Evaluation of FASP, SP3 and iST Protocols for Proteomic Sample

Subscriber access provided by RYERSON UNIVERSITY

Article

Evaluation of FASP, SP3 and iST Protocols for Proteomic Sample Preparation in the Low Microgram Range Malte Sielaff, Jörg Kuharev, Toszka Bohn, Jennifer Hahlbrock, Tobias Bopp, Stefan Tenzer, and Ute Distler J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.7b00433 • Publication Date (Web): 26 Sep 2017 Downloaded from http://pubs.acs.org on September 26, 2017

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 50

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

Evaluation of FASP, SP3 and iST Protocols for Proteomic Sample Preparation in the Low Microgram Range Malte Sielaff†, Jörg Kuharev†, Toszka Bohn†, Jennifer Hahlbrock†, Tobias Bopp†, Stefan Tenzer‡,*,†, Ute Distler‡,*,† †

Institute for Immunology, University Medical Center of the Johannes Gutenberg University Mainz, Langenbeckstr. 1, 55131 Mainz, Germany

* Correspondence should be addressed to: Dr. Ute Distler Phone: +49 (0) 6131 17-6192 E-mail: [email protected]

Prof. Dr. Stefan Tenzer Phone: +49 (0) 6131 17-6199 E-mail: [email protected]

ACS Paragon Plus Environment

1


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 2 of 50

ABSTRACT

Efficient and reproducible sample preparation is a prerequisite for any robust and sensitive quantitative bottom-up proteomics workflow. Here, we performed an independent comparison between single-pot solid-phase-enhanced sample preparation (SP3), filter-aided sample preparation (FASP) and a commercial kit based on the in-StageTip (iST) method. We assessed their performance for the processing of proteomic samples in the low µg-range using varying amounts of HeLa cell lysate (1 µg - 20 µg of total protein). All three workflows showed similar performance for 20 µg of starting material. When handling sample sizes below 10 µg, the number of identified proteins and peptides as well as the quantitative reproducibility and precision drastically dropped in case of FASP. In contrast, SP3 and iST provided high proteome coverage even in the low µg-range. Even when digesting 1 µg of starting material both methods still enabled the identification of over 3,000 proteins and between 25,000 to 30,000 peptides. On average, the quantitative reproducibility between experimental replicates was slightly higher in case of SP3 (R2= 0.97 (SP3); R2= 0.93 (iST)). Applying SP3 towards the characterization of the proteome of FACS-sorted tumor-associated macrophages in the B16 tumor model enabled the quantification of 2,965 proteins and revealed a “mixed” M1/M2 phenotype.

KEYWORDS. Bottom-up proteomics, sample preparation, filter-aided sample preparation (FASP), single-pot solid-phase-enhanced sample preparation (SP3), in-StageTip digestion (iST), tumor-associated macrophages (TAMs).


2

Page 3 of 50

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


INTRODUCTION Bottom-up mass spectrometry (MS)-based proteomics is currently the most widespread workflow to characterize complex proteomes. In traditional bottom-up experiments, proteins are initially extracted and digested by sequence-specific proteases. The resulting peptides are subsequently analyzed by liquid chromatography (LC) typically coupled to the mass spectrometer through an electrospray interface. The sample preparation prior to LC-MS analysis is a critical part of the overall workflow as it highly impacts sensitivity, robustness and reproducibility of any bottom-up proteomics approach. Over the past decades, sample preparation protocols have been optimized to generate conditions that allow efficient protein extraction and enzymatic digestion as well as improved sample recovery. Significant technological advances in MS instrumentation along with optimized sample preparation procedures enable the analysis of protein amounts down to the attomole range1 and ever-smaller samples that were not accessible to MS-based proteomics several years ago can be efficiently analyzed today.2 Several sample processing steps are required to produce LC-MS compatible analytes including protein extraction and denaturation, enzymatic digestion of proteins into peptides followed by peptide purification. Disruption of cells and extraction of proteins is often enhanced by the addition of detergents and chaotropes. The majority of these additives, however, impair enzymatic digestion and LC-MS-analysis. Different strategies have been developed for the removal of cellular debris, salts, detergents and chaotropes including affinity-based methods,3–5 membrane filtration,6,7 electrophoretic approaches5,8 and protein precipitation.5,8 While these approaches allow the efficient removal of detergents such as sodium dodecyl sulfate (SDS), they


3


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 50

often require additional time-consuming processing steps accompanied by sample loss, which may induce unwanted biases in sample composition. Over the past years, multiple workflows have been introduced that address these issues. In particular, methods that facilitate sample manipulations in a single vessel such as filter-aided sample preparation (FASP),7 single-pot solid-phase-enhanced sample preparation (SP3)9 and inStageTip digestion (iST)10 have gained increased popularity. By minimizing sample handling while concomitantly maximizing recovery of analytes they enable the analysis of quantitylimited biological material.7,9–11 FASP, first introduced by Manza6 and further modified by Wisniewski7, as well as variations thereof11–15 have gained widespread popularity as they allow the removal of SDS and other low-molecular weight contaminants (including salts, lipids, etc) by simple centrifugation through a molecular weight cut-off (MWCO) ultrafiltration device. Proteins are retained in the device by a filter membrane and are accessible for subsequent enzymatic digestion. The proteolytically generated peptides are small enough to pass the filter and can be collected by centrifugation. Over the past years, FASP has seen widespread utility for sample processing and has been adapted for many different applications including the analysis of different cell and tissue types,16–19 formalin-fixed paraffin-embedded tissues,11,20 the large-scale characterization of the ubiqitinome,21 mapping of the N-glycoproteome,22 as well as the brain phosphoproteome.23 Typically, after FASP digestion about 50% of starting material can be recovered, which has no adverse effects on proteome coverage when higher amounts of sample are processed.9,11,12 However, incomplete peptide recovery after FASP digestion can be a problem when processing only minute amounts of material. Sample losses associated with FASP have been minimized using for example alternative reagents during sample loading and proteolysis as well as detergents that prepassivate the ultrafiltration unit.11,14 Approaches such as


4

Page 5 of 50

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


enhanced FASP (eFASP)14 report improvements regarding sensitivity, recovery and proteome coverage adding 0.2% deoxycholic acid (DCA) to the digestion buffer. However, in a follow-up study24 no significant differences other than eFASP extracting slightly more basic peptides could be observed between the original protocol and eFASP. SP39 and iST,10 both introduced in 2014, are two more recent “single vessel” approaches that aim at minimizing sample losses by simplifying the proteomic sample preparation workflow. In iST, all steps including cell lysis, processing, and digestion of proteins are performed in stopand-go extraction tips (StageTips) with an inserted C18 disk.10 The iST approach thus enables complete sample preparation in a single reactor. In principle, iST resembles an in-solution digestion with the advantages of a single FASP-like reaction vessel that avoids the use of a filter membrane. The C18 disk serves as a physical barrier for insoluble material and macromolecules. Additionally, it enables final peptide clean-up using solid-phase extraction (SPE). One drawback of iST as compared to FASP is the limitation regarding the use of certain reagents (i.e. iST cannot remove SDS). Like iST, SP39 is a technique for sample preparation in a single tube. In SP3, surface-functionalized (i.e. carboxylate-coated) paramagnetic beads are used to trap proteins and peptides in hydrophilic layers that form around the beads when the organic composition of the buffer is increased and the pH adjusted. The beads can be immobilized within a magnetic field. This allows the efficient removal of contaminating agents including chaotropes and detergents by washing with different organic solvents (i.e. ethanol (EtOH) and acetonitrile (ACN)). After rinsing, bound proteins or peptides can be eluted from the beads using an aqueous solution. This protocol thus allows to perform protein clean-up, enzymatic digestion, desalting, and peptide recovery in in a single tube. The feasibility of SP3 for ultrasensitive proteome analysis has been exemplified by the analysis of HeLa cells where over 15,000 unique peptides


5


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 6 of 50

were identified from as little as 1,000 cells.9 SP3 has been successfully applied in multiple recent studies including, for example, the characterization of lipid binding25 as well as chromatinassociated proteins,26,27 the quantitative proteomic analysis of mouse kidney tissue substructures28 and single-cell human oocytes.29 The SP3 protocol has been further refined and adapted for the in-depth proteome profiling of clinical tumor tissues including formalin-fixed and paraffin-embedded material.30 In the original study, the authors benchmarked SP3 against FASP analyzing 10 µg of yeast whole-cell lysate. Both protocols displayed similar proteome coverage with no bias in the physicochemical properties of identified peptides for either method.9 Moreover, the authors performed a preliminary comparison of SP3 with the data obtained from iST10 indicating a higher sensitivity for SP3 when small-sized samples (down to 1,000 cells) are analyzed. However, these comparisons either focused on a certain starting amount of material or were performed comparing inter-laboratory datasets. To our knowledge, a systematic evaluation of the different protocols for varying amounts of starting material in the low µg-range does not exist yet. In the present study, we performed an independent comparison of FASP, iST and SP3 for the preparation of proteomics samples in the low µg-range. We digested different amounts of HeLa proteins ranging from 1 - 20 µg using the three different protocols and systematically evaluated their performance regarding peptide yields, proteome coverage as well as quantitative reproducibility. We demonstrate that all three approaches performed similarly for starting amounts of 20 µg and showed no bias towards a specific group of peptides and proteins. Whereas no reduction in proteome coverage was observed for SP3 and iST when only minute amounts of starting material were available, the performance of FASP dropped drastically when digesting material below 10 µg. We further tested the three protocols for the processing of bone


6

Page 7 of 50

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


marrow-derived macrophages (BMDMs) isolated by fluorescence-activated cell sorting (FACS). As SP3 outperformed FASP and iST, we subsequently applied SP3 for the proteomic characterization of tumor-associated macrophages (TAMs) from an orthotopic preclinical tumor model, revealing a proteomic signature different to in vitro generated TAMs31. Notably, proteins exclusively detected in our dataset showed a highly significant enrichment of proteins associated with immune system processes and the KEGG pathway “complement and coagulation cascades”.


