5 Discussion

↓97

In the present work essential steps of the replication cycle of influenza A virus were investigated. For this purpose a new imaging technique was established and a recently invented quantitative method for proteome studies applied. The presented results give new insights into the progress of viral replication and provide a deeper understanding of the viral-host interaction.

In the first chapters a sequence specific technique for viral mRNA imaging in living infected cells was introduced. This technique is based on the combination of PNA molecules with intercalating fluorophores. The most crucial advantage is that the fluorescent moiety apart from other probes is inserted as a base substitution. This leads to the highly specific generation of fluorescence signals assigned to viral mRNA molecules. For the first time, viral mRNA was investigated in a sequence specific manner in mammalian cells using FIT-PNAs.

↓98

To demonstrate the wide application spectrum of the presented technique influenza A virus infection was further determined via FACS analysis and a VSV derived transcript was studied in infected cells using a specific PNA probe.

The last chapter focuses on the progression of influenza A virus translation and the viral impact on the host cell proteome. Both, the viral and the cellular proteome were quantitatively assessed using mass spectrometry in concert with SILAC. The non-radioactive labelling of essential amino acids in newly synthesized proteins enables a time-resolved quantitative study upon influenza A virus infection. The results provide detailed information about the impact of viral action on cellular processes and viral-host interactions during the early stages of replication on a systems level.

Collectively, this work implies data on the influenza A virus replication which set a basis for computational modelling, the development of new antiviral strategies and a deeper understanding of the viral host-dependency.

5.1 Hybridisation and Fluorescence Properties of FIT-PNAs

↓99

PNA molecules exhibit a stronger binding affinity to RNA than RNA or DNA to their respective counterparts [150]. Nielson et al. 1994 explained this by pointing at the main structural property of PNA molecules which is the peptide backbone. Apart from DNA or RNA molecules PNAs are uncharged and thus lack the charge repulsion effect [159]. Moreover, the backbone substitution prevents endonuclease mediated degradation [153] and thus provides a stable labelling during the imaging time range. The nuclease and protease resistance of the used FIT-PNA was assessed in cell lysate in the absence of the target sequence (see figure 12). The fluorescence of the FIT-PNA was virtually unchanged indicating that the structure was not destroyed by cleavage. The influence of nucleases was shown by measuring MB 2 under the same conditions. After 5 hours incubation in cell lysate, the fluorescence of this probe increased by more than 40% reflecting a strong decrease in sensitivity. Consequently, this experiment revealed the favoured biostability of PNA molecules over DNA Molecular Beacons.

FIT-PNAs possess a highly specific sequence discrimination shown by Socher et al. 2008 [182] in a study detecting single nucleotide polymorphisms in a real-time PCR approach. This is in good agreement with the results of this work since the applied PNAs exhibited high sequence sensitivity. FIT-PNA 1b which was designed to target a 17 bases in length sequence in the NA mRNA exhibits an 11-fold enhancement in fluorescence intensity upon hybridisation with the artificial RNA target 3b (see 3.1.9). Even a 12-fold increase was reached for the measurement at 60 °C with the DNA target 5 (see 3.1.9.). Previous studies with artificial target sequences reported a maximal possible enhancement by a factor of 20 [170]. However, one has to keep in mind, that the quantum efficiency of the intercalating fluorophore, thiazole orange and its derivates, strongly depends on the environment. The five nucleic bases (thymine, guanine, cytosine, adenine, and uracil) produce distinct spatial properties which influence the position of the heterocycles and thus the basis for maximal fluorescence enhancement [172,173]. Whether there is a preference for the substituted nucleic base remains to be clarified. Jarikote et al. 2007 [180] provides clues for the PNA sequence design. But to date, the ideal sequence cannot be predicted theoretically due to the complexity of altering parameters like the interaction with neighbouring base pairs in combination with the sequence length. This effect is reflected by experiments using the sequence of the swine H1N1/Mexico/2009 strain which is shortened in length but the remaining target is sequence identical and the analogues sequence of the H3N2/X-31 strain which exhibits seven continuous matched base pairs situated in close proximity to the TO base surrogate. While the FIT-PNA 1b hybridised to the H1N1/2009 target confirmed the high enhancement in fluorescence, it stayed virtually unchanged in combination with the H3N2/X-31 target sequence (see figure 11).

