4 Results

↓58

4.1 Studies on the Neuraminidase Transcript of Influenza A Virus 

↓59

In fields of virology and medicine it is of great interest to investigate the virus replication in detail. Designing new and efficient antiviral drugs relies on the complete knowledge of the whole infection cycle. After attachment and fusion the production of viral mRNA is the first critical step for infection establishment.

In the following paragraphs of this chapter a novel strategy to investigate the time-resolved progression of viral mRNA synthesis in vitro and its cellular localisation in living infected MDCK cells using sequence specific FIT-PNAs is described. For initial experiments the neuraminidase transcript of influenza A virus was chosen as target due to its high relevance for the viral replication (please review Introduction chapters 1.2 to 1.4) and the possibility to use an accessible target sequence described in a previous imaging study [121]. All presented results in chapters 4.1.1 to 4.1.6 were published in Kummer et al. 2011 [183].

4.1.1 Identification of a Suitable FIT-PNA 

The solid phase synthesis [178]  and fluorescence spectroscopy was performed by Dr. A. Knoll (Institute of Chemistry, Humboldt University Berlin) for all used PNA molecules.

↓60

The suitable FIT-PNA 1a was chosen out of the PNA oligomers listed in table 4. Keeping the length of the PNA oligomers constant, the TO-containing PNA-monomer was shifted through the sequence.

Table 4: PNA oligomers used to identify suitable FIT-PNA 1a with corresponding enhancement factors upon hybridisation with complementary target at 37 °C. Lys = Lysine

Sequence

Enhancement factor

H-Lys-cagtt-Aeg(TO)-ttatgccgttg-Lys-NH2

3.5

H-Lys-cagtta-Aeg(TO)-tatgccgttg-Lys-NH2

5.4

H-Lys-cagttat-Aeg(TO)-tatgccgttg-Lys-NH2

3.9

H-Lys-cagttatt-Aeg(TO)-tgccgttg-Lys-NH2

2.6

H-Lys-cagttatta-Aeg(TO)-gccgttg-Lys-NH2

1.5

H-Lys-cagttattat-Aeg(TO)-ccgttg-Lys-NH2

0.8

H-Lys-cagttattatg-Aeg(TO)-cgttg-Lys-NH2

1.2

H-Lys-cagttattatgc-Aeg(TO)-gttg-Lys-NH2

1.1

Figure 9: Chemical structure of FIT-PNA 1a.

↓61

Aiming on the development of a probe applicable for mRNA imaging in living cells and for quantification of mRNA expression levels by real-time PCR analysis the probe was tested at both conditions. Accordingly, fluorescence spectra of FIT-PNA 1a in absence of the complementary target and in the hybridised status were recorded. To emulate the physiological condition in living cells RNA 3a (see 3.1.9) was used as target at 37 °C and to maintain RT-PCR conditions DNA 5 was employed as target at 60 °C. FIT-probe 1a provided an 11-fold increase of the TO emission upon hybridisation with complementary RNA target 3a and a 12-fold increase upon hybridisation with DNA 5 at 60 °C (see figure 10).

Figure 10: Fluorescence spectroscopic measurements of FIT-PNA 1a.

Excitation wavelength was set to 485 nm. FIT-PNA 1a in presence (solid) and absence (dotted) of matched H1N1 RNA 3a (left panel) and H1N1 DNA target 5 (right panel) at 37 °C and 60 °C, respectively. Conditions: 1 µM probe and 10 µM target in 100 mM NaCl, 10 mM NaH2PO4, pH 7.0.

The N1 specificity of FIT-PNA 1a was confirmed by employing the sequence of the influenza A swine H1N1/Mexico/2009 variant which is shortened in length but the remaining sequence is identical to the A/PR/8 strain ( see figure 11).

↓62

Additionally, the sub-type specificity was assessed employing the analogous sequence of the X-31 influenza A strain (NA mRNA, H3N2, nt 16–32). The X-31 sequence exhibits seven continuous matched base pairs situated in proximity to the TO base surrogate. Nevertheless, the fluorescence intensity of FIT-PNA 1a remained virtually unchanged upon supplementation of RNA 4 (see figure 11) at 37 °C.

Figure 11: Fluorescence spectroscopic measurement of FIT-PNA 1a.

Excitation wavelength was set to 485 nm. Emission spectra of FIT-PNA 1a in presence (solid, grey) and absence (dotted) of semi-matched H3N2 RNA 4 as well as in presence of the H1N1 2009 RNA (solid, black) was recorded from 500 nm to 700 nm at 37 °C. Conditions: 1 µM probe and 10 µM target in 100 mM NaCl, 10 mM NaH2PO4, pH 7.0.

4.1.2 Measurement of FIT-PNA Biostability in Cell Lysate

One of the main issues in nucleic acid research is the biostability of the detection probes. For instance fluorescent RNA or DNA oligomers for FISH (fluorescence in-situ hybridisation) applications lack nuclease-resistance and therefore their usage is essentially limited to fixed cells.

↓63

The enhanced biostability of FIT-PNAs relies on the peptide backbone (see Introduction chapter 1.6.3). This was demonstrated by comparing the fluorescence intensity of the FIT-PNA 1a with a DNA-Molecular Beacon (MB 2) both bound to the complementary RNA target over a time-scale of 60 min in MDCK cell lysate (see chapter 3.2.2.6).

The fluorescence intensity of MB 2 increased and consequently the sensitivity decreased by more than 40% as a result of just 60 min exposure to the cell lysate. This phenomenon can be caused by nuclease-mediated degradation and/or unselective binding of DNA binding proteins. This separates the chromophores spatially and leads to an increase of F0 in the absence of target. In contrast, the increase of fluorescence F/F0 provided by FIT-probe 1a remained high, regardless of the duration of exposure as shown in figure 12.

Figure 12: Relative fluorescent intensities of FIT-PNA 1a and MB 2 in cell lysate upon addition of complementary target.

Probes were excited at 485 nm and 559 nm and emission was recorded at 530 nm and 593 nm for FIT-PNA 1a and MB 2, respectively. Conditions: 1 µM probe, 10 µM RNA target in 100 µl cell lysate, 37 °C.

↓64

We concluded that FIT-probe 1a is stable in cell lysates and thus, very likely also in living cells.

4.1.3 Real-Time Quantitative PCR

Real-time qPCR is frequently employed in the field of gene expression research. With the help of this sensitive method even very small changes in the expression level are detected. In this work RT-qPCR was used to assess the viral mRNA production during the infection cycle of influenza A/PR/8 in MDCK cells. The applied infection protocol, total RNA purification from MDCK cells, in vitro cDNA synthesis and RT-qPCR were carried out as described (see chapters 3.2.4.1 to 3.2.4.4).

