1 Introduction

↓3

The free online database PubMed administrated by the U.S. National Library of Medicine (NLM) as part of the National Institute of Health (NIH, U.S.A.) hosts the worldwide largest comprehension of abstracts in life sciences and biomedical research. The entry term “influenza” gives more than 61,400 results demonstrating the high significance of influenza virus research.

In addition, other numbers connected to this viral genus, like 3 to 5 million cases of disease in the two annual flu seasons (each per hemisphere) with more than 500,000 deaths or pandemic outbreaks with up to 80 million fatalities (1918), render any further explanation for the importance and urgency of intensive investigation on influenza virus.

Influenza virus belongs to the virus family of Orthomyxoviridae. This Greek term describes the main characteristic phenotype of influenza caused illnesses by combining standard (= ortho) and mucus (myxo) [1]. In general, influenza viruses affect the upper and lower respiratory tract. Traditionally, viruses have been classified by the implemented disease due to the lack of knowledge about viruses themselves. The aforementioned virus family is consisted of influenza A, B and C virus, the Togotovirus and recently suggested the infectious salmon anaemia virus [2].

↓4

This introductory part will focus on influenza A virus first by a short portrait of the genetic and structural properties. In the following, the impact of the virus on the host cell as well as the pandemic risk classification will be discussed.

Particular attention will be paid on the replication cycle of influenza A virus with respect to the topic examined by the present thesis.

1.1 Biology of Influenza A Virus 

The genome of influenza A virus consists of 8 segments composed of single-stranded RNA in negative sense orientation [1]. The viral RNAs (vRNAs) are packed in ribonucleoprotein (RNP) complexes and function as templates for the messenger RNA (mRNA) and the complementary RNA (cRNA) synthesis.

↓5

Antigenic variations in the sequence of the 2 major surface proteins, hemagglutinin (HA) and neuraminidase (NA), are the reason for the influenza A virus standard nomenclature. There are 16 HA variants and 9 NA variants known [3,4]. Due to host restrictions only 6 subtypes (H1N1, H2N2, H3N2, H5N1, H7N7 and H9N2) have been isolated from humans [1].

The virions exhibit a pleiomorphic shape built up by a lipid envelope which is derived from the host`s membrane as a result of the budding process. Virus morphology can range from spherical shape with a diameter of 100 nm to filament with a length in excess of 300 nm. There are indications, that the viral proteins [5-8] and conversely the host cell type [9] influence the physiological appearance of the virion morphology. A common feature are the embedded proteins, the HA, NA and the matrix protein 2 (M2), which are anchored by short hydrophobic amino acid sequences. Beneath the lipid envelope the viral matrix protein 1 (M1) forms a protein layer [10], the virion core. Apart from M1, the nucleoprotein (NP) is the most abundant protein in the virion and interacts with the sugar-backbone of the vRNA molecules in a 1 monomer to 20 nucleotides ratio [11]. It is assumed that the polymerase complex (subunits PB1, PB2 and PA) is linked to the higher ordered structure (e.g. supercoils) formed by the vRNAs [12, 13]. The non-structure protein (NS2, also NEP) plays a minor role inside the virion but is responsible for the nuclear export of the viral genome segments [14]. The last out of ten viral proteins is a non-structural protein (NS1) which has major influence on the regulation of influenza A replication and the cellular biosynthesis machinery.

1.2 Genome and Proteome of Influenza A Virus

Except the segments 2, 7 and 8 all of the influenza A virus genes are in monocistronic organisation. The segments are ordered by their sequence length.

Segment 1: 2341 nucleotides (Basic Polymerase Protein 2 – PB2)

↓6

The polymerase subunit PB2 exhibits a nuclear localisation signal in order to fulfil its function related to the viral transcription and replication inside of the cellular nucleus [15]. The characteristic cap-snatching mechanism is mediated by the PB2 [16, 17]. In particular, the cap-binding site is localised at the carboxyl terminal end. In concert with its endonuclease activity it uses cellular cap-structures to provide primers for the viral transcription [18]. Furthermore, the PB2 contains binding sites for the PB1 subunit at its amino terminus. Additional binding sequences for the NP are supposed to function for regulatory purposes [19].

Segment 2: 2341 nucleotides (Basic Polymerase Protein 1 – PB1)

PB1 contains the 4 conserved amino acid motifs of RNA-dependent RNA polymerases and thus represents the functional protein for RNA polymerization [20]. The binding sites for PA and PB2 are located at the amino- and the carboxyl terminal end, respectively [21], indicating that PB1 exhibits a central role for the assembly of the polymerase complex and its function. Interestingly, in some influenza A strains an alternative open reading frame encoding for the accessory protein PB1-F2 can be found. This exhibits pro-apoptotic activity in host immune cells [22].

