The Infection and Uncoating Mechanism of the Giant Melbournevirus
Sahar Shammakhi
__________________________________________________________
Degree project in biology, Master of science (2 years), 2020 Examensarbete i biologi 30 hp till masterexamen, 2020
Biology Education Centre and The Laboratory of Molecular Biophysics, Department for Cell and Molecular Biology, Uppsala University
Supervisor: Kenta Okamoto
2
Content Page
Abbreviations p. 4
Abstract p. 5
1. INTRODUCTION p. 6
1.1 Rationale p. 6
1.2 Viruses p. 6
1.2.1 History and definition of viruses p. 6 1.2.2 Diversity and classification of viruses p. 8
1.2.3 Life cycles of viruses p. 8
1.3 Giant viruses p. 9
1.3.1 Definition of the term 'giant virus' p. 9 1.3.2 Giant virus genome, proteome and morphologies p. 10 1.3.3 Giant viruses in the environment and host range p. 11 1.3.4 Giant virus entry and uncoating mechanisms p. 12
1.4 Melbournevirus p. 13
1.4.1 Classification, genome and proteome of MelV p. 13 1.4.2 Distinct features and morphology of MelV p. 14 1.4.3 Putative entry and uncoating mechanism of MelV p. 14
1.5 Aims and Objectives p. 15
2. MATERIALS AND METHODS p. 17
2.1 Acanthamoeba castellanii cultivation p. 17
2.2 MelV propagation, purification and titration p. 17 2.3 Infectivity assay with glycan competition p. 17 2.3.1 Sample preparations for the competition infectivity assay p. 17
2.3.2 Virus titration assays p. 18
2.4 MelV SDS-PAGE and lectin blotting p. 19 2.5 Glycan-induced protein secretion in amoeba cells p. 19 2.6 Identification of amoeba phagosomal enzymes p. 20
2.6.1 Induction of phagosomal proteins and purification of
amoeba phagosomes p. 20 2.6.2 Identification of proteins by LC/MS/MS p. 20 2.7 Dissociation assay of MelV particles using purified
phagosome enzymes p. 21
3. RESULTS p. 22
3.1 Infectivity assay with glycan competition p. 22 3.1.1 Amoeba cells appear mobilised when treated with
100 mM mannose p. 22
3.1.2 TCID50/ml titration allude to competitive binding of
Man and MelV to amoeba cells p. 22
3
3.1.3 qPCR titration significantly indicates binding between
Man and MelV to amoeba cells p. 23
3.2 MelV contains at least a single mannose-containing glycoprotein
of approximately ~70 kDa in size p. 25
3.3 Mannose induces the most protein secretion from A. castellanii compared to other glycans, with secretion protein levels
significantly higher than in phagosomes p. 25 3.4 Proteomics samples have different protein composition yet the
phagosome sample shares 70% of its proteins with secretory
protein sample p. 27 3.5 Dissociation assay samples require further optimisation p. 29
4. DISCUSSION p. 30
4.1 Putative entry mechanism of the MelV mediated by a
mannose-presenting protein p. 30 4.2 Glycans could trigger phagocytosis in the amoeba cells p. 31
5. CONCLUSION p. 33
6. APPENDIX p. 34
6.1 Summary of section II of thesis:
Isolation of Viruses from an Environmental Sample p. 34
6.1.1 Introduction p. 34
6.1.2 Materials and methods p. 35
6.1.3 Results and discussion p. 36
6.2 Supplementary tables p. 38
6.3 Supplementary figures p. 39
Acknowledgements p. 41
References p. 42
4
Abbreviations
CPE Cytopathic effect
Cryo-EM Cryogenic electron microscopy DNA Deoxyribonucleic acid
EM Electron microscopy
Gal Galactose
GlcNAc N-acetylglucosamine
ICTV International committee on taxonomy of viruses LC-MS/MS Liquid chromatography tandem mass spectrometry LDB Large and dense body
Man Mannose
MelV Melbournevirus
MOI Multiplicity of infection mRNA Messenger RNA
NCLDV Nucleocytoplasmic large DNA virus ORF Open reading frame
PBS Phosphate buffered saline PBST Phosphate buffered saline tween PPYG Protease peptone yeast glucose
qPCR Quantitative polymerase chain reaction PVDF Polyvinylidene fluoride
Rn Relative fluorescence RNA Ribonucleic acid RPM Revolutions per minute
RT Room temperature
SDS-PAGE Sodium dodecyl sulphate polyacrylamide gel electrophoresis SEM Scanning electron microscope
STEM Scanning transmission electron microscopy TCID50 Tissue culture infectious dose 50%/mL
5
Abstract
Since their 'discovery' at the turn of the 21st century, giant viruses of the amoeba have captured the fascination of virologists. They have raised a plethora of questions regarding their evolution and ecological significance and have greatly defied a century's old definition of viruses. By now, it is understood that a handful of giant viruses enter the amoeba via the phagosomal pathway. This thesis chooses to focus on the giant Melbournevirus (MelV) regarding its entry and uncoating pathway. We now conclude that the initial attachment between MelV and amoeba cells is built upon glycan interactions based on evidence that mannose competitively inhibits MelV binding. This attachment likely entails an
approximately 70 kDa mannose containing glycoprotein on the MelV. Mannose and other glycans induce secretion of proteins including phagosomal enzymes from amoeba. Based on these findings, it is hypothesised that the mannose-induced phagosomal enzymes could play a role in the uncoating of the MelV. The results further reveal isolated phagosomes, also to some extent the glycan-induced protein secretions, to have high levels of proteins involved in cell redox homeostasis. This implies that the highly oxidative environment of the phagosome may be an overlooked physiological factor when regarding the uncoating of the MelV. A deeper understanding of the physiological uncoating conditions can be used for studying internal structures of giant viruses, such as the enigmatic Large and Dense Body (LDB) of the MelV particle.
6
1. INTRODUCTION
1. 1 Rationale
All sciences make use of tools to portray their specialist concepts and this is no different for biology; both the hierarchical categorisation "systema naturae" and the binomial
nomenclature systems originate from Carl Linneas in 18th century Uppsala, Sweden. Both are now further developed and used to classify every single living organism on earth (Basic biology, 2019). The field of taxonomy originally paved the way for phylogeny which has seen a significant shift since genome sequencing methods progressed during the closing decades of the 20th century (Brown, 2002; Wikberg 2010). This inevitably adds rich and complex layers in understanding our biosphere. Indeed, the classification of both the entities from which this project hinges are deemed complex: the giant Melbournevirus (MelV) and its host, an amoeba, Acanthamoeba castellanii (A. castellanii).
Different amoeba species are often studied for their relevance to human health and, also as a trojan horse for fastidious pneumonia agents by taking the place of our own phagocytic immune cells (Khan, 2006; Raoult et al., 2007). It was, in fact, through a study involving isolating such fastidious bacteria in amoeba co-cultures that the first "giant virus" was discovered at the turn of the 21st century (La Scola et al., 2003). Had it not been found then the decades or even century-old concepts regarding viruses, particularly that of size, would have persisted without challenge. The evident delay in discovering giant viruses may reflect the inclination of funding to be directed towards research with medical or economic
implications. Current research efforts on understanding giant viruses are considered to be pure biology, since. these viruses do not impact humans, animals or crops- at least not directly. The discovery, therefore, is an example of how applied and pure biological research can intertwine and become interchangeable. Traditionally, virology has been a disease- focused field, yet, it is giant viruses that are the field's most unexpected discovery this century. It continues to be a mostly pure biology-based research topic, as will be the case for this thesis project. Judging, however, by the deepening understanding of the role of viruses in the web of life and its evolution, along with the contributions to biology that the expanding virology research continues to provide, one can only predict applications would arise upon significant findings (Claverie & Abergel, 2016).
