key: cord-1046202-n9bhrnpv
authors: Herbert, Alan; Shein, Alexander; Poptsova, Maria
title: Z-RNA and the flipside of the SARS Nsp13 helicase
date: 2022-03-03
journal: bioRxiv
DOI: 10.1101/2022.03.03.482810
sha: a54fb20a05976d19b5273e6eebb0afc9a0704d4b
doc_id: 1046202
cord_uid: n9bhrnpv

We present evidence that the severe acute respiratory syndrome coronavirus (SARS) non-structural protein 13 (Nsp13) modulates the Z-RNA dependent regulated cell death pathway of necroptosis (1). We show that Z-prone sequences (called flipons (2)) exist in coronavirus and provide a signature that enables identification of the animal viruses which have become human pathogens. We also identify a potential RHIM in Nsp13. These two observations allow us to suggest a model in which Nsp13 may regulate Z-RNA-initiated RHIM-dependent cell death outcomes at two steps. The first step involves possible new ATP-independent Z-flipon helicase activity in Nsp13, which is distinct from the activity of the canonical A-RNA helicase. This activity unwinds/quenches nascent Z-RNAs, preventing their sensing by ZBP1. The second step involves RHIM-dependent inhibition of ZBP1, RIPK3 and/or RIPK1, preventing cell death downstream of Z-RNA sensing. Together the RHIM and Z-flipon helicase have the potential to alter the host response to the virus and the effectiveness of drugs targeting the NSP13 helicase.

to adopt the Z-conformation can be scored using the ZHUNT3 (ZH3) program (4) . Much of the energy 9 cost for forming Z-DNA and Z-RNA is in the creation of junctions between right-handed and left-handed 10 helices. For Z-RNA, the flip occurs more easily when dsRNA contain basepair mismatches, non-canonical 11 basepairs or unpaired residues at both ends of the Z-forming segment (5) . 12 Z-RNA and innate immunity. Z-DNA and Z-RNA are recognized in a conformation specific manner by 13 protein containing the Zα domain that was first discovered in the dsRNA editing enzyme ADAR1 (6), but 14 is also found in many other proteins, including Z-DNA binding protein (ZBP1) and virally encoded 15 proteins like E3 that is encoded by vaccinia virus (7) . Interaction between Zα proteins regulates the 16 regulated necroptotic cell death pathway (8, 9) . The pathway is triggered when ZBP1 binds to Z-DNA or 17 Z-RNA and activates receptor interacting protein kinases (RIPKs) 1 and 3, through the RIP homotypic 18 interaction motif (RHIM) shared by all three proteins. RIPK3 then phosphorylates Mixed Lineage Kinase 19 Domain Like (MLKL) pseudokinase, leading to necroptosis, while RIPK1 induces Caspase-8-dependent 20 apoptosis, via the adaptor protein FADD. The RHIM is a ~40 aa motif, first identified in RIPK1 (10), which 21 contains an (I/V)Q(I/L/V)G sequence at its core (11) . The pathway plays an important role during 22 infection by the negative RNA stranded influenza virus (12) , as well as upon infection with the 23 herpesviruses murine cytomegalovirus (mCMV) and herpes simplex virus (HSV)-1/2 and the poxvirus 24 vaccinia virus (13) . 25

Viruses and ZBP1 dependent necroptosis. Viruses that are prone to form Z-DNA and Z-RNA have 26 developed strategies to regulate ZBP1-dependent necroptosis. These include encoding ZBP1 homologs, 27 such as E3, that compete with ZBP1 for Z-RNA (14) . Another strategy by which viruses regulate 28 ZBP1/RIPK signaling is by encoding proteins with RHIMs. So far, only RHIMs produced by DNA viruses 29 like the Herpesviridae mCMV, HSV-1 and HSV-2 are known to play an important role in virulence (15) . 30