Segment 3: 2233 nucleotides (Acidic Polymerase Protein – PA)

The smallest subunit PA is a phosphorylated protein [23] which function is restricted to the nucleus and the presence of the whole polymerase complex. It is suggested that PA has helicase and ATP-binding activity and is required for PB1 accumulation in the nucleus [24, 25]. Inside of the nucleus the PA subunit induces proteolysis resulting in an aberrant nuclear morphology and chromatin condensation [26].

Segment 4: 1778 nucleotides (Hemagglutinin – HA)

↓7

HA is the most prominent and best studied influenza A protein. It mediates the binding of the virus to sialic acid residues on the cell surface and the fusion of the viral with the endosomal membrane. Segment 4 encodes for a precursor polypeptide HA0 which is cleaved by serine proteases into the HA1 and the HA2 subunits which remain linked by a disulfide bound. This posttranslational process is essential for the conformational change of the HA. This is triggered by low-pH conditions inside the late endosome and leads to exposition of the fusion peptide which is situated at the amino terminal end of HA2 [27]. Attachment to the host is mediated by the pocket region localised at the globular head of the HA1 subunit which therefore alters the host specificity [28]: a discrimination between ɑ-2,3- (avian) and ɑ-2,6- (human and avian) linkages of the sialic acid and galactose is caused by one amino acid residue (glutamine or leucine at position 226, respectively). On the contrary, the HA2 portion exhibits the fusion activity [29]. HA is organized in the viral lipid-envelope as a homotrimeric transmembrane protein.

The attention for HA is also contributed to the susceptibility for neutralizing antibodies and the effect on the pathogenicity of the virus subtype [30].

Segment 5: 1565 nucleotides (Nucleoprotein – NP)

As an essential component of the ribonucleoprotein complex the NP binds via an RNA-binding domain in a non-sequence dependent manner to the vRNAs [31]. This binding is assumed to regulate the switch from the transcription to the replication mode as NP may act as an anti-termination control factor [32]. Nuclear localisation signals within the NP amino acid sequence indicate that this shuttle protein is involved in the transport of viral RNP complexes into the cellular nucleus and the export of novel vRNAs in association with M1 and NS2 [33-35].

Segment 6: 1413 nucleotides (Neuraminidase – NA)

↓8

The glycosylated homotetrameric [36] NA is not required for virus entry, replication and assembly [37] but plays an important role in the very late stage of the infection cycle which is the detachment of novel synthesized virus particles from the host cell membrane [38]. The cleavage of an ɑ-ketosidic linkage between a terminal sialic acid and a neighbouring D-galactose or D-galactoseamine residue of cell-surface glycoproteins or gangliosides, respectively, prevents virions from aggregation on the cell surface [38]. Likewise the NA removes sialic-acid residues from the viral envelope in order to prevent self-accumulation and thus enhances infectivity [38]. Host anti-NA antibodies or NA inhibitors (e.g. oseltamivir carboxylate) prevent viral replication. Each NA monomer is composed of 4 structural domains: a globular head, a stalk, a transmembrane domain and a cytoplasmic domain [36]. The latter is thought to influence the (neuro-) virulence (glycosylation site at position 130 [39]) and virion morphology [7]. Matrosovich et al. 2004 experimentally evidenced the essential role of NA also for virus invasion of the ciliated human airway epithelium. Removal of decoy receptors prevents strong binding of the virion to non-target cells which would impede virus access to target cells [40].

Segment 7: 1027 nucleotides (Matrix proteins – M1 and M2)

The mRNA of segment 7 is bicistronic: while M1 is a direct transcript, M2 is produced by using an alternative splicing site [10, 41].

Several functions of M1 are described: inhibition of viral transcription [42], assistance of vRNP nuclear export [33, 43], support of the NP self-assembly into the helical structure of the vRNP [44], and involvement in virus assembly and budding process [6, 8]. It is assumed that the amino-terminus where the nuclear localisation signal is encoded is responsible for self-polymerization and membrane binding, in concert with the NS2 protein [45].

↓9

In contrast to M1, M2 appears as a homotetrameric transmembrane protein [46] embedded in the viral envelope. M2 functions as an ion channel [47] and regulats the pH inside the virus particle when incorporated in endosomes. Internal acidification causes the dissociation of the vRNPs from the M1 [33], a process termed uncoating, that allows the viral genome to enter the host nucleus. This step in the viral infection cycle is a target for antiviral therapies blocking further replication [48-50]. The impact of the transmembrane domain of M2 for an efficient replication is still under discussion [51, 52].

Segment 8: 890 nucleotides (Nonstructural proteins – NS1 and NS2)

Analogue to segment 7 the mRNA transcribed from segment 8 is as well biscistronic. Therefore, the NS2 protein is an alternative splicing variant of the NS1 mRNA [53].

