The Potyviridae P1
protease modulates viral replication and host
defense responses
U N I V E R S I D A D
A UTONÓMA DE
M A D R I D
Fabio Pasin
Madrid 2015
FACULTAD DE CIENCIAS
Departamento de Biología Molecular
The Potyviridae P1
protease modulates viral replication and host
defense responses
Fabio Pasin, M.Sc.
Ph.D. Dissertation
Madrid 2015
Doctorado en Bioquímica, Biología Molecular, Biomedicina y Biotecnología
THESIS SUPERVISORS
Carmen Simón-Mateo
Centro Nacional de Biotecnología Madrid, Spain
Juan Antonio García
Centro Nacional de Biotecnología Madrid, Spain
THESIS ADVISOR
Iván Ventoso
Centro de Biología Molecular ‘Severo Ochoa’
Madrid, Spain
U N I V E R S I D A D
A UTONÓMA DE
M A D R I D
FACULTAD DE CIENCIAS
Departamento de Biología Molecular
iv
v
A mia Nonna, Amerigo e Nadia
PAXtibi mar ce evangelistamevs
vi
vii vii
CONTENTS
viii
ix
CONTENTS
SUMMARY xi
RESUMEN xv
INTRODUCTION 1
VIRAL POLYPROTEIN PROCESSING 3
VIRAL PROTEASES 5
THE POTYVIRAL P1 PROTEASE 6
INTRINSIC DISORDER AND PROTEASE ACTIVATION 8
PROTEASES IN HOST-PATHOGEN INTERACTIONS 9
VIRAL PROTEASES AS ANTIVIRAL TARGETS 9
AIM 11
RESULTS 15
PAPER I 19
PAPER II 35
PAPER III 69
DISCUSSION 81
P1 AND ANTIVIRAL RNA SILENCING 83
P1 SELF-CLEAVAGE AND POTYVIRAL RNA SILENCING SUPPRESSION 83 P1 AND P1a PROTEASE SELF-CLEAVAGE AND VIRAL HOST RANGE 84 P1 N-TERMINAL REGION: A NEGATIVE REGULATOR OF P1 SELF-CLEAVAGE 84
P1 AND PLANT IMMUNE RESPONSES 85
REGULATION OF P1 AND P1a PROTEASE SELF-CLEAVAGE AND HOST
ADAPTATION: OUR PROPOSED MODEL 86
FUTURE PERSPECTIVES 89
CONCLUSIONS 93
CONCLUSIONES 97
ACKNOWLEDGEMENTS 101
REFERENCES 105
x
AGO Argonaute protein
CI Cylindrical inclusion protein CP Coat protein
CVYV Cucumber vein yellowing virus DCL Dicer-like protein
ETI Effector-triggered immunity GFP Green fluorescent protein HCPro Helper component protease
IFN Interferon
iTRAQ Isobaric tags for relative and absolute quantification
NIaPro Nuclear inclusion “a” protease NIb Nuclear inclusion “b” protein
PAMP Pathogen-associated molecular pattern
PPV Plum pox virus
PR Pathogenesis-related protein RIN4 RPM1 interacting protein4
ROS Reactive oxygen species
RRL Rabbit reticulocyte lysate translation system
RSS RNA silencing suppressor
RT-qPCR Reverse transcriptase-quantitative PCR SA Salicylic acid
SIS SA-mediated immunity suppressor SUMO Small ubiquitin-like modifier
T2A Thosea asigna virus 2A peptide VPg Viral protein genome-linked
WGE Wheat germ extract translation system
ABBREVIATIONS
xi xi
SUMMARY
xii
xii
xiii
SUMMARY
W ith up to 190 assigned species to date, the Potyviridae is one of the largest families of plant-infecting viruses. The evolutionary success and economic impact of potyviruses is underlined by their wide host range, which includes herbaceous and woody plants, of both monocot and dicot groups. Similarly to picorna-like superfamily members, potyvirids have a single-stranded RNA genome and code for protein precursors that are processed by viral proteases.
L arge polyproteins of all members of the genus Potyvirus present at their N- terminal end the P1 serine protease. The last of the potyviral endopeptidases to be identified, P1 is activated by a yet-unknown host factor and self-cleaves from the remainder of the polyprotein. P1 self-release is indispensable for viral viability; P1 acts, however, as an accessory factor for viral amplification and two Potyviridae genera lack any P1-like protein. P1 participates in defining virus host range, although its specific contribution to potyviral infection is still unclear. We show that inefficient self- cleavage of P1 and a P1-like protease impairs viral RNA silencing suppression and viral infections in a host-dependent way. To study P1 protease regulation, we mapped the core C-terminal catalytic domain by bioinformatic analysis and in vitro activity experiments. We demonstrate that the P1 N-terminal region behaves as a negative regulator of P1 proteolytic activity. In the absence of the P1 protease antagonistic regulator, viral loads were reduced, while symptom severity and salicylate-dependent immune responses were increased.
F inally, we suggest that host-dependent modulation of a viral protease activity has
evolved to contain infection burden, and thus to escape from antiviral defenses
and improve long-term replicative capacity.
xiv
xv xv
RESUMEN
xvi
xvii
RESUMEN
C on 190 especies hasta la fecha, la familia Potyviridae es una de las más n umerosas de los virus de plantas. El éxito evolutivo y el impacto económico de los potyvirus está avalado por su amplio espectro de huésped, que incluye plantas herbáceas y leñosas, y grupos de plantas monocotiledóneas y dicotiledóneas. Al igual que los miembros de la superfamilia del tipo picorna- , los potyvirus tienen un genoma de RNA de cadena sencilla y polaridad positiva que se traduce en poliproteínas que se procesan proteolíticamente por proteasas virales.
L a última proteasa potyviral identificada, P1 es la serín proteasa localizada en el extremo N- terminal de las poliproteínas de todos los miembros del género Potyvirus. Una vez activada, por un factor del huésped aún desconocido, P1 se autoprocesa para liberarse del resto de la poliproteína y este procesamiento es indispensable para la viabilidad viral. P1 actúa como un factor accesorio para la amplificación viral, aunque un par de géneros de la familia Potyviridae carece de una proteína del tipo P1. P1 participa en la definición del espectro de huésped, sin embargo su papel específico en la infección de los potyvirus aún no se conoce.
Hemos comprobado que un auto-procesamiento ineficiente de P1 y de P1a, una proteasa del tipo P1, inhibe la supresión del silenciamiento de RNA y la infección viral de un modo específico de huésped. Para estudiar la regulación de la proteasa P1, hemos mapeado el dominio catalítico C-terminal mediante análisis bioinformáticos y experimentos de actividad proteasa in vitro. Hemos comprobado, que la región N- terminal de P1 se comporta como un regulador negativo de la actividad proteolítica de P1. En ausencia del regulador antagonista de la proteasa P1, la acumulación viral disminuye mientras que la severidad de los síntomas y las respuestas inmunes dependientes de salicilato aumentan.
T odos estos resultados sugieren que la modulación específica de huésped de
la actividad proteasa viral ha evolucionado para contener la infección viral y de
este modo evitar las respuestas de la planta y promover la capacidad de replicación
durante más largo plazo.
xviii
xviii
11
INTRODUCTION
22
INTRODUCTION
P lant viruses are one of the main causes of economic losses in crop production worldwide, and ecological changes, together with intensive agronomical practices, are predicted to favor the emergence of new plant viruses (Elena et al., 2014).