7


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 8 of 50

MATERIALS AND METHODS Reagents and antibodies. Unless otherwise stated, all solvents (Ultra LC-MS grade) were purchased from Carl Roth (Karlsruhe, Germany) and all chemicals were obtained from SigmaAldrich (St. Louis, MO). Carboxylate-modified paramagnetic beads (Sera-Mag SpeedBeads (Hydrophilic), CAT# 45152105050250, and Sera-Mag SpeedBeads (Hydrophobic), CAT# 65152105050250) were purchased from GE Healthcare (Chicago, IL). In-StageTip (iST) sample preparation kits were purchased from PreOmics (Martinsried, Germany). PE/Cy7 antimouse/human CD11b and Brilliant Violet 421TM anti-mouse CD45 antibodies were purchased from BioLegend (San Diego, CA). APC-eFluor 780 anti-mouse F4/80 (BM8) and FITC antimouse MHC class II eBioscienceTM antibodies were obtained from Thermo Fisher Scientific. Ethics statement. All mice used for this study were bred and housed at the animal facility of Johannes Gutenberg University using institutionally approved protocols. Animal procedures were performed under the supervision of the authorized investigators in accordance with the European Union normative for care and use of experimental animals. Permission was obtained from the Landesuntersuchungsamt Koblenz (23 177-07/G 14-1-002). Cell culture. The human cervix carcinoma cell line HeLa was obtained from the German Resource Centre for Biological Material (DSMZ). Cells were cultured in Iscove’s Modified Dulbecco’s Medium (IMDM; PAN Biotech, Aidenbach, Germany) supplemented with 10% (v/v) fetal calf serum (FCS; Thermo Fisher Scientific (Invitrogen), Waltham, MA), 1% (v/v) Lglutamine (Carl Roth), and 1% (v/v) sodium pyruvate (Serva, Heidelberg, Germany) at 37°C in a 5% CO2 environment. Cells were harvested at 70% confluence. After an initial washing step with phosphate buffered saline (PBS; Carl Roth), cells were detached from the culture flasks using


8

Page 9 of 50

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


0.05% Trypsin-EDTA solution (Sigma-Aldrich). Detached cells were transferred into centrifugal tubes and washed three times with PBS. Cell pellets were stored at -80°C until further processing. Generation of bone marrow-derived macrophages (BMDMs). In order to generate BMDMs, C57Bl/6 mice were sacrificed by CO2 asphyxiation. Femurs and tibiae of hind legs were extracted, cleaned from surrounding tissue, and disinfected in isopropyl alcohol. Both ends of the bones were cut off and bone marrow cells were flushed out with Minimal Essential Medium (MEM) supplemented with 2% (v/v) FCS using a syringe and a needle. After pelleting the cells by centrifugation, erythrocytes were lysed in 1 mL of Gey’s solution (155 mM NH4Cl, 10 mM KHCO3, 0.13 mM EDTA) for 1 min. Reaction was quenched by the addition of 2% (v/v) FCS in MEM and the cell suspension was passed through a cell strainer (70 µm, EASYstrainer cell strainer, Greiner Bio-One). The number of living cells was determined using viability staining with trypan blue and a hemocytometer. Afterwards, cells were centrifuged and resuspended in L-929 cell conditioned medium which served as source for macrophage colony stimulating factor (M-CSF). L-929 conditioned medium was generated culturing M-CSF producing murine L-929 cells (DSMZ) for seven days in IMDM medium supplemented with 5% (v/v) FCS, 1% (v/v) L-glutamine and 1% (v/v) sodium pyruvate. After centrifugation and sterile filtration, supernatant was used for BMDM cell culture at 10% (v/v) in IMDM medium containing 10% (v/v) FCS, 1% (v/v) L-glutamine and 1% (v/v) sodium pyruvate. Cells were transferred into 6-well culture plates at a concentration of 5×106 cells/well. At days four, seven and eight cells were fed. At day nine, adherent cells were harvested and subjected to FACS analysis.


9


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 10 of 50

Isolation of tumor-associated macrophages (TAMs). To induce melanoma formation, 100 µL of B16-OVA melanoma cell suspension (200,000 cells in total) was subcutaneously injected into the right flank of C57Bl/6 mice.32 After 17 days of tumor growth, mice were sacrificed by CO2 asphyxiation and the tumor tissue was surgically isolated. In order to prepare single-cell suspensions the tumors were incubated in Collagenase D (1 µg/mL in PBS) at 37°C for 30 min. Afterwards, the cell suspensions were filtered twice through 40 µm cell strainers (EASYstrainer cell strainer, Greiner Bio-One) and subjected to FACS analysis. Tumor experiments were performed in triplicates. FACS analysis. BMDMs and TAMs were isolated by FACS as described before using antibodies against CD11b, F4/80, MHCII and CD11b, F4/80, CD45, respectively.33 In brief, harvested BMDMs and single-cell suspensions of melanomas were incubated with FACS staining solution (PBS containing 0.5% (w/v) bovine serum albumin (BSA), 5 mM EDTA, 0.01% sodium azide and the respective antibodies) at 4°C for 15 min. Afterwards, cells were washed and resuspended in 0.5% (w/v) BSA, 5 mM EDTA in PBS prior to FACS analysis. Cells were sorted on a FACSAria II Cell Sorter (BD Biosciences) into 1.5 mL protein LoBind tubes (Eppendorf AG, Hamburg, Germany). Isolated cells were washed once with PBS prior to lysis. Protein extraction. Lysis of cells was performed using either a urea-based lysis buffer (7 M urea, 2 M thiourea, 5 mM dithiothreitol (DTT), 2% (w/v) CHAPS) or the commercial lysis buffer provided in the iST sample preparation kit (PreOmics). In both cases, lysis was promoted by sonication at 4°C for 15 min using a Bioruptor (Diagenode, Liège, Belgium). Samples were additionally incubated for 10 min at 95°C prior to sonication, when iST lysis buffer was used.


10

Page 11 of 50

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


After cell lysis, protein concentration was determined using the Pierce 660 nm protein assay (Thermo Fisher Scientific) according the manufacturer´s protocol and samples were aliquoted in 20 µL- aliquots containing 1, 2, 5, 10, or 20 µg of protein. Protein digestion. HeLa cell lysates containing 1 - 20 µg of proteins were digested using FASP, SP3 and iST. BMDMs were subjected to FASP, SP3 and iST digest. TAMs were processed by SP3. 1) Filter-aided sample preparation (FASP). FASP was performed as detailed before.34 Briefly, lysates were loaded onto spin filter columns (Nanosep centrifugal devices with Omega membrane, 30 kDa MWCO; Pall, Port Washington, NY) and detergents were removed washing the samples three times with buffer containing 8 M urea. Afterwards, proteins were reduced using DTT, followed by alkylation with iodoacetamide (IAA). Excess IAA was quenched by the addition of DTT. Afterwards, buffer was exchanged washing the membrane three times with 50 mM NH4HCO3. Proteins were digested overnight at 37°C using trypsin (Trypsin Gold, Promega, Madison, WI) at an enzyme-to-protein ratio of 1:50 (w/w). Peptides were recovered by centrifugation and two additional washes with 50 mM NH4HCO3. Flow-throughs were combined. Afterwards, samples were acidified with trifluoroacetic acid (TFA) to a final concentration of 1% (v/v) TFA and lyophilized. Purified peptides were reconstituted in 0.1% (v/v) formic acid (FA) for LC-MS analysis. 2) Single-pot solid-phase-enhanced sample preparation (SP3). Two types of carboxylatemodified paramagnetic beads were used for SP3 digest (see above). Prior to use, beads were combined in a ratio of 1:1 (v/v), rinsed and reconstituted in water at a concentration of 20 µg solids/µL as described by Hughes et al.9 The prepared bead mix was stored at 4°C. Initially, SP3