1.2 Viruses
1.2.1 History and traditional definition of viruses
To this day, a 0.2 µm filter is commonly used as an alternative to heating in order to sterilize liquids by trapping microorganisms such as most bacteria. Viruses are known to be able to pass through these, including the first of such filters, a Chamberland-Pasteur filter. This is
7
how the first virus, Tobacco mosaic virus (TMV), was discovered by Dimitry Ivanovsky in 1892: a filterable disease-causing substance that was unculturable without cells, invisible under light microscope, and simply too small to fit the newly established germ theory of that era that accommodated bacteria so well (Ivanovsky, 1892).
This description of viruses endured for half a century. Several factors cemented the notion of viruses being non-living entities, including subsequent discoveries of other viruses, the first electromagnetic image of TMV made in 1939, along with the developing understanding of how they propagate. The study of bacteriophages, which are viruses which target bacteria, formed the basis of many biological principles and led to the Lwoff's criteria in 1957 which definitively distinguished viruses from cells; the criteria was updated in 1966. It included 5 discriminative characteristics as described in Table 1: while microorganisms had both DNA and RNA, reproduced through binary fission, possessed metabolic enzymes and a translation machinery, viruses lacked all of these. Viruses have only one type of nucleic acid which is all they need to reproduce with no binary fission involved, a lack of metabolic enzymes and no translation machinery. Lwoff's criteria also endured for another half century up until the discovery of the first giant viruses in 2003 (Claverie & Abergel, 2016; Lwoff & Tournier, 1966).
Table 1 A summary of Lwoff's discriminating criteria between viruses and microorganisms (Lwoff &
Tournier, 1966).
Criterion Microorganism Virus
Nucleic acid Both DNA and RNA Either DNA or RNA Re/production From integrated sum of
their constituents
From nucleic acid only
Growth Binary fission No binary fission
Metabolism Possess enzymes for metabolic functions
Lacks system of metabolic enzymes Translation
machinery
Possess translation machinery
Make use of the hosts translation machinery
The last two characteristics describe parasitic behaviour. A comparison could be made between viruses and obligate intracellular parasitic or symbiont bacteria such as Rickettsia or Buchnera, respectively. However, though reductive evolution has resulted in a loss of many metabolic functions for these bacteria, they still maintain enough functions to manifest typical features of cellular life forms, such as energy transduction (Driscoll et al., 2017;
Andersson, 2000; Yewdall et al., 2018). Where the line between viruses and bacteria do blur, however, is with giant viruses. Their discovery has sparked an ongoing debate on the origins of viruses itself and has raised the question of what is a virus? And where is their place in the biosphere? These are questions steeped in scientific history and intertwined with the story of the giant virus itself (Pradeu et al., 2016).
8 1.2.2 Diversity and classification of viruses
A unique feature of many viruses is their symmetrical capsids made up of identical protein subunits termed capsomers. The capsid surrounds and protects the viral genome and itself can be further surrounded by an envelope, which is made up mostly of a lipid bilayer derived from its last host cell. Viruses can possess many morphologies. While there is significant diversity, the structural parameters of capsomers are limited, which means capsids regularly form two common shapes: rod-shaped viruses have helical symmetry and spherical viruses have icosahedral symmetry (Madigan et al., 2015; Wagner et al., 2008).
Viruses are not classified according to binomial nomenclature like other cellular organisms.
Instead, the incredibly diverse virosphere is classified using a myriad of methods and is increasingly based upon genomic properties in addition to phylogenetic relationships and phenotype. Different viruses can harbour significantly diverse genome which all, ultimately, lead to mRNA expression: mainly genomes can be RNA or DNA, single (positive or negative sense) or double stranded, segmented or not. Other factors contributing to the classification of viruses is genome size and how the genome functions in the host. The International
Committee on Taxonomy of Viruses (ICTV) offers the most comprehensive method of classifying viruses by combining many virus attributes as the basis for the designation of viruses into families and genera (Wagner et al., 2008). ICTV has devised a taxonomical system resembling the Linnaean system, which has recently expanded from including 5 to 15 hierarchical ranks. This is to help accommodate the wealth of metagenomic data and the resultant genetic diversity of the virosphere (ICTV executive committee, 2020).
1.2.3. Life cycles of viruses
The diversity of the virosphere creates a myriad of specific molecular interactions between viruses and hosts; this is reflected in the variation within the 'typical' virus life cycle between different viruses (Rothenburg & Brennan, 2020). Viruses have variable 'host ranges' in which they can successfully replicate. The typical virus multiplication cycle can be outlined by three main steps: Virus entry into the cell, viral gene replication, and virion assembly. Many
viruses are lytic in that their production causes the host cell to burst, others are lysogenic in that the viral genome is integrated into the host DNA resulting in chronic infection (Wagner et al., 2008). The viral capsid plays an important roles in the cycle. Viral genes are protected by this protein coat, only to unleash their potential upon infection of their host and the high jacking of their essential systems (Villarreal, 2008). Studies of viral capsids have led to elucidating the initial cell entry mechanism of viruses. This project focuses on the cell entry stages of giant virus MelV into A. castellanii amoebal cells.
9
1.3 Giant viruses
1.3.1 Definition of the term 'giant virus'
Since the first giant virus was discovered in 2003 numerous other giants have been found, many of which are in fact rediscoveries, having previously been mistaken for microbes of an endosymbiotic or parasitic nature. Consecutive findings have spawned new families which have landed as members of the phylogenetic superfamily Nucleocytoplasmic Large DNA viruses (NCLDV). NCLDV includes viral families from different orders and non-classified viruses with common features: notably, relatively large DNA viruses with large genomes and common gene content (Iyer et al., 2006).
The definition of a giant virus is somewhat arbitrary. In fact, depending on the review different viruses tend to be included into the 'giant virus' enclave. Some definitions use two simple parameters: genomes larger than 200 kb, which overlaps with a third of bacterial genome sizes, and particle sizes larger than 200 nm (Wilhelm et al., 2017). Other definitions attempt to find common features in genes, morphology and replication cycles, while theories regarding the evolutionary nature of their unique features remain wide-ranged and elusive (Wilhelm et al., 2017).
Though not formally recognized by the ICTV, the NCLDV superfamily originally included five virus families including Poxviridae which was once considered to include the largest ever virus, the Poxvirus (Yuan, 1999). While many of the smaller NCLDV are not considered giant, the term 'giant virus' was in fact used much earlier upon discoveries of some algae- infecting NCLDV (Van Etten & Meints,1999). In hindsight, many viruses characterized as early as the 1970's are potential 'giant viruses', (Wilhelm et al., 2017). Alas, the term 'giant viruses' is both ambiguous, subjective and not always mutually inclusive with the term NCLDV.
During the last 17 years, interest has peaked, as three significantly 'giant' virus families have been added to the NCLDV superfamily: Mimiviridae, Marseilleviridae and, more
controversially, Pandoraviridae (Brandes & Linial, 2019). Another claim is that there are in fact more putative giant virus families such as Faustoviruses and others with single or few members like Pithovirus and Mollivirus (Aherfi et al., 2016). However, considering the swift expansion of giant virus families, along with the precarious nature of viral taxonomy itself, these classifications are likely to change. There are even recent calls to collect giant viruses into a new order, Megavirales (Colson et al., 2012).
