Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
Quantitative Secretome Analysis of Activated Jurkat Cells using Click Chemistry-Based Enrichment of Secreted Glycoproteins Kathrin Elisabeth Witzke, Kristin Rosowski, Christian Müller, Maike Ahrens, Martin Eisenacher, Dominik A. Megger, Jürgen Knobloch, Andrea R. Koch, Thilo Bracht, and Barbara Sitek J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.6b00575 • Publication Date (Web): 03 Oct 2016 Downloaded from http://pubs.acs.org on October 6, 2016
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Proteome Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
Quantitative Secretome Analysis of Activated Jurkat Cells using Click Chemistry-Based Enrichment of Secreted Glycoproteins
Kathrin E. Witzke1, Kristin Rosowski1, Christian Müller1, Maike Ahrens1, Martin Eisenacher1, Dominik A. Megger1, Jürgen Knobloch2, Andrea Koch2, Thilo Bracht1*# and Barbara Sitek1#
1
Medizinisches Proteom-Center, Ruhr-Universität Bochum, Bochum, Germany
2
Medical Clinic III for Pneumology, Allergology, Sleep and Respiratory Medicine,
Bergmannsheil University Hospital, Ruhr-Universität Bochum, Bochum, Germany
#
These authors contributed equally to the work
*Corresponding author: Dr. Thilo Bracht, Medizinisches Proteom-Center, Ruhr-Universität Bochum, 44801 Bochum, Germany, Tel. +49-(0)-234/32-29985, E-mail:
[email protected]
1 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Abstract (198 words, max 200) Quantitative secretome analyses are a high-performance tool for the discovery of physiological and pathophysiological changes in cellular processes. However, serum supplements in cell culture media limit secretome analyses, but serum depletion often leads to cell starvation and consequently biased results. To overcome these limiting factors, we investigated a model of T cell activation (Jurkat cells) and performed an approach for the selective enrichment of secreted proteins from conditioned medium utilizing metabolic marking of newly synthesized glycoproteins. Marked glycoproteins were labeled via bioorthogonal click chemistry and isolated by affinity purification. We assessed two labeling compounds conjugated with either biotin or desthiobiotin and the respective secretome fractions. 356 proteins were quantified using the biotin probe and 463 using desthiobiotin. 59 proteins were found differentially abundant (adjusted p-value ≤ 0.05, absolute fold change ≥ 1.5) between inactive and activated T cells using the biotin method and 86 using the desthiobiotin approach with 31 mutual proteins cross-verified by independent experiments. Moreover, we analyzed the cellular proteome of the same model to demonstrate the benefit of secretome analyses and provide comprehensive data sets of both. 336 proteins (61.3 %) were quantified exclusively in the secretome. Data are available via ProteomeXchange with identifier PXD004280.
Keywords: Click chemistry, desthiobiotin, biotin, T cell activation, quantitative secretomics, Jurkat cells, label-free proteomics, affinity purification, SPECS
2 ACS Paragon Plus Environment
Page 2 of 37
Page 3 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
1.
Introduction Secreted proteins are of major importance in intercellular communication. They
mediate diverse physiological functions like differentiation, proliferation and immunity1 and often have a significant impact on disease progression, which has been investigated in various entities2
3 4 5 6
. Especially in immune cells, the secretome contains the effector proteins of
extracellular signaling and the immune response. Therefore, the investigation of the secretome can provide novel insights into physiological and pathophysiological mechanisms and processes as well as a better understanding of immunological processes in this context.7
8 9
Secretome analyses using quantitative proteomics are well established and described for diverse cell culture models6 10. However, as the secreted proteins are mostly low abundant in comparison to high abundant serum proteins contained in cell culture media, the analysis is challenging. One approach to address this problem is the incubation of cells in serum-free medium. Most secretome analyses are performed either without serum,7 with labeling strategies or a combination of both approaches.8 The disadvantage of serum starvation, though, is that cell behavior may be massively biased and proteins may be set free into the medium by occurring cell death and lysis, thereby distorting the protein composition of the secretome.1
7 11
Therefore, alternative techniques for secretome analyses that allow serum to
remain in the medium are of great advantage. The SPECS approach (secretome protein enrichment with click sugars) utilizes the high percentage of glycoproteins in the secretome and is based on the click chemistry-based labeling of newly synthesized glycoproteins and their subsequent enrichment from the conditioned medium.12 At first, the metabolic marking of glycoproteins is achieved by incorporation of an azide modified sugar, e.g. ManNAz (tetraacetylated N-azidoacetyl-Dmannosamine), into the biosynthetic pathway of the cell,13 where it is converted into azide modified N-acyl sialic acid (SiaNAz). As sialic acids are incorporated into the terminal positions of glycan moieties of glycoprotein, they are well exposed to possible reaction 3 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 4 of 37
partners.14 The metabolic incorporation into only newly synthesized glycoproteins ensures the differentiation from glycoproteins which are contained in serum. Subsequently, the actual labeling reaction of the azide residues is achieved by using alkyne derivates, for example dibenzocyclooctynes, which are linked to specific conjugates that can be used for purification. A similar approach has also been described by Eichelbaum et al. who used the azide modified amino acid azidohomoalanine that gets incorporated into proteins instead of methionine.1 Biotin is one of the most frequently used labels for affinity-based protein isolation. The extremely strong interaction between biotin and avidin is used for selective capturing of biotinylated molecules. It allows stringent washing conditions and consequently the efficient elimination of unspecifically binding proteins and other contaminants. The recently emerged labeling reagent desthiobiotin is a compound derived from biotin by the removal of a sulfur atom. It can substitute for biotin in the reaction with avidin, making the binding more easily reversible, as the interaction is less strong.15 In contrast to biotin, desthiobiotin allows for elution by displacement with biotin and is thereby well suited for isolation and analysis of interaction partners of the labeled proteins.15 As has been shown previously in the studies performed by Kuhn et al.12
16 17
and
Eichelbaum et al.,1 18 the click chemistry based enrichment of secreted proteins is well suited for the quantitative secretome analysis. With the aim of our analysis being the immune system, we employed the SPECS approach for an unbiased quantitative secretome analysis of a T cell model (Jurkat cells) and monitored the modulation of the secretome that was induced by T cell activation. Secreted glycoproteins were enriched using bioorthogonal click chemistry and two different labels and enrichment strategies were utilized and compared. The complementary nature of the enrichment strategies enabled the analysis of different fractions of the secretome and furthermore allowed methodological and biological cross-verification of the results. Finally, we compared the observed changes of the secretome with those found in a cellular proteome analysis of the same model. We assessed the value of secretome studies 4 ACS Paragon Plus Environment
Page 5 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
regarding their benefit in identification and quantification of proteins when compared with studies focusing on the cellular proteome.
5 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
2.
Materials and Methods
2.1.