11


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 12 of 50

digest was performed according to the protocol of Hughes et al.9 In brief, lysates were reduced and alkylated using DTT and IAA, respectively. After reduction and alkylation, 2 µL of the prepared bead mix was added to the lysate and samples were acidified using FA. Afterwards, ACN was added to a final concentration of 50% (v/v) and samples were allowed to settle at room temperature for 8 min. Subsequently, beads were immobilized by incubation on a magnetic rack for 2 min. The supernatant was discarded and the pellet was rinsed with 70% (v/v) EtOH in water and 100% ACN. Beads were resuspended in 5 µL of 50 mM NH4HCO3 supplemented with trypsin at an enzyme-to-protein ratio of 1:25 (w/w). After overnight digestion at 37°C, ACN was added to reach a final concentration of 95% (v/v). After mixing and incubation, supernatant was removed and beads were rinsed with 100% ACN. Peptides bound to the beads were eluted using 2% (v/v) dimethyl sulfoxide (DMSO) in water. Supernatant containing purified peptides was transferred into a fresh tube. Prior to LC-MS analysis, the peptides were acidified using 1% FA. As proteins did not interact efficiently with the carboxylate beads during the initial binding step, we slightly modified the original protocol in order to maximize protein binding and recovery: After reduction and alkylation, 2 µL of the prepared bead mix was added to the lysate. However, in contrast to the original protocol9, samples were not acidified. Afterwards, ACN was added to a final concentration of 70% (v/v) instead of 50% (v/v). After short vortexing of the sample-bead mix, we further increased the initial incubation time (at room temperature) from 8 to 18 minutes. Additionally, we repeated the SP3 peptide purification step with the supernatant of the initial peptide binding step. Towards this purpose, the collected supernatant was supplemented with 2 µl of carboxylate-functionalized bead mix. ACN wash and recovery step were repeated. Peptides derived from the initial binding step as well as from the second


12

Page 13 of 50

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


purification step were pooled. All other steps were conducted as described above according the protocol of Hughes et al.9 3) In-StageTip digestion (iST). iST was performed according to the manufacturer’s instructions using a commercial sample preparation kit (PreOmics). Briefly, the cell lysate was loaded onto the StageTips and digestion buffer was added. After 3 h of incubation at 37°C, stop buffer was transferred to the samples. Afterwards, the devices were centrifuged at 3,800 × g for 1 min and washed twice using two different wash buffers which were provided in the kit. The StageTips were then placed into fresh collection tubes and peptides were recovered by centrifugation at 2,800 × g for 1 min using the elution buffer. The elution step was repeated once. Eluted peptides were vacuum dried and reconstituted in LC-loading buffer. SDS-polyacrylamide gel electrophoresis (SDS-PAGE). Binding efficiencies of proteins (10 µg) to the carboxylate beads during SP3 protein clean-up was analyzed by SDS-PAGE and Coomassie staining. Prior to loading onto the gel (NuPageTM 10% Bis-Tris Gel, Thermo Fisher Scientific), bead-bound proteins and vacuum-dried supernatants of the clean-up procedure were recovered in NuPage LDS sample buffer (Thermo Fisher Scientific) and incubated for 5 min at 95°C. After electrophoretic separation, proteins were fixed and stained incubating the SDS-gel with Coomassie InstantBlue (Expedeon, San Diego, CA) for 30 min. Gels were recorded with a ChemiDoc XRS system (Bio-Rad, München, Germany). LC-MS Analysis. LC-MS analyses were performed using a NanoAQUITY UPLC system (Waters Corporation, Milford, MA), which was coupled on-line to a Synapt G2-S high definition mass spectrometer (Waters Corporation) via a NanoLockSpray dual electrospray ion source (Waters Corporation). Tryptic digests corresponding to 200 ng of peptides were loaded directly


13


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 14 of 50

onto a HSS-T3 C18 1.8 µm, 75 µm × 250 mm reversed-phase analytical column (Waters Corporation). Mobile phase A was 0.1% (v/v) FA and 3% (v/v) DMSO in water. Mobile phase B was 0.1% (v/v) FA and 3% (v/v) DMSO in ACN. Peptides were separated running a gradient from 5-40% (v/v) mobile phase B at a flow rate of 300 nL/min over 90 min. The column was heated to 55°C. MS analysis of eluting peptides was performed by ion-mobility separation (IMS) enhanced data-independent acquisition (DIA) in UDMSE mode as detailed before.34,35 In brief, precursor ion information was collected in low-energy MS mode at a constant collision energy of 4 eV. Fragment ion information was obtained in the elevated energy scan applying drift-time specific collision energies. The spectral acquisition time in each mode was 0.6 s with a 0.05 sinterscan delay resulting in an overall cycle time of 1.3 s for the acquisition of one cycle of low and elevated energy data. [Glu1]-fibrinopeptide was used as lock mass at 100 fmol/µL and sampled every 30 s into the mass spectrometer via the reference sprayer of the NanoLockSpray source. Data processing and label-free quantification. Raw data processing and database search of LC-MS data were performed using ProteinLynx Global Server (PLGS, ver.3.0.2, Waters Corporation). Data were searched against a custom compiled UniProt Human proteome database (UniProtKB release 2017_02, 20,168 entries) that contained a list of common contaminants. The following parameters were applied for database search: trypsin was specified as enzyme for digestion, up to two missed cleavages per peptide were allowed and peptides had to have a length of at least six amino acids. Carbamidomethyl cysteine was set as fixed, and methionine oxidation as variable modification. The false discovery rate (FDR) for peptide and protein identification was assessed using the target-decoy strategy by searching a reverse database. FDR was set to 0.01 for database search in PLGS.


14

Page 15 of 50

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


Post-processing of data (i.e. of the three experimental replicates of each condition as well as the cross-annotation between methods and conditions) was performed using the software tool ISOQuant ver.1.834,35 including retention time alignment, exact mass retention time (EMRT) as well as IMS clustering, normalization and protein homology filtering. Algorithms and settings have been described in detail.34,35 A FDR of 0.01 at the peptide-level was applied for cluster annotation in ISOQuant. Proteins were only reported if they had been identified by at least two peptides with a minimum length of six amino acids, a minimum PLGS score of 6.0 and no missed cleavages. For each protein absolute in-sample amounts were calculated using TOP3 quantification.36 To assess cleavage efficiencies of the different protocols, data were reprocessed in ISOQuant using the settings as described above but including peptides with missed cleavages. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium (http://proteomecentral.proteomexchange.org) via the PRIDE partner repository37 with the dataset identifier .


15


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 50

RESULTS AND DISCUSSION Quantity-limited biological samples including tumor biopsies, rare (sub-)populations of e.g. immune cells or subcellular fractions and organelles of certain cell types still pose a challenge for MS-based proteomics as it can be very difficult to obtain sufficient material to generate highquality mass spectrometric data. Over the past years, microscale proteomic workflows greatly benefited from the development of novel sample preparation methods such as FASP7 and the more recently introduced SP39 and iST10 approaches, which enable excellent proteome coverage even in the low µg-range. However, despite marked improvements regarding sample loss, robustness and reproducibility, none of the described methods is yet accepted as a ‘gold standard’ for ultrasensitive proteomics. In the present study, we therefore investigated the qualitative and quantitative performance of FASP, SP3 and iST for the processing of mammalian cells in the low µg-range. Optimization of SP3 protein binding conditions Prior to the comparison with the other methods, we optimized the SP3 protocol as we observed substantial loss of protein during the initial protein-binding step (Figure 1A). In the original protocol9 (see also Figure 1A), cell lysates are initially mixed with SP3 beads and acidified with formic acid (FA). Afterwards, ACN is added to the protein bead mixture to a final concentration of 50% (v/v) to immobilize proteins on the beads in a HILIC-type of manner. Once immobilized on the beads, proteins can be purified and lysis buffer substances can be removed efficiently using a combination of EtOH and ACN washes. However, when following the instructions of the original protocol we observed that most proteins remained in the supernatant and did not interact with the carboxylate beads during the initial binding step (Figure 1A). No protein loss was


16

Page 17 of 50

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


observed during the EtOH and ACN rinses. Increased ACN concentrations (up to 95% (v/v) ACN) as well as altered protein-to-bead ratios (up to tenfold the number of magnetic beads) did not improve protein-binding efficiency during the initial step (data not shown). Hence, we decided to systematically evaluate the influence of the pH in combination with different ACN concentrations on the interaction of proteins and carboxylate-functionalized beads (Figure 1B). Experiments with final concentrations of 50-70% (v/v) ACN revealed a clear correlation between low pH and the amount of unbound proteins that could be still detected in the supernatant (Figure 1B, upper panels and lower left panel). Protein recovery was further improved by increasing the incubation times from 8 to 18 min in combination with additional mixing of the sample (Figure 1B, lower right panel). Complete binding of proteins to the carboxylate-functionalized beads was achieved at neutral pH and ACN concentrations of 70% (v/v) (and higher) as well as incubation times of 18 min (Figure 1B). We used these optimized conditions throughout the rest of the study to promote initial binding of proteins to the carboxylate beads. In a recent study from Hughes et al. the authors also applied a slightly modified version of the original protocol omitting the initial acidification step.30 Using optimized conditions for protein immobilization, efficient protein recovery was observed for both SDS and urea-based lysis buffers, different protein amounts (145 µg), as well as concentrated and dilute solutions (Figures S1 and S2). We further assessed the peptide recovery in SP3 using tryptic HeLa peptides that were generated by FASP. Full recovery was observed when incubating the peptides with carboxylate-coated beads at neutral pH, which is in line with the data of Hughes et al.9 However, similar to the observations on protein level, acidic conditions impaired efficient binding of peptides to the carboxylate beads (Figure 1C). When handling the beads, we further observed that the beads aggregated and precipitated under