PNA 1c (M1) showed a minor increase (factor 5.7) in fluorescence hybridised to the complementary artificial target sequence in comparison to the unspecific target sequence (factor 2) (see figure 34). One could argue that the TO and the BO do not share the same sensitivity properties caused by their varying structure as the BO lacks one carbon ring but the results obtained in living cells revealed a higher enhancement factor for the BO probe compared to the TO probe. The difference in sensitivity of PNA 1b and 1c is also not contributed to the nucleic base composition as for both probes the AT:GC ratio is 1.7. It is likely, that the position of the fluorophore plays a more sophisticated role for the sensitivity as demonstrated by the results in table 4, 5 and 6. For example, the estimated maximal enhancement factor of FIT-PNA 1d (L, VSV) was 3.4. But if the TO base surrogate was shifted only one position to the N-terminal end it lost nearly 70% sensitivity (factor 1.1). In general, a centred position for the intercalating fluorophore substitution is recommended whereas variations cannot be excluded.

↓100

Moreover, one has to deal with the given target sequence possibilities in biological applications, e.g. specific mRNA imaging in living cells. This limits the design of the PNA sequence. In this case the problematic issue is the accessibility of the target sequence for the detecting probe. Secondary structures of mRNA molecules exhibit double stranded regions which produce single stranded loops as side effect. All PNAs used in this work were directed to single stranded regions of the corresponding mRNA species.

Fang et al. 2010 [230] reported about a method to generate a native mRNA antisense-accessible sites library (MASL) for designing mRNA imaging probes. This technique is based on the identification of antisense-accessible sites using RT-ROL (reverse transcription with random oligonucleotides libraries). Albeit it is only demonstrated to work for mouse macrophages this technique provides for a powerful tool to simplify the PNA sequence selection.

An essential drawback of FIT-PNAs is the low quantum yield of thiazole orange and its derivates (0.1-0.4) compared to fluorescent proteins (e.g. EGFP 0.6). However, the target specific increase in fluorescence enables the discrimination of bound and unbound probe. Fluorescent proteins in combination with the MS2 technique are suitable for imaging of mRNA in yeast but tend to form aggregates resulting in false-positive signals [116].

↓101

Notably, these assumptions on hybridisation and fluorescence features of PNA probes are based on measurements in buffered solutions excluding interfering factors which are present in living cells. These factors include binding of RNA/DNA binding proteins, false-positive signals evoked by competitive target sequences of cellular mRNA molecules and the cytosol itself creating the fluorophore environment.

5.2 FIT-PNAs in Real-Time Quantitative PCR

In 1993 Kary Mullis was awarded with the Nobel prize in Chemistry for the invention of the polymerase chain reaction [231]. Essentially, this technique allows the amplification of any nucleic acid sequence generating identical copies. But for analytical purposes the original PCR was limited since the amount of the product was the same independent from the template concentration. Higuchi et al. 1992 [232] improved the standard PCR enabling the simultaneous amplification and product monitoring during the course of the reaction. This is based on the detection of the increase in fluorescence of the reporter dye which is proportional with the amount of specific product. Among the several chemistries which have been developed TaqMan probes® and SYBR-Green I® are the most common used detecting agents [233].

↓102

Typical applications for real-time qPCR include analysis of gene expression, chromosome aberration and single nucleotide polymorphism [234]. In this work the amount of influenza A virus NA and M1 mRNA per infected cell was determined based on a specific real-time qPCR. To enhance the detection specificity the reaction was performed utilizing the FIT-PNAs 1a and 1c which were designed to target the NA and M1 mRNA, respectively, and were essentially used in mRNA imaging in living infected cells. This is a great advantage over Molecular Beacons [126,148] or Locked Nucleic Acids [235]: although they can be applied to qPCR or imaging, a single probe is restricted to one application due to the fact that their functionality is temperature-dependent.

