VI. Results
1.1 Establishment, validation and technical monitoring of PAL-‐qLC-‐MS/MS technique
aniline-‐catalyzed oxime ligation) technique for cell surface protein labeling and enrichment156 for the use with primary human T cells and to prepare the samples suitable for mass spectrometry analysis. The PAL technique is based on the fact that most cell surface proteins are glycosylated.184 Periodate is used to oxidize the alcohol groups of the sugar residues to form aldehydes. Aniline is then catalyzing the reaction in which the aldehyde forms a stable oxime-‐linkage to the Biotin-‐derivate Aminooxy-‐Biotin (Fig. 3). After this reac-‐
tion, the glycosylated cell surface proteins are stably labeled with Aminooxy-‐Biotin, the cells are lysed and frozen. To then identify the Biotin-‐tagged cell surface proteins, the PAL-‐
technique was complemented with quantitative liquid chromatography-‐tandem mass spec-‐
trometry (qLC-‐MS/MS). Therefore, the tagged cell surface proteins were purified and en-‐
riched via streptavidin beads and the proteins were digested first with Trypsin, followed by a PNGase F digest. The two enzymatically digested peptide fractions were separately kept and measured via qLC-‐MS/MS (Fig. 5).
Figure 5: Scheme of the PAL-‐qLC-‐MS/MS technique. This overview presents the steps of the PAL-‐qLC-‐MS/MS technique. The cells are oxidized and biotinylated via PAL technique (1.) and then lysed (2.). The biotinylated cell surface proteins are enriched via Streptavidin beads (3.) and purified via centrifugation (4.). The proteins are first enzymatically digested with Trypsin (5.) and the digested peptides are separated as Trypsin fraction.
The Streptavidin beads coupled to the remaining peptides are incubated with PNGase F as a second enzyme for digestion (6.) and the resulting peptides separated as PNGase F fraction. Both fractions are analyzed separately via quantitative liquid chromatography-‐tandem mass spectrometry (qLC-‐MS/MS) (7.).
1.1.1 Influence of oxidation and biotinylation process
Oxidation agents like NaIO4 are known to be critical for the survival of cells, therefore the influence of different concentrations of NaIO4 was tested in terms of cell survival, detection of protein expression and biotinylation efficiency as an adequate labeling technique (Fig. 6).
Figure 6: Establishment and validation of PAL-‐qLC-‐MS/MS. Different NaIO4 concentrations in the one-‐pot reaction of the PAL-‐qLC-‐MS/MS as well as the Biotin labeling efficiency were tested on living human CD4+ T cells, which were activated for 17h with anti-‐CD3/anti-‐CD28, and analyzed via flow cytometry. A) Staining of activated CD4+ T cells with PI to check cell viability upon oxidation treatment (1 mM or 20 mM NaIO4).
B) Staining of activated CD4+ T cells with anti-‐CD69 antibody upon oxidation treatment (1 mM or 20 mM NaIO4). C) Biotinylation efficiency of cell surface proteins, which were either untreated or treated with NaIO4 was checked via Streptavidin-‐PE staining. The graph presents an overlay of two experiments showing the Strep-‐
tavidin staining of untreated and treated cells.
1mM and 20 mM of NaIO4 were tested within the oxidation/biotinylation mix. Naive CD4+ T cells from one human blood donor were activated with anti-‐CD3/anti-‐CD28 for 17h. 8x106 cells each were incubated without NaIO4, with the 1mM or the 20 mM NaIO4 containing oxi-‐
dation/biotinylation one-‐pot mix and stained for flow cytometry analysis. Fig. 6A shows that a concentration of 20 mM NaIO4 decreased the viability of cells about 39.8 % compared to 1 mM NaIO4 treatment, where 20.5 % of the cells are PI positive (untreated cells exhibit around 4.4 % of dead cells). The staining for CD69 was performed as a control, because CD69 is a well characterized T cell activation marker, and this showed that a concentration of 20 mM NaIO4 decreases the CD69 expression about 15 % (Fig 6B). These results pointed to a working concentration of 1 mM NaIO4, but high biotinylation efficiency still needed to be ensured. A staining of cells with Streptavidin-‐PE, which were treated with 1 mM NaIO4 con-‐
taining oxidation/biotinylation mix, confirmed a sufficient biotinylation efficiency of 99 % (Fig. 6C).