Experimental Design Secretome analyses were performed twice (independent experiments), once with each
of the introduced click chemistry reagents containing either biotin or desthiobiotin. Analyses were performed in six experimental replicates of three groups (control: DMSO/DMSO, inactive: ManNAz/DMSO and activated: ManNAz/PMA+Ionomycin). The cellular proteome analysis was performed in seven independent experimental replicates of two groups (inactive: DMSO and activated: PMA+Ionomycin, Figure 1).
2.2.
Cell Culture Experiments The Jurkat cell line (DSMZ, Braunschweig, Germany) was cultivated in RPMI1640
supplemented with 10 % FBS (PAN-Biotech, Aidenbach, Germany). For secretome analyses, glycoproteins were metabolically marked using ManNAz (Life Technologies) at a concentration of 50 µM. 0.25 x 106 cells/ml were added to prepared media and incubated for 72 h. After 48 h cells were activated using 50 ng/ml PMA (Phorbol 12-myristate 13-acetate) and 500 ng/ml Ionomycin calcium salt (Sigma-Aldrich). The 24 hour-activation was initiated during metabolic marking to warrant simultaneous ending. Cells were removed by centrifugation (5 min, 4000 x g) and supernatants were filtered through 0.45 µm filters (Sarstedt, Nümbrecht, Germany) into centrifugal filter units (Amicon Ultra, 10 kDa cut-off, Merck). Excess ManNAz was removed by ultrafiltration (4 °C, 4000 x g) and concentrated media was transferred into smaller Amicons (3 kDa cut-off). Buffer exchange with PBS followed three times (20 min, 4 °C, 16,100 x g). For the cellular proteome analysis, 1 x 106 cells/ml were activated identically. Cells were harvested after 8h, 16h and 24h by centrifugation (4 min, 300 x g), washed twice with ice-cold PBS and pellets were stored at – 80 °C until further processing.
6 ACS Paragon Plus Environment
Page 6 of 37
Page 7 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
2.3.
Click Chemistry Enrichment of Glycoproteins Metabolically marked glycoproteins were biotinylated over night at 4 °C with 20 µM
of either DBCO (Dibenzocyclooctyne)-Sulfo-Link-Biotin Conjugate or DBCO-PEG4Desthiobiotin (Jena Bioscience, Jena, Germany), respectively. Buffer exchange was performed as before to remove free biotin. Enrichment procedures varied slightly for the two reagents as described below.
2.3.1. Enrichment using DBCO-Sulfo-Link-Biotin Conjugate Samples were diluted to 1 ml using 2 % SDS and 2 µM TCEP in PBS. After incubation on High Capacity NeutrAvidin Agarose Resin (Thermo Fisher Scientific) over night at 4 °C, samples were washed on columns (Bio-Rad) with 10 ml of 1 % SDS in PBS, 2 M NaCl in PBS and 0.1 M Glycin (pH 3), respectively. For elution, the resin was transferred into Spin Columns (Life Technologies). Samples were incubated with elution buffer (2 % SDS, 30 mM Biotin, 6 M Urea, 2 M Thiourea in PBS) for 15 min at RT before being heated at 95 °C for 15 min. After cooling, eluates were collected via centrifugation (5 min; 16,100 x g).
2.3.2. Enrichment using DBCO-PEG4-Desthiobiotin Conjugate Samples were diluted to 1 ml with PBS before incubation on resin as described for biotin. Washing was performed solely with PBS (30 ml). Elution was achieved by displacement with free biotin (30 mM Biotin in PBS). Samples were incubated with elution buffer for one hour at RT before eluates were collected via centrifugation (5 min; 16,100 x g).
2.4.
Cell Lysis (Cellular Proteome Analysis) Cell pellets were lysed with 0.1% Na-DOC in TBS supplemented with protease
inhibitor cocktail (complete mini, Roche, Penzberg, Germany) and Benzonase (25 U per 7 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
sample, Merck). Samples were sonicated for 10 min on ice before lysis buffer (30 mM Tris, 7M urea, 2 M thiourea, 0.1% SDS, pH 8.5) was added to solubilize proteins. Samples were centrifuged (10 min; 16,100 x g) to remove insoluble material and supernatant concentrations were determined using the Bradford assay (Bio-Rad).
2.5.
Sample Preparation 15 µg proteins per sample were applied to 18 % Tris-Glycine-Gels (Anamed
Elektrophorese, Rodau, Germany) at 100 V for 10-15 min to collect the proteins in a single band for each sample. Protein bands were stained with Coomassie and dissected. Digestion with trypsin (SERVA, Heidelberg, Germany) was performed over night at 37 °C. Peptides were extracted using 0.05 % TFA, 50 % ACN (1:1), dried in a vacuum centrifuge and dissolved in 0.1 % TFA. Peptide concentration was determined via amino acid analysis as described before 19.
2.6.
LC-MS/MS Analysis LC-MS/MS analyses were performed on a Q-Exactive (Thermo Fisher Scientific)
coupled to an Ultimate 3000 RSLCnano HPLC System (Dionex, Idstein, Germany) as described earlier.20 Briefly, 350 ng of injected peptides were pre-concentrated on a trap column (Acclaim PepMap 100, 100 µm x 2 cm, C18, 5 µm, 100 Å) within 7 min at a flow rate of 30 µl/min with 0.1 % TFA. Peptides were transferred onto an analytical column (Acclaim PepMap RSLC, 75 µm x 50 cm, nano Viper, C18, 2 µm, 100 Å) and separated with a gradient from 5-40 % solvent B over 98 min at 400 nl/min at 60 °C (solvent A: 0.1 % FA, solvent B: 0.1 % FA, 84 % acetonitrile). Acquisition of MS/MS spectra followed in a data-dependent mode. Full scan mass spectra in the Orbitrap analyzer were acquired in profile mode at a resolution of 70,000 at 400 m/z and within a mass range of 350-1400 m/z. For MS/MS
8 ACS Paragon Plus Environment
Page 8 of 37
Page 9 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
measurements, the ten most abundant peptide ions (charge states + 2, + 3, + 4) were chosen and fragmented via higher-energy collisional dissociation (HCD).
2.7.