For the qPCR approach performed in this work the primer pairs were designed to amplify a 101bp in length sequence in which the target region of the PNA was included. Therefore the increase in fluorescence of TO and BO correlated with the amount of specific product and enabled an accurate and sensitive quantitative data analysis.

Assuming a certain reaction efficiency (one doubling of molecules per amplification cycle), the number of specific cDNA molecules which were initially present in the sample at the cycle of threshold was determined. The time course of samples 0 – 8 h p.i. revealed a maximum of NA copy number at 4.5 h p.i. whereas for the M1 the maximal amount was reached 5 h p.i. (see chapters 4.1.3. and 4.2.2.). Despite this slight difference in both cases the amount stayed virtually constant until 7 h p.i. and decreased in the following. The lowering of the mRNA abundance might indicate the switch from the transcription to the replication mode.

↓103

Shapiro et al. 1987 [66] generally concluded that the synthesis of viral mRNAs largely determines the translation rate of the encoded proteins. Thus, alterations in the temporal appearance of individual mRNA molecules are surmised. Apart from the NS1 which was proved to be transcribed in the early phase of the viral replication cycle for NA and M1 only a slight difference in the onset of transcription was determined. Notably, based on a rough estimation of the maximal amount of an mRNA species per infected cell revealed a higher concentration for M1 (~106 copies/infected cell) compared to NA (~105 copies/infected cell). This is in good agreement with the viral particle composition: approximately 3000 copies of M1 form the viral core, while only 100 molecules of NA are embedded in the viral envelope [236].

Summarizing, it can be surmised that the amount of a particular mRNA is more crucial for the encoded viral protein abundance than a temporal control of the transcript synthesis.

These results may not exclude alterations evoked by strain specificities and cell population heterogeneity. The variability of the studied biological systems gives rise to considerable uncertainty. That is why these data cannot be in detail related to other influenza A virus strains or cell lines. In fact the time course gives an impression of the dynamics of the infection and the amount of produced viral mRNA. This point requires further studies concerning the remaining influenza A segments which might reveal larger differences in temporal progression and mRNA concentration in infected cells.

↓104

Moreover the application of single cell PCR approaches to the determination of mRNA copy numbers might reveal variations in infection efficiency and progression due to cell heterogeneity.

5.3 Parallel Imaging of Viral mRNAs in Living Infected Cells

To date, most insights in nucleic acid research in living systems are based on two techniques which focus on a sequence specific detection in concert with a fluorescence signal increase: the MS 2 technique [114,115] and Molecular Beacons [125]. Indeed, these strategies are limited to special applications (e.g. imaging in yeast) or exhibit certain disadvantages (e.g. susceptibility to nucleases), respectively. The introduction of FIT-PNAs [170] into the field of mRNA imaging in living cells improved the method spectrum. The FIT-PNA technique is neither restricted to a certain cell type nor suffering from drawbacks which other probes are faced with [116,147,148].

In the present work for the first time FIT-PNAs were employed for viral mRNA imaging in living infected cells using CLSM [183].

↓105

In living influenza A virus infected MDCK cells FIT-PNA 1b showed a 4.5-fold increase in fluorescence at 4.5.h p.i. upon binding to the NA mRNA target sequence. Control experiments with non-infected and SFV infected cells confirmed the high sensitivity of the probe as in these samples no significant fluorescence was detected (see chapter 4.1.6). These findings revealed that the increase in fluorescence of FIT-PNA 1b was due to target hybridisation and not evoked by the binding of proteins or unspecific hybridisation to any viral mRNA.