In contrast to standard RT-PCR the detection fluorophore SYBR-Green (Invitrogen) was replaced by the NA specific FIT-PNA 1a. The increase in TO fluorescence during the amplification revealed the synthesis of an influenza A/PR/8 NA specific sequence in infected samples, which was not observed for the non-template control (absence of cDNA) and for non-infected MDCK cells ( see figure 13, left panel).

↓65

For quantitative analysis, a 10-fold dilution series of the amplicon was performed (see figure 13, right panel) and from this data a calibration curve was calculated (see figure 14). The cycle numbers needed to furnish threshold fluorescence (threefold above the average of non-template control fluorescence emission at cycles 3 – 12) were plotted against the template concentration. The CT value is linear, inversely proportional to the corresponding logarithmic template concentration: if the concentration of template decreases, the CT value increases, because there is less specific product synthesised. Therefore the coefficient of determination R2 is used to estimate the proportion of variability in a data set. R2 (given in figure 14) showed a high correlation and revealed high amplification efficiency and specificity. Over at least seven orders of magnitude the FIT-probe 1a provided a linear measuring range.

Figure 13: Quantitative Real-time PCR analysis.

Left: Amplification curve on measuring fluorescence of FIT-PNA 1a in response to 1 ng cDNA obtained from influenza A/PR/8 infected MDCK cells 4.5 h p.i. (solid, grey), non-infected control cells (solid, black) and the non-template control (dotted). Right: Amplification curve of a short sequence encoding for the NA (H1N1) in a 10-fold dilution series. For conditions please review chapter 3.2.4.3. The mean of three independent experiments is plotted.

Figure 14: Logarithmic presentation of the calibration curve. Cycle of threshold values plotted against the logarithm of the template concentration in ng/μl to estimate the spec-ificity of the RT-qPCR (R2).

↓66

The results of the qPCR analysis revealed the time-dependency of the expression levels of viral mRNA. In figure 15 the NA copy number per 1 ng cDNA at 1 - 9 h p.i. is given. The graphed results revealed the required time of the virus to attain the maximal level of NA mRNA at 4 – 5 h p.i., which means that approximately 105 copies of NA mRNA per ng cDNA were used as template in qPCR.

In addition, assuming that 106 MDCK cells were used per sample one can roughly estimate the maximum NA mRNA level corresponding to about 104 copies per infected cell. This is in agreement with data reported for MDCK cells infected with an influenza A virus reassortant (A/NWS/33HA-G70cΔNAMviA, H1N9). In this study 105 copies of total viral mRNA per cell 7 h p.i. were detected by a ribonuclease protection assay [209].

Figure 15: Time course of the NA mRNA copy number per ng cDNA of influenza A/PR/8 infected MDCK cells.

The calculation relies on the calibration curve (Figure 14) obtained from the RT-qPCR analysis. Mean ±SEM of three independent experiments is plotted.

↓67

The viral mRNA expression levels decreased at prolonged times after infection illustrating the switch from transcription to replication mode.

4.1.4 Verification of Influenza A Infection

For fluorescence intensity data analysis the cells were selected manually from images (see chapter 4.1.6). Hence, it is important to determine the percentage of infected cells. Infection of MDCK cells with influenza A/PR/8, followed by specific labelling with an anti-NP (H1N1) antibody as well as a secondary goat anti-mouse antibody was performed as described in chapters 3.2.3 and 3.2.5 5 h p.i. The confocal laser scanning microscopy images depicted in figure 16 indicate an infection efficiency of ~98% by the near quantitative staining.

Figure 16: Confocal laser scanning microscopy images of fixed MDCK cells stained with an anti-NP (H1N1) antibody.

(A, B) Non-infected MDCK cells and (C, D) influenza A/PR/8 infected MDCK cells (5 h p.i.) were labelled with an anti-NP (H1N1) antibody to determine the infection efficiency. Non-infected cells showed no fluorescence signal. In contrast the infected sample showed a strong nuclear NP staining. Images were acquired with an inverted confocal laser scanning microscope using a 60x oil-immersion objective at room temperature. Alexa 568 (anti-NP) was excited employing a 559 nm laser. White bars correspond to 10 μm. DIC = differential interference contrast

4.1.5 Verification of Cell Viability

↓68

FIT-PNAs were transferred into living MDCK cells using streptolysin O. As a consequence of the SLO treatment and reversible permeabilisation of the plasma membrane the cell viability might be impaired in cell damage and death. Here, a cell vitality test based on propidium iodide, which is excluded by intact cells, was employed to address this important issue. As demonstrated in figure 17 propidium iodide was excluded from both MDCK cells treated with SLO and untreated MDCK cells showing just minor fluorescence inside the cells. In contrast, cells incubated at acidic conditions exhibited a strong propidium iodide signal.

Figure 17: Confocal laser scanning microscopy images of living MDCK cells stained with propidium iodide (PI).

(A, D) Untreated MDCK cells, (B, E) SLO-treated MDCK cells and (C, F) HCl-treated cells were incubated with PI for 10 min to determine the cell viability. Untreated and SLO-treated cells showed minor fluorescence signals. In contrast, the HCl-treated sample demonstrated a positive PI incorporation. Images were acquired with an inverted confocal laser scanning microscope using a 60x water objective at room temperature. PI was excited employing a 559 nm laser. White bars correspond to 10 μm. DIC = differential interference contrast

4.1.6 Detection of Neuraminidase mRNA in Living Infected MDCK Cells

Detailed information about a certain biological process requires preferably an easy examinable system that closely mimics in vivo conditions. Working in cell culture provides both: a living biological system and easy handling. For influenza A infection MDCK cells are the most prominent cell culture model system [210]. Employing this cell line the formation and localisation of NA mRNA was followed with the help of a sequence specific FIT-PNA. The probe 1a introduced in chapter 4.1.1 was extended by PEGylation to enhance the solubility in the cell cytoplasm and to prevent segregation or nuclear import [211,212]. The linkage of a polyethylene glycol residue had just minor effects on its fluorescence responsiveness [183].

↓69

FIT-PNAs were transferred into living infected (and non-infected) MDCK cells grown on 35 mm glass bottom dishes enabled by SLO mediated plasma membrane permeabilisation (see chapter 3.2.2.3).

Considering the result of maximum NA mRNA level obtained by the above described RT-qPCR (see chapters 3.2.4.3 and 3.2.4.4) images in figure 18 were acquired 4.5 h p.i. to obtain the maximum enhancement of fluorescence. A first visual inspection indicated a strong increase in fluorescence for the influenza A infected cells in comparison to non-infected cells. Neither Semliki Forest Virus (SFV) [213,214] infected cells nor cells infected with influenza A but stained with VSV L protein specific FIT-PNA 1d evinced similar fluorescence patterns. The probe 1d fluoresced upon hybridisation with specific VSV L mRNA in infected BHK-21 cells (see chapter 4.4), but is virtually non-responsive in influenza infected cells. These control experiments proved that the increased fluorescence is due to specific complex formation of FIT-PNA 1b with NA mRNA, but not to virus infection per se or unspecific binding processes mediated by proteins or nucleic acids.