Protein Identification and Quantification Protein identification was performed with Proteome Discoverer 1.4 (Thermo Fisher
Scientific). Spectra were searched against the UniProtKB/Swiss-Prot database (Release 2015_11; 549,832 entries) via Mascot (ver.2.5, Matrix Science). Taxonomy was set to homo sapiens and mass tolerance to 5 ppm for precursor ions and 0.4 Da for fragment ions. Dynamic modifications were considered for cystein (propionamide) and methionin (oxidation). The false discovery rate was calculated using Proteome Discoverer’s percolater function for the secretome studies and the target decoy PSM validator for the cellular proteome study, respectively, and identifications with an FDR greater than 1 % were rejected. Progenesis QI (ver. 2.0.5387.52102, Nonlinear Dynamics) was used for label-free quantification. All .raw files were aligned to a reference run and a master map of common features was applied to all experimental runs to adjust retention time differences. Ions charged 2+, 3+ and 4+ with minimum three isotope peaks were considered. Normalization of raw ion abundances automatically corrected technical or experimental variations between runs (for details see Ref19). Quantified features were identified via Proteome Discoverer identifications. All non-conflicting peptides were considered for protein quantification. Normalized protein abundances were obtained from the software and analyzed using an in-house written R-script applying ANOVA followed by Tukey’s honest significant difference (HSD) method. Fold changes between groups were determined based on normalized abundances while ANOVA was calculated using arcsinh-transformed data for consistency with the Progenesis QI software. The FDR was controlled by adjusting ANOVA p values using the method of Benjamini and Hochberg. For proteins with adjusted ANOVA p values below the significance level of α=0.05 the TukeyHSD method was applied to further characterize the identified 9 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
difference in abundance levels between groups. Proteins were considered differentially abundant between the inactive and activated groups with an absolute fold change ≥ 1.5 and a p TukeyHSD ≤ 0.05. Proteomics data have been deposited as complete submission to the ProteomeXchange Consortium via the PRIDE partner repository. Data set identifier is PXD004280 and DOI 10.6019/PXD004280. Conversion of msf files (result files of Proteome Discoverer) into the mzIdentML standard format happened via the ProCon – PROteomics CONversion tool (ver. 0.9.641).21
2.8.
Protein Characterization Proteins were considered glycoproteins based on the so-termed UniProtKB/SwissProt
keyword (Release 2015_11; 4618 entries) and possibly secreted/shed if tagged with the keyword ‘secreted’ or GO-annotated as ‘extracellular’ (GO: 0005576, 0031012, 0005615 and 0070062), ‘cell surface’ (GO: 0009986) or ‘external side of plasma membrane’ (GO: 0009897). Additionally, softwares SignalP 4.122 and SecretomeP 2.023 were used for detecting proteins secreted classically or via non-classical pathways, respectively. There, proteins containing more than two transmembrane helices for classical secretion or at least one for non-classical secretion were not considered secreted. These predictions were made by the TMHMM server 2.0 (CBS).24
10 ACS Paragon Plus Environment
Page 10 of 37
Page 11 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
3.
Results
3.1.
Quantitative Secretome Analyses Quantitative Jurkat secretome analyses were performed twice, with six replicates for
each of the three groups (control, inactive, activated), respectively and once with each of the two introduced click chemistry reagents, DBCO-Sulfo-Link-Biotin and DBCO-PEG4Desthiobiotin (Figure 1). The two analyses will be henceforth referred to as the ‘Biotin’ or ‘Desthiobiotin’ method. Both methodological and quantitative results were compared, and mutual differentially abundant proteins in both independent experiments were chosen for further evaluation.
3.1.1. Biotin In the Biotin secretome analysis, 644 proteins (507 protein groups) were identified by the Proteome Discoverer software in total and 356 proteins were quantified by Progenesis QI. 59 proteins were differentially abundant between inactive and activated cells. Of those, 51 were higher abundant in activated cells and 8 were higher abundant in inactive cells (Supplementary Table 1).
3.1.2. Desthiobiotin In the Desthiobiotin secretome analysis, 758 proteins (601 protein groups) were identified by the Proteome Discoverer software in total and 463 proteins were quantified by Progenesis QI. 110 proteins were differentially abundant while comparing activated and inactive cells. 83 of these proteins showed higher abundance in activated and 27 in inactive cells (Supplementary Table 2).
3.2.
Comparison of Enrichment Efficiencies
11 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
The results of the Biotin and Desthiobiotin method were analyzed and compared regarding their enrichment efficiency (Figure 2A). After glycoprotein enrichment of inactive and activated samples, the glycoprotein fractions in the Biotin method mounted up to 65.1 % and 68.5 %, respectively, with only 33 % in the control samples. Accordingly, the share of glycoproteins in the Desthiobiotin method reached 53.1 % and 60.6 % in inactive and activated samples, respectively, with only 38.6 % in the controls. While the total number of glycoproteins remained relatively constant, the non-glycoprotein amount increased in the enriched (inactive and activated) samples of the Desthiobiotin method. On the level of identified peptides and peptide spectrum matches, only minor differences between the approaches could be observed (Figure 2B) which were almost in line with the protein results.
3.3.
Comparison of Quantitative Results The comparison of total quantified and differentially abundant proteins between both
methods showed an increase of 30 % in total protein numbers and almost a doubling in differentially abundant protein numbers for the Desthiobiotin method, with 463 quantified proteins compared to 356 and 110 differentially abundant proteins compared to 59 (Figure 4). Possibly, secreted proteins make up most of those numbers with percentages of at least 90 % in both methods, leaving only a small amount of non-secreted glycoproteins and residual proteins, thus being neither secreted nor glycoproteins. Total quantified glycoprotein numbers were fairly constant between the methods, with 286 proteins compared to 271. Furthermore, almost all quantified and differentially abundant glycoproteins are possibly secreted, with over 93 % in all cases (Figure 4). For method verification, the focus was laid on differentially abundant proteins detected by both methods. Figure 3 (C and D) shows Venn diagrams for overlaps of both quantified and differentially abundant proteins. With an overlap of all quantified proteins mounting up to 271 proteins, there are 36 proteins that are also differentially abundant in both 12 ACS Paragon Plus Environment
Page 12 of 37
Page 13 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
methods (Table 1). Working with the mutual proteins, a correlation analysis of the corresponding fold changes was performed (Figure 5). Furthermore, a linear fit has been applied to the mutual differentially abundant proteins showing a good linear correlation with R² = 0.77 and a slope of 0.84.
3.4.
Quantitative Cellular Proteome Analysis In the cellular proteome analysis, 4575 proteins (4034 protein groups) were identified
in total and 3403 proteins were quantified by the Progenesis QI software (Supplementary Table 3). The activation of Jurkat cells was analyzed in a time-dependent manner and three different time points were investigated. The highest number of differentially abundant proteins was observed after 24 h with 521 significantly differentially abundant proteins of which 323 were higher abundant in activated cells and 198 showed a higher abundance in the inactive group, respectively (Supplementary Figure 2).
3.5.
Comparison of Secretome and Cellular Proteome Analyses The results of the secretome and cellular proteome analyses were compared. For the
secretome, all complementary differentially abundant proteins of the two enrichment methods were taken into consideration, for the cellular proteome those differentially abundant after 24 hours of activation. 336 of 548 (61.3 %) proteins were quantified exclusively in the secretome analyses with only 38.7 % also quantified in the cellular proteome analysis. Of the 134 differentially abundant proteins from both secretome analyses, only eleven were also found differentially abundant in the cellular proteome analysis after 24 hours of Jurkat T cell activation (521 differentially abundant proteins, Supplementary Table 3 and 4, Supplementary Figure 2). Of these eleven proteins, almost all are found in both the cell (cytosol, nucleus or membrane) and extracellularly. Merely one protein, the transport protein Equilibrative nucleoside transporter 1 (SLC29A1), is only attributed to the membrane. 13 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
4.