The discovery of these giant viruses was delayed by up to a century because the notion that viruses are 'smaller' had become such an epistemological barrier. Illustrated in Figure 1 is how the 0.2 µm threshold traps a lot of giant viruses as well as microorganisms. Lwoff's criteria did not include size as a specific criterion, but it was an important feature in
10
distinguishing viruses from microorganisms. Today, giant viruses defy Lwoff's criteria for reasons other than just size (Lwoff & Tournier 1966; Abergel et al., 2015).
1.3.2 Giant virus genome, proteome, and morphologies
Owing to the complexity and large size of the genomes, proteome studies of the giant viruses are important for characterizing them. Many open reading frames of the giant viruses’
genome lack any homologues all together and are hence considered ORFans (Open Reading Frames that are not homologous to previously registered ones). Thereby, the level of
uncharacterized protein is significantly high and often forms the majority, ranging from 50%
to over 80%, of a giant virus' proteome. To put this into context, bacterial proteomes harbour 30-40% uncharacterized proteins while this figure is approximately 60% in humans (Hanson et al., 2009). The proportion of ORFans was initially incredibly high with first family
representatives but have decreased as giant virus families rapidly expand upon successive isolations and discoveries (Brandes & Linial, 2019).
Viral genes mostly encode for proteins which either function as structural protein such as capsids, or as non-structural, regulatory and accessory proteins which aid viral replication and assembly (Ulversky & Longhi, 2010). Approximately 70% of viruses encode for less than 10 genes, while only 10% encode for a hundred or more genes with a sliver of just 0.3% of viruses encoding for more than 500 genes. Giant viruses are placed largely at the latter end of this spectrum with genome sizes ranging between 200 kb to 2.5 Mb. Their predicted proteins demonstrate genome complexity (Colson et al., 2017); some giant viruses of the Mimiviridae
Figure 1. An illustration depicting how the commonly used 0.2 µm filter traps microorganisms.
E. coli, a microorganism, and giant viruses such as Mimivirus, Pithovirus, Melbournevirus and Poxvirus (roughly to scale) are trapped by the filter, perpetuating the epistemological barrier. Image created with BioRender.com, (2020).
11
family possess genes associated with translation and transcription machinery, DNA repair and even virophage defence mechanisms (Brandes & Linial, 2019). These features further clash with Lwoff's definition of viruses and shorten the divide between microbes and viruses, where the distinction increasingly appears to take on the form of a spectrum rather than a binary distinction.
There does appear to be a general trend where the larger the virus genome the larger the virus particle size, with Pithovirus bunking this trend (Colson et al., 2017; Okamoto et al., 2017).
Giant viruses also have wide ranging morphologies; while the majority of NCLDVs exhibit an icosahedral or a roughly icosahedral capsid structure, many have ovoid or spherical shapes. Cryo-electron microscopy (cyro-EM) studies of several icosahedral NCLDV reveal that they share common structural features. The protein capsomers making up the external icosahedral capsid appear to be arranged and organized similarly in these NCLDV; they also possess at least one internal membrane encapsulating the nucleocapsid (Fang et al., 2019).
Covering the external capsid, the possession of fibrils adorned with glycoproteins also varies between viruses and prove to be an important feature in the attachment of the virus to its host (Colson et al., 2017; Rodrigues et al., 2015). Fibrils along with the large particle size are thought to facilitate the infiltration of giant viruses into the notoriously selective phagocytic process of amoebas (Korn & Weisman, 1967; Wilhem et al., 2017).
1.3.3 Giant viruses in the environment and host range
Recent metagenomic data allude to ubiquitous environmental distribution patterns and
expansive host ranges and proteomes of NCLDV (Shulz et al., 2020). Most giant viruses have been isolated from environmental samples from a diverse range of geographical locations through co-culturing with amoeba. As the co-culturing strategy evolved over the years, so had the expansion of the so-called 'giant viruses of the amoeba' (Aherfi et al., 2016).
The acanthamoeba genus, and more specifically Acanthamoeba castellanii species, is the most commonly used amoeba in isolation methods.
Acanthamoeba are free living amoeba ubiquitously found in the environment. They phenotypically switch from trophozoite to cyst form depending on the favourability of their surroundings, with the trophozoite form notably exhibiting finger like projections named acanthopodia visible in Figure 2;
these allow adhesion to surfaces, movement and feeding. Acanthamoeba feed on a wide range of microorganisms such as bacteria, yeast and protist.
They feed in the form of pinocytosis and
phagocytosis, the latter being a selective process (Khan, 2006). It has long been known, through experiments with latex beads, that particles larger
Figure 2. A light microscope image of Acanthamoeba castellanii in trophozoite form. Magnification is at 20X and the acanthopodia of these confluent amoeba are evident.
12
than 0.557 µm are ingested individually while particles smaller than this are required to accumulate at the surface first (Korn & Weisman, 1967). The protein profiles of phagosomes of the pathogenic Entamoeba histolytica (E. histolytica) species are largely comparable when using uncoated latex beads and carboxylate-coated beads (Okada et al., 2006). However, amoeba can distinguish between phagosomes with digestible and indigestible particles, since beads are later exocytosed and food particles retained and digested (Bowers & Olszewski, 1983). In nature, phagocytosis usually involves lectins on the amoeba initially binding appropriate glycans on microorganisms (Korn & Weisman, 1967; Sharon & Lis, 1989).
Though it is likely that Mimivirus' natural hosts are amoeba, the majority of giant viruses appear to have different host ranges and their natural hosts remain unknown. The refined co- culturing strategies reflect the initial drive to isolate as many giant viruses as possible, with true diversification seen only though using different amoeba species as hosts. This has resulted in the discoveries of new viruses such as the Tupanvirus. The fixation on discovery and characterisation has initially shifted the focus away from understanding their true role in our ecosystems. There are indications that algae are a common host for NCLDV in which case attachment of the virus would be through adsorption to algae cell wall. This would suggest giant viruses may play a critical role in the wider natural cycles of oceanic and other environments especially since giant viruses have since been found in zooplankton and several other unicellular protists (Rolland et al., 2019; Wilhelm et al., 2017; Fischer et al., 2010).
There is also literature indicating the presence of Marseilleviruses in human samples though these results are controversial (Rolland et al., 2019).
1.3.4 Giant virus entry and uncoating mechanisms
The majority of viruses use the endocytic/phagocytic pathways for entry, but giant viruses are unique in that they enter purely via phagocytosis (Hulo et al., 2017; Yamauchi & Greber, 2016). It is understood that giant viruses demonstrate microbial mimicry in order to enter the amoeba: not only in terms of their large size but also in initial attachments to the host cell.
The largest of the giant viruses of the amoeba, which include Mimiviridae, Pandoraviridae, Mollivirus and Pithovirus, are all 600 nm or larger in diameter so they qualify for the selective phagocytosis route (Abergel et al., 2015). Research on Mimivirus' infection mechanism to its host Acanthamoeba castellanii suggests glycan interactions facilitate its initial adhesion to host cells; the removal of its fibrils significantly reduced attachment but did not affect viral replication (Rodrigues et al., 2015), while the inhibition of phagocytosis did (Andrade et al., 2017).