Notably, no RNA virus has yet been shown to encode a RHIM-containing protein. 31

A corona virus-specific Z-flipon signature. We were interested in whether the SARS family 33 coronaviruses might also regulate Z-RNA dependent ZBP1 activation and cell death. We used a two-34 pronged approach. First, we searched for Z-prone sequences in coronavirus using the program ZH3. We 35 were interested in those sequences that altered as the virus adapted to humans. We found that all 36 examined coronaviruses contained sequences that were Z-prone, and that the Z-signature for each 37 strain of virus is unique (Figure 1 ).Using the signature, it is easy to identify the host animal from which 38 the pathogenic human viruses arose. For example, we show that the signatures for Llama coronavirus 39 conditions. While not experimentally calibrated for the formation of Z-RNA, the scores observed for the 45 CoVs are far in excess of this value and it likely that the sequences in these viruses are flipons. With this caveat in mind, we observed the loss in the 1ab transcript of a strong Z-forming sequence when comparing SARS1 to SARS2. The reduced ZH3 score in SARS2 is due to three synonymous mutation in 48 adjacent codons that preserve a hairpin sRNA structure and the peptide sequence PARAR (residues 335-49 339 of SARS2 Nsp13). This arginine rich sequence has the potential to bind nucleic acids (figure 2A and 50 B) (16) . Two Z-prone sequences were present in the equivalent region of MERS Nsp13. The MERS 51 peptide, PAKAR, has a lysine in place of the first arginine, a conservative substitution. 52

A putative RHIM associated with coronaviruses severely pathogenic in humans. Second, we searched 53 for viral protein containing a RHIM and found that the prototypical VQIG sequence was present in the 54 Nsp13 protein of highly pathogenic human CoVs, but not in other CoVs thought to infect humans ( Figure  55 2 D and E). 56

The structure of NSP13 helicase. Both strategies focused our attention of the Nsp13 helicase and its 57 potential novel role as a Z-flipon helicase. Nsp13 unwinds both DNA and RNA in the 5'->3' direction. It is 58 a member of the helicase superfamily 1B (SF1B) (17) . Along with other CoV encoded proteins, it forms a 59 viral replication-transcription complex. High resolution crystallographic and electron-microscopy 60 structures of NSP13 have recently been published and suggests that a second Nsp13 cooperates with 61 the first to enhance translocation along the CoV genomic RNA (18) (19) (20) . The Nsp13 helicase has two 62 canonical RecA domains between which ATP is bound and through which single-stranded RNA egresses 63 from the complex. Nsp1 also has three N-terminal domains that are unique to nidovirus helicases. The 64 stalk domain connects a zinc-binding (ZincBD) domain to the 1B domain ( Figure 3) . 65

The Nsp13 RHIM is present in the 1B domain, which bridges the RecA1 and RecA2 domains. Rather 66 surprisingly, the peptide PARA motif, which is in the RecA1 domain, interacts with the RHIM when the 67

Nsp13 is in the ATP-free closed conformation ( Figure 3A ). In the active conformation, the RHIM and 68

PARA separate to open a channel through which the single-stranded RNA passes ( Figure 3B ). Further separation of the 1B domain from the RecA2 domain renders Nsp13 unable to translocate on RNA (20) . 70

The cavity created by this separation appears large enough to accommodate a dsRNA structure (Figure  71 3C). 72

The cavity is lined by the residues known to bind single-stranded RNA (ssRNA) (shown by white space fill 74 carbons in figure 4) , both to the phosphate backbone and also in a base-specific manner. Into the newly 75 formed cavity projects the PARA peptide, with R337 and R339 forming a hook. Also, tyrosine Y205, along 76 with W178, faces the interior of the cavity. Potentially Y205 and W178 could engage each other in an 77 edge to face configuration similar to that of Y177 and W195 in the Zα domain and enable the 78 conformation-specific recognition of Z-RNA. 79