Page 14 of 37
Discussion The comparison of two click chemistry compounds, which contained either biotin or
desthiobiotin labels for protein isolation, highlighted their complementary function and specific characteristics that can be taken into consideration for using either or both compounds. The extremely strong interaction of the biotin compound with NeutrAvidin allowed for the usage of harsh washing conditions. Consequently, this resulted in reduced copurification of proteins other than metabolically marked glycoproteins and the removal of unspecifically binding proteins (Figure 4). On the other hand, the rigorous conditions that had to be applied for elution also led to a leakage of monomeric NeutrAvidin from the resin (data not shown). Boiling together with the harsh elution buffer destroys the structure of tetrameric NeutrAvidin, leading to its co-elution in a monomeric form.25
26
For click chemistry reagents
containing a desthiobiotin label, elution is possible simply by biotin displacement as the interaction to NeutrAvidin has a lower dissociation constant (10-11 M vs. 10-15 M). Thus, NeutrAvidin leakage could be prevented. Due to data dependent LC-MS/MS analysis, NeutrAvidin leakage might result in lower detection of the enriched glycoproteins as peptides of interest might be masked. However, the comparison of total identified and quantified glycoprotein numbers, which remained almost constant between methods, does not support this concern. Only the difference in total numbers of differentially abundant glycoproteins hints at a possible limitation of the biotin method in comparison. In an analysis of the glycoprotein enrichment efficiency for both methods (Figure 2), the Desthiobiotin method showed a slight increase in total protein numbers in the enriched samples with additional non-glycoproteins. This increase in non-glycoproteins can be attributed to the gentler washing conditions. Thus, non-glycoproteins associated with the enriched glycoproteins were co-purified. Interestingly, there was almost no difference in total numbers of non-glycoproteins of the controls in both methods (190 vs. 205). This again implies that the huge increase in co-purified non-glycoproteins in the enriched samples of the 14 ACS Paragon Plus Environment
Page 15 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
Desthiobiotin method can indeed be attributed to the capturing of interaction partners of enriched glycoproteins. When looking at the quantified proteins generated with the two different reagents, the amount increased by about 30 % when using the desthiobiotin labeling reagent. The number of differentially abundant proteins even about doubled (Figure 4). This major increase again stems from interacting non-glycoproteins. In fact, most nonglycoproteins enriched due to the gentle washing procedure could be annotated as belonging to the secretome. The percentage of secreted proteins obtained by both enrichment strategies was satisfying with over 90 % providing an excellent basis for secretome analyses. In summary, the biotin approach is useful for focusing especially on glycoproteins while the desthiobiotin approach allows a greater coverage of the secretome. In a comparison with other SPECS-using quantitative secretome studies by Kuhn et al., our obtained numbers of secreted proteins were equally high
16 17
or even higher12, even without fractionation via SDS-PAGE.
The modified method described by Eichelbaum et al. using labeling and enrichment via azidohomoalanin in combination with pulsed SILAC1 gained a slightly higher number of secreted proteins. This might be attributed to the labeling of not only glycoproteins but all proteins containing methionine residues. The predominance of increased protein abundances upon T cell activation is in line with expectations of cell behavior due to stimulation and can also be found in literature.8
27
Having performed two independent analyses provides the opportunity of cross-verification of the results and demonstrates the functionality and usefulness of the method. Overlaps of about 50 % of all quantified and 27 % of all differentially abundant proteins can, again, be attributed to method variations and stresses the complementary factor of the two compounds. A correlation analysis of the fold changes of mutual proteins revealed no conflicts for differentially abundant proteins of either method (Figure 5). The applied linear fit shows an evident trend towards similar regulations with slightly higher fold changes in the Desthiobiotin method. Cross-verification can also be found on the level of protein function. 15 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 16 of 37
Most of the mutual differentially abundant proteins can be set into the context of immunity or T cell activation. The neurosecretory protein VGF, for example, is known to be secreted by neurons with functions in cell growth and survival.28 VGF has also been shown to be expressed mainly in activated T cells and is positively correlated with T cell survival.29 ENPP1 is an ecto-enzyme that can be shed/secreted into serum30 and is associated with the immune system, where it has been suggested to be upregulated upon T cell activation.31 Another example is DMBT1, a secreted glycoprotein of the scavenger receptor cysteine-rich (SRCR) family known to be involved in angiogenesis, innate immunity and inflammation, upon which it is up-regulated.32 DMBT1 is mostly described for endothelial cells, but has also been detected throughout the immune system and the Jurkat cell line.33 Another function of DMBT1 lies in its ability to bind VEGF (vascular endothelial growth factor). Two other VEGF-binding proteins, VEGF receptors (VEGFR) FLT-1 (vascular endothelial growth factor receptor 1) and NRP-1 (Neuropilin-1), are also found upregulated in both analyses, while VEGF itself could not be detected. VEGF, an angiogenesis factor also involved in inflammation processes like immune cell trafficking,34 is produced35 and secreted36 in response to T-cell activation, possibly explaining upregulation of its receptors. Experimental data showed VEGF/VEGFR interactions to induce Akt (protein kinase B) and ERK (extracellular-signal regulated kinase) activation, which are also activated by T cell receptorsignaling pathways,34
35
hinting at a possibly costimulatory function of VEGF and their
receptors in T-cell activation. FLT-1 and NRP-1 have both already been detected in Jurkat cells34 and T-lymphocytes and are both also present in a secreted form.35
37 38
On the other
hand, proteins previously unknown to be differentially secreted upon T cell activation were discovered. One of these proteins, which could be cross-verified in our independent studies, is Calsyntenin-3 (CLSTN3). It is a member of the cadherin superfamily of cell adhesion molecules and is normally found in the nervous system localized to postsynaptic membranes,39 40 where its ectodomain can be shed from the cell surface.41 16 ACS Paragon Plus Environment
Page 17 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
The fact that well known T cell cytokines like Interleukin-2 and Interferon ߛ were not identified via LC-MS/MS can be attributed to their extremely low concentration levels even upon activation when they are higher abundant.42 With described Interleukin-2 concentration levels of about 0.7 ng/ml to 15 ng/ml in Jurkat and T cell supernatants, respectively, which makes up only about 6 ppm of the total protein amount at most,43
44
identification via LC-
MS/MS can only be achieved with severe fractionation to reduce the complexity of the samples43. Identification of Interferon ߛ with Jurkat supernatant concentrations of about only 10 pg/ml42 is, thus, even more difficult. Furthermore, another hindrance for detection of IL-2 is the fact that it is only O-glycosylated and O-glycosylation are not labeled by ManNAz. In a comparison of the secretome results with the performed quantitative cellular proteome analysis, the functionality of the established method could, again, be demonstrated. 336 proteins (61.3 %) could be quantified exclusively in the secretome analyses, with only 38.7 % also quantified in the cellular proteome analysis. Of the differentially abundant proteins of both secretome analyses only eleven proteins were also found differentially abundant in the cellular proteome analysis. All but one of those could be attributed to both the cellular proteome and the secretome. Solely the Equilibrative nucleoside transporter 1 (SLC29A1) was only associated with the membrane. This result emphasizes the complementary roles of cellular proteome and secretome and demonstrates the usefulness of quantitative secretome analyses.