These four viruses further follow similar life cycle patterns. They all have some form of portal, whether it's an apical pore, cork or a stargate; this portal opens to allow the internal lipid membrane to fuse with the phagosome membrane and release its inner contents into the cytoplasm, as is illustrated in Figure 3 (Andrade et al., 2017). Other giant viruses smaller than 600 nm such as Faustoviruses are also likely to enter the amoeba via mechanisms that entail more than just phagocytosis, including endocytosis and micropinocytosis, such is the case with other NCLDV with their hosts (Rodrigues et al., 2016). Notably, an unknown
13
biochemical signal is responsible for the initial portal opening of the capsid within the phagosome. Henceforth, the rest of the replication cycle does vary between different giant viruses, for instance, Mimivirus and Pithovirus are cytoplasmic while Mollivirus and Pandoravirus are nucleocytoplasmic (Abergel et al., 2015).
1.4 Melbournevirus
1.4.1 Classification, genome and proteome of MelV
This project's focus is on the giant virus named MelV that belongs to the Marseilleviridae family, part of the NCLDV group. The Marseilleviridae isolates are further categorized into lineages A to E according to phylogenetic analysis. MelV belongs to lineage A along with the prototype of the family, Marseillevirus (MsV). MelV was found in the environmental waters of Melbourne- Australia, hence its name, and is a giant virus of the amoeba as it was isolated through its ability to infect A. castellanii cultures (Doutre et al., 2014; Rolland et al., 2019).
The MelV capsid forms an icosahedron up to 250 nm in diameter which harbours a double stranded DNA genome of approximately 370 kb in size. In the first study by Doutre et al., bioinformatics analysis found there to be 403 open reading frames (ORFs) which had
extremely similar homologies to other lineage A Marseilleviridae viruses, MsV and Cannes 8 virus, which were isolated in Europe. This was despite the contrasting geological locations in which the isolates were found, prompting further investigation into the selection pressures of the viruses; it was concluded that the genome complexity of the MelV is integral to viral fitness as opposed to the genes being dispensable (Doutre et al., 2014).
An SDS-PAGE gel profile of MelV reveals at least 10 abundant proteins while tandem mass- spectrometry based proteomics analysis identified at least 58 structural and packaged
proteins. The majority are uncharacterized proteins, but many were identified as putative viral and membrane proteins (Okomoto et al., 2018). Also detected were histone-like proteins,
Figure 3. An overview of giant virus entry and uncoating mechanism (represented here by Mimivirus and its stargate portal). Illustrated is 1) entry via phagocytosis into the amoeba and later 2) the opening of the portal for the fusion of inner membrane with phagosome membrane to release inner genome content into the cytoplasm.
Image created with BioRender.com (2020).
1
2
14
which is a notable feature of the Marseilleviridae viruses and a reflection of the chimeric nature of the typical genome of giant viruses of the amoeba (Boyer et al., 2009; Okamoto et al., 2018).
1.4.2 Distinct features and morphology of MelV
The differences between MsV and MelV become more apparent when one considers their distinct features and morphologies. Cryo-EM images of MsV faintly portray the presence of 12 nm long fibrils, whereas no fibrils are detected on MelV cryo-EM images nor on the EM images (Boyer et al., 2009; Okamoto et al., 2018). Despite their many similarities in terms of genome size and morphology the replication times are drastically different: 5 hours for MsV and 12 hours for MelV (Boyer et al., 2009; Doutre et al., 2014).
But what sets MelV apart, and which forms the inspiration for this thesis project, is the unique and reproducible internal microassembly shown within MelV particles. This microassemby is named Large and Dense Body
(LDB) and is approximately 30 nm in diameter and is shown in Figure 4. Each MelV particle appears to contain an LDB and it remains a mystery as to what it may be. Cryo-EM images indicate the presence of a double layered internal membrane and while the position of the LDB is not fixed they do appear closer to the membrane (Okamoto et al., 2018). As mentioned in Okamoto et al. (2018), to decipher the nature of the LDB it is necessary to first break open the outer capsid in a way in which the LDB remains intact.
1.4.3 Putative entry and uncoating mechanism of MelV
The life cycle of MelV is comparable to MsV in that they are both cytoplasmic i.e. genome replication occurs in the cytoplasm of the host (Boyer et al., 2009; Doutre et al., 2014).
Though the MelV entry mechanism has not been studied before, EM images allude to similarities to the MsV particles entry mechanism. The study of the entry mechanism of Marseilleviridae viruses is particularly significant since they are too small to spontaneously trigger phagocytosis. So far, it is understood that MsV particles have two main entry
pathways into the amoeba: they can aggregate to form giant vesicles upon release from a cell in order to trigger phagocytosis in the subsequent infection, or they can enter as single particles through an endosome-stimulated pathway. For the latter entry pathway, details regarding the initial attachment of the virus particle to the amoeba cell surface have yet to be elucidated (Arantes et al., 2016). This finding resonates with the Doutre et al. paper where intracytoplasmic vacuoles filled with mature MelV particles were observed under the EM.
Whether these vacuoles remain intact after cell lysis is not entirely clear (Doutre et al., 2014).
Figure 4. Representative cryo-EM raw image of the MelV particles.
Arrows pointing to the singular Large and Dense Body (LDB) within the capsid. (Image adjusted from Okamoto et al., 2018).
15
Glycan interactions may be involved in the non-elucidated endosome stimulated entry pathway of single MsV particles into amoeba cells. No molecular studies have been done regarding the entry mechanisms of the MelV.
The uncoating of viral capsids occurs after cell binding and entry. As many viruses enter via the endocytic pathways, they have adapted various mechanisms to uncoat their capsids during endocytosis uptake, such as sensing pH and ion levels or being digested by lysosomal
enzymes (Wagner et al., 2008; Yamauchi & Greber, 2016). A recent paper by Schrad et al.
(2020), has found that treatment of Mimivirus and related viruses with a combination of temperatures as high as 100°C and a pH as low as 3 completely opens up the stargate structure on its capsid (Schrad et al., 2020). For the icosahedral giant viruses, it has been suggested by Fang et al. (2019) that they would follow similar principles in reassembling their outer capsid due to the commonality within their structures; so it could be the case that their disassembly also follows similar cues. (Fang et al., 2019). The natural uncoating mechanism of giant viruses, however, is still shrouded in mystery. It may be that the natural trigger for giant virus uncoating within the amoebal phagosomes depends on the unique phagosomal environment and specifically phagosomal proteins, rather than artificial parameters such as temperature and pH.
1.5 Aims and Objectives
This master thesis is directed at apparent knowledge gaps in our understanding of MelV and other giant viruses. The first knowledge gap is the details of viral entry of MelV particles into the amoeba and conversely that of Marseilleviridae viruses. The second is to contribute in understanding the uncoating mechanism of MelV under amoebal physiological conditions;
this is by conducting a proteomics study on the amoebal phagosome. The overall structure of the thesis is illustrated in Figure 5.
Specific aims
1) We primarily concentrate on revealing MelV's entry mechanism. The specific aims are:
1a) to determine whether glycan interactions are involved in the initial attachment between MelV and amoeba, 1b) to decipher the nature of the initial attachment i.e. the function of MelV surface polysaccharide for cell entry.
2) We explore the uncoating strategy of MelV. The specific aims are: 2a) to study A.
castellanii secretion proteins that are stimulated by glycan attachment and probe into its implications on MelV entry pathway and uncoating, 2b) to isolate and study via proteomics A. castellanii phagosomal proteins to explore their nature and the possible functions they may have on MelV uncoating mechanism. An attempt was even made at recreating amoebal physiological conditions to dissociate (uncoat) the virus.