The signal distribution inside influenza A infected MDCK cells appeared non-random. It is conceivably that due to regulatory, developmental and economical aspects the (viral) mRNA is localised in subcellular compartments utilizing cellular transport mechanisms [211].

↓70

Figure 18: Confocal laser scanning microscopy images of living MDCK cells stained with FIT-PNAs.

(A - D) Influenza A/PR/8 infected, (E - H) non-infected and (I - M) Semliki Forest Virus infected MDCK cells stained with FIT-PNA 1b (NA, H1N1) and (N – Q) influenza A/PR/8 infected cells stained with FIT-PNA 1d (L, VSV) at 4.5 h p.i. All control samples (E – Q) showed minor fluorescence signals. In contrast, the influenza A/PR/8 infected and with the NA mRNA specific FIT-PNA treated MDCK cells exhibited high TO fluorescence signals. Images were acquired with an inverted confocal laser scanning microscope using a 60x water objective at 37 °C. TO was excited employing a 488 nm laser. White bars correspond to 10 μm. DIC = differential interference contrast

An evaluation of FIT-PNA 1b performance compared to MB was realized by measuring MB 2 used in a previous viral mRNA imaging study [121] and comparing the fluorescence enhancement upon target binding in living infected MDCK cells. As illustrated in figure 19 the NA specific MB 2 showed just weak fluorescence signals. No significant difference between the infected and the non-infected cells could be visually observed.

Figure 19: Confocal laser scanning microscopy images of living MDCK cells stained with MB 2.

(A - D) Influenza A/PR/8 infected and (E - H) non-infected MDCK cells stained with MB 2 (NA, H1N1) at 4.5 h p.i. A comparison of both samples revealed no clear difference in fluorescence intensity. Images were acquired with an inverted confocal laser scanning microscope using a 60x water objective at 37 °C. TMR was excited employing a 559 nm laser. White bars correspond to 10 μm. DIC = differential interference contrast

↓71

For fluorescence intensity quantification an image analysis was performed using Image J. Regions of interest (ROI) including nucleus and cytoplasm were selected as depicted in figure 20 and the mean fluorescence intensity per area was calculated by the programme. The background fluorescence was subtracted for each image and all values were normalised to the corresponding control.

Images of influenza A infected MDCK cells stained with FIT-PNA 1b revealed a 4.5-fold enhancement of fluorescence compared to non-infected MDCK cells. Control experiments were performed and supported this finding. While SFV infected and FIT-PNA 1b treated cells showed a 1.1-fold increase in fluorescence, for the influenza A infected but stained with FIT-PNA 1d MDCK cells no increase in fluorescence was observed.

Figure 20: Image analysis using Image J for determining the enhancement factor.

Row data images were used to calculate the mean fluorescence intensity of manually selected areas (ROI = region of interest). White bar corresponds to 10 μm.

↓72

Figure 21: Relative fluorescence intensities of FIT-PNAs and MB 2 in living infected MDCK cells.

The integrated fluorescence intensity values of influenza A/PR/8 NA specific FIT-PNA 1b, VSV L protein specific FIT-PNA 1d and influenza A/PR/8 NA specific MB 2 in MDCK cells infected with influenza A/PR/8 or infected with SFV, respectively, are shown. Normalized values are indicated as bars. Scatter bars correspond to SEM. Data represents three independent experiments with five cells per experiment (n=15).

4.1.7 FACS based Detection of Influenza A Infection 

The FIT-PNA technique is applicable to a variety of diagnostic standard methods. This was demonstrated by utilizing the neuraminidase specific FIT-PNA 1b to detect influenza A infected MDCK cells in a FACS based approach (see figure 22). In contrast to the non-infected control, the influenza A/PR/8 infected cells showed an increased TO fluorescence signal. Furthermore, the shape of the intensity curve exhibited an additional maximum, which is called shoulder.

A second control using the X-31 influenza A strain (H3N2) was included into the FACS analysis. This strain lacks completely the target sequence of FIT-PNA 1b. Interestingly, the X-31 infected MDCK cells revealed a decrease in detected TO fluorescence compared to the non-infected control indicating the presence of a partial matching sequence in the cellular transcriptome. For verification of the presented results the experiment was repeated three-times independently. All replicates confirmed the results plotted in figure 22.

↓73

Figure 22: FACS analysis of influenza A infected MDCK cells using sequence specific FIT-PNA 1b.

MDCK cells were infected with the influenza A/PR/8 and the X-31 strain (see 3.2.3). After 4.5 h incubation the cells were trypsinized and fixed in solution. Cell permeabilisation and delivery of FIT-PNA 1b was performed using saponin. The shown histogram presents non-gated data of 10000 counts. TO was excited at 494 nm and emission maximum was set to 519 nm.

4.2 Studies on the Matrix Protein 1 Transcript of Influenza A Virus

To investigate the progress of viral transcription is one of the most relevant issues in influenza A research. This mechanism is highly conserved throughout the different influenza A virus strains and thus holds great promise for new antiviral strategies. During the viral replication cycle the coordinated action of the distinct viral proteins contributes to virus entry, genome release into the cytoplasm, transcription, replication, protein biosyntheses, assembly and budding. Consequently, regulation of transcription in terms of function enforcement is conceivable. Are there any temporal differences in viral mRNA progression during the time course of replication?

For this purpose a second FIT-PNA specific to the matrix protein 1 (M1) of influenza A/PR/8 carrying a TO derivate called pyridinium benzothiazole (BO) as intercalating fluorophore was designed and examined.

4.2.1 Identification of a Suitable PNA

↓74

Based on the experience gained from the design of FIT-PNA 1b specific to NA mRNA the PNAs specific to M1 mRNA were composed with a centred base surrogate and a polyethylene glycol chain (PEG) to enhance the specific fluorescence signal and the solubility of the probe inside the cytosol, respectively.

The FIT-PNA probe with the maximum enhancement (third PNA sequence in table 5, figure 23) was chosen and named FIT-PNA 1c in the following.