5.
Conclusion The secretome enrichment method previously described by Kuhn et al.12 for unbiased
quantitative analyses of the secretome was successfully applied in a T cell model system and showed a high potential for the use in the investigation of immunological processes. The assessment of two different click chemistry labeling compounds provided methodological insights for future secretome studies. Observed secretome changes were cross-verified by 17 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
independent experiments, providing highly confident data under unbiased conditions that add additional information on the biology of the T cell activation. Moreover, we analyzed the cellular proteome of the same model and provide comprehensive data sets of the secretome and cellular proteome of activated T cells. The almost exclusive identification and quantification of proteins in the secretome fractions compared to the cellular proteome demonstrated the benefit of the applied approaches.
18 ACS Paragon Plus Environment
Page 18 of 37
Page 19 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
6.
References
1.
Eichelbaum, K.; Winter, M.; Berriel Diaz, M.; Herzig, S.; Krijgsveld, J., Selective enrichment
of newly synthesized proteins for quantitative secretome analysis. Nature biotechnology 2012, 30 (10), 984-90. 2.
Kawanishi, H.; Matsui, Y.; Ito, M.; Watanabe, J.; Takahashi, T.; Nishizawa, K.; Nishiyama,
H.; Kamoto, T.; Mikami, Y.; Tanaka, Y.; Jung, G.; Akiyama, H.; Nobumasa, H.; Guilford, P.; Reeve, A.; Okuno, Y.; Tsujimoto, G.; Nakamura, E.; Ogawa, O., Secreted CXCL1 is a potential mediator and marker of the tumor invasion of bladder cancer. Clinical cancer research : an official journal of the American Association for Cancer Research 2008, 14 (9), 2579-87. 3.
Kulasingam, V.; Diamandis, E. P., Proteomics analysis of conditioned media from three breast
cancer cell lines: a mine for biomarkers and therapeutic targets. Molecular & cellular proteomics : MCP 2007, 6 (11), 1997-2011. 4.
Ma, Y.; Visser, L.; Roelofsen, H.; de Vries, M.; Diepstra, A.; van Imhoff, G.; van der Wal, T.;
Luinge, M.; Alvarez-Llamas, G.; Vos, H.; Poppema, S.; Vonk, R.; van den Berg, A., Proteomics analysis of Hodgkin lymphoma: identification of new players involved in the cross-talk between HRS cells and infiltrating lymphocytes. Blood 2008, 111 (4), 2339-46. 5.
Planque, C.; Kulasingam, V.; Smith, C. R.; Reckamp, K.; Goodglick, L.; Diamandis, E. P.,
Identification of five candidate lung cancer biomarkers by proteomics analysis of conditioned media of four lung cancer cell lines. Molecular & cellular proteomics : MCP 2009, 8 (12), 2746-58. 6.
Wu, C. C.; Hsu, C. W.; Chen, C. D.; Yu, C. J.; Chang, K. P.; Tai, D. I.; Liu, H. P.; Su, W. H.;
Chang, Y. S.; Yu, J. S., Candidate serological biomarkers for cancer identified from the secretomes of 23 cancer cell lines and the human protein atlas. Molecular & cellular proteomics : MCP 2010, 9 (6), 1100-17. 7.
Brown, K. J.; Formolo, C. A.; Seol, H.; Marathi, R. L.; Duguez, S.; An, E.; Pillai, D.;
Nazarian, J.; Rood, B. R.; Hathout, Y., Advances in the proteomic investigation of the cell secretome. Expert review of proteomics 2012, 9 (3), 337-45.
19 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
8.
Bonzon-Kulichenko, E.; Martinez-Martinez, S.; Trevisan-Herraz, M.; Navarro, P.; Redondo, J.
M.; Vazquez, J., Quantitative in-depth analysis of the dynamic secretome of activated Jurkat T-cells. Journal of proteomics 2011, 75 (2), 561-71. 9.
Meissner, F.; Mann, M., Quantitative shotgun proteomics: considerations for a high-quality
workflow in immunology. Nature immunology 2014, 15 (2), 112-7. 10.
Makridakis, M.; Vlahou, A., Secretome proteomics for discovery of cancer biomarkers.
Journal of proteomics 2010, 73 (12), 2291-305. 11.
Roper, S. M.; Zemskova, M.; Neely, B. A.; Martin, A.; Gao, P.; Jones, E. E.; Kraft, A. S.;
Drake, R. R., Targeted glycoprotein enrichment and identification in stromal cell secretomes using azido sugar metabolic labeling. Proteomics. Clinical applications 2013, 7 (5-6), 367-71. 12.
Kuhn, P. H.; Koroniak, K.; Hogl, S.; Colombo, A.; Zeitschel, U.; Willem, M.; Volbracht, C.;
Schepers, U.; Imhof, A.; Hoffmeister, A.; Haass, C.; Rossner, S.; Brase, S.; Lichtenthaler, S. F., Secretome protein enrichment identifies physiological BACE1 protease substrates in neurons. The EMBO journal 2012, 31 (14), 3157-68. 13.
Saxon, E.; Bertozzi, C. R., Cell surface engineering by a modified Staudinger reaction.
Science 2000, 287 (5460), 2007-10. 14.
Dube, D. H.; Bertozzi, C. R., Metabolic oligosaccharide engineering as a tool for
glycobiology. Curr Opin Chem Biol 2003, 7 (5), 616-25. 15.
Hirsch, J. D.; Eslamizar, L.; Filanoski, B. J.; Malekzadeh, N.; Haugland, R. P.; Beechem, J.
M.; Haugland, R. P., Easily reversible desthiobiotin binding to streptavidin, avidin, and other biotinbinding proteins: uses for protein labeling, detection, and isolation. Analytical biochemistry 2002, 308 (2), 343-57. 16.
Kuhn, P. H.; Voss, M.; Haug-Kroper, M.; Schroder, B.; Schepers, U.; Brase, S.; Haass, C.;
Lichtenthaler, S. F.; Fluhrer, R., Secretome analysis identifies novel signal Peptide peptidase-like 3 (Sppl3) substrates and reveals a role of Sppl3 in multiple Golgi glycosylation pathways. Molecular & cellular proteomics : MCP 2015, 14 (6), 1584-98. 17.
Kuhn, P. H.; Colombo, A. V.; Schusser, B.; Dreymueller, D.; Wetzel, S.; Schepers, U.;
Herber, J.; Ludwig, A.; Kremmer, E.; Montag, D.; Muller, U.; Schweizer, M.; Saftig, P.; Brase, S.; 20 ACS Paragon Plus Environment
Page 20 of 37
Page 21 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
Lichtenthaler, S. F., Systematic substrate identification indicates a central role for the metalloprotease ADAM10 in axon targeting and synapse function. eLife 2016, 5. 18.