The second section entails an attempt at isolating a new giant virus from environmental
16
Figure 5. An overview of the thesis aims and objectives. The thesis consists of two parallel parts with the main aim being to further understand the entry and uncoating mechanisms of MelV, the other to isolate a novel giant virus, by conducting a range of experiments (written in the arrows).
waters in Uppsala, Sweden and was run in parallel to the main project using algae isolates. A summary is annexed to the appendix.
Melbournevirus Environmental
water sample Infectivity
assay + Western blot
of MelV proteins
Proteomics analysis of phagosome and secretory
enzymes
Sample filtration and
co-culturing with algae, isolation and EM imaging
Entry mechanism
Uncoating mechanism
Novel giant virus
Understanding the nature of the LDB
17 2. MATERIALS AND METHODS
2.1 Acanthamoeba castellanii cultivation
This study used a seed of the A. castellanii strain (ATCC 30010D) which were axenically cultured using autoclaved proteose peptone yeast glucose (PPYG) basal medium, for which each 900 mL was complemented with 10 mL of each 400 mM of MgSO4, 5 mM of
Fe(NH4)2(SO4)2, 250 mM of Na2HPO4 and 250 mM of KH2PO4; along with 8 mL of 50 mM CaCl2 and 50 mL of 2 M glucose. The pH was adjusted to 6.5 and the complete culture medium was filtered with a 0.22 µm filter. A. castellanii cells were incubated and regularly passaged in TC-flasks (Sarsted str) at 28°C and 5% CO2 level.
2.2 MelV propagation, purification and titration
The MelV strain was kindly given by Abergel and Claverie groups, Aix-Marseille University.
The propagation and purification protocol for MelV particles was followed and modified from a previously reported paper (Doutre et al., 2014). MelV solution was inoculated in each of ten T75 flasks with 100% confluent amoeba (A. castellanii) and left for over 48 h. The amoeba/virus supernatant was harvested and centrifuged for 20 min at 600 g, 4°C to remove cell debris. The infected cell fluid was centrifuged for 1 h at 10,000 g, 4°C to pellet down the virus. The pellet was washed with 30 ml of PBS (-) for 1 h at 10,000 g, 4°C. The supernatant was removed, and the pellet was dissolved in 1 ml of PBS (-). The dissolved virus solution was added to the end of a continuous 40%/10% sucrose gradient in a 14 ml ultra-clear
centrifuge tube. The tube was ultra-centrifuged for 45 min at 6,500 g, 4°C (Optima XPN-100- Beckman Coulter). Each of two visible bands were extracted and washed with 30 ml PBS (-) buffer for 1 hour at 10,000 g, 4°C to remove sucrose. The virus pellet was resuspended in 200 µl of PBS (-) buffer. The purified MelV was confirmed through titration and the structure was observed with a scanning transmission electron microscope (STEM) detector using scanning electron microscopy (SEM) with Quanta FEG 650 (ThermoFisher Scientific). For the titration of the purified MelV particles, two 24-well plates containing amoeba of 50%
confluence were used. Each lane was used for one level of the MelV particle dilution series 10-4 to 10-14 (made with PPYG culture medium), with one lane kept as a negative control; for each well 200 µl of viral dilution was inoculated into the corresponding lane. Cytopathic effects (CPE) were observed over 48 hours and TCID50 per ml (Tissue culture infectious dose 50%/mL) was calculated. The virus purification/titration was performed twice to produce two MelV solutions: the first was used for the preliminary optimization experiments, and the second to produce the results presented in this thesis, which had a TCID50/ml of 2.5 X 1010 . 2.3 Infectivity assay with glycan competition
2.3.1. Sample preparations for the competition infectivity assay
The following experiment was repeated to create a biological replicate. A 6-well plate was
18
prepared to contain 100% confluent attached amoeba cells. The culture medium was removed and replaced with 1 ml of PBS (-) which was then removed. MelV particles were inoculated at a Multiplicity of Infection (MOI) of 1 along with 250 µl of glycans to three of the wells:
100 mM of mannose (Man), 100 mM of galactose (Gal) or 100 mM of N-Acetylglucosamine (GlcNAc). For a positive control MelV was inoculated with culture medium alone in one of the well. The negative controls included one well with only culture medium and another with only 100 mM of Man. The plate was shaken manually for 1 min before incubation for 30 min on ice after which the amoeba were observed by light microscopy (Nikon eclipse TS100). For each well, 1 ml of PBS (-) was added and the amoeba/solution was resuspended, transferred to 1.5 ml tubes and centrifuged for 10 min at 1000 g, 4°C. The supernatants were discarded, and the pellets resuspended in 50 µl PBS (-) and stored at -20°C: these competition infectivity assay samples are to be titrated.
2.3.2. Virus titration assays
The titration of viruses was conducted using two different methods: TCID50 calculation and quantitative (q)PCR. For TCID50 titration, the MelV-infected amoeba samples (the
competition infectivity assay samples) freeze-stored in 50 µl of PBS (-) were frozen and thawed three times, to lyse the amoeba cells and hence release any MelV particles which may have entered. The volumes were made up to 100 µl of PBS (-) before being used to create serial dilutions of 10-2 to 10-6 with PPYGculture medium. The titration of each of the
duplicated samples were carried out in 48-well plates. The plates contained 30-50% confluent amoeba and the culture medium was removed before inoculation with 100 µl of sample serial dilutions; one lane of wells was kept as a negative control. The CPE was observed over 2 days before calculating TCID50 for each of the competition infectivity assay samples.
Statistical tests were carried out using Excel and GraphPad.
For qPCR titration, the 10-2 serial dilution of each of the MelV-infected amoeba samples were used (omitting the Man-based positive control) for genomic (g)DNA extraction (total DNA including viral dsDNA) using PureLink Genomic DNA MiniKit (Invitrogen) by following the product's protocol to prepare Gram negative bacterial cell lysate (Doutre et al., 2014). The final gDNA concentration was measured using a NanoDrop (ND-1000
Spectrophotometer v3.8, Saveen Werner) to confirm their applicability for qPCR. Forward and reverse primers for the major capsid proteins (MCP, obtained from Eurofins) were used, as shown in Table 2. These primers were designed by my peer Dylan Valli, whose master thesis was a parallel study establishing RNA silencing in MelV (Valli, 2020).
Primer Primer sequence Tm (°C) Amplicon size (bp)
Forward ACCACTGAACCAGCCTATGC 59.4 118
Reverse AATACCACGGCACCATCAGG 57.3 118
Table 2 DNA sequence of MelV major capsid protein (MCP) primers for qPCR.
To clarify, qPCR was performed for two biological replicates of the competition infectivity assay, where for each biological replicate 5 µl and 9 µl template gDNA were tested; for each
19
of these samples a technical replicate was made. The qPCR was carried out using
QuantStudio 6 Flex applied system (Thermofisher scientific), with SYBR green as a passive reference. The method used to calculate the 'relative concentration of MelV mRNA' within the samples as a percentage was based on their cycle threshold (Ct) values: [2 (Ct of positive control sample) - (Ct of sample)] * 100.