17


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 18 of 50

neutral conditions along with the bound peptides indicating successful peptide-bead interaction (Figure S3). Comparison of FASP, SP3 and iST for the digest of cell lysates in the low µg-range After the optimization of SP3 protein binding conditions, we focused on the comparison between FASP, SP3 and iST for the processing of proteomic samples in the low µg-range. Towards this purpose, HeLa cell lysate (1 µg – 20 µg) was processed in triplicates by FASP, SP3 and iST. In case of iST, a commercial kit was used for sample preparation. After proteolytic digestion, tryptic peptides were analyzed by LC-MS using ion-mobility enhanced DIA.34,35 When processing 20 µg of starting material, all three approaches showed similar performance (Figure 2). We were able to detect 3,600 to 4,400 proteins and 28,000 to 36,900 peptides using the different protocols. Regarding number of identified peptides and proteins iST slightly outperformed both, FASP and SP3. On average, we detected 4,448 proteins and 36,671 peptides in 20 µg HeLa cell lysate using iST (Figure 2A, Tables S1 and S2). In case of FASP and SP3, the average number of identified proteins was lower (3,739 and 3,608 proteins, respectively). We further assessed the precision of each method by calculating the coefficient of variation (CV) for the protein abundances between the three sample preparation replicates (Figure 2B, Table S2). FASP and iST showed similar median CVs of 8.9% and 7.7%, respectively. SP3 displayed the highest median CV of (13.2%) for the 20 µg sample. In quantitative proteomics experiments, acquired MS data are typically normalized during data processing to increase comparability and reliability of downstream analysis. Data normalization aims to correct systematic biases derived from non-biological sources such as small variations in the experimental conditions during the course of sample preparation and LC-MS analysis. We tested to which extent normalization of


18

Page 19 of 50

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 data could correct biases derived from the different sample preparation protocols. After normalization, the precision of each method was increased (Figure 2B). SP3 benefited most from the normalization process (median CV: 13.2% before and 8.4% after normalization). However, as compared to the other protocols, iST still showed the lowest variation between experimental replicates with a median CV of 6.0 % after normalization. Despite differences in the CV distributions, all three protocols showed a high quantitative reproducibility between sample preparation replicates (Pearson correlation, R2 = 0.93 – 0.99, no normalization) for the 20 µg samples. However, the reproducibility of FASP was slightly lower as compared to SP3 or iST (Figure 2C). The lower reproducibility observed in FASP might be attributed to a higher rate of missed cleavages, which we observed in FASP as compared to the other two protocols (Figures S4 and S5). When digesting material below 20 µg, we observed a considerable drop in the performance of FASP. Whereas around 3,739 proteins were identified on average from 20 µg of starting material, only 353 proteins were detectable in 1 µg of HeLa lysate (Figure 2A). Along with a decreased number of identified proteins, the precision of the method dropped markedly (median CV of 17.5%) and we also observed lower correlation between technical replicates (R2 = 0.93 (20 µg) and R2 = 0.83 (1 µg)) (Figure 2B,C). Processing 10 µg of protein with FASP already resulted in a significantly lower number of identified proteins as compared to the 20 µg sample (3,739 proteins versus 3,150 proteins). In contrast, both, SP3 and iST, showed high proteome coverage even when handling material in the low µg-range (1-2 µg). Whereas the number identified proteins remained in the same range for SP3 independent of the amount of starting material (20 – 1 µg), a small drop in detectable proteins could be observed for iST comparing the 20 µg and the 1 µg sample. On average, around 23,500 peptides and 3,020 proteins could be


19


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 20 of 50

identified with iST in 1 µg of HeLa lysate and approx. 27,000 peptides corresponding to 3,300 proteins with SP3. Even in the low µg-range, both methods displayed high quantitative reproducibility between technical replicates. However, SP3 (R2 = 0.97 (1 µg)) slightly outperformed iST (R2 = 0.93 (1 µg)). In addition, precision was high for both protocols with median CVs of 8.5% (SP3, no normalization) and 7.4% (iST, no normalization) (Figure 2B). Interestingly, precision of SP3 was higher in the low µg-range (i.e. 1 µg) as compared to the 20 µg sample. The poor performance of FASP observed in the present study for samples below 20 µg can be most likely attributed to the fact that only about 50% of starting material can be recovered after the digest.9,11,12 Although in previous studies no adverse effects on proteome coverage could be observed for higher amount of starting material,9,11,12 in our hands the poor recovery seems to affect sample processing in the low µg-range. Sample loss associated with FASP digestion has been observed before for quantity-limited samples.5,18,38 For higher amounts of starting material peptide yield in FASP can be increased, for example, by sequential digestion with multiple enzymes, also referred to MED-FASP.13,39 Further, it has been shown that the type and shape of filter units can have an impact on the performance of the FASP protocol.12,39,40 Peptide yields, for example, increase with higher MWCO filters (30 K MWCO as compared to 3 K and 10 K MWCO) while concomitantly reducing centrifugation time. Based on the data of the respective studies and own prior observations we also chose 30 K MWCO filters for the present study. Initially, to assess the performance of each method (raw) data derived from the different protocols and starting amounts were processed and handled individually. Comparing the datasets for the 20 µg samples, around 52% of proteins were identified with all three protocols (Figure 2D). Typically, in MSE-based workflows, where all precursors are fragmented in parallel across


20

Page 21 of 50

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 full mass range during the elevated energy scan, fragment ion spectra are highly multiplexed, which may impair peptide identification in complex samples. In previous studies we could show that cross-annotation between runs markedly improves data quality and reduces missing values.35 To determine if proteins that were exclusively detected by only one or two sample preparation methods can be attributed to the differences in sample preparation or derive from the way the data is acquired in MSE-based workflows, we reprocessed the datasets and performed a crossannotation of features between the runs. Feature cross-annotation revealed that 86% of identified proteins were present in all three digests of the 20 µg samples (Figure 2D, Table S3). Even in the 1 µg sample, the overlap of proteins identified in both, SP3 and iST, was still 90%. As discussed in detail before, the number of proteins identified by FASP was markedly reduced in the 1 µg sample. Although cross-annotation between the different methods allowed to recover 872 proteins in the 1 µg FASP digest, the overall number of proteins in FASP still lacked far behind the numbers of proteins identified by SP3 and iST (Figure 2D). We further tested if the different sample preparation strategies showed any biases towards certain protein groups. Although we detected proteins that displayed significant changes between the different protocols, gene ontology (GO) analysis of those candidates did not reveal any bias towards a specific protein class or subcellular localization (Figure S6, Table S4). Moreover, there was no visible bias regarding the physicochemical properties of the proteins and peptides captured with each protocol in case of the 20 µg samples, as indicated by the molecular mass, isoelectric point and the grand average hydropathy (GRAVY)41 distributions of identified peptides and proteins (Figure S7A,B). In case of FASP and SP3 these observations are in line with the findings of Hughes et al.,9 who found no discernible differences in the properties of identified peptides between the two methods when analyzing 10 µg of whole yeast lysate. Regarding the FASP


21


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 22 of 50

protocol the observed drop in performance for 1 µg samples is accompanied by a bias towards proteins and peptides with lower molecular weight and proteins that display either a low or a high isoelectric point (Figure S7C,D). For quantity limited samples, such as FACS-sorted cells or laser-capture microdissected tissues, determination of total protein content is sometimes difficult or even impossible. Therefore, samples from an experiment series may often contain varying amounts of starting material. As these unexpected variations in the amounts of starting material may negatively affect the results of label-free quantification experiments, we assessed the performance of each protocol regarding its repeatability, precision and quantitative reproducibility when different starting amounts are used. All three protocols displayed a high overlap between the proteins that were identified when processing cell lysates ranging from 2 to 10 µg (Figure 3, Table S5). Similarly to the comparison between the methods, cross-annotation recovered protein identifications in case of FASP in the low µg sample. However, around 14.5% of the proteins were only detectable in the 5 µg and 10 µg samples. The drop of performance for samples in the low µg-range in case of FASP is also reflected in the high CVs of measured protein abundances (median CV of 51.3% without normalization), when comparing samples with different starting amounts. Also, quantitative reproducibility dropped markedly in particular when comparing the 2 µg and 5 µg samples (Figure 3A). In contrast, SP3 and iST both display high precision and quantitative reproducibility in the low µg-range even when the amount of starting material differs prior to proteolytic digestion (Figure 3B,C). Analysis of FACS sorted macrophages by FASP, SP3 and iST


