Doctor of Philosophy / Department of Diagnostic Medicine/Pathobiology / Ying Fang / Viral papain-like proteases (PLPs)-related proteins have been shown to be actively involved in host innate immunity manipulation and virus replication. In this dissertation, the research were focused on the elucidation of biological roles of nidoviral PLPs-related proteins in innate immunity suppression and viral RNA transcription regulation. Porcine Respiratory and Reproductive Syndrome Virus (PRRSV) is the most prominent swine diseases worldwide. Understanding PRRSV pathogenesis and development efficient vaccines are highly required in swine industry. PRRSV nsp2-related proteins including nsp2 and two ribosomal frameshifting products-nsp2TF and nsp2N share the same N-terminal PLP2 domain. In chapter 2, nsp2TF and nsp2N were demonstrated to be critical for host innate immunity suppression at least through the PLP2 domain-mediated deubiquitination and deISGylation effects. The infection of nsp2TF/nsp2N knockout mutants significantly upregulated antiviral innate immune responses in vitro. Furthermore, manipulating the expression of nsp2TF/nsp2N could enhance innate and adaptive immunity in pigs, providing potential basis for modified live vaccine development. In addition to the PLP2 domain of PRRSV nsp2-related proteins, the biological roles and biochemical nature of the poorly investigated long mysterious PLP2 downstream region was also characterized in Chapter 3. This long unknown region is also shared by nsp2-related proteins. At first, the hyper-phosphorylation nature of nsp2-related proteins was demonstrated. Physical features of this uncharacterized region was then delineated, including two intrinsically disordered hypervariable regions spaced by a structured inter-species conserved domain. One critical phosphorylated residues in the conserved domain were later proved to be of great importance in recombinant virus rescue and subgenomic RNAs accumulation. Collectively, our investigations underline the pleiotropic effects exerted by nsp2-related proteins in virus life cycle and potential contributions with pathogenesis. In Chapter 4, potential functions of PLP encoded by other nidovirus was also investigated. We discovered a unique cross-order recombination event, in which the chimeric picornavirus-enterovirus G expresses the PLP gene, homologous to torovirus (ToV) PLPs. Like other nidoviral PLPs, the recombinant ToV-PLP was proved to be a highly active deubiquinase/deISGylase and potent innate immune antagonist. After PLP knockout, viral fitness is significantly decreased and the suppression on host ubiquitination/ISGylation is largely reduced. Furthermore, host antiviral innate immune responses have been greatly upregulated post PLP knockout mutant infection. Our study underscores potential contributions of PLP domain in viral pathogenicity, and further provides an ideal example for how recombination shapes virus evolution. In summary, studies in this dissertation highlight the critical roles of nidoviral PLPs-related proteins in host immunity manipulation and virus replication, and more importantly, potential links with viral pathogenicity and application in vaccine development.

<p>Coronaviruses (CoVs) are single-stranded, positive-sense RNA viruses in the Coronaviridae family. Within this family are four different genera, Alpha-, Beta-, Gamma-, and Deltacoronaviruses with human-infecting CoVs spanning the Alpha- and Beta-CoV genera. Most notably, Severe Acute Respiratory Syndrome Coronavirus-1 (SARS-CoV-1) and SARS-CoV-2 are Betacoronaviruses that spread worldwide in their outbreaks from 2002-2003 (SARS-CoV-1) and 2019-2020 (SARS-CoV-2). Human-infecting Alphacoronaviruses, NL63-CoV and 229E-CoV, have caused milder infections involving respiratory disease, gastroenteritis, and in more severe cases, death. Despite milder disease, Alphacoronaviruses are the cause of 15-30% of severe upper and lower respiratory tract infections each year. There have been recent efforts in the development of potent, small-molecule inhibitors to treat SARS-CoV-2 infection but there is an ongoing need to develop new and effective anti-coronavirus therapeutics to treat other human-infecting CoVs circulating society. Coronaviruses encode two essential proteases, the papain-like protease (PLP) and the 3C-like protease. PLPs are cysteine proteases located in non-structural protein 3 (nsp3). PLPs processes the viral polyprotein, releasing the first three nonstructural proteins encoded in the virus, and also are involved in evading the innate immune response through deubiquitinating (DUB) and deISGylating activity. </p> <p><br></p> <p>This study compares the substrate specificity and catalytic function of multiple human-infecting PLPs from both Alpha- and Beta-CoVs including NL63-CoV PLP2, 229E-CoV PLP2, Canine-CoV PLP2, FIPV-CoV PLP2, PEDV-CoV PLP2, SARS-CoV-1 PLpro, and SARS-CoV-2 PLpro. Interestingly, Alphacoronavirus PLP2s have a >400-fold greater