Novel Nsp13 flipon helicase activity. Docking Nsp13 to Z-RNA in silico leads us to propose a model 80

where Nsp13 acts as a flipon helicase, preventing Z-RNA formation by capturing single-stranded RNA 81 formed during the flip to and from A-RNA. The strand separation generated is powered by the free 82 energy stored in Z-RNA rather than through ATP hydrolysis. The process of strand capture would be 83 enhanced if Nsp13 had specificity for Z-prone sequences. Indeed, the recent Chen at al structures detail 84 the binding of Nsp13 to an alternating pyrimidine/purine sequence CAUGU substrate (20) . Recognition 85 of Z-RNA by Nsp13leading to strand separation would then prevent binding by ZBP1 and activation of 86

Binding to Z-RNA would also lead to the exposure of the Nsp13 RHIM, which is on the other side of the 88 1B domain ( Figure 3C ). Once the RHIM is sprung lose, it is free to contact ZBP1, RIPK1 and RIPK3 and 89 modulate their activities. 90 Disease Implications .At this stage, we only know of the possibilities and hope that the above 92 hypotheses will provide a framework for further investigation of Nsp13 effects on necroptosis in 93 properly qualified BSL3 laboratories. If it is found that pathogenic CoVs do indeed produce Z-RNA (as the 94 ZH3 algorithm predicts) and that the RHIM in Nsp13 regulates ZBP1, RIPK3 and/or RIPK1 signaling, then 95 we suggest that effects will depend on the Coronavirus and Human Pathology. From our analysis, it seems that the RHIM domain is a requirement 105 for CoV to cause severe pathology in humans. Acquiring a RHIM may be necessary for the successful 106 jump the virus made from its natural host to humans, although not a sufficient one as other mutations 107 that enable engagement of human cell surface receptors are required. The mutations that lead to loss of 108 the strong Z-prone sequence present in SARS1 may have favored spread of SARS2 (Figure 2A and 2B) . 109

The potential combination of the RHIM domain with the Z-flipon helicase to modulate the cell death 110

pathways so far appears unique among human viruses and may account for its extreme virulence. 111

The high frequency in humans of protein variants in the necroptotic pathway that affect its function (22) 112 leave some individuals more vulnerable to SARS2 induced cell death. While Nsp13 inhibitors that target 113 the ATP-binding site are likely to be effective in early infection by inhibiting the classical helicase function, they are unlikely to be effective against the ATP-independent Z-flipon helicase clusters that 115 potentially produce severe pathology at later stages of the disease. 116

The Z-forming potential of the Coronavirus sequences were assessed using ZHUNT3 using the following 118 command line "zhunt3nt 12 6 MP where AS is a student who identified the strong Z-forming segment in SARS1. AH conceptualized and 126

wrote the paper with edits from MP. 127 

Nsp13 and Z-dependent necroptosis A. The high scoring ZH3 peak in SARS (NC_004718.3) maps to Nsp13 and encodes an arginine rich peptide with potential to bind to Z-RNA. Potential Z-RNA forming sequences are highlighted in yellow and form the Z-RNA stem within the red box identified using RNAfold (23) . B. In SARS2 (NC_045512.2), three non-synonymous mutations conserve the Z-RNA stem and the peptide sequence, but diminish the Z-forming potential C. The equivalent Nsp13 region in MERS has a different peptide sequence. The Z-RNA forming element is encoded immediately after this sequence block, with another Z-RNA element close-by (in dashed box) D. Phylogram showing the evolutionary distance between coronavirus strains with high (in red) and low (in blue) human pathogenicity. E. The sequence aligner MUSCLE (24) reveals that highly pathogenic coronaviruses have a conserved RHIM domain in NSP13 (in red)

The interaction between the Nsp13 RHIM (residues 193-6) and tPARAR (residues 335-9) are conformation dependent. A. the ATP-free apo form of Nsp13, the RHIM (yellow space fill carbons) contacts PARAR339 (crimson space fill carbons) B. In the active state, RHIM and PARAR separate, opening the 5´ end of the single-strand RNA channel, that has the 3´end marked by N516 (white space fill carbons). C. The open complex in which the RHIM separates from the RecA2 domains to create a cavity that is large enough to accommodate dsRNA. This opening is associated with a rotation of the Zinc binding domain (ZincBD) relative to N516. PARAR also rotates, changing the position of R337 and R339. The structures are from PDB files 7NIO (18), 6XEZ (19) and 7RDX (20) as labeled, with images rendered using the NGLViewer (25).