Table 5: PNA oligomers used to identify suitable FIT-PNA 1c with corresponding enhancement factors upon hybridisation with complementary target RNA at 37 °C. Lys = Lysine

Sequence

Enhancement factor

H-Lys(PEG)-catg-Aeg(BO)-ctgattagtg

1.8

H-Lys(PEG)-catgt-Aeg(BO)-tgattagtg

2.1

H-Lys(PEG2)-catgtctg-Aeg(BO)-ttagtg

5.7

H-Lys(PEG)-catgtctga-Aeg(BO)-tagtg

2.0

↓75

Figure 23: Chemical structure of PNA 1c.

Measurements with the synthetic target sequence (H1N1) and a mismatch control (H3N2) revealed an enhancement factor of 5.7 and 2.0, respectively.

Figure 24: Fluorescence spectroscopic measurements of FIT-PNA 1c.

Excitation wavelength was set to 440 nm. The fluorescence of FIT-PNA 1c in presence (solid, grey) and absence (dotted) of matched H1N1 RNA 3c and mismatched RNA target 4 (H3N2) was recorded from 450 nm to 600 nm at 37 °C. Conditions: 1 µM probe and 10 µM target in 100 mM NaCl, 10 mM NaH2PO4, pH 7.0.

↓76

4.2.2 Real-Time quantitative PCR

Similarly to the neuraminidase mRNA progression studies using quantitative Real-time PCR the synthesis of M1 mRNA during the replication cycle was estimated (for details please see chapter 4.1.3).

With one exception the RT-qPCR was performed following the described protocol (see chapter 3.2.4.3): the M1 specific FIT-PNA 1c was used as the detecting probe. Therefore excitation and emission conditions had to be adjusted to the BO fluorescence properties. The probe was excited using a 450/25 nm excitation filter and emission was recorded with the help of a 485/20 nm emission filter.

↓77

In this case, the increase in BO fluorescence directly correlated to the amplification of an M1 specific sequence in influenza A/PR/8 infected samples. For the non-template and the non-infected control samples no specific sequence amplification was observed.

Using the same principle as described above a 10-fold dilution series (see figure 25) of the amplicon was analysed to generate a calibration curve (see figure 26) and to quantify the amount of M1 mRNA molecules in infected MDCK cells (see figure 27).

Figure 25: Amplification curve of a 101 bp in length sequence encoding for the M1 (H1N1) in a 10-fold dilution series.

For conditions please review chapter 3.2.4.3. The mean of three independent experiments is plotted.

↓78

Figure 26: Logarithmic presentation of the calibration curve. Cycle of threshold values plotted against the logarithm of the template concentration ng/µl to estimate the specificity of the RT-qPCR (R2).

Figure 27: Time course of the M1 mRNA copy number per ng cDNA of influenza A/PR/8 infected MDCK cells.

The calculation relies on the calibration curve (see figure 18) resulted from the RT-qPCR analysis. Mean ±SEM of three independent experiments is plotted.

4.2.3 Detection of Matrix protein 1 mRNA in Living Infected MDCK Cells

The applied staining protocol for the SLO-mediated delivery of FIT-PNA 1b was adjusted to the time requirements of the M1 mRNA progression as determined by the above described RT-qPCR analysis. Therefore imaging was performed at 5 h p.i.

↓79

Living influenza A/PR/8 infected MDCK cells 5 h p.i. showed an intense cytosolic fluorescence signal (see figure 28). The low fluorescence signals in the non-infected and SFV infected MDCK cells supported the sequence specificity of PNAs, especially of PNA 1c.

Interestingly, the intracellular fluorescence pattern differs from that of the FIT-PNA 1b based staining of NA mRNA. This phenomenon remains to be clarified by simultaneous detection of M1 mRNA and NA mRNA in living infected cells (see chapter 4.3).

Figure 28: Confocal laser scanning microscopy images of living MDCK cells stained with FIT-PNA 1c.

(A - D) Influenza A/PR/8 infected, (E - H) non-infected and (I - M) Semliki Forest Virus infected MDCK cells stained with FIT-PNA 1c (M1, H1N1) at 5 h p.i. The control samples (E-M) showed minor fluorescence signals. In contrast, the influenza A/PR/8 infected and with the M1 mRNA specific FIT-PNA treated MDCK cells exhibited high BO fluorescence signals. Images were acquired with an inverted confocal laser scanning microscope using a 60x water objective at 37 °C. BO was excited at 440 nm. White bars correspond to 10 μm. DIC = differential interference contrast

↓80

Image analysis for fluorescence intensity calculations was performed according to the procedure described for FIT-PNA 1b. The mean fluorescence intensity of BO in influenza A infected MDCK cells showed an enhancement factor of 6.97 compared to the non-infected control. In contrast, SFV infected MDCK cells revealed only a 1.2-fold increase of intensity (see figure 29).

Figure 29: Relative fluorescence intensities of FIT-PNA 1c in living infected MDCK cells.

The integrated fluorescence intensity values of influenza A/PR/8 M1 specific FIT-PNA 1c in MDCK cells infected with influenza A/PR/8 or infected with SFV, respectively, are graphed. Normalization was performed referring to the corresponding mean of the control (non-infected). Scatter bars correspond to SEM. Data represents three independent experiments with five cells per experiment (n=15).

4.3 Simultaneous Study on the Neuramindase and Matrix Protein 1 Transcripts

Another crucial benefit of the introduced probes using intercalating fluorophores for a sequence specific detection of viral mRNA in living infected cells is the possibility of a simultaneous visualisation of several mRNA species.

4.3.1 Imaging Conditions

↓81

Considering the emission spectra (see figure 30) of both fluorophores, thiazole orange and pyridium benzothiazole, the emission range was adjusted in order to prevent cross-talk between both channels (TO: 530 nm - 600 nm, BO: 460 nm - 490 nm).

Figure 30: Relative fluorescence spectra of intercalating fluorophores.

BO = pyridium benzothiazole organge, YO = oxazole yellow, TO = thiazole orange, MO = thiazole pyridine

However, the emission of BO could lead to an excitation of the TO fluorophore and thus to false positive signals. Therefore BO and TO were measured in the other respective channel (see figure 31). This control revealed that the cross-talk effect is negligible in the case of localisation studies. But it is recommended to perform the experiments for fluorescence enhancement calculations upon sequence specific hybridisation in living cells separately.

↓82

Figure 31: Confocal laser scanning microscopy images of living influenza A/PR/8 infected MDCK cells measured in the BO (Ex = 440 nm) and the TO (Ex = 484 nm) channel.

MDCK cells were stained with (A-C) PNA 1c (BO, M1 specific) or (D-F) FIT-PNA 1b (TO, NA specific) and imaged at 5 h p.i. Each probe was excited by both channels sequentially. Images were acquired with an inverted confocal laser scanning microscope using a 60x water objective at 37 °C. White bars correspond to 10 μm. DIC = differential interference contrast

4.3.2 Simultaneous Detection of Neuraminidase and Matrix protein 1 mRNA in Living Infected MDCK Cells 

As discussed earlier, the pattern of fluorescence signal due to specific binding of PNA 1c displayed variations in signal intensity as highlighted by white arrows in figure 32 (A, E, I). In contrast, the fluorescence signal of FIT-PNA 1b seemed to be as well distributed heterogeneously but lacks such regions of signal accumulation.