The Potyviridae family is a major group of plant viruses and its members are characterized by single-stranded positive-sense RNA genomes, with filamentous flexuous virus particles (Adams et al., 2012). The Potyviridae family includes 190 assigned species (International committee on taxonomy of viruses, 2015), which are distributed in seven genera whose viruses have monopartite genomes and one genus with bipartite genome members (Revers and García, 2015).
The representative Potyvirus and Ipomovirus genera contain viruses with monopartite genomes (Figure 1). Viral genomes are translated into a large polyprotein that is cleaved into at least 10 different products (Revers and García, 2015); a frameshift in the P3 cistron originates an additional polyprotein and the processed product P3N-PIPO (Chung et al., 2008). Controlled limited proteolysis of precursor polypeptides is a major process in the production of biologically active proteins (Krausslich and Wimmer, 1988); the release of functional polyprotein subunits in Potyviridae is mediated by virus-encoded endopeptidases (Adams et al., 2005a).
VIRAL POLYPROTEIN PROCESSING
Due to the limited coding capacity of their genome, viruses undertake distinct translational strategies (Walsh et al., 2013).
One of most widely employed by RNA viruses involves polyprotein synthesis from a long open reading frame (Wellink and van Kammen, 1988). Large polypeptide precursors are synthesized, and release of functional viral gene products is mediated by post-translational cleavages (Wellink and van Kammen, 1988); co-translational events might also take part in polyprotein processing (Ryan and Drew, 1994; Donnelly et al., 2001). Initially identified in poliovirus (Summers and Maizel Jr., 1968), polyprotein expression strategy recurs in the picorna-like superfamily, which includes the Potyviridae family as well as single-stranded RNA viruses that infect animals and diverse unicellular eukaryotes (Koonin et al., 2008).
Viral polyprotein processing for the release of
mature non-structural or structural subunits
is a strategy common in double-stranded
RNA viruses (Hudson et al., 1986; Birghan
et al., 2000), retroviruses (Dunn et al., 2002)
and also in DNA viruses (Torruella et al.,
1989; Webster et al., 1993; Simón-Mateo et
al., 1993). Putative protease domains were
identified in DNA mobile genetic elements
such as Helitron (Kapitonov and Jurka,
2007) and Polinton/Maverick transposons
(Kapitonov and Jurka, 2006; Pritham et al.,
44
INTRODUCTION
Potyviruses, Rymoviruses
poly A P1 HCPro P
K1CI
K2NIb CP
NIa VPgPro
Ipomoviruses
PN+PIPO NIa
CVYV, SqVYV
P1a
SPMMV CBSV CBSV Tritimoviruses, Poaceviruses
Brambyviruses Brambyviruses
Macluraviruses
Bymoviruses
Figure 1. Genome maps and virus-encoded proteases of Potyviridae members. Viral polyproteins and their mature subunits are represented as boxes. The ipomoviral P1a protein and potyviral, rymoviral proteins are indicated. Hatched boxes beneath P3 regions indicate PIPO frameshift products. The terminal viral protein genome-linked (VPg) is represented as a black ellipse. Virus-encoded proteases are colored: MEROPS clan PA, family S30 chymotrypsin-like serine proteases are shown in red, PA-C4 chymotrypsin-like cysteine proteases in orange, and CA-C6 papain-like cysteine proteases in yellow.
CBSV, Cassava brown streak virus; CVYV, Cucumber vein yellowing virus; SPMMV, Sweet potato mild
mottle virus; SqVYV, Squash vein yellowing virus; modified from (Revers and García, 2015).
55
INTRODUCTION
2007), suggesting a protease role in the release of functional transposon subunits from large precursor polyproteins, and a link with DNA virus evolution (Fischer and Suttle, 2011; Krupovic and Koonin, 2015).
Most eukaryotic mRNAs are monocistronic;
in contrast, viruses exploit many translational strategies to maximize their coding capacity (Firth and Brierley, 2012). One of the most used involves polyproteins, since it confers a clear advantage for viruses to encode precursors that deliver a set of functionally diverse proteins. Mature subunits are physically independent and free to move and distribute into different cell compartments (Krausslich and Wimmer, 1988). A full-length polyprotein is theoretically derived from a single translational event, compromising the timely expression of the individual viral cistrons. To overcome this possible drawback and successfully regulate replication, assembly and spreading stages, various post-translational mechanisms have evolved to modulate the spatial-temporal availability of functional viral proteins. For instance, it is not uncommon that the same polypeptide precursor is hydrolyzed by several endopeptidases. Cleavage kinetics are thus linked to enzyme processivity and, in trans- acting proteases, to the distinct affinity for the target cleavage sites (Dougherty and Dawn Parks, 1989; García et al., 1992; Revers and García, 2015). Low affinity protease target sites can generate cleavage intermediates with long half-lives and biological functions
different from those of the mature viral subunits (Wellink and van Kammen, 1988).
Mature subunits are themselves targets of degradation, and their different half-lives might have role in regulating viral replication and assembly stage succession (Wellink and van Kammen, 1988; Choi et al., 2013).
In summary, viral polyprotein maturation appears to be a highly regulated process that plays a relevant role in the control of pathogen infections (Revers and García, 2015).
VIRAL PROTEASES
Many viruses code for one or more proteases in their genomes. This allows viruses to coordinate their own highly specific substrate- enzyme systems, limiting the effects of possible cellular constraints (Krausslich and Wimmer, 1988). Due to size restrictions of viral genomes and compared to their cellular counterparts, is not surprising that viral proteases are generally the smallest proteases currently known (Tong, 2002). Even when cellular and viral proteases share the same backbone fold, they have little sequence similarity that might be restricted only to active site residues. This extensive divergence is underlined by the viral evolution of different substrate specificities and the emergence of new enzyme classes.
Examples are the cysteine proteases found
in several RNA viruses (e.g., the picornaviral
3C
Proand the potyviral NIaPro), which were
predicted (Gorbalenya et al., 1989), and later
confirmed by mutagenesis and structural
INTRODUCTION
studies (Seipelt et al., 1999; Phan et al., 2002), to be derived from a chymotrypsin-like serine protease ancestor. Apart from chymotrypsin- like serine and chymotrypsin-like cysteine proteases and based on their backbone folds, papain-like cysteine and pepsin-like aspartic proteases are other classes of these enzymes found in viruses (Tong, 2002). Viral enzymes with folds related to deubiquitinases and SUMO-1-specific cysteine protease were identified (Andrés et al., 2001; Weber, 2013;
Byrd et al., 2013; Bailey-Elkin et al., 2014), as well as proteases with unique structures such as the assemblin serine protease from Herpesviridae (Gibson, 2013). A more comprehensive list and classification of (viral and cellular) proteases can be found at the MEROPS database (Rawlings et al.,
2012). Many proteases were found in plant virus genomes (García et al., 1998), and a classification according to MEROPS is summarized in Table 1.
