Umeå University's logo

Enteroviruses are abundant pathogens of humans and animals. Their replication is strictly dependent on the conserved, viral AAA+ ATPase 2C. 2C is an oligomerizing, peripheral membrane protein, and its low solubility as recombinant protein has hampered functional studies of the full-length, recombinant protein bound to a membrane. Here, we describe a modification of the classical, ultracentrifugation-based liposome flotation assay optimized to study the interaction of recombinant 2C with membranes and the functions of membrane-bound, full-length recombinant 2C. The assay takes advantage of the high solubility of recombinant 2C while fused to a maltose-binding protein. Removing this solubility-enhancing tag by specific protease cleavage in the presence of liposomes allows 2C to associate with membranes prior to aggregating. Fluorophore labeling of protein and liposomes allows rapid and precise quantitation of 2C’s association with membranes. This assay is adaptable to any peripheral membrane protein that can be fluorophore-labeled and expressed as a solubility-enhancing fusion protein.

Key features

• This protocol extends widely used liposome flotation assays to low-solubility peripheral membrane proteins,such as the enteroviral protein 2C

• 2C is expressed and purified as a fusion protein with a solubility-enhancing tag, which is cleaved off in the presence of liposomes.

• Fluorophore-labeling of liposomes and protein facilitates quantitative readout of protein association with membranes.

• The protein-conjugated liposomes can also be used for other studies using, e.g., dynamic light scattering, cryo-EM, and enzymatic activity assays.

The endosomal sorting complex required for transport (ESCRT) machinery constitutes multisubunit protein complexes that play an essential role in membrane remodeling and trafficking. ESCRTs regulate a wide array of cellular processes, including cytokinetic abscission, cargo sorting into multivesicular bodies (MVBs), membrane repair, and autophagy. Given the versatile functionality of ESCRTs, and the intricate organizational structure of the ESCRT machinery, the targeted modulation of distinct ESCRT complexes is considerably challenging. This study presents a pseudonatural product targeting IST1-CHMP1B within the ESCRT-III complexes. The compound specifically disrupts the interaction between IST1 and CHMP1B, thereby inhibiting the formation of IST1-CHMP1B copolymers essential for normal-topology membrane scission events. While the compound has no impact on cytokinesis, MVB sorting, or biogenesis of extracellular vesicles, it rapidly inhibits transferrin receptor recycling in cells, resulting in the accumulation of transferrin in stalled sorting endosomes. Stalled endosomes become decorated by lipidated LC3, suggesting a link between noncanonical LC3 lipidation and inhibition of the IST1-CHMP1B complex.

The Enterovirus genus of the Picornaviridae family includes non-enveloped, positive-sense single-stranded RNA (ssRNA) viruses. This genus of viruses causes many diseases such as poliomyelitis by poliovirus (PV), cardiomyopathy by coxsackievirus B3 (CVB3), common cold by rhinoviruses (RVs) and meningitis by Enterovirus 71 (EV 71). The 7.5 kb enterovirus genome encodes a polyprotein, which is subsequently cleaved to yield viral proteins. These viral proteins hijack and modify the membranes of the Golgi and ER to give rise to replication organelles (ROs).

In the first paper of my thesis, we showed that membranes of the ROs act as an assembly line for the assembly of enteroviruses. Using cryo-electron tomography we studied poliovirusinfected cells. The one feature that we found most interesting was that a protein was tethering viral capsid to membranes as well as two membranes together. We hypothesized that viral protein 2C was a strong candidate. Therefore, I tried to mimic the interaction between 2C and viral capsid using purified Poliovirus and 2C protein. However, after trying several biochemical conditions I could not mimic this interaction. This strongly indicated that I might need to have a membrane in the system and this led me to study the membrane binding activity of 2C in the second paper.

In the second paper, I studied an Enteroviral multi-functional protein: 2C, a highly conserved AAA+ ATPase that plays an important role in the biogenesis of ROs and virus assembly. One of the most interesting features of 2C is its N-terminal membrane-binding domain consisting of 40 amino acids. Using in vitro reconstitution methods and biochemistry, I investigated the association of full-length 2C with lipid vesicles and how this affects the function of the protein. I showed that amino acids 12 to 40 are important not only for membrane binding but also for hexamerization. Moreover, truncation of the first 11 amino acids leads to loss of membrane tethering activity of the protein, which is essential for the formation of ROs. I was also able to demonstrate that the protein is sufficient to recruit RNA to the membrane as it was not previously known how RNA replication is localized to the membrane and here, we found the possible mechanism. In this realistic reconstituted system, I have shown that 2C is not a helicase but an ATP-independent RNA chaperone. Collectively, these discoveries offer a biochemical foundation for various functions of 2C in enterovirus replication. This sets up a more practical biochemical framework for future research, which could potentially contribute to the development of drugs aimed at thwarting enterovirus infections.