Figure 32: Confocal laser scanning microscopy images of living influenza A/PR/8 infected MDCK cells measured in the BO (Ex = 440 nm) and the TO (Ex = 484 nm) channel.

MDCK cells were stained with PNA 1c (BO, M1 specific) and FIT-PNA 1b (TO, NA specific) and imaged at 5 h p.i. Each probe was excited for both channels sequentially. Arrows point at regions of high fluorescence intensity. Images were acquired with an inverted confocal laser scanning microscope using a 60x water objective at 37 °C. White bars correspond to 10 μm. DIC = differential interference contrast

4.4 Studies on the RNA-Dependent RNA Polymerase Transcript of Vesicular Stomatitis Virus

↓83

The sequence specific detection of viral mRNA inside the infected cells was implemented to investigate the Vesicular Stomatitis Virus (VSV) replication. This demonstrates the versatile application options of the FIT-PNA technique.

VSV belongs to the group of negative-sense single-stranded non-segmented RNA viruses (mononegavirales), precisely to the Rhabodoviridae family. It causes serious zootic vesicular disease in cattle, pigs, horses and numerous other species [215-218]. The genome is composed of a sequential arrangement of genes which are separated by non-coding intergenic regions. The RNA-dependent RNA polymerase (= L protein) is the most crucial protein of VSV [219]. It mediates the transcription of the negative ssRNA into positive strands mimicking cellular mRNA molecules. Mammalian cells lack such an enzyme making the L protein essential for VSV replication. The VSV specific FIT-PNA was designed to target the highly conserved region of the L protein situated at the 5´end of the viral genome [220-222].

4.4.1 Identification of a Suitable FIT-PNA

First, to design appropriate probes, mfold calculations were carried out with the help of the prediction tools on the Zuker webpage (see 3.1.10) to assess the secondary structure of the respective mRNA molecule (see figure 33).

↓84

To identify conserved regions, the L gene of experimentally used virus particles was sequenced by the Meixner GmbH (Berlin). The obtained sequence was aligned to the one given in the PubMed database (NC_001560, Indiana strain). Matched regions were assumed to be conserved or at least less susceptible for mutations.

Figure 33: Secondary structure of the L protein mRNA (VSV) predicted by mfold.

The structure was obtained using the mfold calculator. The inset shows a magnification of the target region used as template for the FIT-PNA design (sequence marked with red bars).

A single stranded sequence inside a conserved region was chosen and ten PNA oligomers (listed in table 6) were synthesized with changing position of the TO base surrogate. Furthermore, for solubility improvement, different amino acid residues were tested. The PNA with the maximum increase in fluorescence upon hybridisation with the target sequence (3.4 at 37 °C) will be named FIT-PNA 1d in the following.

↓85

Table 6: PNA oligomers used to identify suitable FIT-PNA 1d with corresponding enhancement factors upon hybridisation with complementary target at 37 °C. Lys = Lysine, Glu = Glutamic acid, Gly = Glycine

Sequence

Enhancement factor

H-Glu-cgttt-Aeg(TO)-taattcgtctc-Gly-NH2

3.0

H-Lys-cgttt-Aeg(TO)-taattcgtctc-Lys-NH2

3.1

H-Lys-cgttt-Aeg(TO)-taattcgtctc-Lys-NH2

3.1

H-Glu-cgtttc-Aeg(TO)-aattcgtctc-Gly-NH2

1.1

H-Glu-cgtttct-Aeg(TO)-attcgtctc-Gly-NH2

3.4

H-Glu-cgtttcta-Aeg(TO)-ttcgtc-Gly-NH2

3.2

H-Glu-cgtttcta-Aeg(TO)-ttcgtctc-Gly-NH2

2.1

H-Glu-cgtttctaa-Aeg(TO)-tcgtc-Gly-NH2

2.4

H-Glu-cgtttctaa-Aeg(TO)-tcgtctc-Gly-NH2

2.0

H-Glu-cgtttctaat-Aeg(TO)-cgtctc-Gly-NH2

0.45

Figure 34: Fluorescence spectroscopic measurement of FIT-PNA 1d.

Excitation wavelength was set to 485 nm. FIT-PNA 1c was measured in presence (solid) and absence (dotted) of matched target RNA 6 at 37 °C. Conditions: 1 µM probe and 10 µM target in 100 mM NaCl, 10 mM NaH2PO4, pH 7.0.

4.4.2 Verification of Vesicular Stomatitis Virus Infection

To determine the VSV infection the glycoprotein G of VSV [216] was visualized using a polyclonal antibody in combination with a Cy3 conjugated secondary goat anti-rabbit IgG antibody. The ICC was performed 3 h p.i. as described above (see chapters 3.2.3 and 3.2.5). The near quantitative staining of VSV infected BHK-21 cells revealed ~100% VSV G positive cells (see figure 35). Because the image analysis was performed manually, the percentage of infected cells is of great relevance.

↓86

Figure 35: Confocal laser scanning microscopy images of fixed BHK-21 cells stained with an anti-G (VSV) antibody.

(A, B) Non-infected BHK-21 cells and (C, D) VSV infected BHK-21 cells (3 h p.i.) were labelled with an anti-G (VSV) polyclonal antibody to determine the percentage of infected cells. Non-infected cells showed no fluorescence signal. In contrast the infected sample showed a strong G (VSV) staining. Images were acquired with an inverted confocal laser scanning microscope using a 60x oil-immersion objective at room temperature. Cy3 (goat anti-rabbit IgG) was excited employing a 559 nm laser. White bars correspond to 10 μm. DIC = differential interference contrast

4.4.3 Detection of L Protein mRNA in Fixed BHK-21 Cells

To apply the SLO-mediated FIT-PNA delivery failed for BHK-21 cells. In contrast to the intercellular adhesion of epithelial MDCK cells, fibroblasts lack strong connections to neighboring cells and thus were not able to survive the plasma membrane permeabilisation. Therefore BHK-21 cells were subjected to fixation and Triton X-100 treatment to transfer FIT-PNA probes into the cytoplasm.

Hence, it was not possible to carry out a continuous time-resolved measurement of the fluorescence progression during the VSV replication cycle due to the aforementioned delivery problem. The time course illustrated in figure 36 represents samples where the replication cycle was interrupted at various time points (0 min, 30 min, 60 min, 90 min, 120 min, 150 min) post infection by fixation.