To date, three classes of proteases have been found in potyvirid genomes (Figure 1 and Table 1), (i) a chymotrypsin-like cysteine protease NIaPro present in all Potyviridae members and classified as MEROPS clan PA, family C4. (ii) A papain-like cysteine protease HCPro is classified as MEROPS clan CA, family C6 (Guo et al., 2011); HCPro is generally present in one copy, but absent in some ipomoviruses (Janssen et al., 2005;
Valli et al., 2007; Mbanzibwa et al., 2009), and an HCPro-like protease (P2-1) is found in bymoviruses (Adams et al., 2005a). (iii) A chymotrypsin-like serine protease P1 is classified as MEROPS clan PA, family S30. P1 is generally present in one copy, although some ipomoviruses have two P1- like proteins [i.e., P1a and P1b (Valli et al., 2007)] with different proteolytic specificities (Rodamilans et al., 2013), while Bymovirus and Macluravirus genera members lack any P1 homologue (Adams et al., 2005a; Kondo and Fujita, 2012).
THE POTYVIRAL P1 PROTEASE
P1 is located at the beginning of the polyprotein of all potyviruses, and comparisons between distinct viruses show that it is the most divergent of potyviral proteins in length and amino acid sequence (Shukla et al., 1991;
Table 1. MEROPS classification of proteases of plant viruses
Taxonomy
family MEROPS
Clan Family Catalytic type
Caulimoviridae AA A3 Aspartic
Benyviridae C- C36 Cysteine
Betaflexiviridae C- C23 Cysteine Closteroviridae C- C42 Cysteine
Tymoviridae C- C21 Cysteine
Secoviridae PA C3 Cysteine
Potyviridae PA C4 Cysteine
CA C6 Cysteine
PA S30 Serine
Luteoviridae* PA S39 Serine
Sobemovirus** PA S39 Serine
*An additional clan SK, family S14 protein is annotated [MEROPS:
MER000484], but its existence is unlikely since it corresponds to Potato leafroll virus 5’UTR [GenBank:D00530.1].
**Genus with unassigned family
INTRODUCTION
Adams et al., 2005b). Recombination and gene duplication events have contributed to P1 evolution, and were suggested to have facilitated successful potyviral adaptation to a wide range of host species (Valli et al., 2007).
Despite overall large variability of potyviral P1 proteins, the P1 C-terminal region is relatively well conserved. It harbors a serine protease domain that is responsible for cis-cleavage of the P1-HCPro junction and thus, P1 self- release from the remainder of the polyprotein (Adams et al., 2005a). Once released, the mature P1 C-terminal end is thought to be trapped in the active cleft, leading to self-inhibition of the trans cleavage activity (Verchot et al., 1992).
P1 was the last of the potyviral endopeptidases identified, and its proteolytic activity was hypothesized based on the finding that in planta cleavage at the N-terminal end of HCPro is catalyzed by a protease other than HCPro itself or NlaPro (Carrington et al., 1990;
Berger et al., 1989). The hypothesis was later confirmed by in vitro cleavage assays using the P1 of the potyviruses Tobacco etch virus (Verchot et al., 1991) and Tobacco vein mottling virus (Mavankal and Rhoads, 1991), and by protoplast transfection for the Potato virus Y P1 (Yang et al., 1998). From these experiments and compared to the rest of the potyviral proteases, a peculiarity emerged for P1. Whereas both NIaPro and HCPro cleave in the rabbit reticulocyte lysate (RRL) system,
no P1 protease activity could be detected in this translation system. In contrast, P1 self- cleaves in wheat germ extracts (WGE), and this evidence led the authors to suggest that P1 requires a host factor for activation (Carrington et al., 1990; Verchot et al., 1991;
Mavankal and Rhoads, 1991; Verchot et al., 1992). Although some P1 host interactors have been reported (Shi et al., 2007; Martínez and Daròs, 2014), the identity of the factor necessary for P1 protease activation remains elusive. Interestingly, in one study, P1 was reported to self-cleave in insect cells, albeit with relatively low processivity (Thornbury et al., 1993).
P1 acts as a viral accessory factor, since it contributes to potyviral amplification, but it is dispensable for infection onset and viral movement (Verchot and Carrington, 1995a). The P1 cistron is not essential, although inactivating mutations of its catalytic domain preclude virus viability (Verchot and Carrington, 1995b), making P1 an attractive target for the development of antiviral tools.
P1 involvement in the definition of virus
host range was highlighted (Salvador et
al., 2008; Maliogka et al., 2012). Whereas
many functions have been attributed to
P1, its contribution to potyviral infection is
still unclear [reviewed in (Rohožková and
Navrátil, 2011)]. P1 was described to interact
and co-localize with viral CI helicase, and
based on this finding, it was suggested to
form part of potyviral replication complexes
INTRODUCTION
(Arbatova et al., 1998; Merits et al., 1999;
Guo et al., 2001; Rohozková et al., 2014). P1 was shown to strengthen the RNA silencing suppressor activity of HCPro (Anandalakshmi et al., 1998; Kasschau and Carrington, 1998;
Rajamäki et al., 2005; Valli et al., 2006), but a recent report pointed to P1-mediated enhancement of HCPro synthesis rather than a specific P1 activity (Tena Fernández et al., 2013). In agreement with a stimulatory role in viral protein translation, P1 was found to bind to the 60S ribosomal subunit (Martínez and Daròs, 2014).
INTRINSIC DISORDER AND PROTEASE ACTIVATION
To perform its biological function(s), a protein is expected to fold into a defined tertiary structure. This paradigm loosens when we consider proteins with signaling, recognition, and regulation activities, that are thus involved in establishing numerous protein- protein/protein-ligand interactions (Xue et al., 2014). This kind of protein is often intrinsically disordered or enriched in disordered regions that do not have a stable or unique fold in solution. Viral proteins show a high occurrence of structurally flexible segments that are suggested not only to facilitate interaction with host components but also to increase viral adaptability (Tokuriki et al., 2009).
Compared to rigid protein structures, amino acid changes in intrinsic disordered proteins are predicted to be better tolerated and have
reduced functional drawbacks (Tokuriki et al., 2009). Moreover, mutations in intrinsically disordered regions might drive new domain evolution that enables viral proteins to target novel partners and to acquire new activities (Xue et al., 2014).
Disordered regions also contribute to the regulation of protein function, and flexible loops in peptidase precursors act in many cases as protease activation switches (Khan and James, 1998). An example is trypsin, which is synthesized as the inactive precursor trypsinogen, with a disordered loop partially obstructing the substrate-binding cleft (Khan and James, 1998). In phytopathogens, the AvrRpt2 cysteine protease, a bacterial effector, is synthesized and delivered to the host in an unfolded, inactive state (Coaker et al., 2006). Once in the host, AvrRpt2 is activated by a host cyclophilin peptidyl- prolyl cis/trans isomerase, and cleaves its host substrate RPM1 INTERACTING PROTEIN4 (RIN4) (Coaker et al., 2005).
Disordered regions are also relevant for the
maturation of viral polyproteins (Mathur et al.,
2012; Satheshkumar et al., 2005). Disorder
in proteases is thus a means to contain
detrimental protein degradation, to enable
spatial/temporal regulation of proteolytic
activity, and to promote multifunctionality
(Khan and James, 1998; Satheshkumar et
al., 2005; Marín and Ott, 2014).