A second influenza A virus mRNA species was targeted by a PNA molecule carrying BO as intercalating fluorophore. This M1 specific PNA 1c reached an enhancement of 6.97 at 5 h p.i. in comparison to the non-infected control. The result deviated to the measurement with the complementary artificial target RNA furnishing only a 5.7-fold increase. As discussed afore previous studies provided evidence for the impact of the environment on the fluorescence properties of the intercalating dyes. Svanvik et al. 2001 [237] determined the quantum yield influencing factors of TO using light-up probes which demands to be further assessed for the remaining TO derivates. Importantly, both probes show comparable fluorescence enhancement factors upon specific target binding enabling simultaneous detection of the NA and M1 mRNA species in the same cell (see chapter 4.1.6 and 4.2.3). To date, imaging of mRNA in living cells was restricted to one sort of mRNA molecules. With the help of the FIT-PNA technique alterations in localisation, transport or turnover-rates of different RNA molecules can be investigated including dynamic processes in living cells. In particular, this is of great relevance for the understanding of viral-host interaction and the mechanisms behind viral hijacking of host cell processes.

For this purpose, the emission range of TO and BO were adjusted to prevent cross-talk between the two emission channels. TO was measured from 510 nm to 540 nm individually and from 530 nm to 600 nm in combination with the BO probe. Because TO has its emission peak at 530 nm, the shift did not dramatically influence the brightness of the probe. BO emission was recorded from 470 nm to 500 nm in individual measurements but was shifted 10 nm to lower wavelengths with negligible impact on the fluorescence intensity. 

↓106

Whether the derivates of TO besides BO, oxazole yellow (YO) and thiazole pyridine (MO) (see chapter 4.3.1, figure 30), provide for the same applicability in PNA molecules remains to be clarified. It is likely, that several combinations might be problematic due to fluorescence interference effects. Therefore the adjustment of the recorded emission range has to be performed careful to prevent cross-talk between the emission channels.

Simultaneous CLSM imaging of the NA mRNA and the M1 mRNA upon influenza A virus infection 5 h p.i. in living cells revealed alterations in the fluorescence patterns. Despite the observation of a certainly not homogenous distribution in both cases, for the M1 mRNA distinct regions of increased fluorescence were detected (see figure 32, white arrows).

This leads to the assumption that the heterogenous distribution of the NA and M1 mRNA is based on a certain localisation due to regulatory and economical reasons. Martin et al. 2009 [211] reviewed the localisation of cellular mRNA molecules in distinct subcellular compartments. The viral replication is a fine tuned process which requires temporal and spatial control of gene expression. Therefore a controlled localisation of viral transcripts at the rough endoplasmatic reticulum or polysomes to enhance the translation rate of viral transcripts could be surmised.

↓107

This is also reflected by the work of Davey et al. 2011 [80] who described a mechanism how viruses hijack for example the host intracellular transport machinery (see Introduction). It is assumed that viruses mimic cellular motifs, named SLiMs (short linear motifs), as proved for the influenza A protein PB2 mimicking the nuclear localisation signal of the cellular importin ɑ to enter the host nucleus and for NS1 to alter the host immune response [80]. 

In respect to the described binding activity of the NS1 to a wide range of RNA molecules [54] and the influence on the cellular mRNA export as well as viral mRNA translation [57,58] it is conceivable that the NS1 controls also the transport and localisation of the viral mRNA molecules by entering host cell proteins using SLiMs.

Figure 42: Overview of SLiMs roughly classified by function.

(modified from Davey et al. 2011) The numbering scheme corresponds to the examples provided in Davey et al. 2011. Motifs which are used by influenza A protein PB2 and NS1 to interact with cellular proteins are highlighted in red.

↓108

Besides that, viral mRNA is assumed to hold a key position in the influenza A virus infection induced host cell apoptosis as described previously by Morris et al. 1999 [238]. Induction of apoptosis after the last round of progeny virus synthesis may prevents reinfection of cells already been infected [209]. The ability to adjust the efficiency of infection by regulating transcription and replication gives important clues on the virulence and pathogenicity of pandemic influenza A strains. The FIT-PNA technique can be used to study the temporal and spatial progression of transcription and is therefore the method of choice.