In the third project, I studied deoxyadenosine kinase (dAK) from Giardia intestinalis, a protozoan responsible for severe diarrhea spread by the fecal-oral route. Giardia is completely dependent on the host for deoxyribonucleosides as it lacks a de novo pathway for their synthesis. dAK uses ATP as a phosphate donor to phosphorylate deoxyadenosine, this activity is essential for genome replication of the protozoan. In this project, we have biochemically and structurally characterized dAK. Here, I used cryo-electron microscopy (cryo-EM) single particle analysis to show that dAK exists as a homo-tetramer in solution. This is an important finding because this is the first example of a non-thymidine kinase1-like deoxyribonucleoside kinase with a tetrameric structure and subsequent mutagenesis analysis showed that tetramerization allows the enzyme to achieve a higher affinity for its deoxyadenosine substrate.

In summary, in my thesis, I have biochemically and structurally characterized proteins supporting the genome replication of enteroviruses and Giardia intestinalis.

Enteroviruses are a vast genus of positive-sense RNA viruses that cause diseases ranging from common cold to poliomyelitis and viral myocarditis. They encode a membrane-bound AAA+ ATPase, 2C, that has been suggested to serve several roles in virus replication, e.g. as an RNA helicase and capsid assembly factor. Here, we report the reconstitution of full-length, poliovirus 2C’s association with membranes. We show that the N-terminal membrane-binding domain of 2C contains a conserved glycine, which is suggested by structure predictions to divides the domain into two amphipathic helix regions, which we name AH1 and AH2. AH2 is the main mediator of 2C oligomerization, and is necessary and sufficient for its membrane binding. AH1 is the main mediator of a novel function of 2C: clustering of membranes. Cryo-electron tomography reveal that several 2C copies mediate this function by localizing to vesicle-vesicle interfaces. 2C-mediated clustering is partially outcompeted by RNA, suggesting a way by which 2C can switch from an early role in coalescing replication organelles and lipid droplets, to a later role where 2C assists RNA replication and particle assembly. 2C is sufficient to recruit RNA to membranes, with a preference for double-stranded RNA (the replicating form of the viral genome). Finally, the in vitro reconstitution revealed that full-length, membrane-bound 2C has ATPase activity and ATP-independent, single-strand ribonuclease activity, but no detectable helicase activity. Together, this study suggests novel roles for 2C in membrane clustering, RNA membrane recruitment and cleavage, and calls into question a role of 2C as an RNA helicase. The reconstitution of functional, 2C-decorated vesicles provides a platform for further biochemical studies into this protein and its roles in enterovirus replication.

The protozoan parasite Giardia intestinalis is one of only a few organisms lacking de novo synthesis of DNA building blocks (deoxyribonucleotides). Instead, the parasite relies exclusively on salvaging deoxyadenosine and other deoxyribonucleosides from its host environment. Here, we report that G. intestinalis has a deoxyribonucleoside kinase with a 1000-fold higher catalytic efficiency (k_cat/K_M) for deoxyadenosine than the corresponding mammalian kinases and can thereby provide sufficient deoxyadenosine triphosphate levels for DNA synthesis despite the lack of de novo synthesis. Several deoxyadenosine analogs were also potent substrates and showed comparable EC₅₀ values on cultured G. intestinalis cells as metronidazole, the current first-line treatment, with the additional advantage of being effective against metronidazole-resistant parasites. Structural analysis using cryo-EM and X-ray crystallography showed that the enzyme is unique within its family of deoxyribonucleoside kinases by forming a tetramer stabilized by extended N- and C-termini in a novel dimer–dimer interaction. Removal of the two termini resulted in lost ability to form tetramers and a markedly reduced affinity for the deoxyribonucleoside substrate. The development of highly efficient deoxyribonucleoside kinases via oligomerization may represent a critical evolutionary adaptation in organisms that rely solely on deoxyribonucleoside salvage.

Viruses are masters at using cellular pathways to aid their replication. Cryo-electron tomography of poliovirus-infected cells revealed how it utilizes macroautophagy to its advantage. Assembly of these non-enveloped virions takes place directly on membranes and requires PIK3C3/VPS34 activity to be completed, whereas the canonical autophagy inducer ULK1 restricts virus assembly. The tomograms further revealed that enterovirus-induced autophagy is selective for RNA-loaded virions, which may help ensure maximum infectivity of the virus-laden vesicles released through secretory autophagy.

Enteroviruses are non-enveloped positive-sense RNA viruses that cause diverse diseases in humans. Their rapid multiplication depends on remodeling of cytoplasmic membranes for viral genome replication. It is unknown how virions assemble around these newly synthesized genomes and how they are then loaded into autophagic membranes for release through secretory autophagy. Here, we use cryo-electron tomography of infected cells to show that poliovirus assembles directly on replication membranes. Pharmacological untethering of capsids from membranes abrogates RNA encapsidation. Our data directly visualize a membrane-bound half-capsid as a prominent virion assembly intermediate. Assembly progression past this intermediate depends on the class III phosphatidylinositol 3-kinase VPS34, a key host-cell autophagy factor. On the other hand, the canonical autophagy initiator ULK1 is shown to restrict virion production since its inhibition leads to increased accumulation of virions in vast intracellular arrays, followed by an increased vesicular release at later time points. Finally, we identify multiple layers of selectivity in virus-induced autophagy, with a strong selection for RNA-loaded virions over empty capsids and the segregation of virions from other types of autophagosome contents. These findings provide an integrated structural framework for multiple stages of the poliovirus life cycle.