INTRODUCTION
PROTEASES IN HOST-PATHOGEN INTERACTIONS
Several pathogens code for proteases that participate in promoting infectivity and virulence. They regulate pathogen replication or target specific host factors for inactivation or direct degradation. In animal viruses, selective promotion of viral RNA translation by proteolysis of factors needed to translate host transcripts is a common strategy [reviewed in (Lloyd, 2006; Komarova et al., 2009)]. Examples are poliovirus (Etchison et al., 1982), foot-and-mouth disease (Devaney et al., 1988) and human immunodeficiency viruses (Ventoso et al., 2001), whose proteases inhibit cap-dependent translation by cleaving eukaryotic initiation factor 4G.
Several viral proteases also process poly(A)- binding protein to provide shut-off of host translation (Lloyd, 2006; Komarova et al., 2009).
As the type I interferon (IFN) system has a central role in animal innate immune responses to viruses, it thus seems reasonable that viruses evolved strategies to neutralize IFN signaling components (Taylor and Mossman, 2013). Viral proteases also participate in counteracting innate immune responses by cleaving or inactivating key host IFN signaling elements (Foy et al., 2003;
Meylan et al., 2005; Neznanov et al., 2005;
Clementz et al., 2010). Host protein turnover and signaling rely largely on ubiquitin- proteosome-mediated degradation, and viral
proteases evade ubiquitin-dependent immune responses by removing the ubiquitin molecule from substrates (Frias-Staheli et al., 2007;
Isaacson and Ploegh, 2009). In an alternative strategy to escape immune responses, dependence of viral protease activation on the presence of defined host factors can restrict viral polyprotein maturation (Lackner et al., 2006). This mechanism was suggested to control viral virulence and to contain infection burden and defense activation (Lackner et al., 2004).
In potyviruses, polyprotein processing was shown to be an important regulatory mechanism, since non-lethal mutations in NIaPro or its recognition sites are critical for alternative host adaptation (Chen et al., 2008; Calvo et al., 2014). Whereas several effectors of plant-infecting bacteria are known to have protease activity and to promote the degradation of host defense components (Shao et al., 2002; Göhre and Robatzek, 2008;
Coaker et al., 2005; Gimenez-Ibanez et al., 2014), such evidence is still lacking for plant viruses. For instance, potyvirid proteases were found to interact with several host factors (Revers and García, 2015), but involvement of protease catalytic activity remains elusive, since none of these interactors were reported to be actively cleaved.
VIRAL PROTEASES AS ANTIVIRAL TARGETS
Viral proteases are central players in the
10 10
INTRODUCTION
replication of many viruses, and disturbance of their activity causes marked viral debilitation that could lead to loss of infectivity. Virus- encoded proteases are thus important antiviral targets (Krausslich and Wimmer, 1988); they are generally characterized by well-defined substrate specificities, and identification of viral protease-specific inhibitors with no/minimal interference with host-cell metabolism could provide clinically or agronomically useful antiviral agents (Krausslich and Wimmer, 1988; Tong, 2002).
One of the major successes in this field was the identification of protease inhibitors for human immunodeficiency virus therapy (Wu et al., 2014) and NS3-4A protease inhibitors for the treatment of hepatitis C virus. Such drugs were recently approved for medical use, and second-generation protease inhibitors aimed at a better dosage schedule and improved tolerance have reached clinical development (Clark et al., 2013). Emergence of protease inhibitor resistant variants was reported (Condra et al., 1995; Sarrazin et al., 2007; Kieffer and George, 2014), although combination therapies with a non-overlapping resistance profile is expected to increase the genetic barrier to resistance (Rong et al., 2010; Clark et al., 2013).
In contrast to medically relevant pathogens, identification of antiviral drugs against agronomical threats suffers a considerable research gap. With regard to plant-infecting viruses and specifically potyviruses, efforts
to use protease inhibitors as antiviral tools showed only limited success (García et al., 1993; Gutierrez-Campos et al., 1999; Wen et al., 2004). Development of fluorometric protease assays suitable for high-throughput screens (Yoon et al., 2000) and availability of small-molecule collections with a high degree of structural diversity (Schreiber, 2000; Galloway et al., 2010) could overcome the technical limitations of the early studies.
The identification of protease inhibitors is
thus an “old concept”, but still an attractive
antiviral strategy against plant viruses. This
is especially the case for viruses that affect
high-value crops such as fruit trees and
vines, and when host resistance genetic traits
have not been identified or their introgression
into varieties of commercial interest is too
laborious.
11 11
AIM
12
1
AIM
T he aim of the present work was to elucidate the role of the Potyviridae P1 protease in viral infections and host interactions.
To fulfill this aim, Plum pox virus (PPV), a representative member of the genus Potyvirus, was used as model system. PPV codes for a canonical P1 protease at the N-terminal end of the viral polyprotein (Šubr and Glasa, 2012), and is the causative agent of sharka disease, which affects stone-fruit production worldwide (Cambra et al., 2006; García et al., 2014).
In addition, PPV-based chimeras were generated to study the effect of a P1- like protease from a genus Ipomovirus member, P1a protein from Cucumber vein yellowing virus (CVYV). CVYV P1a is phylogenetically related to the P1 sequence of the potyviruses Papaya ringspot virus and Zucchini tigré mosaic virus (Valli et al., 2007; Romay et al., 2014). Although P1a function is not known, its proteolytic activity shares similarity with PPV P1 (Rodamilans et al., 2013).
The specific objectives were:
• To develop a method to facilitate quantitative comparisons of different viral clone accumulations,
• To characterize PPV P1 proteolytic activity in in vitro cleavage assays, and to define its contribution in potyviral infections, and
• To elucidate the relevance of CVYV P1a self-cleavage activity in host
adaptation.
14
15 15
RESULTS
1
1
1
1
RESULTS
T he materials and methods used, and
the results obtained in this study are
presented in the three publications and
associated supporting materials included
below.
1
1
1
RESULTS
ABSTRACT
Background: Fluorescent proteins are extraordinary tools for biology studies due to their versatility; they are used extensively to improve comprehension of plant-microbe interactions. The viral infection process can easily be tracked and imaged in a plant with fluorescent protein-tagged viruses. In plants, fluorescent protein genes are among the most commonly used reporters in transient RNA silencing and heterologous protein expression assays. Fluorescence intensity is used to quantify fluorescent protein accumulation by image analysis or spectroscopy of protein extracts; however, these methods might not be suitable for medium- to large-scale comparisons.
Results: We report that laser scanners, used routinely in proteomic studies, are suitable for quantitative imaging of plant leaves that express different fluorescent protein pairs. We developed a microtiter plate fluorescence spectroscopy method for direct quantitative comparison of fluorescent protein accumulation in intact leaf discs. We used this technique to measure a fluorescent reporter in a transient RNA silencing suppression assay, and also to monitor early amplification
dynamics of a fluorescent protein-labeled potyvirus.
Conclusions: Laser scanners allow dual- color fluorescence imaging of leaf samples, which might not be acquired in standard stereomicroscope devices. Fluorescence microtiter plate analysis of intact leaf discs can be used for rapid, accurate quantitative comparison of fluorescent protein accumulation.