Nsp13 opens to expose a hook and platen. A The arginine rich PARAR hook (crimson carbons) and the tyrosine (Y205, grey carbons) platen create a surface for docking to dsRNA. Tryptophan (W168)(blue carbons) has the potential to orientate a to create a Z-RNA specific recognition element like that present in the Zα domain (26) . The RHIM domain (yellow carbons) is free to engage other RHIM proteins. The space fill with white carbons show the residues identified as making base-specific contacts with single-stranded RNA in the active conformation shown in Figure 3B (residues 178, 179, 230, 233 in domain 1B, residues 311, 335,361, 362:E,363, 390, 408, 410:E in RecA1). C. In addition to the ATP-dependent helicase activity, Nsp13 has the potential to capture single stranded RNA when Z-RNA flips to and from A-RNA. The higher energy Z-RNA powers the strand-separation, providing the ΔG needed to fuel Z-flipon helicase activity.

Molecular mechanisms of cell death: recommendations of the Nomenclature Committee on Cell Death

A Genetic Instruction Code Based on DNA Conformation

Z-DNA and Z-RNA in human disease

Thermogenomics: thermodynamic-based approaches to genomic analyses of DNA structure

Recognition of non-CpG repeats in Alu and ribosomal RNAs by the Z-RNA binding domain of ADAR1 induces A-Z junctions

A Z-DNA binding domain present in the human editing enzyme, doublestranded RNA adenosine deaminase

Zα-domains: At the intersection between RNA editing and innate immunity

Herpes simplex virus suppresses necroptosis in human cells

RIP1/RIP3 binding to HSV-1 ICP6 initiates necroptosis to restrict virus propagation in mice

Identification of a novel homotypic interaction motif required for the phosphorylation of receptor-interacting protein (RIP) by RIP3

Programmed necrosis in the cross talk of cell death and inflammation

Influenza Virus Z-RNAs Induce ZBP1-Mediated Necroptosis

Viral Z-RNA triggers ZBP1-dependent cell death

Vaccinia virus E3 prevents sensing of Z-RNA to block ZBP1-dependent necroptosis

True grit: programmed necrosis in antiviral host defense, inflammation, and immunogenicity

Arginine-rich motifs present multiple interfaces for specific binding by RNA

Mechanism of Nucleic Acid Unwinding by SARS-CoV Helicase

Structure, mechanism and crystallographic fragment screening of the SARS-CoV-2 NSP13 helicase

Structural Basis for Helicase-Polymerase Coupling in the SARS-CoV-2 Replication-Transcription Complex

Ensemble cryo-electron microscopy reveals conformational states of the nsp13 helicase in the SARS-CoV-2 helicase replication-transcription complex. bioRxiv

The Structure of the Necrosome RIPK1-RIPK3 Core, a Human Hetero

Evolutionary Profile for (Host and Viral) MLKL Indicates Its Activities as a Battlefront for Extensive Counteradaptation

ViennaRNA Package 2.0

MUSCLE: multiple sequence alignment with high accuracy and high throughput

NGL Viewer: a web application for molecular visualization

Crystal structure of the Zα domain of the human editing enzyme ADAR1 bound to left-handed Z-DNA

We that Sid Balachandran for suggesting the possibility of RHIM domains in SARS2 128and for helpful edits to the manuscript. 129