2.4 MelV SDS-PAGE and lectin blotting
A master mix consisting of 13 µl purified MelV sample, 5 µl LDS sample buffer (4X) and 2 µl sample reducing agent (10X) was prepared for SDS-PAGE. Three sets of wells were loaded with 3 µl of this master mix each alongside a well loaded with 8 µl of PageRuler Plus Pre-stained Protein Ladder (Thermo Scientific). The samples were separated by SDS-PAGE (NuPAGE 4-12% Bis-Tris Gel by Invitrogen) at 80 V for 0.5 h gradually increasing to 140 V.
One set of MelV sample plus ladder was cut away for Coomassie-blue staining (InstantBlue by Expedeon) and imaged using the Bio-rad ChemiDoc MP imaging system. The two other sets were cut for separate transfer onto PVDF membranes pre-soaked in methanol; the filter papers used were soaked in transfer buffer (25 mM of Tris base, 192 mM of glycine, 10%
(v/v) of methanol). The Bio-rad Trans-Blot Turbo Transfer System was used at 25 V constant for 1 h. The membranes were then soaked in blocking buffer Bovine Serum Albumin (BSA) 3% (w/v) in PBS overnight at 4°C. One membrane was blotted with 100 µg/ml of
concanavalin A (con A), Alexa Fluor 488 conjugate and the other with 5 µg/ml of wheat germ agglutinin (WGA), fluorescein conjugate for 1 h at RT. The membranes were washed thrice with PBST for 30 min each time. The two membranes were developed using the using the Bio-rad ChemiDoc MP imaging system under the appropriate settings.
2.5 Glycan-induced protein secretion in amoeba cells
This method, studying the nature of inducing amoeba secretions, has been optimized to quantify amoeba response to different glycans and understand its possible implication on MelV entry route. Amoeba cells of 80-100% confluence from two T75 flasks were resuspended and centrifuged for 10 min at 600 g, room temperature (RT). The pellet was washed with 10 ml of 200 mM Bis tris buffer (pH 6.5) for 10 min at 600 g, RT. The pellet was resuspended to starve overnight in 18 ml of PPY medium (no glucose) and equally distributed into two 6-well plates. The amoeba cells were then detached from the wells into 1.5 ml Eppendorf tubes, centrifuged for 10 min at 600 g, 4°C and washed with 1 ml of 200 mM Bis tris buffer for 10 min at 600 g, 4°C. The 12 separate pellets were resuspended by 130 µl of the following 12 variables: 1 mM, 10 mM and 100 mM concentration solutions of glycans Man, Gal and GlcNAc, 200 mM Bis tris (pH 6.5), PPY medium and PPYG medium (which has 100 µM glucose concentration). The tubes were incubated for 2 h, 35°C at 180 rpm, then centrifuged for 10 min at 1000 g, 4°C to ensure removal of cells. The pellets were checked for cell lysis using a light microscope and the supernatants were collected of which 13 µl of each were separated by SDS-PAGE (NuPAGE 4-12% Bis-Tris Gel by Invitrogen) at 80 V for 0.5 h, gradually increasing to 150 V for the next 1.5 h, along with PageRuler Plus Pre-stained Protein Ladder (Thermo Scientific). The gel was stained with Coomassie based
20
staining solution (InstantBlue by Expedeon) overnight before imaging with Bio-rad ChemiDoc MP imaging system. The gel was analysed with ImageJ. The remaining
supernatant containing secretory proteins induced by 100 mM Man was prepared and sent for Mass Spectrometry (MS) proteomics analysis for identifying the secretory proteins. This work was supported by the Mass Spectrometry Based Proteomics Facility in Uppsala.
2.6 Identification of amoeba phagosomal enzymes
2.6.1 Induction of phagosomal proteins and purification of amoeba phagosomes
The method of purifying amoebal phagosomes was modified from the previously reported methods (Okada et al., 2006; Reiner et al., 2009). This method consists of two steps. Step 1:
Induction of amoeba phagosomes by latex beads of polystyrene, 10% (w/v), and 0.8 micron in diameter (Sigma). The particle number was estimated to be 3.5 x 1011 particles/mL (Spherotech, 2019). Amoeba cells from three confluent 175 cm3 flasks were harvested, pelleted down and resuspended in 1 mL of PPYG culture medium. A cell count was conducted; an approximate calculation found there to be 2.32 x 106 amoeba cells/ml. A volume of 1 ml of beads were added with a concentration to amount to three beads per cell (approximately 7 x 106 beads). The amoeba and beads mix are incubated at RT for 10 min at 100 rpm and then placed on ice. The mix is centrifuged for 10 min at 600 g, 4°C. The cell/bead pellet is washed with PBS (-) for 10 min at 600 g, 4°C; the pellet is resuspended in 3 mL of HB+ (250 mM of sucrose, 3 mM of imidazole, 1 mM of EDTA, pH 7.4 with protease inhibitor cocktail from Sigma). Step 2: Isolation and purification of amoeba
phagosomes. The amoeba/bead HB+ suspension was homogenized through ball bearing using Isobiotec cell homogenizer with a 28-micron clearance ball though 40 strokes. The
homogenate was observed through phase contrast microscopy then centrifuged for 10 min at 2,000 g, 4°C and the post nuclear supernatant (PNS) was collected. A sucrose step gradient was set up with first 3.12 mL of 62% (w/w) sucrose in common buffer (CB: 3 mM of imidazole, 1 mM of EDTA, pH 7.4) mixed with the PNS at the bottom of a 14 mL
ultracentrifuge tube. This was followed by 3 mL of 35% (w/w) sucrose in CB, 2 mL of 25%
(w/w) sucrose in CB topped with approximately 4 mL of HB+. The step gradient was centrifuged for 1.5 h at 34,400 rpm (Sw40 Ti rotor, 210,000 g), 4°C.
Fourteen fractions were collected from the top to the bottom and 13 µL of each was used for SDS-PAGE analysis. The phagosomal fractions were collected from the first interphase between the HB+ and the 25% (w/w) sucrose. The purified phagosomes were sent for MS proteomics analysis for identifying the phagosomal enzymes as mentioned in section 2.6.2.
2.6.2 Identification of proteins by LC/MS/MS (Liquid chromatography tandem mass spectrometry)
From the purified phagosome fraction three separate samples were prepared. The first sample was simply 50 µL of the original phagosome fraction. The second sample was PBS (-) added to 650 µL of the phagosome fraction and ultra-centrifuged for 1 h at 100,000 g, 4°C. The pellet was dissolved in 20 µL PBS (-). The third sample was total protein precipitation using
21
methanol and chloroform extraction of 150 µL phagosome fraction. For the mannose induced enzyme secretion sample, total protein precipitation using methanol and chloroform
extraction was also performed with 117 µL of this secretion sample.
Methanol/chloroform protein precipitation entails the sample being mixed with 4 times its volumes in methanol, then with the equal amount of sample volume in chloroform and vortexed, then 3 times its volume in deionized water, vortexing in between. The sample mix is centrifuged for 5 min at 14,000 g, RT. The upper fraction was removed and 4.3 times the origin sample volume in methanol is added and the tube is inverted 3 times. The mix is centrifuged again for 5 min at 14,000 g, RT. The supernatant was discarded, and the protein pellet is obtained and not resuspended.
