

Figure 6. Characteristics of dsRNA that lead to catalytic activation of OAS1/OAS2. (A) Schematic representation of in vitro transcribed 43bp dsRNA with either a 17 nt 3’-overhang (ss- dsRNA) or a 3’-monouracil (U-dsRNA; indicated by arrows where RNAse A cleavage occurs). (B) ss-dsRNA (red) of indicated length was used to activate OAS1. OAS1 (300nM) and either ss- dsRNA (red), dephosphorylated ss-RNA (green), U-dsRNA (purple), or dephosphorylated U- dsRNA (orange) (10ug/mL) were incubated at 37°C in presence of ATP and MgCl2 and assayed for PPi after 1 hr. Data is plotted relative to activation of OAS1 by poly (I:C). Error bars represent the standard deviation of three replicates. (C). Identical experiments were performed as in (B) for OAS2 activation. Statistical analysis (t test) was performed comparing either (i) ss-dsRNA to U- dsRNA or (ii) ss-dsRNA (5’OH) to U-dsRNA (5’-OH) for all dsRNA lengths simultaneously. p values are shown on top of columns. Draft

Figure 7. Kinetic analysis of OAS2 activation by ss-dsRNA and U-dsRNA. (A). Initial rate (Vo) determination for OAS2 (400nM) over a range of ss-dsRNA concentration. OAS2 activity was determined by calculating initial rates (Vo) in triplicate at the specified RNA concentrations (see Materials & Methods for details). Error bars represent the variation in nonlinear regression analysis. (B) OAS1 (400 nM) activation over a range of ss-dsRNA lengths. (C) Identical experiments were performed as in (B) for OAS2 activation. (D) Denaturing TBE RNA gel (10%) showing the purity of in vitro transcribed dsRNA after RNAse A cleavage. 1.5 µg of RNA was mixed with the RNA denaturing load dye (Materials and Methods), heated for 5 minutes at 95 °C and run on the gel at 120 V for 45 min. (E) OAS2 (400nM) activation over a range of U-dsRNA lengths. OAS2 activity was determined by calculating initial rates in triplicates at specified RNA concentration. Error bars represent the variation in nonlinear regression analysis.

Figure 8. Comparison of OAS1/OAS2 activation by in vitro transcribed short viral dsRNAs. (A) Denaturing TBE gel (10%) showing the purity of viral dsRNA. 2 µg of each RNA was loaded on a denaturing TBE gel and run at 120 V for 45 minutes and detected by toluidine blue staining. (B) OAS1 (300 nM) was incubated with viral RNA (300nM) in presence of ATP and MgCl2 at 37 °C, quenched at specific time points and assayed for the PPi production. Error bars represent the standard deviation from three replicates. (C) OAS2 (300 nM) was incubated with viral RNA (300nM) in presence of ATP and MgCl2 at 37 °C, quenched at specific time points and assayed for the PPi production. Error bars represent the standard deviation from three replicates.

31 https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 32 of 39

Draft

https://mc06.manuscriptcentral.com/bcb-pubs Page 33 of 39 Biochemistry and Cell Biology

Draft

https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 34 of 39

Draft

https://mc06.manuscriptcentral.com/bcb-pubs Page 35 of 39 Biochemistry and Cell Biology

Draft

https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 36 of 39

Draft

https://mc06.manuscriptcentral.com/bcb-pubs Page 37 of 39 Biochemistry and Cell Biology

Draft

https://mc06.manuscriptcentral.com/bcb-pubs Biochemistry and Cell Biology Page 38 of 39

Draft

https://mc06.manuscriptcentral.com/bcb-pubs Page 39 of 39 Biochemistry and Cell Biology

Draft

https://mc06.manuscriptcentral.com/bcb-pubs