The only absolute non-structural viral protein NS1 binds to a wide range of RNA molecules [54] indicating its functional influence on splicing of mRNAs, cellular and viral transcription [55], cellular mRNA export [54] and adenylation [56] as well as viral mRNA translation [57, 58]. NS1 activity might result in decreased susceptibility to cellular antiviral defence mechanisms [59].

↓10

The description “non-structural protein NS2” is misleading because it exhibits a binding domain specific to M1 at its carboxyl-terminal end and is incorporated at low amounts into the virion [60]. O´Neill et al. 1998 [14] suggested to rename it in “nuclear export protein (NEP)” with reference to the presence of a nuclear export signal and the indispensability for vRNA nuclear export [61].

1.3 Replication of Influenza A Virus 

Figure 1: Replication cycle of influenza A virus (taken from www.qiagen.de).

Recognition of ubiquitous exposed sialic acid residues on the cell surface by the viral HA allows attachment of the virus particle to the host cell surface. This is followed by the uptake of the virus into the endosome. After uncoating, the viral ribonucleoparticles are delivered into the host cytoplasm and subsequently into the nucleus where transcription and replication take place. Translation occurs on free and membrane-associated ribosomes. Self-assembly of viral proteins leads to the budding of progeny influenza A virus particles from distinct regions of the host plasma membrane.

Due to the requirement of cellular resources, the vRNPs are trafficked to the nucleus hijacking the cellular import machinery. For this purpose, the viral proteins carry nuclear localisation signals, as aforementioned.

↓11

The nucleus represents the place of action for both, the transcription of capped and polyadenylated viral mRNA and the replication producing antigenomic viral cRNA molecules and progeny vRNP complexes [62]. The templates are still the negative-stranded vRNAs. They all exhibit conserved motifs at their endings: 12 bases at the 3´end and 13 bases at the 5´end showing partly inverted complementation. These sequences are thought to interact with each other forming hairpin shaped structures [63]. Whereas cRNA molecules are exact and full length copies of the complementary vRNAs [64], the viral mRNA lacks circa 17 nucleotides at the 3´end (complementary to 5´end of the corresponding vRNA) [65]. The afore described nuclear export mechanisms involving the NP, M1 and NS2 [35,45,61] are restricted to mRNA and vRNA whereas the cRNA stays in the nucleus for the entire replication process [66].

Transcription of influenza A virus

One fundamental aspect of influenza A virus transcription initiation is the cap-snatching from cellular mRNA molecules by the endonuclease activity of the PB2 polymerase [18]. These 10 to 15 nucleotides are used as primers for the complementary strand synthesis [67] and protect the viral mRNA from endonucleolytic degradation [68]. In addition, a successful initiation of transcription also requires a specific secondary structure of the promoter sequence [11,69] indicating that sequences at both ends of the cRNA molecule are necessary.

Simultaneously to the cellular termination an uracil rich sequence in close proximity to the template`s 5´ end produces the synthesis of a poly(A) tail finishing transcription [65]. Plotch et al. 1977 demonstrated the host-independency of viral mRNA polyadenylation by in vitro experiments [70]. The poly(A) tail and the cap structure are essential features of mRNA molecules with respect to nuclear export and biostability [71].

↓12

As transcription is a selective process, the appearance of individual viral mRNA species varies during the time course of infection. An early and a late transcription can be distinguished. While in the first step the synthesis of all 8 mRNA species occurs equally, preferentially the NS1 and NP mRNAs are transcribed at the beginning of the late transcription. Shapiro et al. 1988 evidenced the early synthesis of NS1 mRNA with respect to the delayed presence of M1 mRNA [66].

Replication of influenza A virus

The formation of copy RNA in full length requires neither primer sequences nor polyadenylation [64]. Generally, the key question how the virus facilitates the regulation of transcription versus replication was under strong discussion suggesting viral as well as cellular factors [66,72]. Even the degradation of novel synthesized cRNA molecules without bound viral RNA polymerase or the viral NP [73] might result in the switch to the replication mode.

↓13

Recently, Olson et al. 2010 presented a comprehensive description of the underlying regulative interactions. The process involves the polymerase, the three viral RNA species, and the cap structure, consequently ending up in a complex interaction cycle [74].

Figure 2: How influenza A virus controls the switch from transcription to replication mode.