According to the LC/MS/MS results obtained from the three sample preparations the third phagosome sample, which underwent methanol/chloroform protein precipitation, provided the highest number of identified proteins and thus it was this sample's data which was further analysed. The experimental procedure, as described in the sample report from the Mass Spectrometry Based Proteomics Facility in Uppsala (Manoilov & Shevshenko, 2020), was largely similar for both the three phagosomal samples and the protein secretion sample. The whole volume of the sample was reduced, alkylated and in-solution digested with trypsin according to the standard operating procedure. For the three phagosomal samples purification was by Pierce C18 Spin Columns (Thermo Scientific) and the samples dried. For the enzyme secretion sample purification was by ZipTip® C18 tips and the sample dried. Dried peptides were resolved in 30 μL of 0.1% formic acid prior to nano-LC-MS/MS. The resulting peptides were separated in reversed-phase on a C18-column and electrosprayed on-line to a QExactive Plus Orbitrap mass spectrometer (Thermo Finnigan) with 35 min gradient. Tandem mass spectrometry was performed applying HCD. Sequest algorithm for database search was used, embedded in Proteome Discoverer 1.4 (Thermo Fisher Scientific) against the database
consisted of Acanthamoeba castellanii proteome extracted from Uniprot, Release March 2020. The search parameters were set to Enzyme: Trypsin (Fixed modification was
Carbamidomethyl (C), and variable modifications were Oxidation (M), Deamidated (NQ).
The search criteria for protein identification were set to at least two matching peptides of 95% confidence level per protein (Manoilov & Shevshenko, 2020).
2.7 Dissociation assay of the MelV particles using purified phagosomal enzymes
To elucidate the uncoating mechanism of MelV, a volume of 500 µl of the purified amoebal phagosomal enzymes was treated with 0.5% (v/v) NP-40 (Nonidet P-40 Substitute, Sigma) and aliquoted into 100 µl volumes where the pH was adjusted, using pH paper, to pH 5 using 8 µl of 10 mM citric acid. For this 100 µl mixture, 1 µl of purified MelV particles was added.
The final mixture was incubated at 32°C with 10 µl samples taken at 1 h, 3 h and 24 h from start of incubation. These samples were visualised on carbon coated grids with SEM (Quanta 650 FEG (Thermofisher Scientific) with and without 2% gadolinium staining.
22
3. RESULTS
3.1 Infectivity assay with glycan competition
3.1.1. Amoeba cells appear most mobilized when treated with 100 mM mannose
During the competition-infectivity assay, an important observation was made which led to the optimization of the protocol (see section 2.3.1). As shown in Figure 6, the amoeba cell
morphology and level of attachment to the wells differed after 30 min of incubation in different glycan-containing solutions. In the wells in which glycans were added, the amoeba appeared less attached and this was particularly the case when the glycan was Man.
3.1.2. TCID50/ml titration allude to competitive binding of Man and MelV to amoeba cells The TCID50/mL values here reveal which concentrations of MelV are required to cause 50%
CPE when amoeba cells are treated with different glycans. For this titration only one biological replicate was used. The mean MelV TCID50/ml values were 1 x 105, 1.05 x 106, 1.50 x 105, 5.00 x 104 and 3.50 x 102 for the positive control, Gal, GlcNAc, Man and the negative control, respectively (Figure 7). It is important to note the standard error bars are negligible for all variables except for Gal and Man where they are relatively large. The results indicate that only treatment with Man gave a TCID50/ml value smaller, by a factor of 2, than with the positive control where there was no glycan treatment. However, an unpaired T test between the two variables gave a two-tailed P value of 0.42. A one-way ANOVA test of the whole data set resulted in a P-value of 0.44; this included the result of another negative control using mannose which had a TCID50/ml of 0 which is not shown in figure 7.
Figure 6. Mobilised amoeba cells after glycans' treatment. A light microscopy image of the amoeba cells in wells during the competition-infectivity assay following 30 min incubation with 250 µl of different solutions: Positive control (No glycans, only MelV particles in PPYG medium), 100 mM of Gal, GlcNAc, Man or negative control (PPYG medium).
23
3.1.3. qPCR titration significantly indicates competitive binding between Man and MelV to amoeba cells
One of the two biological replicates of the competition infectivity assay completely failed to
Figure 8. qPCR amplification plot of samples from the competitive infectivity assay. The plot obtained from no glycan (positive control), Gal, GlcNAc and Man treatments with MelV particles are shown in blue, green, orange and grey. The plot obtained from the negative control without adding Mel particles is shown in yellow.
The sample gDNA was run with MelV major capsid protein (MCP) primers. The graph shows the average of raw fluorescence (Rn) curves for each sample based on two technical replicates.
0 1 2 3 4 5 6
20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40
Rn /Fluorescence
Cycle number
Positive Ctrl Gal GlcNAc Man Negative Ctrl Figure 7. MelV TCID50/ml values from the competition infectivity assay. The results from 100 mM Gal, GlcNAc and Man are shown. Positive control refers to where MelV particles only were inoculated whereas in the control (PPYG) only culture medium was added to amoeba cells. The P value for the whole data set and P value between the positive control and Man are shown.
24
amplify during qPCR, as was also the case when 9 µl gDNA templates were used. Hence, the following results are based on qPCR of two technical replicates of a single set of competition infectivity assay samples of which 5 µl gDNA was used as a template. MCP primers were set as a marker for the presence of MelV in the sample, hence, they also represent an endogenous control.
Figure 8 shows curves for the average raw fluorescence (Rn) values for each sample. The positive control has the highest curve followed by Gal, while GlcNAc and Man overlap. The negative control as expected has the lowest fluorescence curve. Notably, Supplementary Figure 1 shows that the technical replicates gave relatively harmonious amplification curves, where the curves represent the relative fluorescence (ΔRn) based on Sybr green. There are slight discrepancies in the curves representing the positive and negative control but especially in Man treated samples.
Averages of the technical replicates were used to calculate raw mean Cycle thresholds (Ct) at fluorescence Δ Rn threshold value of 0.75 (value was chosen for crossing with the linear phases of the curves), which is highlighted in Supplementary Figure 1. As shown in Supplementary Figure 2, the mean Ct value increases from 'no glycan' treatment (positive control) through to Gal, GlcNAc, Man treatment and negative control: 32.3, 34.6, 37.0, 37.8 and 38.6. A one-way ANOVA of this data set gave a P value of 0.0083. This corresponds to a 2.3, 4.8, 5.6 and 6.4 cycles difference between no glycan treatment than with Gal, GlcNAc, Man and negative control treatment (no virus/no glycan): that translates to approximately 5, 28, 50 and 80 times difference in the relative amount of MelV RNA. This is presented in Figure 9 which shows the relative concentration of MelV mRNA in these different samples.
Figure 9. The relative concentrations of MelV mRNA of different samples treated with different glycans. The values are based on the Ct values of a qPCR titration of the competition infectivity assay.
0 20 40 60 80 100 120
Positive ctrl Gal GlcNAc Man Negative ctrl Relative concentration of MelV mRNA
Condition used for competition infectivity assay
25 3.2 MelV contains at least a single
mannose-containing glycoprotein of approximately ~70 kDa in size
Following SDS-PAGE of purified MelV particles, Coomassie-blue staining revealed 11 prominent protein bands on the gel as shown in Figure 10.The darkest band
correlates to the 55 kDa major capsid protein (MCP) highlighted in the same figure. No fluorescence was detected following blotting with GlcNAc and sialic acid-binding WGA (Fig. 10, panel C).
However, a single band was detected when blotted with mannose-binding Concanavalin A (Con A) lectin correlating to a prominent band seen in the Coomassie-stained gel (Fig.
10, Panel A and B). The size of this band is approximately 70 kDa.