A comparison of the images acquired with FIT-PNA 1b in influenza A infected MDCK cells 4.5 h p.i. (see chapter 4.1.6) and 5 h p.i. (see chapter 4.3.2) revealed a difference in the fluorescence signal localisation. At 4.5 h p.i. the signal was mainly concentrated in the nucleus whereas at 5 h p.i. it appeared more cytosolic. This demonstrates the dynamics of the mRNA localisation and will be further assessed in ongoing studies in terms of real-time monitoring of single infected cells. In contrast, MBs enter the nucleus but accumulate there due to binding of RNA/DNA binding proteins. Therefore they are not longer capable for further target detection in the cytosol and are not suitable to monitor viral mRNA transport in living infected host cells.

Whether the binding of the PNA has an impact on the lifetime of the viral transcript or the translation efficiency into viral proteins requires further investigation. Previous studies in nuclear medicine described antisense-strategies with PNA molecules [152,154] and thus, demonstrated the applicability of PNAs as antisense agents. It is likely, that the underlying mechanism of the eventually produced gene knock-down is comparable to that of small interfering RNAs or micro RNAs [83,88].

↓109

Nevertheless, the risk to modify the native localisation, transport and dynamics of (viral) mRNA is not given as it occurs by using GFP tags in tandem repeats (MS2 technique). Hence, the fluorescent probe is very small and does not require molecular modifications of the target or the transgenic expression of the selected gene leading to a non-physiological over-expression. With the help of the FIT-PNA technique the mRNA progression can be followed in real-time in living cells.

A crucial aspect in working with PNA molecules is the transfer into the cytosol of living cells. Due to their uncharged character [150] PNAs do not interact with standard transfection reagents in which the binding to the shuttle agent is based on electrostatical effects (e.g. Lipofectamine, Invitrogen). In this work a sufficient PNA delivery was carried out utilizing streptolysin O mediating a reversible plasma membrane penetration (see chapter 3.2.2.3). For MDCK cells this method did not reduce cell viability as shown with a cell viability test using propidium iodide (see figure 17).

5.4 Imaging of Viral mRNA in Fixed Cells 

Apart from epithelial derived cell lines (MDCK) exhibiting strong connections to neighbouring cells, fibroblasts lack intercellular adhesion and thus were not able to survive the SLO mediated plasma membrane permeabilisation. For the present study of VSV L mRNA, the BHK-21 cells were subjected to fixation and Triton X-100 treatment to deliver FIT-PNA 1d into the cytoplasm. Cell-penetrating peptides (CPPs) might be an attractive tool to solve the delivery problem into living cells concerning fibroblast or SLO sensitive cell lines. Liu et al. 2010 [239] suggested CPPs for the transfer of quantum dots (luminescent semiconductor nanocrystals) in biomedical approaches whereas Shim and Kwon et al. 2010 [240] described CPPs for the targeted delivery of siRNAs in vivo. CPPs are short peptides enriched in basic amino acids that penetrate the plasma membrane in a receptor- and energy-independent manner. Octa-arginine or nona-arginine achieved the highest delivery efficiency in this context [241]. The peptide can be linked covalently or non-covalently to the cargo and showed minor cytotoxicity [242]. Due to their peptide backbone PNA molecules can be easily modified with additional peptide chains. Thus, it is highly recommended to test the applicability of CPPs to the delivery of PNA molecules.

↓110

Besides the drawback that imaging in living cells could not be carried out, interesting information about the progression of the VSV L mRNA were acquired (see chapter 4.4.2., figure 36). At 60 min p.i. the fluorescence of TO increased in comparison to the starting point (0 min p.i.). The quantitative analysis revealed that the high fluorescence signal stayed constant for approximately 30 min and decreased in the following samples (until 150 min p.i.). Although the enhancement factor was low the high sensitivity of the probe generated a clear discrimination to the non-infected and the SFV infected controls showing no significant increase in fluorescence (see chapter 4.4.2., figure 36).