RELEVANCE
A fluorometer-based method and a strand- specific RT-qPCR assay were developed to monitor early amplification dynamics and viral RNA accumulation of a GFP-tagged PPV clone.
CONTRIBUTIONS
F. Pasin performed the experiments, contributed to experiment design, data analysis, and paper writing.
PAPER I
Rapid fluorescent reporter quantification by leaf disc analysis and its application in plant-virus studies
Plant Methods. 2014 Jul 5;10:22. doi: 10.1186/1746-4811-10-22
Fabio Pasin, Satish Kulasekaran, Paolo Natale, Carmen Simón-Mateo and Juan Antonio
García
20
21
RESULTS
PAPER I. Full-length Publication
M E T H O D O L O G Y Open Access
Rapid fluorescent reporter quantification by leaf disc analysis and its application in plant-virus studies
Fabio Pasin
*, Satish Kulasekaran, Paolo Natale, Carmen Simón-Mateo and Juan Antonio García
Abstract
Background: Fluorescent proteins are extraordinary tools for biology studies due to their versatility; they are used extensively to improve comprehension of plant-microbe interactions. The viral infection process can easily be tracked and imaged in a plant with fluorescent protein-tagged viruses. In plants, fluorescent protein genes are among the most commonly used reporters in transient RNA silencing and heterologous protein expression assays.
Fluorescence intensity is used to quantify fluorescent protein accumulation by image analysis or spectroscopy of protein extracts; however, these methods might not be suitable for medium- to large-scale comparisons.
Results: We report that laser scanners, used routinely in proteomic studies, are suitable for quantitative imaging of plant leaves that express different fluorescent protein pairs. We developed a microtiter plate fluorescence spectroscopy method for direct quantitative comparison of fluorescent protein accumulation in intact leaf discs. We used this technique to measure a fluorescent reporter in a transient RNA silencing suppression assay, and also to monitor early amplification dynamics of a fluorescent protein-labeled potyvirus.
Conclusions: Laser scanners allow dual-color fluorescence imaging of leaf samples, which might not be acquired in standard stereomicroscope devices. Fluorescence microtiter plate analysis of intact leaf discs can be used for rapid, accurate quantitative comparison of fluorescent protein accumulation.
Keywords: Fluorescent protein, Fluorescence spectroscopy, Microtiter plate, RNA silencing, Plant virus
Background
Reporter genes and their products are valuable tools for plant studies, due to the ease of imaging and quantifica- tion of the proteins encoded [1]. Fluorescent proteins are widely employed as reporters, since they have no require- ments for exogenous substrate/co-factors and do not interfere with cell growth or function [2]. These proteins can be detected and imaged in live tissue without cell lysis or biochemical analysis, and they allow optical exploration of cell structures and molecule dynamics as well as patho- gen monitoring with minimal sample preparation [3].
Use of fluorescent protein as a quantitative reporter in- cludes evaluation of new vectors for heterologous protein expression and of promoter activity, translational regula- tion and transient RNA silencing [4-8]. In plant pathology and symbiosis studies, fluorescent proteins are an
important aid for monitoring infection/colonization onset and spreading, and thus facilitate comprehension of host- microbe interactions. Since the first demonstrations that plant viruses are useful vectors for foreign sequence trans- fer to their hosts [9-12], several genes were shown to be suitable RNA virus reporters; they include those that en- code chloramphenicol acetyltransferase, firefly and Renilla luciferases, β-glucuronidase, anthocyanin biosynthesis transcription factors, and Aequorea victoria green fluores- cent protein (GFP) [12-18].
Compared to other markers, fluorescent protein genes inserted into viral genomes offer good reporter stability [19], viral localization to individual cells, and monitoring of co-infection with differently-labeled viruses [20,21]. A further advantage of these proteins is that their fluores- cence intensity is directly proportional to protein amount and can be used for quantification [22,23]. Although GFP fluorescence can be quantified by image analysis [24,25], this involves time-consuming steps that can be overcome
* Correspondence:[email protected]
Centro Nacional de Biotecnología (CNB-CSIC), Darwin 3, Madrid 28049, Spain
PLANT METHODS
© 2014 Pasin et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Pasin et al. Plant Methods 2014, 10:22 http://www.plantmethods.com/content/10/1/22
22
RESULTS PAPER I. Full-length Publication
by spectrofluorometric measurement of intact plant or- gans or protein extracts from GFP-expressing samples [23,26,27].
A microplate assay was recently described that measures luciferase activity in intact leaf discs [28]. In a similar ap- proach, here we evaluated the use of 96-well plate readers for rapid quantification of two A. victoria GFP variants, the ultraviolet (UV)-excitable mGFP5 [29] and a mutant with enhanced brightness sGFP(S65T) [30]. The method was applied in viral RNA silencing suppressor studies and in accumulation monitoring of GFP-labeled Plum pox virus (PPV) clones. A palette of engineered monomeric fluorescent proteins was expressed transiently in plants (Table 1) and shown to be easily quantifiable by direct leaf disc analysis.
Results and discussion
Laser scanner imaging of Nicotiana benthamiana leaves
GFP variants such as mGFP5 [29], which can be excited by long-wavelength ultraviolet (UV) light, are used fre- quently in plant studies of species other than the small- sized Arabidopsis, since fluorescence imaging of whole specimens is constrained by objective lens size of fluores- cence (stereo)microscopes. The need for fluorescence mi- croscopes is overcome by use of UV lamps as excitation sources, although this restricts fluorophore choice and limits multi-fluorescence imaging. Scanners with excitation lasers at 457, 488, 532, and 633 nm are used for fluores- cence imaging in two-dimensional difference gel analysis systems [36] and have a relatively large glass platen (for ex- ample, 35 cm × 43 cm, in the Typhoon 9400). As a 633 nm laser might be unsuited to leaf tissue imaging due to inter- ference from chlorophyll autofluorescence [37], we tested whether 457, 488 and 532 nm lasers can be used for im- aging N. benthamiana leaves that transiently express fluor- escent proteins. Plant expression vectors bearing coding sequences for mGFP5 or a monomeric red fluorescent pro- tein TagRFP-T [35] were delivered to plants by Agrobacter- ium infiltration. Tomato bushy stunt virus p19 RNA silencing suppressor was co-expressed to increase yield of
the heterologous proteins delivered [38]. At 6 days post- agro-infiltration (dpa), N. benthamiana leaf fluorescence was acquired after excitation with 488 nm and 532 nm la- sers. A strong signal was detected in leaf patches expressing the fluorescent proteins. Only background signal was detected in non-infiltrated leaf areas and when non- optimal excitation/emission conditions were used, i.e., mGFP5-expressing patches imaged with TagRFP-T set- tings (Ex532/Em580) and TagRFP-T-expressing patches im- aged with mGFP5 settings (Ex488/Em526) (Figure 1A, C).
To expand fluorophore choice, we tested a cyan (mTFP1;
[32]) and a yellow (mPapaya1; [34]) fluorescent protein, and found them to be easily imaged in agro-infiltrated leaves (Figure 1B, D). These results support the suitability of mGFP5/TagRFP-T and mTFP1/mPapaya1 pairs for laser scanner bicolor imaging in plants.