3.3 Mannose induces the most protein secretion from A. castellanii compared to other glycans, with secretion protein levels significantly higher than in phagosomes Glycans from amoeba food sources are exposed to both amoeba protein secretion and amoeba phagosomal proteins upon phagocytosis. The same can be said for MelV particles upon infection. It is, therefore, important to observe and study both the nature and quantity of these amoeba proteins to which both glycans and MelV particles are exposed to; these amoeba proteins could play a role in MelV uncoating (Yamauchi & Greber, 2016). The SDS-PAGE gel profile (Figure 11A) displays that A. castellanii does indeed release many proteins upon treatment to different glycans. There is a general trend where increased protein release is correlated with treatment of A. castellanii's to higher glycan concentrations. The quantified band intensity shown in Figure 11B seem to also indicate this.
At 100 mM concentration it was mannose which induced the highest levels of protein release, closely followed by GlcNAc and then Gal. Mannose still induces the most protein release even at lower concentrations, however, the amount induced by lower concentrations of Gal and GlcNac fluctuates. Interestingly, Bis-tris also induces protein release at levels similar to that of 1 mM GlcNAc; for the other controls, culture mediums PPY and PPYG, relatively low protein release is induced.
The SDS-PAGE profile of the phagosomal fraction collected (lane 3) displays significantly fainter protein bands then in the lower fractions, as shown in Figure 12.
A B C
Figure 10. Gels following SDS-PAGE of purified MelV, A) stained with Coomassie-blue, B) blotted with 100 µg/ml Concanavalin A (Con A) and C) 5 µg/ml Wheat Germ Agglutinin (WGA) lectins. The black arrow is highlighting the position of the major capsid protein on the Coomassie-stained gel.
26
0 5 10 15 20 25 30
Relative band density
Glycan/control 1 mM 10 mM 100 mM Control
Figure 12. SDS-PAGE of Acanthamoeba castellanii post nuclear supernatant (PNS) fractions. Fractions 1 to 13 correspond to lanes 1 to 13, following sucrose gradient ultracentrifugation. Lane 3/fraction 3 contains phagosomes as described previously by Okada et al., 2006 and Reiner et al., 2009. Protein ladder is BioRad Prestained Protein Standards Dual Color.
Figure 11. Secreted proteins from amoebal cells by glycan treatment. A) SDS-PAGE gel profile
(Coomassie blue stained) revealing protein secreted from A. castellanii following 2 h incubation with different concentrations (1 mM, 10 mM, 100 mM) of glycans Man, Gal and GlcNAc and controls Bis tris, PPY and PPYG mediums. B) The relative density of correlating bands (SDS-PAGE) of proteins secreted from A.
castellanii following 2 h incubation with different concentrations of glycans and controls.
27
3.4 Proteomics samples have different protein composition, yet the phagosome sample shares 70% of its proteins with secretory protein sample.
The estimated proteome size of A. castellanii is 14,939 proteins (UniProt Consortium, 2020).
Data analysis of the LC/MS-MS data (available in supplementary file) identified 310 proteins for the 100 mM Man induced protein secretions and 154 proteins from the isolated
phagosome fraction with proteins scores ranging from 852 to 1.77 and 225 to 1.89, respectively. These numbers reflect to some extend the band intensities seen in their prospective lanes/gels. A protein score is the sum of the scores of the individual peptides, where the Proteome Discoverer uses only the highest scoring peptide for each spectrum and sequence (Tania, 2014). The identified proteins are ranked in order of 'protein score' which represents the level of understanding on the protein (Thermo Fisher Scientific, 2012). Most of these proteins are poorly annotated, most either predicted or inferred from homology, others labelled putative, ambiguously characterized or completely uncharacterized. Only a small minority were annotated with experimental data present at the transcriptional level.
Approximately 70% of the phagosome proteins overlap with the mannose induced secreted proteins. All proteins were categorized according to affiliation to different biological or molecular roles and a comparison is made between the two samples, as shown in Figure 13.
The proportions of identified proteins associated with certain categories are largely similar between the two samples; this is the case with the cytoskeleton category which takes up 11.6% and 11.0% of glycan-induced and phagosomal proteins, respectively. The putative actin-1 protein, affiliated with the cytoskeleton, has the highest protein score for both
Figure 13. The graph highlights the different percentages of proteins identified from the samples.
The mannose-induced protein secretion and isolated phagosomes samples had a total number of identified proteins of 310 and 154, respectively. The identified proteins are categorised according to different biological and molecular roles.
0 5 10 15 20 25 30 35
Percentage of proteins
Biological/molecular role
Glycan induced protein secretion Isolated phagosomal proteins
28
samples. The proteins with the top fifteen scores for both the samples are listed in Supplementary Tables 1A and 1B. Similarly, the proportions of proteins categorized as 'uncharacterized and ambiguously characterized' in the two samples are alike. However, this 'uncharacterized and ambiguously characterized' category takes up only a fifth of the top- listed for the glycan-induced sample (Supplementary Table 1A) while up to a third of the top- listed for the phagosome sample (Supplementary Table 1B).
The percentage of proteins identified and categorized as 'markers' in glycan-induced and phagosome samples are also similar at 6.5% (20 proteins) and 7.1% (11 proteins),
respectively. They include proteins involved in cell redox homeostasis, proton transport and peroxidase activity. Further to this, the glycan-induced sample includes proteins that take on a wider range of molecular roles including peptidases and hydrolases. These are presented in Tables 3A and 3B which lists all the identified proteins categorized as 'Markers' in the two samples.
Notably, the proportion of identified proteins associated with translation, protein synthesis and signal transduction are significantly higher in the phagosome sample. On the other hand, the proportion of proteins associated with the stress response is moderately higher in the glycan-induced sample which also sees significantly higher proportion of metabolic and biosynthetic proteins.
Table 3A: LC-MS/MS identified proteins categorized as markers from a sample of 100 mM Mannose- induced protein secretion from A. castellanii.
Accession No. Protein name Biological/molecular role Protein score
% sequence coverage L8GWD2 Peroxiredoxin 2 Cell redox homeostasis 63.98 57.79 L8HH66 VATPase subunit
A, putative
Proton transport 42.05 28.01 L8H9V1 Vacuolar proton
pump subunit B
Proton transmembrane transport
37.33 28.63 L8GMP8 Thioredoxin Cell redox homeostasis 24.89 41.90 L8GT16 Nethylammeline
chlorohydrolase
Hydrolase 24.67 27.31
L8H0T8 Peptidase M20, putative
Peptidase activity 22.50 19.46 L8H083 Peroxidase Peroxidase activity 16.19 27.73 L8GXM1 Glutaredoxin,
putative
Cell redox homeostasis 16.08 44.66 L8H9M3 V-type proton
ATPase subunit
Proton transmembrane transport
15.96 60.91 L8GVQ1 Vacuolar proton
ATPase, putative
Proton transport 13.76 22.57 L8GXZ0 ATP synthase
subunit beta
Proton transport 12.91 7.78 L8HLS8 Peroxiredoxin Cell redox homeostasis 12.59 30.77 L8GFB7 X7, putative Oxidoreductase activity 11.41 23.84 L8HMI7 Survival protein
SurE subfamily protein
Hydrolase activity 9.75 27.24
L8H3T6 Oxidoreductase, aldo/keto reductase, putative
Oxidoreductase activity 9.36 13.64
L8HB23 Oxidoreductase, putative
Oxidoreductase activity 7.95 7.51