22

Page 23 of 50

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


Immune cells show a high level of complexity marked by the vast number of different (sub)populations of B and T lymphocytes, phagocytic and antigen-presenting cells, cytotoxic natural killer cells and granulocytes. Many of the different immune cell populations occur only in low numbers including, for example, rare subsets of B and T cells or tumor-infiltrating leukocytes. Flow cytometry and cell sorting are widely used techniques for the analysis of these cells and have enabled the identification, characterization and isolation even of rare immune cell subsets. As the study of these quantity-limited cell types is of high biological relevance, we tested the applicability of FASP, SP3 and iST for the analysis of limited numbers of FACS sorted immune cells. To this end, we subjected in vitro generated BMDMs to FACS analysis collecting aliquots containing 25,000 BMDMs each (corresponding to approximately 1 µg of total protein). Aliquoted BMDMs were processed in triplicates with FASP, SP3 and iST (Figure 4A, Tables S6 and S7). In terms of proteome coverage SP3 showed the best results identifying on average 3,152 proteins and 22,202 peptides. Whereas on average 2,343 proteins and 17,674 peptides were still detectable with iST, we found that FASP is not feasible for low numbers of FACS sorted cells. On average, we identified 109 proteins and 663 peptides in 25,000 BMDMs accompanied by a poor precision and low quantitative reproducibility using FASP (Figure 4A). In line with the previous experiments, both SP3 and iST show high precision with median CVs of 8.1% (SP3) and 7.7% (iST) after normalization and a high correlation of label-free quantification data between experimental replicates (R2= 0.93 – 0.96, Figure S8). Since SP3 outperformed both FASP and iST when analyzing low numbers of macrophages, we applied SP3 for the ex vivo characterization of TAMs that were isolated by FACS from a B16 melanoma mouse model. TAMs typically account for a major fraction of the leukocytic tumor infiltrate and experimental data indicate that these macrophages enhance tumor progression and


23


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 24 of 50

macrophage infiltration correlates with poor prognosis of many cancer types.42 Albeit their importance in tumor growth and progression the limited access to high numbers of these immune cells has prevented in-depth proteome analysis. As a proof-of-principle we inoculated mice with B16F10 melanoma cells43 and analyzed TAMs upon ex vivo isolation by FACS from an orthotopic preclinical tumor model. In total, we were able to obtain between 22,000 to 30,000 TAMs from the tumor of a single mouse. The optimized SP3 protocol enabled us to identify on average 2,965 proteins and 22,070 peptides in the isolated TAMs (Figure 4B, Tables S6 and S7). In the present dataset we could detect many different marker proteins that have been associated with a TAM phenotype, including for example, Arginase-1, C-X-C motif chemokine 10, CD206, Interleukin-1 receptor antagonist protein (IL-1ra), TGF-beta-1 as well as matrix metalloproteases (e.g. MMP-9) and cathepsins44,45 (see Figure 4C). Moreover, in line with many previous studies45, detected proteins indicate a “mixed” M1/M2 phenotype for the macrophage population isolated from the B16 tumor model. A recent study from Zhu et al.31 focused on the proteomic characterization of exosomes derived from TAMs that were generated in vitro by culturing Ana-1 mouse macrophage cells in CT26 conditioned media. The authors report a threefold increase in released exosomes in their TAM tumor model as compared to normal Ana-1 cells. Interestingly, in our dataset we find a significant enrichment of proteins associated with the GO term “extracellular exosomes” (pvalue = 9.7 x 10-316) (Table S8). We compared our ex vivo TAM dataset to the in vitro generated TAMs from Zhu et al. The overlap of identified proteins from the two studies is close to 50% (Figure 4D). Among the proteins identified in both datasets, we found several proteins associated with MHC class I, TAP-dependent antigen processing and presentation, including tapasin, different proteasome as well as immunoproteasome subunits. Interestingly, Zhu et al.31 report


24

Page 25 of 50

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


higher proteasome activity in TAM exosomes as compared to the corresponding in vitro generated TAMs. In contrast to the data of Zhu et al.31, and the BMDM data of the present study, we exclusively detected the proteasome subunits beta2i (MECL-1) and beta5i (LMP7), both constituents of the immunoproteasome, in the ex vivo TAMs from the B16 melanoma but not the corresponding beta2 and beta5 subunits of the constitutive proteasome (see Figure S9). suggesting that TAMs contain intermediate-type proteasomes46 with an unusual beta1, beta2i, beta5i composition. In combination with the presence of the subunit of the PA28 regulatory particle, this may influence cross-presentation of tumor-derived MHC class I epitopes47. Other proteins associated with TAM differentiation and function, which are present in the dataset of Zhu et al. and were detected in the ex vivo TAMs in the present study include the macrophage colony-stimulating factor 1 (CSF1) receptor, cathepsins, Perilipin-2 (PLIN2), macrophage migration inhibitory factor (MIF) and Sialoadhesin (SIGLEC1). The CSF1 receptor acts as cell-surface receptor for the major macrophage lineage regulator CSF1. High CSF1 expression in cancer is typically associated with poor prognosis and has a major impact on the recruitment and polarization of TAMs44,48. Whereas PLIN2 and MIF have been associated with the pro-tumoral M2 phenotype49, expression of SIGLEC1 has been linked to a tumoricidal immune response50. PLIN2, MIF and SIGLEC1 were also up-regulated in TAMs that were generated in vitro from tumor-extract stimulated BMDMs, also referred to as TES-TAMs51. Unfortunately, no full list of identified proteins was provided in the respective study, only a listing of differentially regulated proteins between TES-TAMs and BMDMs. Hence, the data could not be included in the present manuscript for the comparison of the proteomes of in vitro generated TAMs versus ex vivo TAMs. The authors further report significantly increased levels of Fibrinopeptide B (FGB) promoting invasion ability of tumor cells when expressed in TAMs as


25


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 26 of 50

well as increased quantities of the protumorigenic protein NGAL (Neutrophil gelatinaseassociated lipocalin) in TES-TAMs51. Both proteins were also present in our ex vivo TAM dataset. Moreover, the authors report a metabolic reprogramming of TAMs by the tumor microenvironment towards aerobic glycolysis. In line with their findings, functional annotation analysis revealed an enrichment of proteins associated with the glycolysis/gluconeogenesis pathway in our TAM dataset and we were able to identify all key players associated with aerobic glycolysis that were differentially regulated in the TES-TAM dataset of Liu et al.51 (see Figure S10 and Table S9). Other proteins of interest that were identified in the dataset of Liu et al. and the present study are the proteins associated with positive regulation of angiogenesis including thrombospondin-1 and the myeloid-derived growth factor52. Additionally, we performed a functional annotation analysis of the proteins that were exclusively detected in our dataset as compared to the in vitro generated TAMs derived from Ana-1 cells (Figure 4D). Among the top ten enriched “biological processes”, immune system processes, small GTPase mediated signal transduction as well as innate immune response were the most significant. Moreover, proteins exclusively detected in our TAM data as compared to the in vitro generated TAMs by Zhu et al. point towards different pathways that are associated with basic macrophage function such as endo-/phagocytosis including phagosomal and lysosmal maturation (Figure S11). Additionally, we found many proteins associated with complement and coagulation cascades, including complement components C3 and C5, as well as the C5a receptor 1 mostly present in M2 macrophages. Activation of complement mediators typically promotes angiogenesis and plays a critical role in tumor development. It has been described that complement resistant cancer cells, for example, are hugely infiltrated with high levels with C3a and C5a, TAMs as well as complement regulatory proteins53. Despite the advantages of in vitro


26

Page 27 of 50

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


models our data indicate the necessity to characterize cells, i.e. TAMs, derived from a more “natural” in vivo environment to draw a more comprehensive picture of the underlying mechanisms of TAM polarization and function. In summary, we found that both protocols, SP3 and iST performed well for minute sample amounts. Further, as compared to FASP, which involves many washing and centrifugation steps, they require less hands-on time. With sample preparation times of approximately five hours including vacuum concentration of the peptides after digest and purification the commercial iST kit outperformed the other two protocols in terms of speed. However, as we used the commercial kit for iST, it was the most expensive protocol at about 40 € per sample. In the present study, we focused on the comparison of “single vessel” protocols for the processing mammalian cells. However, the efficiency of a sample preparation method also relies on the investigated sample type and the subsequent LC-MS workflow. Thus, the workflows might need to be adapted when architecturally different organisms, as for example bacteria are analyzed.54 Further, other protocols have been introduced that might require additional equipment but provide excellent performance for sample limited material concomitantly reducing sample handling and increasing speed of analysis, such as immobilized enzyme reactors and pressure cycling technology.55–59

CONCLUSION Reproducible sample preparation is a challenging task that limits the performance of quantitative proteomics experiments, especially when only limiting amounts of starting material are available. In the present study, we compared three “single vessel” proteomic protocols, FASP, SP3 and iST, for the preparation of samples in the low microgram range and assessed


27


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 50

their performance when handling protein amounts ranging from 1 µg to 20 µg. Whereas all three approaches performed well for starting amounts of 20 µg of protein, we observed a massive drop in performance in case of FASP when processing material below 20 µg. As compared to the other two methods, proteome coverage in FASP was markedly lower. The drop of protein identifications was accompanied by decreased quantitative reproducibility as well as a bias towards low molecular weight proteins and peptides. In contrast, both SP3 and iST showed high proteome coverage as well as precision and reproducibility even when handling only 1 µg of protein. To our knowledge, our study provides the first detailed analysis and independent comparison between FASP, SP3 and iST for the processing of samples in the low µg-range and will enable researchers to select the optimal sample processing workflow for their specific application. Applying the optimized SP3 workflow for the analysis of TAMs isolated by FACS sorting from an orthotopic preclinical tumor model, we provide a detailed characterization of the proteome of this clinically relevant cell population60, indicating that the TAMs display a “mixed” M1/M2 phenotype. With recent improvements regarding the sensitivity of LC-MS platforms, analysis of samples with limiting amounts of starting material will become more common thus requiring high reproducible sample preparation in the low µg-range.