catalytic efficiency for ubiquitin compared to Betacoronaviruses PLpro. This work also identifies a non-covalent scaffold of inhibitors that has pan-CoV inhibition; however, the IC50 values are >30-fold higher for NL63-CoV PLP2 than for SARS-CoV-1 PLpro. The X-ray structures of NL63 PLP2 and 229E PLP2 were determined to 2.1 Å and 1.8 Å, respectively, and provide structural information about the substrate and inhibitor binding region that could be the result in the differences in Alpha- and Betacoronavirus PLP function. Since PLP does not function as a single-domain in vivo, it is critical to understand the function of PLP when tethered to other domains of nsp3. This study also investigates nine different constructs of SARS-CoV-2 nsp3 with increasing domains, ranging from the single PLpro domain to Ubl1-Ydomain ΔTM1-TM2. Interestingly, the longer constructs of SARS-CoV-2 nsp3 show less catalytic efficiency for Ub-AMC and greater affinity for ISG15-AMC, with 8-fold lower Km values compared to PLpro alone. Lastly, each SARS-CoV-2 nsp3 construct was inhibited by a known PLpro inhibitor, GRL-0617, with reported IC50 values ranging from 0.91 μM to 1.9 μM. These data show that GRL-0617 still remains a lead compound to be optimized for cellular potency. </p> <p><br></p> <p>Overall, this dissertation advances the understanding of the kinetic and structural differences between Alphacoronavirus PLP2 and Betacoronavirus PLpro enzymes in the efforts of developing a pan-CoV inhibitor. Additionally, these data provide initial kinetic and biophysical characterization of PLpro within the larger context of nsp3 to elucidate the function of PLpro in its most native context during coronaviral infection.</p>

<div>The papain-like protease (PLP) domain of nonstructural protein 3 (nsp3) of the coronavirus (CoV) genome promotes viral replication by processing the CoV polyprotein (protease) and also antagonize innate immune responses by deubiquitinating (DUB) and deISGylating (deISG) host substrates. Selectively removing the DUB/deISG activities of PLP while keeping the protease activity intact is a potential strategy for designing a live attenuated virus. However, it is unclear in the literature the precise mechanism by which PLPs support CoV evasion of the innate immune system. Deciphering the substrate specificity of PLPs for host ubiquitin (Ub) and interferon stimulated gene 15 (ISG15) can therefore help in the design of PLP mutants that selectively lack one activity for evaluating the DUB and deISG mechanism in CoV pathogenesis and replication. </div><div> In this dissertation, we investigate the structure and function of the single PLP (PLpro) from beta-CoVs, severe acute respiratory syndrome coronavirus (SARS-CoV) and Middle East respiratory syndrome coronavirus (MERS-CoV), which are dangerous viral pathogens that emerged from a zoonotic source to cause infectious disease in the human population. Additionally, we translate the knowledge gained to the equivalent PLP2 from alpha-CoV porcine epidemic diarrhea virus (PEDV) and feline infectious peritonitis virus (FIPV), which cause fatal disease in suckling piglets on industrial pork farms and household cats, respectively. The primary objective of this work is to rationally design PLP mutants across beta- and alpha-CoVs to help attenuate CoV infection, as no antiviral or vaccine exist for human CoVs and the efficacy of PEDV vaccines are an ongoing research topic. </div><div><br></div><div>In Chapter 1, different human, animal, and the bat origin CoV strains are introduced. The CoV life-cycle and virion structure are outlined, along with the replicase complex for viral replication. The multidomain nsp3 from alpha- and beta-CoV genomes are also described with a focus on the PLP domain and its proposed cleavage sites of the viral polyprotein. The discovery of the first viral protease DUB and the multiple activities of PLPs are defined, which includes a proposed model of how DUB versus deISG activities may act in the innate immune response. This leads into the therapeutic potential of PLP for an antiviral or live attenuated vaccine, which is followed by the introduction of live attenuated vaccines and the reverse genetics system. Next, proof of concept studies on PLP2 mutants are described and the introduction is concluded by stating the ultimate goal for the design of PLP mutants.