To date, apart from the pathological characteristics and clinical signs less is known about the temporal progression of VSV replication. Investigation of the L mRNA in BHK-21 cells gave first insights into the time course of VSV transcription. Interestingly, the production of this particular VSV mRNA was detected approximately 1 h after infection. This is much faster compared to the results obtained for influenza A virus where the mRNA synthesis started approximately 3 h p.i. This brings up two questions: Why is the VSV transcription faster compared to influenza A virus? And what are the mechanisms behind?

The present data just allow speculations as detailed information about the temporal progression of VSV replication are lacking. Whether the VSV exhibits a more efficient entry or uncoating process leading to an earlier onset of transcription or the polymerases differ in their function remains to be clarified. Thus, since in general the genome composition of VSV (non-segmented) and influenza A virus (segmented) differ from each other it is conceivable that this causes the alteration in transcription rate. Another aspect one has to keep in mind is the amount of novel synthesized virus particles determining the abundance of viral proteins and thus the transcription rate.

↓111

Generally, although BHK-21 and MDCK cells are both of mammalian origin it has to be assumed that these cell lines differ in their biological status which might result in variations in infection progression.

5.5 Global Proteome Analysis of Influenza A Virus Infected MDCK Cells Using SILAC

SILAC is a simple and accurate approach for expression proteomics on a systems level enabling an automated simultaneous identification and quantification of complex protein mixtures [186]. In contrast to ICAT (isotope-coded affinity tag) labelling which is the most well established method in quantitative proteomics by mass spectrometry to date, SILAC does not require multiple fractionation and affinity purification steps and is therefore suitable to small amounts of proteins and the direct comparison of varying states [186]. This is crucial for the presented work as the starting material (number of cells, virus stock) was limited.

Assuming that after a certain number of cell doublings, each argine and lysine will have been replaced by its isotopic analogue, MDCK cells were tested for the efficiency of incorporation. The measurement revealed an incorporation of ~99.8% for cells maintained seven passages in SILAC medium (data not shown). As there is no chemical difference between the labelled and the natural amino acids the cells behaved exactly like a control cell population.

↓112

The SILAC technique was readily applied to the determination of muscle cell differentiation [186], phenotypic comparison of immortalized cell lines with their cognate primary cells [194] and investigations of the impact on the host cell proteome upon virus infection [200,202].

In the present work SILAC was applied to compare the gene expression profile of influenza A virus infected and non-infected MDCK cells on a proteome level. In particular, the viral impact on the host proteome was investigated in the first stages of the infection in an 8 h time-scale. Hence, this study focuses on the influence on viral transcription and replication as well as cellular antiviral processes it is not expected to identify host factors involved in virus assembly, budding or release.

The functional phenotyping of the MDCK proteome after 8 h p.i. provided an unbiased global portrait of representative biological functions (see figure 39). The most prominent category of proteins expressed at higher levels in influenza A infected MDCK cells were related to inflammatory response, induction of apoptosis, immune effector processes, intracellular protein transport and actin cytoskeleton organisation (see also figure 40, cluster 3). Biologically, these reactions coincide with the expected processes in an infected cell and are in good agreement with the findings of a previous study on influenza A virus (A/PR/8) infection in human lung A549 cells at 24 h p.i. [243]. On the one hand the virus hijacks the cellular transport machinery and induces cell death as assumed by Davey et al. 2011 and Morris et al. 1999, respectively [80,238]. On the other hand, the attacked cells recruit defence mechanisms. These proteins responsible for antiviral response were graphed in figure 41 C illustrating the abundance progression in detail. Surprisingly, most of the proteins showed no drastic alterations in protein level over the whole time-scale indicating that they are either activated after 8 h p.i. or completely not included in this specific defence reaction. Further, it is conceivable to surmise that they were blocked by viral host-shutoff processes.