28

Page 29 of 50

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


ACKNOWLEDGEMENTS We thank Sandra Fischer for expert technical help. We thank the German Research Foundation (DFG) (TE599/2-1, TE599/3-1 to S.T.), the Research Center for Immunotherapy (FZI) and the Focus Program Translational Neurosciences (FTN) of the Johannes Gutenberg University Mainz for funding. U.D. was supported by the University Medical Center of the Johannes Gutenberg University Mainz (Internal University Research Funding (Stufe I)).

ASSOCIATED CONTENT Supporting Information. The following files are available free of charge. Supplementary File 1. File containing Supplementary Figures S1 to S11. (PDF) Supplementary Figure S1. Comparison of SDS and urea-based lysis buffers for SP3 digest. Supplementary Figure S2. SP3 protein clean-up using higher amounts of starting material and volumes. Supplementary Figure S3. Scheme of peptide binding to carboxylate-coated paramagnetic beads at neutral and acidic pH. Supplementary Figure S4. Comparison of proteolytic cleavage efficiencies using different digestion protocols.


29


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 30 of 50

Supplementary Figure S5. Analysis of the sequence context for miscleaved peptides in FASP, SP3 and iST. Supplementary Figure S6. Volcano plot representation of differentially “regulated” proteins between SP3, iST and FASP. Supplementary Figure S7. Physicochemical properties of identified proteins and peptides. Supplementary Figure S8. Correlation of experimental replicates of FASP, SP3 and iST digests of FACS-sorted BMDMs. Supplementary Figure S9. Identified proteasome subunits in TAMs derived from a B16 melanoma mouse model. Supplementary Figure S10. Proteins involved in glycolysis/gluconeogenesis identified in TAMs derived from a B16 melanoma mouse model. Supplementary Figure S11. Proteins associated with the phagosome identified in TAMs derived from a B16 melanoma mouse model. Supplementary Table S1. Numbers of proteins and peptides identified by LC-MS in HeLa cell lysate processed by FASP, SP3, and iST using differing amounts of starting material. Accompanying data to Figure 2. (XLSX) Supplementary Table S2. List of identified proteins in HeLa cell lysate prepared by FASP, SP3, and iST using different amounts of starting material. Accompanying data to Figure 2. (XLSX) Supplementary Table S3. List of identified proteins in HeLa cell lysate prepared by FASP, SP3, and iST after cross-annotation between methods. Accompanying data to Figure 2. (XLSX)


30

Page 31 of 50

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


Supplementary Table S4. Functional annotation analysis of proteins that were found to be significantly enriched by one sample preparation over the other. (XLSX) Supplementary Table S5. List of identified proteins in HeLa cell lysate prepared by FASP, SP3, and iST after cross-annotation of the 10 µg, 5 µg and 2 µg samples. Accompanying data to Figure 3. (XLSX) Supplementary Table S6. Numbers of proteins and peptides identified by LC-MS in lysates of FACS-isolated BMDMs and TAMs after processing with FASP, SP3, and iST. Accompanying data to Figure 4. (XLSX) Supplementary Table S7. List of identified proteins in lysates of FACS-isolated BMDMs and TAMs after processing with FASP, SP3, and iST. Accompanying data to Figure 4. (XLSX) Supplementary Table S8. GO analysis of proteins expressed by TAMs isolated from a B16 tumor model. Accompanying data to Figure 4. (XLSX)

AUTHOR INFORMATION Corresponding Authors * Ute Distler; phone: +49 (0) 6131 17-6192; e-mail: [email protected] * Stefan Tenzer; phone: +49 (0) 6131 17-6199; e-mail: [email protected]

Author Contributions ‡ S.T. and U.D. contributed equally.


31


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 32 of 50

Notes The authors declare no competing financial interests.


32

Page 33 of 50

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


REFERENCES (1)

Thakur, S. S.; Geiger, T.; Chatterjee, B.; Bandilla, P.; Fröhlich, F.; Cox, J.; Mann, M. Deep and highly sensitive proteome coverage by LC-MS/MS without prefractionation. Mol. Cell. Proteomics 2011, 10 (8), M110.003699.

(2)

Altelaar, A. F. M.; Heck, A. J. R. Trends in ultrasensitive proteomics. Curr. Opin. Chem. Biol. 2012, 16 (1–2), 206–213.

(3)

Bereman, M. S.; Egertson, J. D.; MacCoss, M. J. Comparison between procedures using SDS for shotgun proteomic analyses of complex samples. Proteomics 2011, 11 (14), 2931–2935.

(4)

Hengel, S. M.; Floyd, E.; Baker, E. S.; Zhao, R.; Wu, S.; Paša-Tolić, L. Evaluation of SDS depletion using an affinity spin column and IMS-MS detection. Proteomics 2012, 12 (21), 3138–3142.

(5)

Kachuk, C.; Stephen, K.; Doucette, A. Comparison of sodium dodecyl sulfate depletion techniques for proteome analysis by mass spectrometry. J. Chromatogr. A 2015, 1418, 158–166.

(6)

Manza, L. L.; Stamer, S. L.; Ham, A.-J. L.; Codreanu, S. G.; Liebler, D. C. Sample preparation and digestion for proteomic analyses using spin filters. Proteomics 2005, 5 (7), 1742–1745.

(7)

Wiśniewski, J. R.; Zougman, A.; Nagaraj, N.; Mann, M. Universal sample preparation method for proteome analysis. Nat. Meth. 2009, 6 (5), 359–362.


33


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)

Page 34 of 50

Feist, P.; Hummon, A. Proteomic Challenges: Sample Preparation Techniques for Microgram-Quantity Protein Analysis from Biological Samples. Int. J. Mol. Sci. 2015, 16 (2), 3537–3563.

(9)

Hughes, C. S.; Foehr, S.; Garfield, D. A.; Furlong, E. E.; Steinmetz, L. M.; Krijgsveld, J. Ultrasensitive proteome analysis using paramagnetic bead technology. Mol. Syst. Biol. 2014, 10 (10), 757–757.

(10)

Kulak, N. A.; Pichler, G.; Paron, I.; Nagaraj, N.; Mann, M. Minimal, encapsulated proteomic-sample processing applied to copy-number estimation in eukaryotic cells. Nat. Methods 2014, 11 (3), 319–324.

(11)

Wiśniewski, J. R.; Ostasiewicz, P.; Mann, M. High recovery FASP applied to the proteomic analysis of microdissected formalin fixed paraffin embedded cancer tissues retrieves known colon cancer markers. J. Proteome Res. 2011, 10 (7), 3040–3049.

(12)

Wiśniewski, J. R.; Zielinska, D. F.; Mann, M. Comparison of ultrafiltration units for proteomic and N-glycoproteomic analysis by the filter-aided sample preparation method. Anal. Biochem. 2011, 410 (2), 307–309.

(13)

Wiśniewski, J. R.; Mann, M. Consecutive proteolytic digestion in an enzyme reactor increases depth of proteomic and phosphoproteomic analysis. Anal. Chem. 2012, 84 (6), 2631–2637.

(14)

Erde, J.; Loo, R. R. O.; Loo, J. A. Enhanced FASP (eFASP) to increase proteome coverage and sample recovery for quantitative proteomic experiments. J. Proteome Res. 2014, 13 (4), 1885–1895.


34

Page 35 of 50

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


(15)

McDowell, G. S.; Gaun, A.; Steen, H. IFASP: Combining isobaric mass tagging with filter-aided sample preparation. J. Proteome Res. 2013, 12 (8), 3809–3812.

(16)

Tanca, A.; Biosa, G.; Pagnozzi, D.; Addis, M. F.; Uzzau, S. Comparison of detergentbased sample preparation workflows for LTQ-Orbitrap analysis of the Escherichia coli proteome. Proteomics 2013, 13 (17), 2597–2607.