(taken from Olson et al. 2010) The transcriptional mode is maintained by increased concentrations of cap source, whereas increased concentrations of vRNA and viral polymerase will lead to a switch toward replication. Encircled plus signs indicate stimulation while T bars indicate repression. RdRp = RNA-dependent RNA polymerase

Translation of influenza A virus

The hypothesis that the amount of translated viral proteins is mostly influenced by the amount of the corresponding mRNA species is widely accepted. As a consequence of the unevenly synthesis of the different mRNA species, the NS1 and the NP are over-represented in the early phase of influenza A protein translation. With progression NS1 production is reduced and HA, NA and M1 are predominantly translated [66]. Transmembrane proteins (HA, NA and M2) are translated by membrane-bound ribosomes into the endoplasmic reticulum (where folding takes place) and maturate in the secretory pathway of the trans-Golgi network. An apical sorting signal ensures the trafficking to the plasma membrane and incorporation into the budding site [75].

↓14

The replication cycle is finished by the self-assembly of the newly synthesised viral proteins followed by the budding and release from the host plasma membrane.

1.4 Pandemic Danger Caused by Influenza A Virus

Epidemics among the human population, in terms of sustained, widespread, person-to-person transmission, are caused only by three HA (H1, H2 and H3) and two NA (N1 und N2) subtypes.

The generation of influenza A virus subtypes that have pandemic potency is contributed to an antigenic shift. This term describes the reassortment of genome segments as a result of double infection. Consequently, these reassortants may encode for novel antigenic proteins with new immunogenic properties [75]. The most prominent case of such an antigenic shift produced the influenza A (H1N1) virus, called “Spanish flu” of 1918-1919, with a high rate of mortality [76]. Low infectivity combined with high pathogenicity of avian derived influenza A subtypes is caused by the host receptor discrimination. Sialylated proteins containing a terminal ɑ-2,3-linkage are localised in the lower regions of the human respiratory tract. Therefore, infections with avian viruses are relatively rare and the human immune system is not adapted. That is why human infections with avian influenza A subtypes cause fatal progressive pneumonia [77].

↓15

Additionally, a mechanism named antigenic drift, describes the accumulation of multiple mutations in antigenic sites (for instance around the receptor binding region at the globular head of HA) and may lead to the formation of a drifting strain which is not longer inhibited by neutralizing antibodies against the parental virus. The host might be infected and produces clinical signs of disease [75].

1.5 Host Interaction and Infection Heterogeneity

Variation among individuals is a common concept in population biology. Recently this idea of heterogeneity became more and more prominent for cell biologists and virologists. In a given cell population the environment of individual cells and the specific intracellular conditions lead to heterogeneities in their status. Virus infection strongly depends on the biochemical, physiological and physical status of the host cell as several distinct cellular processes and resources are involved. Consequently, the individual cell status is critical for influenza A virus infection and for its efficient replication and reassortment [1,9,78].

Influenza A virus infection can be seen as a concert of viral protein activity and cellular protein interaction which is an evolving process. Pathogen sensors (like RIG-I) of the host cell detect viral RNA and induce antiviral defence (e.g. type I interferons). On the other hand, viral proteins (like NS1) suppress antiviral actions [79] or hijack cellular mechanisms and resources [80]. There is strong evidence that viruses mimic cellular motifs, named SLiMs (short linear motifs), to regulate and control cellular proteins. For instance, the influenza A protein PB2 mimics the nuclear localisation signal of the cellular importin ɑ [80].

↓16

The underlying cellular networks have been reviewed by Watanabe T., Watanabe S. and Kawaoka in 2010 demonstrating the wide spectrum of applied approaches and the inconsistency of the findings in different labs [81].

Systematic and genome-wide RNA interference screens [82-85] are contributed to great efforts in the identification of virus-host interaction partners. These studies revealed hundreds of factors involved in all steps of the replication cycle. They have to be interpreted with care for experimental (e.g. efficiency of knock-down, influence on general cell viability) and analytical reasons (e.g. false negative/positive results).

Employing yeast-two-hybrid assays and computational approaches several human interaction partners were identified. These data include RNA-binding proteins, transport proteins, transcription factors, and proteins of the intra-cellular signalling pathways (NFκB, apoptosis, MAPK, and WNT) [79].

↓17

However, all these in vitro or in silico analysis of host-virus interaction partners do not display the dynamic process of individual interactions and/or the impact on the viral replication. Validation and detailed information on the dynamics of the existing network are required. Moreover, investigation of different virus strains and/or cell types will expand the knowledge about general mechanisms and thus, uncover novel antiviral compounds.

Figure 3: Influenza Virus Life Cycle and Host Factors (taken from Watanabe et al. 2010).

The light-orange rectangles indicate individual steps of the influenza virus life cycle. The light-blue rectangles indicate host cellular processes that may be involved in the virus life cycle. Red circles indicate host factors identified in the screens discussed in the named publication and yellow circles indicate host factors identified in other previous studies. For details please review Watanabe et al. 2010.