Eichelbaum, K.; Krijgsveld, J., Rapid temporal dynamics of transcription, protein synthesis,
and secretion during macrophage activation. Molecular & cellular proteomics : MCP 2014, 13 (3), 792-810. 19.
Megger, D. A.; Bracht, T.; Kohl, M.; Ahrens, M.; Naboulsi, W.; Weber, F.; Hoffmann, A. C.;
Stephan, C.; Kuhlmann, K.; Eisenacher, M.; Schlaak, J. F.; Baba, H. A.; Meyer, H. E.; Sitek, B., Proteomic differences between hepatocellular carcinoma and nontumorous liver tissue investigated by a combined gel-based and label-free quantitative proteomics study. Molecular & cellular proteomics : MCP 2013, 12 (7), 2006-20. 20.
Chen, J.; Shinde, S.; Koch, M. H.; Eisenacher, M.; Galozzi, S.; Lerari, T.; Barkovits, K.;
Subedi, P.; Kruger, R.; Kuhlmann, K.; Sellergren, B.; Helling, S.; Marcus, K., Low-bias phosphopeptide enrichment from scarce samples using plastic antibodies. Scientific reports 2015, 5, 11438. 21.
Mayer, G.; Stephan, C.; Meyer, H. E.; Kohl, M.; Marcus, K.; Eisenacher, M., ProCon -
PROteomics CONversion tool. Journal of proteomics 2015, 129, 56-62. 22.
Petersen, T. N.; Brunak, S.; von Heijne, G.; Nielsen, H., SignalP 4.0: discriminating signal
peptides from transmembrane regions. Nature methods 2011, 8 (10), 785-6. 23.
Bendtsen, J. D.; Jensen, L. J.; Blom, N.; Von Heijne, G.; Brunak, S., Feature-based prediction
of non-classical and leaderless protein secretion. Protein engineering, design & selection : PEDS 2004, 17 (4), 349-56. 24.
Krogh, A.; Larsson, B.; von Heijne, G.; Sonnhammer, E. L., Predicting transmembrane protein
topology with a hidden Markov model: application to complete genomes. Journal of molecular biology 2001, 305 (3), 567-80. 25.
Rybak, J. N.; Scheurer, S. B.; Neri, D.; Elia, G., Purification of biotinylated proteins on
streptavidin resin: a protocol for quantitative elution. Proteomics 2004, 4 (8), 2296-9.
21 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
26.
Willems, L. I.; van der Linden, W. A.; Li, N.; Li, K. Y.; Liu, N.; Hoogendoorn, S.; van der
Marel, G. A.; Florea, B. I.; Overkleeft, H. S., Bioorthogonal chemistry: applications in activity-based protein profiling. Accounts of chemical research 2011, 44 (9), 718-29. 27.
Lin, Z.; Fillmore, G. C.; Um, T. H.; Elenitoba-Johnson, K. S.; Lim, M. S., Comparative
microarray analysis of gene expression during activation of human peripheral blood T cells and leukemic Jurkat T cells. Laboratory investigation; a journal of technical methods and pathology 2003, 83 (6), 765-76. 28.
Busse, S.; Steiner, J.; Micheel, J.; Dobrowolny, H.; Mawrin, C.; Krause, T. J.; Adamaszek,
M.; Bogerts, B.; Bommhardt, U.; Hartig, R.; Busse, M., Age-related increase of VGF-expression in T lymphocytes. Aging 2014, 6 (6), 440-53. 29.
Busse, S.; Steiner, J.; Glorius, S.; Dobrowolny, H.; Greiner-Bohl, S.; Mawrin, C.; Bommhardt,
U.; Hartig, R.; Bogerts, B.; Busse, M., VGF expression by T lymphocytes in patients with Alzheimer's disease. Oncotarget 2015, 6 (17), 14843-51. 30.
Li, L.; Yin, Q.; Kuss, P.; Maliga, Z.; Millan, J. L.; Wu, H.; Mitchison, T. J., Hydrolysis of 2'3'-
cGAMP by ENPP1 and design of nonhydrolyzable analogs. Nature chemical biology 2014, 10 (12), 1043-8. 31.
Abbasi, S.; Shin, D. M.; Beaty, N.; Masiuk, M.; Chen, S.; Gonzalez-Garcia, I.; Zhao, M.;
Goding, J.; Morse, H. C., 3rd; Wang, H., Characterization of monoclonal antibodies to the plasma cell alloantigen ENPP1. Hybridoma (Larchmt) 2011, 30 (1), 11-7. 32.
Muller, H.; Nagel, C.; Weiss, C.; Mollenhauer, J.; Poeschl, J., Deleted in malignant brain
tumors 1 (DMBT1) elicits increased VEGF and decreased IL-6 production in type II lung epithelial cells. BMC pulmonary medicine 2015, 15, 32. 33.
Kang, W.; Reid, K. B., DMBT1, a regulator of mucosal homeostasis through the linking of
mucosal defense and regeneration? FEBS letters 2003, 540 (1-3), 21-5. 34.
Basu, A.; Hoerning, A.; Datta, D.; Edelbauer, M.; Stack, M. P.; Calzadilla, K.; Pal, S.;
Briscoe, D. M., Cutting edge: Vascular endothelial growth factor-mediated signaling in human CD45RO+ CD4+ T cells promotes Akt and ERK activation and costimulates IFN-gamma production. Journal of immunology 2010, 184 (2), 545-9. 22 ACS Paragon Plus Environment
Page 22 of 37
Page 23 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
35.
Zhang, J.; Silva, T.; Yarovinsky, T.; Manes, T. D.; Tavakoli, S.; Nie, L.; Tellides, G.; Pober, J.
S.; Bender, J. R.; Sadeghi, M. M., VEGF blockade inhibits lymphocyte recruitment and ameliorates immune-mediated vascular remodeling. Circulation research 2010, 107 (3), 408-17. 36.
Gilat, D.; Hershkoviz, R.; Goldkorn, I.; Cahalon, L.; Korner, G.; Vlodavsky, I.; Lider, O.,
Molecular behavior adapts to context: heparanase functions as an extracellular matrix-degrading enzyme or as a T cell adhesion molecule, depending on the local pH. The Journal of experimental medicine 1995, 181 (5), 1929-34. 37.
Shibuya, M., Vascular endothelial growth factor receptor-1 (VEGFR-1/Flt-1): a dual regulator
for angiogenesis. Angiogenesis 2006, 9 (4), 225-30; discussion 231. 38.
Shibuya, M., Differential roles of vascular endothelial growth factor receptor-1 and receptor-2
in angiogenesis. Journal of biochemistry and molecular biology 2006, 39 (5), 469-78. 39.