Spectral properties and quantification of plant-expressed fluorescent proteins
A fluorescence signal acquired by laser scanner imaging is suitable for quantitative comparisons (Figure 1C, D), as is done routinely in proteomic studies [36]. Image analysis can be a lengthy process, however, and signal quantification can be affected if leaf lamina occupy dif- ferent focal planes during the acquisition step. As mi- crotiter plate readers are available for medium-high throughput analysis, we used a monochromator-based plate reader to analyze the fluorescence signal from in- tact leaf discs collected from agro-infiltrated patches (Figure 2A). We found that fluorescence properties of mGFP5 could be measured without extract preparation (Figure 2B), and excitation and emission spectra closely resembled those reported [29]. Five-fold dilutions of the mGFP5-Agrobacterium strain were used in a transient expression assay. Fluorescence intensity values were consistent with the amount of bacteria delivered (Pearson R
2= 0.9855; n = 4; Infinite M200 values were considered) and independent of the fluorescent plate reader used (Figure 2C).
Table 1 Reporter proteins and fluorescence analysis conditions evaluated
Reporter Laser scanner imaging Plate reader FI quantification Species Structure Ref.
Laser (nm) Em (nm) Ex (nm) Em (nm)
mTagBFP2 n.a.1 n.a. 400/9 455/20 Entacmaea quadricolor Monomer [31]
mTFP1 457 526SP 450/9 480/20 Clavularia sp. Monomer [32]
mGFP5 488 526SP 485/9 535/20 Aequorea victoria Weak dimer [29]
sGFP(S65T) 488 526SP 485/9 535/20 Aequorea victoria Weak dimer [30]
mNeonGreen n.t.2 n.t. 500/9 530/20 Branchiostoma lanceolatum Monomer [33]
mPapaya1 532 555/20 520/9 550/20 Zoanthus sp. Monomer [34]
TagRFP-T 532 580/30 560/9 595/20 Entacmaea quadricolor Monomer [35]
1n.a., not applicable.
2n.t., not tested.
Pasin et al. Plant Methods 2014, 10:22 Page 2 of 11
http://www.plantmethods.com/content/10/1/22
2
RESULTS
PAPER I. Full-length Publication
Rapid fluorometer GFP quantification in transient RNA silencing assays
To determine whether leaf disc fluorescence intensity can be used for quantitative analysis of GFP accumula- tion in leaf tissue, we co-expressed mGFP5 with PPV si- lencing suppressor constructs. These included HCPro with the parent sequence (WT), with the L134H substitu- tion (LH; which abolishes RNA silencing suppression activity [40,41]), and HCPro into which amino acids REN- 239, 240, 241 were replaced by alanines (AS9). The AS9 construct was tested since the corresponding HCPro mu- tants in Tobacco etch virus (TEV) and Turnip mosaic virus (TuMV) are silencing suppression-defective [42-44], but no data are available for PPV. The red TagRFP-T was also included to test for interference with mGFP5 fluorescence analysis (Figure 3A). At 6 dpa, laser scanner imaging de- tected bright fluorescence in patches in which mGFP5 was co-delivered with wild-type HCPro (WT, Figure 3A). Ana- lysis on a 96-well plate reader showed a significantly higher fluorescence signal in WT samples than in those of
the other constructs tested, i.e., LH, AS9 and red fluores- cent protein samples (Figure 3B). Fluorescence intensity in AS9 samples was equivalent to that in silencing suppres- sion mutant L134H samples. These results suggest that the PPV HCPro AS9 (REN-239, 240, 241 replacement) construct behaves like the TEV and TuMV HCPro AS9 mutants. In immunoblot analysis, mGFP5 protein accu- mulation correlated positively with fluorescence signal quantification values (Pearson R
2= 0.9989; Figure 3C). In a parallel experiment, transient delivery of HCPro proteins was confirmed by anti-PPV HCPro immunoblot analysis of samples co-infiltrated with p19 (Figure 3D). We de- tected no TagRFP-T interference in mGFP5 quantification assays (Figure 3A, B).
Monitoring of plant viral amplification dynamics by fluorometer analysis
We used sGFP(S65T), a synthetic GFP version with en- hanced brightness [30], as a sensitive reporter to follow PPV early amplification in plant tissue. The pSN-PPV
Figure 1 Laser scanner imaging of fluorescent protein-expressing leaves. Fluorescent proteins were transiently expressed by co-infiltrating N. benthamiana leaf tissue with an Agrobacterium pSN.5 p19 culture plus cultures of Agrobacterium containing pBin-35S-mGFP5 (mGFP5), pSN.5 TagRFP-T (TagRFP-T), pSN.5 mTFP1 (mTFP1) or pSN.5 mPapaya1 (mPapaya1). Fluorescence was imaged by leaf laser scanning. (A) Signal acquired at 6 dpa for TagRFP-T (red) and mGFP5 (green); green and red channel images were merged. Scale bar, 2 cm. (B) Signal at 3 dpa for mTFP1 (cyan) and mPapaya1 (yellow); cyan and yellow channel images were merged. Scale bar, 2 cm. (C,D) Surface plots of infiltrated patches from above images.
Pasin et al. Plant Methods 2014, 10:22 Page 3 of 11
http://www.plantmethods.com/content/10/1/22
24
RESULTS PAPER I. Full-length Publication
binary vector [45] was used to deliver sGFP(S65T)-tagged PPV by agro-inoculation (Figure 4A). As anticipated, sGFP(S65T) fluorescence was readily detected in infected leaves (Figure 4B). Fluorophore spectra were confirmed by analysis of leaf discs from inoculated leaves. Compared to mGFP5, sGFP(S65T) retained the blue light excitation peak but lacked the UV peak (Figure 4C).
We further compared GFP fluorescence intensity (FI) signal dynamics of leaves agro-inoculated with pSN-PPV (wtPPV) or with pSN-PPV P1-S (S259A), a cDNA clone of a non-infectious PPV mutant with silencing suppres- sion defects [45]. Whereas the FI of the PPV S259A clone peaked at 2 dpa, FI of wtPPV continued to increase over the 6-day time course (Figure 4D). In agro-inoculated leaves, fluorescence quantification results were corrobo- rated by immunoblot analysis of GFP and PPV coat pro- tein (CP, Figure 4E). We developed a strand-specific quantification of PPV RNA by RT-qPCR assay (Additional file 1), and viral RNA amounts at 6 dpa were consistent with protein determinations (Figure 4F). β-glucuronidase and luciferase genes can be used to analyze potyviral accu- mulation, genome amplification rates and cell-to-cell movement [14,46-48]; here we show that detection of a
GFP-tagged virus is quite straightforward, since no sub- strates/co-factors are needed and sample preparation re- quirements are minimal.