1.6 Detection and Imaging of Nucleic Acids

The ability to study nucleic acids during their biosynthesis, intracellular transport, subcellular localisation, and degradation is of great interest in various fields of research like cell biology, medicine, pharmacology or virology. In basic and diagnostic sciences specific detection methods are required.

↓18

Indeed the underlying molecular mechanism of certain cancer subtypes [86-90] and several human diseases, like asthma [91], cutaneous, gastrointestinal and liver disease [92-94], is thought to be connected to dysfunction of microRNAs which act as posttranscriptional modulators of gene expression [88]. Furthermore, the identification of single nucleotide polymorphisms [95] or the specific identification of infection are just three out of many examples of nucleic acid research in clinical diagnostics [96-102].

Studies on endogenous as well as viral nucleic acid variants in the cellular context via microscopy imaging are one of the most relevant issues in nucleic acid research [103-105]. Particularly, virology studies on the replication cycle require specific detection methods to analyse the different steps of viral replication.

There is already a great spectrum of techniques available dealing with the visualisation of nucleic acids in fixed [106,107] and in living [105,108,109] cell systems. However, most of these strategies are limited to special applications or exhibit various disadvantages making imaging of nucleic acids in living systems still a challenging task.

↓19

This introductory part gives an overview of the three exemplary techniques in nucleic acid research namely Fluorescent Proteins (FPs), Molecular Beacons (MBs) and finally Peptide Nucleic Acids (PNAs) which are applicable to living cells.

Figure 4: Schematic depiction of the functional mechanism of (A) Fluorescent Proteins, (B) Molecular Beacons and (C) Peptide Nucleic Acids in nucleic acid detection.

(A) Fluorescent Proteins like the green fluorescent protein (GFP) are used in the MS2 protein system which employs introduced hairpin structures for a sequence specific detection of mRNA molecules in living cells.
(B) Molecular Beacons are hairpin shaped structures carrying a fluorescent moiety and an equivalent quencher which keeps the probe in the non-fluorescent ground state. Spatial separation leads to enhancement of fluorescence.
(C) Peptide Nucleic Acids are nucleic acid analogues in which the backbone is a pseudopeptide rather than a sugar. Sequence specific fluorescent signals are produced with the help of an intercalating dye upon target hybridisation.

↓20

1.6.1 Fluorescent Proteins

All insights in molecular biology and cell biology based on the usage of fluorescent proteins have to be contributed to the efforts of Roger Tsien, Martin Chalfie, and Osamu Shimomura [110]. In 2008 they were honoured with the Nobel Prize in Chemistry for their work on the green fluorescent protein (GFP) and the gathered technical achievements. This fluorescent protein, isolated and cloned from the jellyfish (Aequorea victoria) belongs to the standard methods in nearly all fields of biology and medicine with thousands of applications [111].

Besides the versatile utilization in protein studies, recent developments made FPs suitable for the visualisation of gene expression, nuclear transport and mRNA dynamics in living cells [112]. One of the key benefits of the FP technique compared to in situ hybridisation or oligonucleotide labelling is that the probes are directly expressed as fusion proteins in genetically manipulated organisms. This allows imaging in real time.

↓21

Sensitivity is the biggest limitation of GFP as a detection reagent of small molecules and motivated researchers to make remarkable progress towards new strategies for single-molecule detection [113]. Recently, the MS2 protein technique for mRNA imaging in yeast was established simultaneously in the Singer lab [114] and by Bloom and colleagues [115]. The principle of this method lies in two components: (i) expression of the bacteriophage MS2 RNA coat protein (MCP) fused to the FP and (ii) expression of the mRNA of interest (including the MS2 hairpin motifs). Due to genetic modifications the expression can occur globally or in a particular tissue depending on the chosen promoter [116].

Golding et al. 2005 applied the MS2 technique to Escherichia coli with an influential supplementation: The authors introduced 96 copies of a specific RNA hairpin into the untranslated region of the mRNA of interest enabling the binding of 96 MS2-GFP fusion proteins. With this the fluorescent signal of single molecules was enhanced dramatically [117]. This had a crucial effect on the ability of counting mRNA molecules using conventional fluorescence microscopy.

Figure 5: Depiction of MS2-MCP labelling of endogenous mRNA in living cells.

(adapted from Weil et al. 2010) GFP = green fluorescent protein, MCP = MS2 coat protein, NLS = nuclear localisation signal, RNP = ribonuleoparticle

↓22

Apart from the impressive achievements in transcription research [118,119] the MS2 method exhibits several drawbacks: (i) the lack of knowledge concerning the best position of the hairpin insertion and the difficulties to clone the hairpin repeats, (ii) the requirement of transgenic expression of the selected gene leading to a non-physiological over-expression and (iii) the tendency of the fusion proteins to form aggregates without binding the target sequence leading to false-positive cytosolic signals [116].