↓113

Other proteins like SERPIN B6, CAD and SF1 increased by more than 2 magnitudes in protein abundance until 8 h p.i. SERPIN B6 belongs to the serpin superfamily and controls inflammation and tumor suppression. Cathepsin G, an antiviral agent in neutrophilic polymorphonuclear granules, is one of its targets [226]. CAD (carbamyl phosphate synthetase II, aspartate transcarbamylase, and dihydro-orotase) catalysis the first 3 steps of the pyrimidine biosynthesis and SF1 regulates the splicing of distinct cellular mRNA molecules. The most impressive pattern in the recorded abundance change was shown for the ZER 1 protein with a fast decrease after 2 h p.i. over 4 magnitudes. It recruits the ubiquitin ligase complex while interacting with CUL-2 inducing ubiquitination of special substrates including transcription factors. Whether this has any antiviral effect cannot be ruled out by the given data set but enrichment of proteins related to ubiquitination was already identified by König et al. 2010 [82] employing an arrayed short-interfering RNA screen of 19,000 human genes.

The reduction of specific host proteins as shown in the functional cluster 1 and 6 (see figure 40) representing mainly cellular lipid metabolism and ion homoeostasis, respectively, might be contributed to the enhanced abundance and thus, activity of proteasomal proteins which were assembled in cluster 4.

The host cell gene expression is critical for influenza A virus replication since the cap structures of polymerase II transcripts are required for priming of viral transcription and the cellular splicing machinery is crucial to produce the alternative splicing variants M2 and NS2 (NEP) [243]. This is reflected by the strongly regulated abundance of proteins assigned to cluster 2 (see figure 40) during the infection time range. Assuming that the maximal level of proteins responsible for positive regulation of transcription from polymerase II promoters correlates with the availability of cap structures this finding is in perfect accordance with the 30 – 60 min delayed maximal concentration of NA and M1 mRNA molecules (please review chapter 4.1.3. and 4.2.2.) estimated for influenza A virus infected MDCK cells.

↓114

Moreover, the influenza A virus proteins (except the M2 protein) were included in this proteomic study. Generally, the level of a certain viral protein logically correlates with its function during the replication cycle or the required copy number present in progeny virus particles: For the NP and the NS1 a strong increase in abundance was revealed in the first 2 hours of infection. This is in good agreement with the corresponding function as the level of NP is assumed to regulate the switch from the transcription to the replication mode and the NS1 activity is surmised to impede cellular antiviral defence mechanisms [32,59]. Contrary, the PB1 and NA protein increased slightly in abundance over the whole measured time-range. Caused by the fact that these proteins are not responsible for regulatory mechanisms during the viral replication there might be no requirement for a highly controlled translation rate (for details please review Introduction). Highly abundant structural proteins like the HA and the M1 [236] showed a comparable linear progression until 8 h p.i. but with larger slope resulting in a three-times higher protein amount.

It has to be admitted that the impact of physiological alterations in the cell`s protein expression cannot be excluded for all presented results. This effect is assessed in ongoing studies investigating also samples of influenza A virus infected MDCK cells 8 – 12 h p.i. and non-infected cells over the same time-range. Besides, it is likely, that by far the majority of the discussed issues concerning influenza A virus induced modifications of the host cell proteome will be verified in further studies.

Summarizing, SILAC is the method of choice for detailed investigation of dynamic protein progression during infection as it provides for reliable results and includes also less abundant proteins which are difficult to detect in less sensitive detection methods like ICAT, ICC or Western Blotting. Besides, no chemical modifications or specific antibodies are required enabling a high throughput analysis of thousands of proteins in parallel. Thus, abundance alterations caused by biological variations of samples are minimized making a direct comparison of differently treated samples possible. The presented experimental strategy is improvable by choosing the time point 8 h p.i. as reference. It is likely, that the usage of 1 h p.i. as reference value implied aberrations concerning the exact viral protein amount caused by inaccurate protein determination. Therefore the above presented data interpretation is based on the overall relation of the viral proteins to each other and not on detailed numerical values.

↓115

The field of proteomics is becoming more quantitative than qualitative [205] enlarging the obtained data sets wherein thousands of details are included complicating the causal interpretation. Therefore, if specific antibodies are available targeted studies on specific virus-host interaction partners identified in previous siRNA screens in combination with immunoprecipitation assays give more direct information.


© 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