(17)

Maurer, M.; Müller, A. C.; Parapatics, K.; Pickl, W. F.; Wagner, C.; Rudashevskaya, E. L.; Breitwieser, F. P.; Colinge, J.; Garg, K.; Griss, J.; et al. Comprehensive Comparative and Semiquantitative Proteome of a Very Low Number of Native and Matched Epstein– Barr-Virus-Transformed B Lymphocytes Infiltrating Human Melanoma. J. Proteome Res. 2014, 13 (6), 2830–2845.

(18)

Maurer, M.; Müller, A. C.; Wagner, C.; Huber, M. L.; Rudashevskaya, E. L.; Wagner, S. N.; Bennett, K. L. Combining filter-aided sample preparation and pseudoshotgun technology to profile the proteome of a low number of early passage human melanoma cells. J. Proteome Res. 2013, 12 (2), 1040–1048.

(19)

Li, S.; Cao, X.; Wang, Y.; Zhu, Z.; Zhang, H.; Xue, S.; Tian, J. A Method for Microalgae Proteomics Analysis Based on Modified Filter-Aided Sample Preparation. Appl. Biochem. Biotechnol. [Online early access] 2017. 10.1007/s12010-017-2473-9. Published Online: Apr 11, 2017. https://link.springer.com/article/10.1007/s12010-017-2473-9 (accessed Apr 11, 2017).

(20)

Wiśniewski, J. R.; Duś, K.; Mann, M. Proteomic workflow for analysis of archival formalin-fixed and paraffin-embedded clinical samples to a depth of 10000 proteins.


35


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 36 of 50

Proteomics - Clin. Appl. 2013, 7 (3–4), 225–233. (21)

Casanovas, A.; Pinto-Llorente, R.; Carrascal, M.; Abian, J. Large-Scale Filter-Aided Sample Preparation Method for the Analysis of the Ubiquitinome. Anal. Chem. 2017, 89 (7), 3840–3846.

(22)

Zielinska, D. F.; Gnad, F.; Wiśniewski, J. R.; Mann, M. Precision Mapping of an In Vivo N-Glycoproteome Reveals Rigid Topological and Sequence Constraints. Cell 2010, 141 (5), 897–907.

(23)

Wiśniewski, J. R.; Nagaraj, N.; Zougman, A.; Gnad, F.; Mann, M. Brain phosphoproteome obtained by a fasp-based method reveals plasma membrane protein topology. J. Proteome Res. 2010, 9 (6), 3280–3289.

(24)

Nel, A. J. M.; Garnett, S.; Blackburn, J. M.; Soares, N. C. Comparative reevaluation of FASP and enhanced FASP methods by LC - MS/MS. J. Proteome Res. 2015, 14 (3), 1637–1642.

(25)

Höglinger, D.; Nadler, A.; Haberkant, P.; Kirkpatrick, J.; Schifferer, M.; Stein, F.; Hauke, S.; Porter, F. D.; Schultz, C. Trifunctional lipid probes for comprehensive studies of single lipid species in living cells. Proc. Natl. Acad. Sci. 2017, 114 (7), 1566–1571.

(26)

Zee, B. M.; Alekseyenko, A. A.; McElroy, K. A.; Kuroda, M. I. Streamlined discovery of cross-linked chromatin complexes and associated histone modifications by mass spectrometry. Proc. Natl. Acad. Sci. U. S. A. 2016, 113 (7), 1784–1789.

(27)

Rafiee, M.-R.; Girardot, C.; Sigismondo, G.; Krijgsveld, J. Expanding the Circuitry of


36

Page 37 of 50

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


Pluripotency by Selective Isolation of Chromatin-Associated Proteins. Mol. Cell 2016, 64 (3), 624–635. (28)

De Graaf, E. L.; Pellegrini, D.; McDonnell, L. A. Set of Novel Automated Quantitative Microproteomics Protocols for Small Sample Amounts and Its Application to Kidney Tissue Substructures. J. Proteome Res. 2016, 15 (12), 4722–4730.

(29)

Virant-Klun, I.; Leicht, S.; Hughes, C.; Krijgsveld, J. Identification of Maturation-Specific Proteins by Single-Cell Proteomics of Human Oocytes. Mol. Cell. Proteomics 2016, 15 (8), 2616–2627.

(30)

Hughes, C. S.; McConechy, M. K.; Cochrane, D. R.; Nazeran, T.; Karnezis, A. N.; Huntsman, D. G.; Morin, G. B. Quantitative Profiling of Single Formalin Fixed Tumour Sections: proteomics for translational research. Sci. Rep. 2016, 6 (October), 34949.

(31)

Zhu, Y.; Chen, X.; Pan, Q.; Wang, Y.; Su, S.; Jiang, C.; Li, Y.; Xu, N.; Wu, L.; Lou, X.; et al. A Comprehensive Proteomics Analysis Reveals a Secretory Path- and StatusDependent Signature of Exosomes Released from Tumor-Associated Macrophages. J. Proteome Res. 2015, 14 (10), 4319–4331.

(32)

Overwijk, W. W.; Restifo, N. P. B16 as a Mouse Model for Human Melanoma. In Current Protocols in Immunology; John Wiley & Sons, Inc.: Hoboken, NJ, USA, 2001; Vol. Chapter 20, p Unit 20.1.

(33)

Gautier, E. L.; Shay, T.; Miller, J.; Greter, M.; Jakubzick, C.; Ivanov, S.; Helft, J.; Chow, A.; Elpek, K. G.; Gordonov, S.; et al. Gene-expression profiles and transcriptional regulatory pathways that underlie the identity and diversity of mouse tissue macrophages.


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

Page 38 of 50

Nat. Immunol. 2012, 13 (11), 1118–1128. (34)

Distler, U.; Kuharev, J.; Navarro, P.; Tenzer, S. Label-free quantification in ion mobility– enhanced data-independent acquisition proteomics. Nat. Protoc. 2016, 11 (4), 795–812.

(35)

Distler, U.; Kuharev, J.; Navarro, P.; Levin, Y.; Schild, H.; Tenzer, S. Drift time-specific collision energies enable deep-coverage data-independent acquisition proteomics. Nat. Methods 2014, 11 (2), 167–170.

(36)

Silva, J. C.; Gorenstein, M. V; Li, G.-Z.; Vissers, J. P. C.; Geromanos, S. J. Absolute quantification of proteins by LCMSE: a virtue of parallel MS acquisition. Mol. Cell. Proteomics 2006, 5 (1), 144–156.

(37)

Vizcaíno, J. A.; Côté, R. G.; Csordas, A.; Dianes, J. A.; Fabregat, A.; Foster, J. M.; Griss, J.; Alpi, E.; Birim, M.; Contell, J.; et al. The PRoteomics IDEntifications (PRIDE) database and associated tools: status in 2013. Nucleic Acids Res. 2013, 41 (Database issue), D1063-9.

(38)

Liebler, D. C.; Ham, A.-J. L. Spin filter–based sample preparation for shotgun proteomics. Nat. Methods 2009, 6 (11), 785–785.

(39)

Wiśniewski, J. R. Quantitative Evaluation of Filter Aided Sample Preparation (FASP) and Multienzyme Digestion FASP Protocols. Anal. Chem. 2016, 88 (10), 5438–5443.

(40)

Lipecka, J.; Chhuon, C.; Bourderioux, M.; Bessard, M.-A.; van Endert, P.; Edelman, A.; Guerrera, I. C. Sensitivity of mass spectrometry analysis depends on the shape of the filtration unit used for filter aided sample preparation (FASP). Proteomics 2016, 16 (13),


38

Page 39 of 50

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


1852–1857. (41)

Kyte, J.; Doolittle, R. F. A simple method for displaying the hydropathic character of a protein. J. Mol. Biol. 1982, 157 (1), 105–132.

(42)

Noy, R.; Pollard, J. W. Tumor-Associated Macrophages: From Mechanisms to Therapy. Immunity 2014, 41 (1), 49–61.

(43)

Overwijk, W. W.; Restifo, N. P. B16 as a mouse model for human melanoma. Curr. Protoc. Immunol. 2001, Chapter 20, Unit 20.1.

(44)

Noy, R.; Pollard, J. W. Tumor-Associated Macrophages: From Mechanisms to Therapy. Immunity 2014, 41 (1), 49–61.

(45)

Biswas, S. K.; Mantovani, A. Macrophage plasticity and interaction with lymphocyte subsets: cancer as a paradigm. Nat. Immunol. 2010, 11 (10), 889–896.

(46)

Vigneron, N.; Van den Eynde, B. J. Proteasome subtypes and regulators in the processing of antigenic peptides presented by class I molecules of the major histocompatibility complex. Biomolecules 2014, 4 (4), 994–1025.

(47)

Barrio, M. M.; Abes, R.; Colombo, M.; Pizzurro, G.; Boix, C.; Roberti, M. P.; Gélizé, E.; Rodriguez-Zubieta, M.; Mordoh, J.; Teillaud, J.-L. Human macrophages and dendritic cells can equally present MART-1 antigen to CD8(+) T cells after phagocytosis of gamma-irradiated melanoma cells. PLoS One 2012, 7 (7), e40311.