The most critical aspect in tagging mRNA with GFP molecules is the risk to modify their native behaviour and dynamics.

1.6.2 Molecular Beacons

Molecular Beacons have been invented as alternative RNA labelling technique to FISH (fluorescent in situ hybridisation) [116] as they provide the possibility to detect individual RNA molecules in living cells [105,120-124].

↓23

Figure 6: Depiction of the structural properties of Molecular Beacons. (modified from www.eurogentec.com)

In 1996 Tyagi [125] established the Molecular Beacon technique for RNA imaging. The hairpin structured MBs are constructed of 3 main components: (i) the 15-30 nucleotides spanning loop region, (ii) the 5-8 base pairs in length stem portion and (iii) the fluorophore-quencher pair. The fluorescent moiety is typically connected to the 5´end and the quenching group is linked to the 3´end of the molecule.

The double-stranded stem enables the looped conformation of the MB and thus the non-fluorescent ground state while the quencher is in close proximity (7-10 nm) to the fluorophore. The loop region is responsible for the sequence specific target binding and confers on the MBs target sensitivity. Upon hybridisation with the target sequence the binding energy leads to an opening of the hairpin structure and subsequent dissociation of the stem region. Hence, the thermodynamic equilibrium between the hybridisation energy of the loop sequence and the melting temperature of the stem region is enormously important for the functionality of MBs. This relation is very sensitive and requires a precise adjustment. Low melting temperatures of the MB can already provide for a hairpin structure and hybridisation to the target. However, temperatures over the threshold-set point destroy the hairpin structure and turn it into a linear oligomer. In this situation the distance between fluorophore and quencher is too long for quenching and fluorescence is emitted giving a false-positive signal [124,126]. In contrast, if the melting temperature is too high the stem region is too stable to allow specific hybridisation of the loop sequence leading to false-negative results. In addition, the stem should be designed in order to prevent unfavourable secondary structure formation [125].

↓24

The prevention of fluorescence emission in the unbound state represents the most notable advantage as it allows the discrimination between bound and unbound probe [127]. Moreover, the MB technique profits from the possibility to use a wide range of fluorophore-quencher pairs [127]. All this contributed to the various application of MBs in (quantitative) real time polymerase chain reaction approaches [128-132], the detection of gene mutations or single nucleotide polymorphisms [133-136], and even DNA/RNA binding protein detection [137].

The combination with microscopy techniques enlarges the application spectrum of MBs by (live) cell imaging of nucleic acids, particularly RNA molecules.

Initially, the use of MBs for investigating mRNA in cells was limited to fixed conditions due to the vulnerability of MBs towards nuclease cleavage and the false-positive signals produced by binding of RNA/DNA-binding proteins [138]. For this purpose, MBs have been modified variously to enhance the sensitivity. For example, additional FRET systems or wavelength shifts of the fluorophore were inserted and the native ribonucleotides were replaced by 2-O-methylribonucleotides to achieve nuclease resistance [125,139-142].

↓25

Bratu et al. 2003 [123] demonstrated the use of such an modified MB to visualize the distribution and localisation of an endogenous mRNA (oskar mRNA) which plays an important role in Drosophila melanogaster development. Oocytes were microinjected with up to 106 MB molecules followed by immediate imaging. This method implies an invasive disruption of the plasma membrane which is apparently working with oocytes but causes cell damage or death to most cell types leading to an inefficient delivery of the probe.

A second example of an mRNA study in living cells using MBs was provided by Lennon et al. 2010 [120]. Here the influence of beta 1 integrin on the proliferation and differentiation of osteoblasts was investigated. The MB molecules were delivered into the living cells using the reversible permeabilisation of the plasma membrane with activated streptolysin O. The drawback of this technique is the lack of knowledge about the exact concentration of loaded probe per cell.

Furthermore, MBs have been used for imaging viral mRNA in living host cells [143]. Wang et al. 2008 [121] presented the sequence specific detection of the neuraminidase and the matrix protein 1 and 2 mRNA molecules inside living influenza A infected MDCK cells. Employing FRAP (fluorescence recovery after photobleaching [144]) they investigated the viral mRNA nuclear export.

↓26

The MBs used by Wang et al. were designed carrying at their 5´ends TAMRA (tetramethyl-6-carboxyrhodamine) which belongs to the group of heterocyclic dyes. TAMRA is quenched by DABCYL ((4-(4-dimethylaminophenyl) diazenylbenzoic acid) linked to the 3´ends of the oligomers. Fluorescence of TAMRA in the closed state of the MBs is hindered by non-radiative energy transfer (FRET, Förster resonance energy transfer [145,146]) to the quencher. DABCYL can be used as a universal quencher for a wide range of fluorophores. The underlying mechanism is a dipole-dipole interaction which is strongly influenced by the donor-acceptor distance and the transition dipole moments. Thus, minimal changes in the MB structure may cause false-positive or false-negative results by the aforementioned problems concerning the MB design [147,148].