↓87

After staining with the VSV L protein specific FIT-PNA 1d, an increase in TO fluorescence was observed for the samples at 60 min p.i. and 90 min p.i. (see figure 36 A) compared to the starting point (0 min p.i.). At 120 min p.i. the fluorescence signal returned to its basic intensity. A comparison to non-infected control cells (see figure 36 B) or SFV infected cells (see figure 36 C) showed minor non-significant changes in fluorescence intensities over the same time scale indicating that probe 1d selectively responded to viral mRNA and fluorescence enhancement was not due to unexpected degradation and/or binding processes.

It is known, that during the fixation, preferentially the lysine residues of the proteins were cross-linked by paraformaldehyde preserving secondary and even tertiary structures. Whether there is an influence on the secondary structure of the viral mRNA which has an impact on the accessibility of the target sequence cannot be ruled out.

Figure 36: Confocal laser scanning microscopy images of fixed BHK-21 cells stained with FIT-PNA 1d in a 150 min time course.

(A) VSV infected, (B) non-infected and (C) Semliki Forest Virus infected BHK-21 cells stained with FIT-PNA 1d (L Protein, VSV). The samples were fixed and stained at the indicated time point p.i. All control samples (B, C) showed minor variation in fluorescence signals. In contrast, the VSV infected BHK-21 cells exhibited an enhancement of TO fluorescence intensity 60 – 90 min p.i. Images were acquired with an inverted confocal laser scanning microscope using a 60x oil-immersion objective at 37 °C. TO was excited employing a 488 nm laser. White bars correspond to 10 μm. DIC = differential interference contrast

↓88

For quantitative analysis the method described in chapter 4.1.6 was applied to determine the fluorescence enhancement. Importantly, in this study the nucleus was not included in the calculation of the mean fluorescence intensity.

Data analysis verified the visual observation. For the non-infected and SFV infected BHK-21 cells stained with the FIT-PNA 1d specific to VSV L protein mRNA no significant change in fluorescence intensity was determined (figure 37, upper right and lower left graph, respectively). An additional control was included, namely VSV infected BHK-21 cells stained with FIT-PNA 1b (NA mRNA, H1N1). These cells exhibited no significant fluorescence intensity over the time course as the aforementioned controls demonstrating again the specificity of the FIT-PNA 1d.

In contrast, upon normalization to the corresponding basic fluorescence intensity (0 min p.i.) the VSV infected BHK-21 cells which were treated with the specific FIT-PNA 1d revealed a 1.6 fold enhancement in fluorescence 90 min p.i.

↓89

Figure 37: Fluorescence intensity of VSV L specific FIT-PNA 1c in fixed BHK-21 cells.

The absolute fluorescence intensity of VSV infected (upper left), non-infected (upper right), and SFV infected (lower left) fixed BHK-21 cells stained with FIT-PNA 1d as well as VSV infected BHK-21 cells stained with FIT-PNA 1b (NA, H1N1) is graphed for various time points p.i. from 0 to 150 min p.i. CLSM images were analysed following the procedure described in chapter 4.1.6 excluding the cell nucleus. The mean fluorescence intensity per area was normalized to the intensity at time 0 min p.i. and plotted with SEM. Data represent three independent experiments with three cells per time point and sample. (p* = 0.0174)

4.5 Proteomic Studies of Influenza A Infected MDCK Cells on a Systems Level using Stable Isotope Labelling of Amino Acids in Cell Culture (SILAC) 

Global quantitative analysis of host cell proteomes among others is destined to revolutionize the understanding of viral host interaction. The impact on the host cell proteome upon influenza A infection was analysed using SILAC in combination with LC-MS (see chapter 3.2.8). Samples of infected MDCK cells 0 to 8 h p.i. were investigated.

As a result of an alignment to the dog protein data set, GO annotations and an applied hypergeometric test (see chapter 3.2.9) the identified proteins were clustered by their specific abundance variation (see figure 38). In general, the protein abundance increased for the majority of identified proteins comparing the cell proteome status at 8 h p.i. with the starting situation (0 h p.i.).

↓90

Figure 38: Fold-change distribution of the MDCK proteome upon 8 h infection by influenza A virus.

The total protein amount is plotted against the Log2 fold change in protein abundance for each cluster. The histogram is divided into four quantiles according to the relative protein expression (0 – 5 %, 5 – 25 %, 25 -75 %, 75 – 100 %). 
*5-25%

The presented clusters were assigned to higher-order groups according to their relative expression into 4 quantiles: 0 - 5 %, 5 – 25 %, 25 – 75 % and 75 – 100 %. A classification into quantiles enabled the evaluation with respect to the magnitude of abundance variation. Each quantile was assessed separately for over-represented biological processes based on the gene ontology pathway analysis and the resulting categories were clustered concerning their z-transformed p values in a so called heat map (see figure 39). This enabled the visual interpretation of the infected MDCK cell phenotype after 8 h p.i. in terms of functional modules on a systems level.

Selection of the most prominent biological processes of each quantile resulted in a simplified visual interpretation. In the first category (0 – 5 %), proteins which are responsible for lipid metabolism, like lipoprotein biosynthesis, amino acid lipidation as well as phospholipid, membrane lipid and glycolipid metabolism, were over-represented. Interestingly, also cell adhesion mediating proteins were registered. These proteins were stronger down-regulated than the remaining 95 %. The second category (5 – 25 %) is dominated by proteins involved in ion homeostasis of the cell. The two first categories indicated that 25 % of all proteins were down-regulated. Remarkably, for the main fraction of the identified proteins the estimated abundance remained constant or showed only moderate up-regulation upon influenza A virus infection. Proteins related to RNA biosynthetic processes, like mRNA processing, RNA localisation and splicing, RNA metabolism, as well as factors responsible for positive regulation of gene expression were over-represented in this third quantile.

↓91

The fourth quantile “75-100%” included proteins which were higher up-regulated than the remaining 75 % of all proteins. Considering their GO annotation these proteins were categorized to the following cellular functions: programmed cell death, immune response and protein transport/localisation/modification.

Figure 39: Functional phenotyping of the influenza A/PR/8 infected MDCK cell proteome.

Quantiles resulting from the quantification histogram (see figure 38) are indicated at the top of the heat map. Each quantile was separately analysed for gene ontology pathways and clustered for the z-transformed p values. The most prominent representatives of all over-represented biological processes of each quantile were selected and annotated. At the bottom left a histogram of the z-transformed p-values is plotted.

↓92

Further, an mfuzz clustering was performed resulting in cellular protein groups showing the same dynamics in protein level change during the influenza A virus infection. This method relies on the characterization of single GO terms over the whole time course based on their membership value. (see figure 40, cluster 1-6). While the heat map enabled an absolute evaluation of the protein abundance change in MDCK cells after 8 h p.i., the mfuzz clustering provided for a detailed time-resolved analysis.