</div><div><br></div><div>In Chapter 2, we hypothesize that the flanking ubiquitin-like (Ubl2) domain of MERS-CoV PLpro is not required for its enzymatic function. We characterize the specific activity, kinetics, substrate specificity, and inhibition of the PLpro enzyme with and without the Ubl2 domain and reveal that the Ubl2 domain does not significantly alter PLpro function. We determine the structure of the core PLpro, smallest catalytic unit to 1.9 Å resolution and observed no structural changes compared to the wild-type. Additionally, we demonstrate that a purported MERS-CoV PLpro inhibitor is nonselective in non-reducing conditions and should not be pursed for therapeutic use. We show that the core PLpro enzyme i.e. without the Ubl2 domain is a stable and robust construct for crystallization and is also thermally stable based on thermal melting studies with utility for structure-based drug design. </div><div><br></div><div>In Chapter 3, we shed light on the specificity of SARS-CoV PLpro towards Ub versus ISG15 by characterizing the specific activity and kinetic parameters of SARS-CoV PLpro mutants. In addition, the structure of SARS-CoV PLpro in complex with the C-terminal domain of ISG15 is determined and compared with the Ub-bound structure. Based on the structure and kinetic results, the altered specificities of SARS-CoV PLpro mutants Arg167Glu, Met209Ala, and Gln233Glu are compared with the wild-type. Arg167Glu mutant exhibits DUB hyperactivity and is expected to adopt a more favorable interaction with the Arg42 of Ub. At the same time, ARG167GLU contains a shorter side-chain that hinders interaction with the unique Trp123 of ISG15 for deISG activity compared to the wild-type. These results aid in the development of SARS-CoV PLpro mutants that have directed shifts in substrate specificity for Ub versus ISG15. </div><div><br></div><div>In Chapter 4, the process and antiviral activity of ISGylation is reviewed and how viruses can modulate host-derived versus virus-derived machineries to counteract ISGylation for viral infection. MERS-CoV PLpro is cross-reactive for Ub, but less is known about its specificity towards ISG15. In this study, we determine the structure of MERS-CoV PLpro bound with ISG15 to 2.3 Å resolution and reveal a small hydrophobic pocket of ISG15 that consists of P130 and W123, which differs from Ub hydrophobic patch. We design and determine the kinetic parameters for 13 PLpro mutants and reveal that MERS-CoV PLpro only has a single ubiquitin recognition (SUb1) site. Kinetic studies show that removing the charge of the R1649 greatly enhances DUB/protease activity while mutating in an Arg near R42 of Ub or ISG15 hydrophobic region is detrimental to both DUB/deISG activities. Kinetic experiments and probe-reactivity assays showed that Val1691Arg, Val1691Lys, and His1652Arg mutants are drastically reduced DUB/deISG activities compared to the wild-type. Overall, MERS-CoV PLpro mutants with alter kinetic profiles will be useful for discovery tools and DUB/deISG deficient mutants are great candidates for removing host cell antagonism activity by PLpro for live attenuated vaccines.</div><div><br></div><div>In Chapter 5, the goal is to translate the knowledge gained in Chapters 2-4 on beta-CoVs PLpro and evaluate the substrate specificity of alpha-CoVs FIPV and PEDV PLP2 for mutagenesis experiments. First, we design and purify the core PLP2 enzymes for kinetics. PLP2s are efficient DUBs that prefer Ub to ISG15 in vitro, and this preference is conserved in beta-CoV MHV PLP2 as well as alpha-CoV NL63 PLP2. We determine the structure of alpha-CoV PEDV PLP2 to 1.95 Å resolution and reveal the unique Zn-finger coordinating Cys3-His arrangement of the alpha-CoV genus that differs from past beta-CoV PLP crystal structures. To determine residues of the SUb1 site, we generate a homology model of FIPV PLP2 and overlay our PLP2 structures with MERS-CoV PLpro bound with Ub. In addition, we create electrostatic surface maps across coronaviral PLP subfamilies to evaluate the charge distribution of the SUb1 for the rational design of several FIPV and PEDV PLP2 mutants. We evaluate the turnover of PLP mutants for FRET-based substrates and reveal that His101ArgFIPV and Asn101ArgPEDV are drastically reduced in Ub-AMC activity while their peptide activities are within 2-fold of the wild-type. These mutants show delayed reactivity for Ub probes and no longer cleave Ub-chains displaying isopeptide bonds compared to the wild-type. Results from this study reveal a hot spot in both alpha- and beta-CoVs that can be used to selectively remove DUB activity of PLPs for generating a DUB deficient PLP enzyme. </div><div><br></div><div>In this dissertation, we investigate the substrate specificity of PLPs across alpha- and beta-CoVs and develop a fingerprint for Ub and also shed light on ISG15 recognition. Specifically, hot spots were identified in the SUb1 site of different PLPs, which recognize R42 and hydrophobic Ile44 of Ub. Position 97-98 of PLPs can be used to remove DUB activity by substituting an Arg, but usually effect protease function. Substituting an Arg at position 101 and 136 of coronaviral PLPs serve as the best strategy to remove DUB function while not hindering active site functionality. The DUB/deISG deficient mutants described will be useful for inhibiting the ability of PLPs to function in the innate immune response. Ultimately, this work provides a guide for identifying attenuating mutants in existing CoVs for live attenuated vaccines and also a blueprint for engineering PLPs from new emerging CoVs. </div>