1.6.3 Peptide Nucleic Acids

In 1991 Nielsen and colleagues [149] synthesised a completely artificial analogue to nucleic acids in which the backbone was a pseudopeptide rather than monosaccharide rings and phosphodiester linkages. The structural properties gained the designation of this new molecule: peptide nucleic acid (PNA).

The nucleic bases (purine and pyrimidine) in a PNA molecule are directly attached to the glycine units by methylene carbonyl bounds, exhibiting stronger binding and more specific hybridisation properties than DNA/DNA or DNA/RNA hybrids or other common DNA derivates due to the lack of charge repulsion [150]. Moreover, they lack off-target effects due to non-specific binding to proteins or degradation by RNase H cleavage of the bound mRNA [151,152]. PNAs are not vulnerable to nucleases or proteases [153].

↓27

Figure 7: Depiction of the molecular structure of DNA and PNA molecules. (modified from Ratilainen et al. 1998)

The next section will present examples to emphasize the large variety of the application spectrum and practical flexibility of PNAs.

PNAs have been widely used as antisense agents in nuclear medicine [152,154-157]. In Bonham et al. 1995 the described microinjection of a 15mer PNA targeted towards the untranslated region of the SV40 T antigen mRNA inhibited the expression of the T antigen in CV-1 cells with an efficiency of 99% [158]. Unwanted impact on the phenotype of cells treated with antisense oligomers may not occur while working with antisense PNAs due to the lack of a negative charge and thus excluded protein binding inside and outside of cells. Furthermore the stability of PNA/RNA hybrids is higher than PNA bound to DNA [159]. Hence, PNAs were thought to expand the potential of antisense research although predicting the susceptible mRNA target sequence has proven to be an elusive goal [160].

↓28

Recently, in the Wickstrom lab the usage of PNAs as radiohybridisation probes gave new impacts in cancer diagnostic medicine [161-164]. Chakrabarti and colleagues 2007 [165] demonstrated the detection of pancreas cancer before its physical, chemical or anatomical appearance applying PNAs to non-invasive PET (positron emission tomography) imaging in mice of the KRAS mRNA carrying a disease specific (95 % of all ductal pancreas cancer patients) single nucleotide polymorphism (point mutation in codon 12). This enables an early intervention and promising treatment of cancer patients.

In addition, PNAs in complex with the contrast agent gadolinium (Gd3+) enable intravital magnetic resonance imaging [166,167]. The proof of principle was realized by Heckl et al. 2003 in the form of the identification of tumour cells with the help of PNA- Gd3+ labelling of c-myc over-expression [168].

An innovative step forward in PNA applications was realised by the chemical introduction of an intercalating fluorophore into a PNA molecule as an artificial base-surrogate [169,170]. Thiazole orange which is commonly used in reticulocyte flow cytometry analysis to stain residual RNA in immature blood cells [171] exhibits favourable (fluorescence) properties in order to advance this technology.

↓29

Thiazole orange (TO) belongs to the group of asymmetric cyanine dyes and is essentially non-fluorescent in solution. The cationic chromophore intercalates with high affinity to polynucleic acids resulting in enhancement of quantum yield (0.1-0.4) which is a temperature dependent process [172,173].

Initially, Privat et al. 2001 microinjected modified DNA oligomers with an alkyl linkage chain to thiazole orange into fixed human osteosarcoma (HOS) cells to localise mRNA molecules in the cytoplasm and nucleus [174].

The applicability of PNAs to RNA imaging in CHO cells was already demonstrated by Berndl et al. in 2010 using confocal laser scanning microscopy for visualizing RNA delivery. In this work the splicing and thus shortening in length of mRNA molecules is imaged with the help of the FRET pair TO- Alexa-594 [175].

↓30

The intercalation of TO as a base substitute within a PNA molecule was a further improvement. This method was developed in the sense of FIT (forced intercalation of thiazole orange)- PNAs by Seitz, Bergmann and Heindl [170] and tested with respect to hybridisation, fluorescence, sensitivity and stability properties [176-179]. 

The increase of fluorescence is reported to occur in combination with each of the nucleic bases with a preference for thymine [180] and is able to reach a 20-fold enhancement with discrimination to the neighbouring base pairs [177]. The mechanism of the interaction with the bases in close proximity is generated by the structural architecture of two heterocyclic moieties linked by a (poly-) methine bridge. In the electronic ground state (S0) these two aromatic ring systems are in peripendicular position to each other. Twisting of the moieties around the bond avoids fluorescence even when the molecule is optically excited. Monointercalation (one dye per two base pairs at saturation) leads to a forced torsional motion of the two rings into a coplanar position. This mechanism enables the formation of a π-electron system and thus the increase in fluorescence [181,182].