Direct leaf disc analysis of engineered monomeric fluorescent proteins
There is a wide variety of engineered fluorescent proteins with improved optical and stability properties and many spectral variants were obtained by evolution of the A.
victoria GFP sequence. For multicolor experiments, how- ever, fluorescent proteins with minimal sequence similarity are desirable, to reduce post-transcriptional gene silencing events and assure immunodetection specificity. We evalu- ated the novel bright fluorescent proteins blue mTagBFP2 [31], cyan mTFP1 [32], green mNeonGreen [33], yellow mPapaya1 [34] and red TagRFP-T [35], all derived from species other than A. victoria (Table 1), for transient ex- pression in plants. Fluorophore spectral properties and fluorescence intensity were easily determined using intact leaf discs collected from tissue agro-infiltrated with the corresponding constructs (Figure 5). We also show that the FI of different fluorophores can be measured simultan- eously and, in multicolor experiments, the choice of
Figure 2 Spectral properties and fluorescence quantification of GFP. Fluorescent protein was transiently expressed by co-infiltrating N.
benthamiana leaf tissue with an Agrobacterium pSN.5 p19 culture plus a strain with no expression vector (Φ), or 5-fold dilutions of Agrobacterium containing pBin-35S-mGFP5 (mGFP5 at OD6000.50, 0.10 and 0.02). (A) At 3 dpa, mGFP5 (green) fluorescence was imaged by leaf laser scanning.
(B) In a plate reader, excitation (dotted lines) and emission spectra (solid lines) were measured from leaf discs of tissue agro-infiltrated with mGFP5 (green) or no expression vector (black). Relative fluorescence intensity (RFI) was plotted using mGFP5 peaks equal to 100. Ultraviolet (UV) wavelengths are in gray, visible spectrum colors were assigned as described [39]. (C) Box-plot graphs show quantification values from n = 8 samples/condition. Fluorescence intensity of leaf discs agro-infiltrated with mGFP5 strain dilutions, no expression vector (Φ) or non-treated samples (N) was acquired in monochromator-based (Infinite M200) and two filter-based (Appliskan and Victor X2) plate readers. Fluorescence intensity is expressed in arbitrary units (a.u.).
Pasin et al. Plant Methods 2014, 10:22 Page 4 of 11
http://www.plantmethods.com/content/10/1/22
25
RESULTS
PAPER I. Full-length Publication
Figure 3 Quantification of GFP accumulation in transient RNA silencing assay. (A) GFP was transiently expressed by co-infiltrating N. benthamiana leaf tissue with an Agrobacterium pBin-35S-mGFP5 culture plus cultures of Agrobacterium containing pSN.5 TagRFP-T (RFP), pSN.5 HC-L134H (LH, producing PPV HCPro L134H mutant), pSN.5 HC-AS9 (AS9, producing a PPV HCPro mutant in which amino acids REN-239, 240, 241 were replaced by alanines) or pSN.5 wtHC (WT, producing wild-type PPV HCPro). At 6 dpa, leaf fluorescence was acquired by laser scanning using Ex488/Em526 (green) and Ex532/
Em580 (red); the image overlay is shown (Merged). (B) GFP fluorescence intensity of the agro-infiltrated leaf patches was quantified in a 96-well plate reader.
RFI was plotted using WT mean value equal to 100. Bar graph shows mean ± SD (n = 14 biological replicates from two independent Agrobacterium cultures); the difference between the results marked with different letters is statistically significant, p < 0.01, one-way Anova and Tukey’s HSD test. (C) GFP protein accumulation in infiltrated leaves at 6 dpa was assessed by immunoblot analysis. Relative GFP signal intensities are indicated using average WT equal to 100; the difference between the values marked with different letters is statistically significant, p < 0.01, one-way Anova and Tukey’s HSD test. Each lane represents a pool of 3 or 4 leaf samples infiltrated with two independent Agrobacterium cultures. N, non-treated leaf sample. Ponceau red-stained blot as loading control. (D) HCPro expression by the binary vectors tested was assessed by HCPro immunoblot analysis of leaf co-infiltrated with an Agrobacterium pSN.5 p19 culture (6 dpa). Each lane represents a pool of infiltrated leaf samples. N, non-treated leaf sample. Ponceau red-stained blot as loading control.
Pasin et al. Plant Methods 2014, 10:22 Page 5 of 11
http://www.plantmethods.com/content/10/1/22
2
RESULTS PAPER I. Full-length Publication
reporters with minimal spectral overlap assures signal specificity (Additional file 2).
Conclusions
We present laser scanning as an alternative method for fluorescence imaging of plant samples that, due to their size, cannot be acquired in their entirety in standard fluorescence stereomicroscopes. Dual-color fluores- cence imaging of leaf samples is achieved using fluoro- phore combinations with minimal spectral overlap, such
as mGFP5/TagRFP-T and mTFP1/mPapaya1, and image analysis can be used for raw quantitative comparisons.
We show that fluorescence plate readers are extremely powerful tools for medium-high throughput analysis of fluorescent proteins expressed in plant tissue, making it feasible to collect data from a 96-well plate in a few mi- nutes. Fluorescence intensity is readily quantified in leaf discs, with no need to prepare protein extracts. A large number of improved fluorescent proteins have been de- veloped, and proteins with reduced biological half-life,
Figure 4 Monitoring of GFP-tagged virus amplification dynamics by fluorescence spectroscopy. GFP-tagged viral cDNA clones pSN-PPV (wtPPV, wild-type PPV) and pSN-PPV P1-S (S259A, in which P1 protease catalytic amino acid S259 was replaced by alanine) were delivered to plants by agro-infiltration. (A) Diagram of wild-type PPV (wtPPV) genome originated following pSN-PPV agro-infiltration. Hatched box indicates P3N-PIPO protein. The reporter sGFP(S65T) gene is inserted between NIb and CP coding sequences. (B) N. benthamiana plants were challenged with pSN-PPV, and fluorescence of systemically infected leaves was detected by laser scanning (10 dpa; green). (C) Excitation (dotted lines) and emission spectra (solid lines) of sGFP(S65T) were measured from pSN-PPV agro-inoculated leaf discs (green); leaves infiltrated with an Agrobacterium culture without expression vectors were used as control (black). Relative fluorescence intensity (RFI) was plotted using sGFP(S65T) peaks equal to 100. UV wavelengths are in gray, visible spectrum colors were assigned as described [39]. (D) GFP fluorescent intensity (RFI) from infiltrated leaves was quantified in a 96-well plate reader and plotted using average wtPPV value at 6 dpa equal to 100. Line graph shows mean ± SD (n = 16 samples/condition, from two independent Agrobacterium cultures). (E) Amount of GFP protein and PPV CP in infiltrated leaves at 6 dpa was assessed by immunoblot analysis. Each lane represents a pool of 3 or 4 leaf samples infiltrated with two independent Agrobacterium cultures. Ponceau red-stained blot is shown as loading control. (F) Amount of viral (+)RNA and (−)RNA from inoculated leaves at 6 dpa was quantified by RT-qPCR and plotted using average wtPPV value equal to 100. Bar graph shows mean ± SD (n = 4 biological replicates, from two independent Agrobacterium cultures); ***p < 0.001, Student’s t-test.