Rindler, M. J.; Xu, C. F.; Gumper, I.; Cen, C.; Sonderegger, P.; Neubert, T. A., Calsyntenins
are secretory granule proteins in anterior pituitary gland and pancreatic islet alpha cells. The journal of histochemistry and cytochemistry : official journal of the Histochemistry Society 2008, 56 (4), 381-8. 40.
Lu, Z.; Wang, Y.; Chen, F.; Tong, H.; Reddy, M. V.; Luo, L.; Seshadrinathan, S.; Zhang, L.;
Holthauzen, L. M.; Craig, A. M.; Ren, G.; Rudenko, G., Calsyntenin-3 molecular architecture and interaction with neurexin 1alpha. The Journal of biological chemistry 2014, 289 (50), 34530-42. 41.
Pettem, K. L.; Yokomaku, D.; Luo, L.; Linhoff, M. W.; Prasad, T.; Connor, S. A.; Siddiqui, T.
J.; Kawabe, H.; Chen, F.; Zhang, L.; Rudenko, G.; Wang, Y. T.; Brose, N.; Craig, A. M., The specific alpha-neurexin interactor calsyntenin-3 promotes excitatory and inhibitory synapse development. Neuron 2013, 80 (1), 113-28. 42.
Wang, L.; Zhu, J.; Shan, S.; Qin, Y.; Kong, Y.; Liu, J.; Wang, Y.; Xie, Y., Repression of
interferon-gamma expression in T cells by Prospero-related homeobox protein. Cell research 2008, 18 (9), 911-20. 43.
Finoulst, I.; Vink, P.; Rovers, E.; Pieterse, M.; Pinkse, M.; Bos, E.; Verhaert, P., Identification
of low abundant secreted proteins and peptides from primary culture supernatants of human T-cells. Journal of proteomics 2011, 75 (1), 23-33.
23 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
44.
Zhu, P.; Jiang, W.; Cao, L.; Yu, W.; Pei, Y.; Yang, X.; Wan, B.; Liu, J. O.; Yi, Q.; Yu, L., IL-
2 mRNA stabilization upon PMA stimulation is dependent on NF90-Ser647 phosphorylation by protein kinase CbetaI. Journal of immunology 2010, 185 (9), 5140-9.
Acknowledgement: This work was supported by FoRUM (Forschungsförderung Ruhr-Universität Bochum Medizinische Fakultät, F784-12) and PURE (Protein research Unit Ruhr within Europe) funded by the Ministry of Science, North Rhine-Westphalia, Germany. Deutsche Forschungsgemeinschaft (DFG) provided funding under grant number TRR60 (project Z3). This work was also supported by de.NBI (FKZ 031 A 534A), a project of the BMBF (Bundesministerium für Bildung und Forschung). The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium (http://proteomecentral. proteomexchange.org) via the PRIDE partner repository with the data set identifier PXD004280 and DOI 10.6019/PXD004280. The authors thank the PRIDE Team for their assistance during the data upload.
Abbreviations in the order of appearance: SPECS
-
Secretome protein enrichment with click sugars
ManNAz
-
Tetraacetylated N-azidoacetyl-D-mannosamine
PMA
-
Phorbol 12-myristate 13-acetate
DBCO
-
Dibenzocyclooctyne
PSM
-
Peptide spectrum match
TukeyHSD
-
Tukey’s honest significant difference
24 ACS Paragon Plus Environment
Page 24 of 37
Page 25 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
Conflict of interest statement: The authors have declared no conflict of interest.
25 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
LC-MS/MS data deposit We deposited the quantitative LC-MS/MS data of our experiments to the PRIDE archive (full submission). We provide extensive data sets of both the cellular proteome and the secretome of activated Jurkat T cells. Data set identifier is PXD004280 and DOI 10.6019/PXD004280. The data can be accessed via the reviewer account using the following login details: Username:
[email protected] Password: wntxQjgP
26 ACS Paragon Plus Environment
Page 26 of 37
Page 27 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
SUPPORTING INFORMATION: The following files are available free of charge at ACS website http://pubs.acs.org: Supplementary Table 1 - Secretome analyses of activated Jurkat cells - Biotin and Desthiobiotin method - Significantly differentially abundant proteins. Supplementary Table 2: Time-dependent cellular proteome analysis of activated Jurkat cells - Significantly differentially abundant proteins. Supplementary Results: Quantitative analysis of the cellular proteome. Supplementary Figure 1: Volcano plots illustrating the time-dependent analysis of the cellular proteome of activated Jurkat cells. Supplementary Figure 2: Comparison of sets of quantified and differentially abundant proteins between the secretome and cellular proteome analyses. Supplementary Table 3: List of mutual differentially abundant proteins in the secretome complement and the cellular proteome.
27 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 28 of 37
Tables: Table 1: List of proteins mutual differentially abundant in both secretome analyses Higher abundance in activated Jurkat T-cells Ratio of Means a) Biotin
Ratio of Means a) Desthiobiotin
UniProt Accession
Gene Name
Protein Name
P05067
APP
Amyloid beta A4 protein
1.51
1.80
Q06481
APLP2
Amyloid-like protein 2
4.17
13.60
P15291
B4GALT1
Beta-1,4-galactosyltransferase 1
1.51
3.40
P43251
BTD
Biotinidase
1.67
1.89
Q9BQT9
CLSTN3
Calsyntenin-3
1.93
2.64
Q92187
ST8SIA4
CMP-N-acetylneuraminate-poly-alpha-2,8sialyltransferase
2.88
5.90
P12109
COL6A1
Collagen alpha-1(VI) chain
3.96
5.39
P18850
ATF6
Cyclic AMP-dependent transcription factor ATF-6 alpha
1.95
3.35
Q9UGM3
DMBT1
Deleted in malignant brain tumors 1 protein
8.15
32.12
Q01459
CTBS
4.38
2.42
P22413
ENPP1
Di-N-acetylchitobiase Ectonucleotide pyrophosphatase/phosphodiesterase family member 1
24.69
26.60
P14625
HSP90B1
Endoplasmin
1.65
1.97
P42892
ECE1
Endothelin-converting enzyme 1
1.63
1.67
Q16610
ECM1
Extracellular matrix protein 1
4.85
6.95
P07093
SERPINE2
Glia-derived nexin
3.29
3.01
Q92896
GLG1
Golgi apparatus protein 1
3.78
4.00
P48723
HSPA13
Heat shock 70 kDa protein 13
2.00
3.45
O94898
LRIG2
Leucine-rich repeats and immunoglobulin-like domains protein 2
8.99
6.73
P01033
TIMP1
Metalloproteinase inhibitor 1
2.24
2.09
Q15223
PVRL1
Nectin-1
2.34
2.46
Q92859
NEO1
Neogenin
3.02
2.80
O14786
NRP1
Neuropilin-1
2.86
6.20
O15240
VGF
Neurosecretory protein VGF
43.03
25.77
Q6UXB8
PI16
Peptidase inhibitor 16
1.65
2.87
P10586
PTPRF
Receptor-type tyrosine-protein phosphatase F
2.48
2.94
Q13332
PTPRS
Receptor-type tyrosine-protein phosphatase S
3.11
3.75
Q92729
PTPRU
Receptor-type tyrosine-protein phosphatase U
1.85
2.32
28 ACS Paragon Plus Environment
Page 29 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
O75326
SMA7A
Semaphorin-7A
2.78
4.48
Q92673
SORL1
Sortilin-related receptor
4.98
1.92
O14773
TPP1
Tripeptidyl-peptidase 1
2.33
2.37
P17948
FLT1
Vascular endothelial growth factor receptor 1
3.23
2.65
P98155
VLDLR
Very low-density lipoprotein receptor
14.23
16.00
Q5VU97
CACHD1
VWFA and cache domain-containing protein 1
3.61
4.06
Higher abundance in inactive Jurkat T-cells Ratio of Means a) Biotin
Ratio of Means a) Desthiobiotin
UniProt Accession
Gene Name
Protein Name
P19022
CDH2
Cadherin-2
-1.81
-1.54
Q8WUM4
PDCD6IP
Programmed cell death 6-interacting protein
-1.55
-2.00
Q9P2B2
PTGFRN
Prostaglandin F2 receptor negative regulator
-2.05
-1.77
a)
Statistical significance of differential abundance between the activated and inactive groups
was tested by ANOVA followed by Tukey’s HSD method. All displayed proteins passed the significance threshold of ANOVA p value ≤ 0.05, p TukeyHSD ≤ 0.05.