1.1.2 Validation of protein expression via flow cytometry in parallel to PAL-‐qLC-‐MS/MS sam-‐
ple preparation
To ensure that the results of the quantitative protein expression measurements via PAL-‐qLC-‐
MS/MS were comparable to the outcome of another validated and widely used technique, the samples for the generation of the surface glycoproteome (D1-‐D4) were stained for flow cytometry analysis, in parallel to sample processing for PAL-‐qLC-‐MS/MS. As targets for this validation staining the known T cell markers CD11a, CD62L and CD69 were chosen, because they already appeared in the mass spectrometry results during the establishment phase of the PAL-‐qLC-‐MS/MS technique. Fig. 7A shows that the expression pattern of the three pro-‐
teins during the T cell activation obtained via MS (protein abundance) equals to the pattern measured via flow cytometry (MFI) (shown for one representative donor). Only a slight dif-‐
ference between the results of the techniques can be seen for CD69. The expression change between 12h and 24h of stimulation detected by flow cytometry showed a constant increase in contrast to the MS result, which described more a static expression state of the proteins between 12 and 24h of activation.
Figure 7: Technical monitoring of T cell marker expression during sample preparation for PAL-‐qLC-‐MS/MS via flow cytometry. For the validation of the PAL-‐qLC-‐MS/MS technique, a flow cytometry staining of the surface antigens CD11a, CD62L and CD69 was performed in parallel to sample preparation for PAL-‐qLC-‐MS/MS with T cells of the same donor. A) Protein abundance (PAL-‐qLC-‐MS/MS) values and mean fluorescence intensities (MFI; flow cytometry) are shown for the respective cell surface proteins at the respective time points (one representative donor). The expression pattern obtained via both techniques over the time course is compara-‐
ble to each other. B) Histograms of the fluorescence intensity obtained by flow cytometry for the staining of the respective proteins at the respective time points are shown (one representative donor). (modified after 157)
1.1.3 Assessment of donor variability by comparing the protein expression patterns
To be able to combine the results of the PAL-‐qLC-‐MS/MS technique of the four single differ-‐
ent blood donors, expression patterns needed to be checked for similarity during the activa-‐
tion process. Therefore, the measured protein abundances of the different blood donors at the different stimulation time points were subjected to a principal component analysis and revealed highly concordant protein abundances (Fig. 8A).
Figure 8: Comparability of the donor samples for PAL-‐qLC-‐MS/MS. A) Protein abundances of the samples of the four donors (D1-‐D4) at all time points (0, 3, 6, 12, 24, 48h), obtained by PAL-‐qLC-‐MS/MS, were subjected to a Principal Component Analysis (PCA). This analysis grouped the donor samples at the time points (0h, 3-‐6-‐12h, 24h, 48h). Principal Component 1 (PC1), explaining 67.31 % of the data variance, divides all 48h samples from the other time points and PC2, explaining 11.09 % of the variance, separates the 0h samples from the samples of the remaining time points.157 B) The donor comparability was also assessed via flow cytometry staining dur-‐
ing the stimulation time course for the surface antigens CD69, CD11a, CD25 and CD62L for three donors (D1-‐
D3) at the indicated time points (0, 3, 6, 24, 48h), demonstrating that the obtained expression patterns for the selected surface antigens are comparable between the different donors.
For selected T cell surface markers (CD69, CD11a, CD25, CD62L) the comparability of three of the donors was in addition assessed via flow cytometry (Fig. 8B), which also proved the
possibility to combine the datasets obtained for the different blood donors. Fig. 8B shows a continuous increase of the CD69 signal for the depicted donors, being able to distinguish a positive and a negative population at 24h. The CD11a expression was low until the increase between 24 and 48h, as well as the CD25 expression until 24h, also shown for the three do-‐
nors. CD62L was highly expressed on the naive T cells of all examined donors, rapidly down-‐
regulated after the start of the activation process, but increased again around the 24h time point. The results of the flow cytometry staining also showed high concordance between the donors included in the surface glycoproteome.
1.2 PAL-‐qLC-‐MS/MS-‐based cell surface glycoproteome of human naive and activated CD4+