Socher et al. 2008 described a DNA–PNA hybridisation assay based on real-time PCR that allows sensitive quantification of specific nucleic acids in solution and simultaneous detection of single base mutations. This approach demonstrates the biological applicability of FIT-PNA probes.

↓31

To the best of the author`s knowledge, this work was the first application of FIT-PNAs to detect (viral) mRNA molecules in the living cell [183].

Although originally designed to function as antisense and antigene reagent, PNAs, in particular FIT-PNAs, hold great potential for the investigation of dynamic processes in real time.

1.7 Quantitative Proteomics: SILAC (Stable Isotope Labelling of Amino Acids in Cell Culture) 

During viral infection, interactions between viral and host cell proteins are major checkpoints for en efficient replication of the virus or respectively survival strategies of the host cell. Thus, these interaction pathways and the involved components offer strong potential for new antiviral drug development and deeper understanding of the entire infection mechanism.

↓32

Traditionally, for qualitative and quantitative proteomics two-dimensional (2D) gel electrophoresis has been employed. In (2D) gel electrophoresis proteins are separated in the first dimension according to their isoelectric point and in the second dimension by their molecular weight [184]. The underlying mechanism of a semi-quantitative protein analysis using 2D gel electrophoresis is the comparison of protein staining intensities with respect to a protein standard of known protein concentration [185].

The simultaneous identification and quantification of proteins was realized by introducing mass spectrometry (MS) into the field of proteomics [186]. This generic term implies a description of the complete protein set expressed by the whole genome in the lifetime of a cell including posttranscriptional and posttranslational modifications [187]. Mass spectrometry is per se not quantitative due to variations in the detector response or differential ionization yields for different substances. Accuracy in peak ratios can be realized by using isotopic analogues caused of their chemical uniformity [186].

Initially, isotopic labelling was performed by incorporating 18O (oxygen) atoms at the C-terminus of a peptide [188] in protein chemistry or by 15N (nitrogen) substitution of all N atoms for quantifying variations of microorganisms [189,190]. These methods have crucial drawbacks as the limitation of the 15N substitution to bacteria, incomplete labelling efficiency or difficulties in data interpretation caused by varying numbers of N atoms in peptides. Further attempts to improve the sensitivity including the ICAT (isotope-coded affinity tag) method [191] required the necessity of chemical modifications of the proteins making the whole process even more complicated.

↓33

A more sensitive method which guarantied an easy labelling strategy was urgently requested. For this purpose, Ong et al. 2002 presented an inexpensive, robust and convenient technique which was termed SILAC (stable isotope labelling of amino acids in cell culture). Precisely, cells are propagated in cell culture media supplemented with isotopic variants of essential amino acids. All newly synthesized proteins contain the labelled amino acids without performing chemical modifications or affinity purification procedures [186]. By altering the incorporated nonradioactive isotopes several protein species are distinguishable allowing determination of protein abundance based on relative MS signal intensities [192,193].

SILAC was already used in comparative proteomic phenotyping to compare cell lines to their primary counter-parts [194], in specific investigations of histone posttranslational modification patterns and their correlation to particular tumor features [195], and for the identification of Tyr kinase substrates [196].

One of the first large-scale proteomic studies was performed in the lab of Macek [197]: two physiologically different cultures of Bacillus subtilis have been investigated covering 75 % of the expressed genes in the log growth phase (1928 identified proteins). Besides studies on cellular and microbial level, SILAC is also applicable to in vivo approaches as shown by the Selbach group (SILAC fly) [198] and by Walther and Mann (SILAC mouse) [199].

↓34

In addition, the SILAC technique is useful in virology in order to investigate virus-host interaction during the replication. Studies on the coronavirus infectious bronchial virus [200,201] and on the respiratory syncytial virus (RSV) [202] demonstrate the importance and high information output gained from SILAC based approaches.

Figure 8: Depiction of the schematic procedure of a typical SILAC experiment and the mass spectrometry output.

(adapted from the pSILAC scheme of Schwanhäußer et al. 2009)


© Die inhaltliche Zusammenstellung und Aufmachung dieser Publikation sowie die elektronische Verarbeitung sind urheberrechtlich geschützt. Jede Verwertung, die nicht ausdrücklich vom Urheberrechtsgesetz zugelassen ist, bedarf der vorherigen Zustimmung. Das gilt insbesondere für die Vervielfältigung, die Bearbeitung und Einspeicherung und Verarbeitung in elektronische Systeme.
DiML DTD Version 4.0Zertifizierter Dokumentenserver
der Humboldt-Universität zu Berlin
HTML-Version erstellt am:
03.05.2012