Pasin et al. Plant Methods 2014, 10:22 Page 6 of 11
http://www.plantmethods.com/content/10/1/22
2
RESULTS
PAPER I. Full-length Publication
rapid choromophore maturation and photoactivable var- iants [3,49-51] might be used to increase assay sensitivity and temporal resolution for kinetic studies. We show that co-expression of TagRFP-T has no appreciable effect
on fluorescence intensity quantification of mGFP5. A battery of fluorescent proteins that have minimal se- quence identity with the widely used A. victoria GFP se- quence was quantified easily in a monochromator-type
Figure 5 Direct leaf disc analysis of engineered monomeric fluorescent proteins. Fluorescent proteins were transiently expressed by co-infiltrating N. benthamiana leaves with an Agrobacterium pSN.5 p19 culture plus cultures of Agrobacterium containing pSN.5 mTagBFP2 (mTagBFP2), pSN.5 mTFP1 (mTFP1), pSN.5 mNeon (mNeon), pSN.5 mPapaya1 (mPapaya1), pSN.5 TagRFP-T (TagRFP-T) or a strain with no expression vector (Φ).
(A) At 6 dpa, cell fluorescence was imaged by confocal microscopy. Fluorophore excitation (dotted lines) and emission spectra (solid lines) from agro-infiltrated leaf discs were measured in a 96-well plate reader. Leaves infiltrated with an Agrobacterium culture without expression vectors were used as control (black lines). Relative fluorescence intensity (RFI) was plotted using fluorophore peaks equal to 100. UV wavelengths are in gray, visible spectrum colors were assigned as described [39]. (B) Box-plot graphs show quantification values from n = 8 samples/condition. Fluorescence intensity of the leaf discs agro-infiltrated with the indicated fluorescent protein-expressing plasmid or without expression vector (Φ) was measured in a monochromator-based plate reader. RFI was plotted using each fluorophore mean value equal to 100.
Pasin et al. Plant Methods 2014, 10:22 Page 7 of 11
http://www.plantmethods.com/content/10/1/22
2
RESULTS PAPER I. Full-length Publication
plate reader. We anticipate that the method presented will aid in the design of fluorescence-based experiments with single and multiple reporter genes and facilitate comparisons of fluorophore amounts.
Methods
DNA plasmids and constructs
The binary vector pSN-PPV bearing a full-length cDNA copy of a PPV isolate and its variant pSN-PPV P1-S were reported [45]. An Agrobacterium strain GV3101 containing the binary vector pBin-35S-mGFP5 was kindly provided by D. Baulcombe (University of Cambridge, Cambridge, UK).
For the remaining transient expression vectors, genes of interest were inserted into XbaI/PmlI-digested pSN2-ccdB [45] by Gibson assembly [52]. Briefly, to obtain pSN.5 TagRFP-T (encoding a mutant red TagRFP), the TagRFP se- quence was amplified from pSITEII-6C1 [53] and the S158T mutation [35] was inserted by the overlap extension method [54]. For pSN.5 mTagBFP2, blue mTagBFP se- quence was amplified from pGGC024 [55] (kindly provided by J. Forner, Universität Heidelberg, Heidelberg, Germany), and the I174A mutation [31] was inserted. For pSN.5 mTFP1, the cyan mTFP1 sequence was synthesized de novo (GeneArt, Life Technologies). For pSN.5 mNeon, green the mNeonGreen sequence was amplified from pICSL80019 [56], kindly provided by M. Youles (The Sainsbury Labora- tory, Norwich, UK). For pSN.5 mPapaya1, the yellow mPa- paya1 sequence was synthesized de novo (GeneArt, Life Technologies). For pSN.5 wtHC, PPV HCPro was amplified from pSN-PPV ΔP1 [45]; for pSN.5 HC-L134H, PPV HCPro was amplified from pSN-PPV ΔP1 and the L134H mutation inserted, whereas for pSN.5 HC-AS9, PPV HCPro was amplified from pSN-PPV ΔP1 and amino acids REN- 239,240,241 were replaced by alanines. For pSN.5 p19, tomato bushy stunt virus p19 was amplified from pBIN61- P19 [38]. In all the newly-generated constructs, coding sequences are driven by a double enhancer Cauliflower mo- saic virus 35S promoter, flanked by PPV 5’UTR and 3’UTR, followed by a nopaline synthase terminator.
Plant agro-infiltration
Nicotiana benthamiana and N. clevelandii were grown in a greenhouse maintained at a 16 h light/8 h dark photoperiod, temperature range 19-23°C. Agro-infiltration of N. benthamiana and N. clevelandii plants was as described [6]; whenever possible, tested constructs were delivered in individual patches of the same leaf.
The viral replication assay was conducted in three- week-old N. clevelandii plants following agroinfiltration and sampling guidelines [14], with the exception that a saturating concentration of Agrobacterium (OD
6001.0) was used.
Laser scanner imaging
Plant leaves were sandwiched between two low- fluorescence glasse plates and fluorescence was acquired in a laser scanner (Typhoon 9400, GE Healthcare). Set- tings used were normal sensitivity, focal plane +3 mm and 50–100 μm pixel resolution; excitation lasers and emission filters used are summarized in Table 1. Signal saturation was avoided by adjusting photomultiplier tube voltage. Ty- phoon data were exported to 16-bit .tiff files. ImageJ soft- ware [57] was used to produce false-color images and overlays, and to generate 3D-projections through the Interactive 3D Surface Plot plug-in.
Fluorescence intensity measurements
Black 96-well flat-bottom plates (Nunc) with 50 μL water/
well (to limit sample dehydration) were used for the assay.
A cork borer was used for tissue sampling; individual 5.0 mm-diameter leaf discs, collected at the same distance from the infiltration point, were placed upside down in the prepared plates. Top reading measurements were used to acquire fluorescent protein excitation, emission spectra and intensity quantification in a monochromator-based plate reader (Infinite M200, Tecan Group). Gain value was adjusted manually to avoid signal saturation. RFI was quantified using the excitation and emission bands indi- cated in Table 1. Top reading GFP fluorescence intensity was alternatively quantified in an Appliskan (Thermo Fisher Scientific) and/or Victor X2 (PerkinElmer) filter- based plate readers.
Western blot assays
Liquid nitrogen-frozen plant tissue was homogenized in a TissueLyzer bead mill (Qiagen). Total proteins were ex- tracted, separated by glycine-SDS-PAGE and electroblotted onto a nitrocellulose membrane, as reported [45]. Proteins were detected using rabbit anti-PPV CP and -PPV HCPro sera, and mouse anti-GFP monoclonal antibody (clones 7.1 and 13.1, Roche) as primary antibodies; horseradish peroxidase-conjugated goat anti-rabbit IgG (Jackson) or sheep anti-mouse IgG (GE Healthcare) were used as sec- ondary antibody. For signal quantification, chemilumines- cence was acquired in a ChemiDoc XRS imager (BioRad) and analyzed with ImageJ.
RT-qPCR
Total RNA was extracted with the FavorPrep Plant Total RNA Mini kit (Favorgen), including on-column DNAseI treatment. Purified RNA was quantified spectrophotomet- rically by NanoDrop (Thermo Fisher Scientific) and con- centration adjusted to 50 ng/μL. Strand-specific cDNA for PPV RNA was synthesized for at least three biological rep- licates per condition using tagged cDNA primers in the RT step [58]. The 10-μL RT reactions contained 100 ng of total RNA and (at final concentrations) 1x Superscript III
Pasin et al. Plant Methods 2014, 10:22 Page 8 of 11
http://www.plantmethods.com/content/10/1/22