Figure 40: Fuzzy c-means clustering of the cellular proteins over the infection time course (0 – 8 h p.i.).

Proteins were analysed according to their expression pattern. The time range is not displayed linear due to program restrictions with respect to plotting.

Generally the mfuzz clustering in the first instance revealed 6 groups with a certain expression pattern. Interestingly, for most of the clusters the important time point seemed to be 2 h p.i. indicating the viral impact on cellular processes or the cellular response to viral infection.

↓93

For the transcription of the viral genome, the usage of cap structures originating from cellular mRNA molecules as initiation primers is essential. This mechanism realized by the viral polymerase complex is named cap snatching and results in the premature degradation of cellular mRNA molecules. This may hint at the causal correlation of the expression pattern graphed as cluster 1. Here, processes of the lipid metabolism and mitochondrion organization (inter alia, ATP syntheses coupled electron transport chain) were over-represented and decreased in abundance over the time course of infection.

Cluster 2 showed a comparable expression pattern like cluster 3 but with greater value alterations of expression. Even on the functional level these two clusters seemed to be connected. Proteins of cluster 2 were involved in gene expression processes (e.g. positive regulation of gene expression, negative and positive regulation of transcription from RNA polymerase II promoter and transcription in general). All these proteins control or establish protein expression and thus are required for all regulated proteins during the analysed time range.

A consistent increase in protein level 0 – 8 h p.i. was found for all GO terms assigned to cluster 3. They represent three main functional subgroups: processes related to apoptosis (ubiquitination, proteolysis, actin filament and microtubule organization), to RNA metabolism (RNA splicing, localisation, transport) and to antiviral response (defence response to virus, regulation of immune effector response). The finding is in accordance with the expected reaction of the host cell faced with a viral infection.

↓94

Proteins involved in chromosome organization, DNA packaging, chromatin assembly and disassembly, and nucleosome assembly were over-represented in cluster 4. Despite the causal correlation to the physiological reaction of the induced apoptosis in influenza A virus infected cells, Garcia-Robles et al. 2005 [223] revealed that in purified nucleosomes the RNP complexes bind to the histone tails and M1 to the globular domain of the histone octamer. The hypothesis that viral proteins alter the chromatin structure is supported by the findings of Takizawa et al. 2006 [224] who observed that vRNPs associate partially with the dense chromatin where viral transcription and replication takes place.

Although due to the cap snatching mechanism a wide range of cellular mRNA molecules might be destroyed, the cell integrity seemed to be intact until 8 h p.i. This conclusion is based on the protein composition of cluster 5 including mainly house-keeping genes.

Cluster 6 clearly demonstrated the demand on a time-dependent analysis to reveal the detailed information about the protein abundance changes upon influenza A virus infection. The heat map (see figure 39) indicated a reduction in proteins responsible for ion homoeostasis while the time-dependent functional cluster analysis showed an increase in abundance until 2 h p.i. followed by a drastic decrease which is evinced by the heat map.

↓95

For detailed information concerning the cluster composition please review table 7.

Figure 41 elucidates the dynamic process of protein abundance change upon influenza A virus infection for specified protein groups which are the influenza A virus proteins, virus-host interaction partners identified in a previous RNAi screen [83] and proteins exhibiting GO annotations assigned to antiviral response.

The curves of the diagrams were generated with values which were normalized to the mean of each data set. This enables the logarithmic presentation of the protein abundance change. The abundance change of the viral proteins was referred to 1 h p.i. due to the fact that these proteins are not present in the control sample (0 h p.i.). Whereas the NP and the NS1 protein showed a strong increase in the first 2 hours the PB1 and NA protein only increased slightly. Proteins like M1 and HA exhibited a more linear protein progression during the recorded time scale (Figure 33 A).

↓96

Karlas et al. 2010 [83] estimated 6 host factors crucial for influenza A virus replication as verified by an RNA interference approach in combination with an indirect NP-luciferase-reporter assay. The results of this study are limited to the absolute determination of negative effects on viral replication after 24 h. In the presented work 5 out of these virus host interaction partners were identified in the SILAC experiment enabling a time-resolved analysis of the dynamic processes in the early stage of infection. For example, the COPG (coat protein gamma) protein abundance strongly increased starting at 2 h p.i. by nearly 1 magnitude. COPG belongs to the COP proteins which mediate the biosynthetic transport from the endoplasmatic reticulum to the golgi apparatus, up to the trans-golgi network [225]. The transmembrane viral proteins require this transport mechanism for an efficient posttranslational modification of the HA and the NA protein.

In figure 41 C proteins which were annotated to the GO term antiviral defence and/or response were analysed in their abundance change during the time course 0 – 8 h p.i. The protein level of the majority of these antiviral proteins remained essentially constant over the whole recorded time-scale. While proteins like SERPIN B6 (serin protease inhibitor) [226], CAD (carbamyl phosphate synthetase II, aspartate transcarbamylase, and dihydroorotase; pyrimidine biosynthesis pathway) [227] and SF1 (splicing factor 1) [228] increased in abundance, the ZER 1 protein (recruitment of ubiquitin ligase complex) [229] dramatically decreased after 2 h p.i. for more than 4 magnitudes until 4 h p.i. and recovered in the following.

Figure 41: Logarithmic presentation of the protein abundance fold change of functional clusters upon the time course of influenza A virus infection (0 – 8 h p.i). in MDCK cells.

(A) Influenza A proteins, (B) cellular interaction partners crucial for efficient replication and (C) proteins involved in antiviral defence. The time range is not displayed linear due to program restrictions with respect to plotting.

↓97

Table 7: Gene ontology pathways enriched in fuzzy c-means clusters illustrated in figure 40.