(48)

Ao, J.-Y.; Zhu, X.-D.; Chai, Z.-T.; Cai, H.; Zhang, Y.-Y.; Zhang, K.-Z.; Kong, L.-Q.; Zhang, N.; Ye, B.-G.; Ma, D.-N.; et al. Colony-Stimulating Factor 1 Receptor Blockade


39


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 40 of 50

Inhibits Tumor Growth by Altering the Polarization of Tumor-Associated Macrophages in Hepatocellular Carcinoma. Mol. Cancer Ther. 2017, 16 (8). (49)

Yaddanapudi, K.; Putty, K.; Rendon, B. E.; Lamont, G. J.; Faughn, J. D.; Satoskar, A.; Lasnik, A.; Eaton, J. W.; Mitchell, R. A. Control of tumor-associated macrophage alternative activation by macrophage migration inhibitory factor. J. Immunol. 2013, 190 (6), 2984–2993.

(50)

O’Neill, A. S. G.; van den Berg, T. K.; Mullen, G. E. D. Sialoadhesin - a macrophagerestricted marker of immunoregulation and inflammation. Immunology 2013, 138 (3), 198–207.

(51)

Liu, D.; Chang, C.; Lu, N.; Wang, X.; Lu, Q.; Ren, X.; Ren, P.; Zhao, D.; Wang, L.; Zhu, Y.; et al. Comprehensive Proteomics Analysis Reveals Metabolic Reprogramming of Tumor-Associated Macrophages Stimulated by the Tumor Microenvironment. J. Proteome Res. 2016, 16 (1), acs.jproteome.6b00604.

(52)

Korf-Klingebiel, M.; Reboll, M. R.; Klede, S.; Brod, T.; Pich, A.; Polten, F.; Napp, L. C.; Bauersachs, J.; Ganser, A.; Brinkmann, E.; et al. Myeloid-derived growth factor (C19orf10) mediates cardiac repair following myocardial infarction. Nat. Med. 2015, 21 (2), 140–149.

(53)

Khan, M. A.; Assiri, A. M.; Broering, D. C. Complement and macrophage crosstalk during process of angiogenesis in tumor progression. J. Biomed. Sci. 2015, 22 (1), 58.

(54)

Glatter, T.; Ahrné, E.; Schmidt, A. Comparison of different sample preparation protocols reveals lysis buffer-specific extraction biases in gram-negative bacteria and human cells.


40

Page 41 of 50

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


J. Proteome Res. 2015, 14 (11), 4472–4485. (55)

Huang, E. L.; Piehowski, P. D.; Orton, D. J.; Moore, R. J.; Qian, W. J.; Casey, C. P.; Sun, X.; Dey, S. K.; Burnum-Johnson, K. E.; Smith, R. D. Snapp: Simplified nanoproteomics platform for reproducible global proteomic analysis of nanogram protein quantities. Endocrinology 2016, 157 (3), 1307–1314.

(56)

Zhou, H.; Ning, Z.; Wang, F.; Seebun, D.; Figeys, D. Proteomic reactors and their applications in biology. FEBS J. 2011, 278 (20), 3796–3806.

(57)

Guo, T.; Kouvonen, P.; Koh, C. C.; Gillet, L. C.; Wolski, W. E.; Röst, H. L.; Rosenberger, G.; Collins, B. C.; Blum, L. C.; Gillessen, S.; et al. Rapid mass spectrometric conversion of tissue biopsy samples into permanent quantitative digital proteome maps. Nat. Med. 2015, 21 (4), 407–413.

(58)

Shao, S.; Guo, T.; Gross, V.; Lazarev, A.; Koh, C. C.; Gillessen, S.; Joerger, M.; Jochum, W.; Aebersold, R. Reproducible Tissue Homogenization and Protein Extraction for Quantitative Proteomics Using MicroPestle-Assisted Pressure-Cycling Technology. J. Proteome Res. 2016, 15 (6), 1821–1829.

(59)

Clair, G.; Piehowski, P. D.; Nicola, T.; Kitzmiller, J. A.; Huang, E. L.; Zink, E. M.; Sontag, R. L.; Orton, D. J.; Moore, R. J.; Carson, J. P.; et al. Spatially-Resolved Proteomics: Rapid Quantitative Analysis of Laser Capture Microdissected Alveolar Tissue Samples. Sci. Rep. 2016, 6, 39223.

(60)

Mantovani, A.; Marchesi, F.; Malesci, A.; Laghi, L.; Allavena, P. Tumour-associated macrophages as treatment targets in oncology. Nat. Rev. Clin. Oncol. 2017, 14 (7), 399–


41


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 42 of 50

416.


42

Page 43 of 50

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. SP3 protein and peptide recovery. (A) Schematic workflow of the SP3 protein binding and clean-up procedure as described by Hughes et al.9 Ten µg of HeLa cell lysate were subjected to SP3 treatment according to the original protocol. Proteins recovered after SP3 protein clean-up procedure (green) as well as supernatants derived from initial binding (orange) and purification steps were analyzed by SDS-PAGE. Untreated HeLa cell lysate served as control. No protein loss was observed during the EtOH and ACN washes. However, most of the proteins did not bind to the carboxylate-coated beads during the initial binding step of SP3. (B) Influence of pH and ACN (in v/v %) concentration was tested for SP3 protein binding using HeLa cell lysate. Proteins recovered from the beads (green) as well as supernatants from the initial binding step (orange) were analyzed by SDS-PAGE. Complete protein binding was observed at neutral pH and ACN concentrations of 70% and higher. (C) Peptide recovery was assessed for SP3 peptide clean-up at acidic and neutral pH. Five µg of tryptic HeLa peptides either treated with SP3 or untreated (control) were analyzed by LC-MS. As described in the original protocol complete binding of peptides could be observed at neutral pH.

Figure 2. Comparison of FASP, SP3 and iST regarding numbers of identified peptides and proteins and quantitative reproducibility. Different amounts of HeLa cell lysate ranging from 1 µg to 20 µg of protein were processed using either FASP (green), SP3 (purple) or iST (orange). Sample preparation was performed in triplicates followed by peptide analysis with LC-MS (see also Supplementary Tables 1 and 2). (A) As compared to FASP and SP3, the commercial iST kit showed highest numbers of identified peptides and proteins when processing 10 to 20 µg of


43


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 44 of 50

protein. In the lower range (1 to 2 µg) SP3 and iST displayed similar results regarding numbers of identified peptides and proteins. (B) Coefficients of variation (CVs) for protein abundances were calculated for all three methods and different starting amounts (20 µg and 1 µg) before (w/o N) and after normalization (N). Panel (C) depicts correlation plots of protein intensities for sample preparation replicates of each method. Whereas all three approaches show high quantitative reproducibility between technical replicates for starting amounts of 20 µg of protein, performance of FASP dropped massively when analyzing 1 µg of protein. (D) Venn diagrams of identified proteins in FASP, SP3 and iST for 1 µg and 20 µg of starting material. Before crossannotation: FASP, SP3 and iST data were analyzed and processed individually and no crossannotation was performed. After cross-annotation: Overlap of the different methods after processing and cross-annotation in ISOQuant (Supplementary Table 3).

Figure 3. Reproducibility and robustness of (A) FASP, (B) SP3 and (C) iST comparing different starting amounts. Venn diagrams show the overlap of identified proteins between 10 µg, 5 µg and 2 µg of starting material after processing and cross-annotation in ISOQuant for all three approaches (Supplementary Table 5). Further, coefficients of variation (CVs) for protein abundances between different starting amounts (10 µg, 5 µg and 2 µg) before (w/o N) and after normalization (N) as well as correlation plots are depicted for each method. Whereas SP3 and iST show very similar results regarding overall number of identified proteins as well as robustness and reproducibility when comparing the 10 µg, 5 µg and 2 µg samples, the performance of FASP decreased with lower starting amounts. In particular, the CVs between protein abundances of different starting amounts are poor in case of FASP and can be partially


44

Page 45 of 50

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


overcome by normalization of the data (median CV: 51.3% before and 22.2% after normalization).

Figure 4. Analysis of FACS-sorted macrophages. (A) The left panel shows the comparison of FASP, SP3 and iST regarding numbers of proteins (bars) and peptides (circles) identified in 25,000 FACS-sorted BMDMs. The right panel depicts the CVs for protein abundances for all three methods before (w/o N) and after normalization (N). (B) Numbers of proteins (bars) and peptides (circles) identified in TAMs isolated from the B16 melanoma mouse model after SP3 digestion. (C) Identified proteins indicate a “mixed” M1/M2 phenotype for the macrophage population. (D) Overlap of proteins identified in the ex vivo TAMs of the present study and the in vitro generated TAMs from Zhu et al.31 along with the Top10 enriched “biological processes” and KEGG pathways of the proteins that were exclusively detected in the ex vivo TAMs isolated from the B16 melanoma mouse model.


45


Page 46 of 50

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 ACS Paragon Plus Environment

46

Page 47 of 50

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




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 48 of 50

Page 49 of 50

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




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 50 of 50

Evaluation of FASP, SP3 and iST Protocols for Proteomic Sample

Recommend Documents