29 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Figure Legends:
Figure 1: Schematic representation of the experimental workflow.
Figure 2: Comparison of enrichment efficiency of the biotin and desthiobiotin click chemistry reagents. Inactive samples were only metabolically labeled, activated samples were treated with PMA and Ionomycin in addition and control samples were neither. In A, MS measurements and protein identification provided the protein lists that were compared to a UniProtKB keyword ‘glycoprotein’ list. Mean total glycoprotein numbers of control, inactive and activated samples labeled with the biotin reagent were 63 of 190, 317 of 487 and 343 of 501, respectively. Accordingly, for control, inactive and activated samples labeled with the desthiobiotin reagent mean glycoprotein numbers added up to 79 of 205, 300 of 565 and 349 of 576, respectively. B shows the corresponding comparisons of identified peptides and peptide spectrum matches. Error bars represent the standard deviation.
Figure 3: Results of differential analyses in the Biotin (A) and Desthiobiotin (B) method and comparison of sets of quantified (C) and differentially abundant (D) proteins for the two methods. Threshold for differentially abundant proteins for the p TukeyHSD is 0.05 and 1.5 for the absolute fold change. Mutual differentially abundant proteins are depicted in red and annotated with corresponding gene names. Proteins with a p TukeyHSD = 0 are displayed above the red dashed line (A and B). Positive fold changes indicate higher protein abundance in the activated groups while negative values indicate higher abundance in the corresponding control groups, respectively.
30 ACS Paragon Plus Environment
Page 30 of 37
Page 31 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
Figure 4: Comparison of quantified (A) and differentially abundant (B) proteins obtained with biotin and desthiobiotin click chemistry reagents. The proteins were analyzed regarding their amounts of glycoproteins and possibly secreted proteins. Proteins with a p Tukey HSD lower or equal to 0.05 and an absolute fold change greater or equal to 1.5 between activated and inactive samples were considered to be differentially abundant. Glycoproteins and possibly secreted proteins were annotated as described in the methods section.
Figure 5: Correlation between the fold changes of mutual proteins of both enrichment methods (n=271). Fold change barriers of 1.5 (absolute value) are shown as grey dashed lines. Mutual proteins only differentially abundant in the Biotin method are depicted in green, those of the Desthiobiotin method in blue. Mutual differentially abundant proteins are depicted in red. A linear fit has been applied to mutual differentially abundant proteins (red line) with an R² = 0.77 and a slope of 0.84.
31 ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
for TOC only 82x44mm (300 x 300 DPI)
ACS Paragon Plus Environment
Page 32 of 37
Page 33 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
Figure 1: Schematic representation of the experimental workflow. 177x146mm (300 x 300 DPI)
ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Figure 2: Comparison of enrichment efficiency of the biotin and desthiobiotin click chemistry reagents. Inactive samples were only metabolically labeled, activated samples were treated with PMA and Ionomycin in addition and control samples were neither. In A, MS measurements and protein identification provided the protein lists that were compared to a UniProtKB keyword ‘glycoprotein’ list. Mean total glycoprotein numbers of control, inactive and activated samples labeled with the biotin reagent were 63 of 190, 317 of 487 and 343 of 501, respectively. Accordingly, for control, inactive and activated samples labeled with the desthiobiotin reagent mean glycoprotein numbers added up to 79 of 205, 300 of 565 and 349 of 576, respectively. B shows the corresponding comparisons of identified peptides and peptide spectrum matches. Error bars represent the standard deviation. 177x123mm (300 x 300 DPI)
ACS Paragon Plus Environment
Page 34 of 37
Page 35 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
Figure 3: Results of differential analyses in the Biotin (A) and Desthiobiotin (B) method and comparison of sets of quantified (C) and differentially abundant (D) proteins for the two methods. Threshold for differentially abundant proteins for the p TukeyHSD is 0.05 and 1.5 for the absolute fold change. Mutual differentially abundant proteins are depicted in red and annotated with corresponding gene names. Proteins with a p TukeyHSD = 0 are displayed above the red dashed line (A and B). Positive fold changes indicate higher protein abundance in the activated groups while negative values indicate higher abundance in the corresponding control groups, respectively. 177x132mm (300 x 300 DPI)
ACS Paragon Plus Environment
Journal of Proteome Research
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Figure 4: Comparison of quantified (A) and differentially abundant (B) proteins obtained with biotin and desthiobiotin click chemistry reagents. The proteins were analyzed regarding their amounts of glycoproteins and possibly secreted proteins. Proteins with a p Tukey HSD lower or equal to 0.05 and an absolute fold change greater or equal to 1.5 between activated and inactive samples were considered to be differentially abundant. Glycoproteins and possibly secreted proteins were annotated as described in the methods section. 82x85mm (300 x 300 DPI)
ACS Paragon Plus Environment
Page 36 of 37
Page 37 of 37
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Journal of Proteome Research
Figure 5: Correlation between the fold changes of mutual proteins of both enrichment methods (n=271). Fold change barriers of 1.5 (absolute value) are shown as grey dashed lines. Mutual proteins only differentially abundant in the Biotin method are depicted in green, those of the Desthiobiotin method in blue. Mutual differentially abundant proteins are depicted in red. A linear fit has been applied to mutual differentially abundant proteins (red line) with an R² = 0.77 and a slope of 0.84. 82x64mm (300 x 300 DPI)
ACS Paragon Plus Environment