Cluster 1

sphingolipid biosynthetic process

ATP synthesis coupled electron transport

electron transport chain

transmembrane transport

lipid biosynthetic process

membrane lipid metabolic process

behavioral fear response

sphingomyelin biosynthetic process

photorespiration

response to endoplasmic reticulum stress

learning

ER-nucleus signaling pathway

neutral lipid metabolic process

protein targeting to mitochondrion

cation transport

mitochondrial respiratory chain complex assembly

energy derivation by oxidation of organic compounds

ceramide metabolic process

triglyceride metabolic process

mitochondrial electron transport, NADH to ubiquinone

glycerol ether metabolic process

sphingosine biosynthetic process

sphinganine biosynthetic process

Cluster 2

regulation of BMP signaling pathway

translational elongation

mRNA metabolic process

rRNA processing

transcription

DNA metabolic process

regulation of long-term neuronal synaptic plasticity

regulation of cellular biosynthetic process

regulation of cell cycle

positive regulation of gene expression

nucleocytoplasmic transport

negative regulation of post-embryonic development

positive regulation of gene-specific transcription from RNA polymerase II promoter

positive regulation of transcription, DNA-dependent

regulation of metabolic process

macromolecule biosynthetic process

positive regulation of nucleobase, nucleoside, nucleotide and nucleic acid metabolic process

regulation of gene expression, epigenetic

regulation of macromolecule biosynthetic process

mRNA processing

ribosomal large subunit biogenesis

negative regulation of transcription from RNA polymerase II promoter

negative regulation of nitrogen compound metabolic process

negative regulation of transcription

negative regulation of biosynthetic process

regulation of RNA stability

positive regulation of cellular biosynthetic process

gene expression

cellular protein metabolic process

mitotic spindle elongation

nucleic acid metabolic process

cellular macromolecule metabolic process

cell cycle phase

spindle organization

primary metabolic process

microtubule-based process

ribosome assembly

mRNA stabilization

RNA biosynthetic process

ribosomal small subunit assembly

RNA splicing

mitotic cell cycle

regulation of nucleobase, nucleoside, nucleotide and nucleic acid metabolic process

regulation of gene expression

regulation of transcription, DNA-dependent

negative regulation of RNA metabolic process

cellular nitrogen compound metabolic process

positive regulation of macromolecule biosynthetic process

RNA processing

Cluster 3

positive regulation of ubiquitin-protein ligase activity involved in mitotic cell cycle

positive regulation of ligase activity

positive regulation of protein ubiquitination

negative regulation of ubiquitin-protein ligase activity involved in mitotic cell cycle

negative regulation of ligase activity

teasomal ubiquitin-dependent protein catabolic process

anaphase-promoting complex-dependent pro

regulation of protein catabolic process

regulation of actin filament length

cellular protein complex assembly

cellular macromolecule metabolic process

cellular macromolecule localisation

positive regulation of macromolecule metabolic process

regulation of apoptosis

monosaccharide metabolic process

negative regulation of protein complex disassembly

imaginal disc development

establishment or maintenance of cell polarity

macromolecular complex subunit organization

actin filament polymerization

glucose metabolic process

protein localisation

cellular macromolecular complex disassembly

negative regulation of macromolecule metabolic process

protein import into nucleus

hexose catabolic process

regulation of cell death

protein depolymerization

regulation of multi-organism process

regulation of mRNA processing

glycolsis

alcohol catabolic process

regulation of cellular metabolic process

actin filament-based process

vesicle coating

cellular component disassembly

regulation of cellular component biogenesis

response to acid

response to DNA damage stimulus

regulation of cellular component organization

regulation of actin filament depolymerization

cardiac myofibril assembly

germ cell development

protein transmembrane transport

microtubule cytoskeleton organization

response to stimulus

protein localisation in organelle

alpha-beta T cell differentiation

negative regulation of protein transport

cell differentiation

response to other organism

regulation of cellular catabolic process

primary metabolic process

actomyosin structure organization

negative regulation of microtubule polymerization or depolymerization

activation of pro-apoptotic gene products

leukocyte mediated cytotoxicity

regulation of RNA splicing

RNA localisation

sexual reproduction

anatomical structure formation involved in morphogenesis

positive regulation of programmed cell death

actin filament capping

organ development

nuclear mRNA splicing, via spliceosome

RNA splicing, via transesterification reactions

death

negative regulation of protein polymerization

regulation of response to biotic stimulus

protein complex biogenesis

cell motility

regulation of cell shape

leukocyte mediated immunity

cell division

positive regulation of cellular component organization

cardiac cell development

regulation of mitosis

production of molecular mediator involved in inflammatory response

post-embryonic morphogenesis

regulation of establishment of protein localisation

cell projection assembly

ameboidal cell migration

tissue development

intracellular protein transport

regulation of alternative nuclear mRNA splicing, via spliceosome

defence response to virus

regulation of defence response to virus by virus

regulation of biological process

cellular carbohydrate catabolic process

translational initiation

negative regulation of organelle organization

proteolysis

regulation of actin filament-based process

negative regulation of cellular process

negative regulation of molecular function

regulation of protein metabolic process

regulation of cytoskeleton organization

protein modification by small protein conjugation

cell cycle process

mitotic cell cycle

regulation of protein modification process

negative regulation of protein ubiquitination

regulation of ubiquitin-protein ligase activity

proteasomal protein catabolic process

post-translational protein modification

ubiquitin-dependent protein catabolic process

macromolecule catabolic process

negative regulation of cellular protein metabolic process

regulation of catalytic activity

actin filament organization

cellular protein catabolic process

positive regulation of molecular function

positive regulation of cellular process

modification-dependent macromolecule catabolic process

macromolecule modification

positive regulation of cellular protein metabolic process

Cluster 4

nucleosome assembly

DNA packaging

chromatin assembly or disassembly

chromosome organization

peptidyl-proline modification

peptidyl-proline hydroxylation to 4-hydroxy-L-proline

skin development

acidic amino acid transport

glycoprotein metabolic process

anterior/posterior pattern formation

regulation of embryonic development

organic anion transport

cholesterol metabolic process

Cluster 5

cellular ketone metabolic process

organic acid metabolic process

glutamine family amino acid catabolic process

branched chain family amino acid catabolic process

isocitrate metabolic process

succinate metabolic process

regulation of acetyl-CoA biosynthetic process from pyruvate

cellular respiration

lipid oxidation

protein homotetramerization

regulation of cofactor metabolic process

glutamine family amino acid biosynthetic process

metabolic process

carboxylic acid catabolic process

cellular amino acid metabolic process

small molecule catabolic process

oxidation reduction

generation of precursor metabolites and energy

carboxylic acid metabolic process

glutamate metabolic process

amine catabolic process

fatty acid metabolic process

coenzyme catabolic process

cellular lipid catabolic process

cofactor biosynthetic process

2-oxoglutarate metabolic process

tricarboxylic acid cycle

acetyl-CoA biosynthetic process

fatty acid beta-oxidation

succinyl-CoA metabolic process

cofactor metabolic process

lipid metabolic process

acetyl-CoA metabolic process

amine metabolic process

cellular catabolic process

Cluster 6

ion transport

metal ion transport

response to copper ion

regulation of smooth muscle cell proliferation

chloride transport

iron ion transport

regulation of necrotic cell death

response to retinoic acid

drug transmembrane transport

positive regulation of endothelial cell migration

indole and derivative metabolic process

intermediate filament organization

indolalkylamine metabolic process

negative regulation of cytokine production

positive regulation of calcium ion transport


© 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