Reproducible Tissue Homogenization and Protein ... - ACS Publications

Subscriber access provided by SUNY DOWNSTATE

Article

Reproducible Tissue Homogenization and Protein Extraction for Quantitative Proteomics using MicroPestle-assisted Pressure Cycling Technology Shiying Shao, Tiannan Guo , Vera Gross, Alexander Lazarev, Ching Chiek Koh, Silke Gillessen, Markus Joerger, Wolfram Jochum, and Ruedi Aebersold J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.5b01136 • Publication Date (Web): 21 Apr 2016 Downloaded from http://pubs.acs.org on April 26, 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 24

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

Reproducible Tissue Homogenization and Protein Extraction for Quantitative Proteomics using MicroPestle-assisted Pressure Cycling Technology Shiying Shao1,2, Tiannan Guo 1,Vera Gross3, Alexander Lazarev3, Ching Chiek Koh1, Silke Gillessen4, Markus Joerger4, Wolfram Jochum5, Ruedi Aebersold1,6* 1 Department of Biology, Institute of Molecular Systems Biology, ETH Zurich, Switzerland 2 Division of Endocrinology, Tongji Hospital, Huazhong University of Science & Technology, Wuhan, 430030, PR China 3 Pressure BioSciences, Inc. South Easton, Ma. USA 4 Department of Oncology/Hematology, Kantonsspital St. Gallen, St. Gallen, Switzerland 5 Institute of Pathology, Kantonsspital St. Gallen, St. Gallen, Switzerland 6 Faculty of Science, University of Zurich, Zurich, Switzerland

Table

Figure

0

4

Supplemental

Supplemental

Table

Figure

17

5

* Corresponding author Address for Correspondence: Ruedi Aebersold Institute: Department of Biology, Institute of Molecular Systems Biology, ETH Zurich, Address: Wolfgang-Pauli-Str. 16, 8093, Switzerland Tel: +41 446333170 Fax: +41 446331051 E-mail：[email protected] Abbreviations DDA, data-dependent acquisition; DIA, data-independent acquisition; FDR, false discovery rate; PCT, pressure cycling technology; SWATH, Sequential Windowed Acquisition of all Theoretical fragmentions; TPP, Trans-Proteomic Pipeline Keywords: mass spectrometry, pressure cycling technology, PCT-MicroPestle, PCT-MicroCap, SWATH

S-1

ACS Paragon Plus Environment


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 24

Abstract The reproducible and efficient extraction of proteins from biopsy samples for quantitative analysis is a critical step in biomarker and translational research. Recently, we described a method consisting of pressure cycling technology (PCT) and SWATH-MS for the rapid quantification of thousands of proteins from biopsy-size tissue samples. As an improvement of the method, we have incorporated the PCT-MicroPestle into PCT-SWATH workflow. The PCT-MicroPestle is a novel, miniaturized, disposable mechanical tissue homogenizer that fits directly into the microTube sample container. We optimized the pressure cycling conditions for tissue lysis with the PCT-MicroPestle, and benchmarked the performance of the system against the conventional PCT-MicroCap method, using mouse liver, heart, brain and human kidney tissues as test samples. The data indicate that the digestion of the PCTMicroPestle extracted proteins yielded 20-40% more MS-ready peptide mass from all tissues tested with a comparable reproducibility when compared to the conventional PCT method. Subsequent SWATH-MS analysis identified a higher number of biologically informative proteins from a given sample. In conclusion, we have developed a new device that can be seamlessly integrated into PCTSWATH workflow, leading to increased sample throughput and improved reproducibility, at both protein extraction and proteomic analysis levels, when applied to quantitative proteomic analysis of biopsy-level samples.

1


Page 3 of 24

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 Tissue homogenization and cell lysis are the first steps in virtually all molecular biology and molecular diagnostic techniques.1, 2 They serve to disrupt the tissues/cells and release nucleic acids, proteins and other biomolecules that can be collected in soluble form for downstream processing and analysis. Since almost all clinical specimens are of limited quantity, enabling protein preparation with maximal recovery and reproducibility from minimal sample amounts is essential. Even if relatively larger tissue samples are available, it is frequently advantageous to process subsections of similar morphological appearance to minimize the effects of intra-tissue heterogeneity at the molecular level.3 This is particularly the case in solid tumours. However, efficient and consistent homogenization and lysis of small tissue samples for the reproducible extraction of proteins remain challenging. Traditional and widely used methods for tissue homogenization typically rely on tissue grinders (e.g. Potter-Elvehjem, Wheaton, USA or Dounce, Corning, USA) or mortar and pestle (e.g. Micro Pestle, Sigma, USA) for tissue disruption.4 These methods rely on manual sample processing, often one sample at a time, and suffer from several critical drawbacks, specifically the need for relatively large tissue samples and elevated technical variation when processing smaller tissues. For example, tissue grinders, which depend on the forces generated by pestle motion, both rotation and up-and-down pestle strokes, require fine control and much practice to get consistent protein yields. The electric TissueRuptor (Qiagen, USA) has a rotating blade under controllable forces, which allows for more reproducible and less labor-intensive sample tissue homogenization. However, similar to the other traditional methods mentioned above, the TissueRuptor generally requires relatively large tissues and it is challenging to obtain reproducible protein samples from micro-scale tissue specimens. These limitations impede the concurrent processing of large sample cohorts required for many clinical and basic science studies and contribute to the difficulties in reproducing experimental results over time and across laboratories. Pressure cycling technology (PCT) was developed to perform tissue homogenization and protein extraction from biopsy-size tissue samples by alternating cycles of ultrahigh and low hydrostatic pressures.5-8 PCT is different from conventional mechanical methods in which the disruptive forces are inhomogenous, and largely confined to the sample surface. In addition, the method is automated, programmable, and can be used for semi-automated sample preparation with lower person-to-person variability due to the well-controlled, non-shearing disruptive physical forces.9 Furthermore, in the described PCT method, up to 16 samples can be processed in parallel, thus increasing throughput, reducing processing time and minimizing sample degradation and variability. Using PCT, we have developed a protocol that combines tissue homogenisation, protein extraction, and pressure-accelerated digestion in the same microTube in a single operation. When the PCT method was integrated with SWATH-mass spectrometry (MS), several thousand 2



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 24

proteins could be identified and quantified from a single tissue biopsy.10-12 We further evaluated the minimal amount of wet tissue required for PCT-SWATH analysis,13 and found that as few as 50,000 human cells and 0.2 ~ 0.5 mg of wet mouse and human tissues reliably produced peptide amounts that were sufficient for multiple SWATH-MS analyses at optimal sample loading. That study also demonstrated that the reproducibility of this protocol for smaller tissue samples was higher than that of other protocols tested.13 Collectively, these data suggest that PCT can improve extraction and digestion using minute amounts of samples. However, limitations of this method include the relatively poor yield from fibrous and hard tissues, and incomplete homogenization of relatively large tissue samples.13 To further improve the performance of the method described above, we explored a novel miniaturized device, called PCT-MicroPestle, a disposable mechanical tissue homogenizer that fits inside the microTube sample container. Under repeated pulses of high hydrostatic pressure, the PCT-microTube walls collapse around the essentially incompressible Teflontm MicroPestle, resulting in a homogenizerlike effect. Pressure cycling also acts directly on the sample, helping to disrupt membranes and solubilize proteins. These two concurrent actions contribute to more efficient tissue lysis and protein extraction, even for challenging tissue types. In this study, we further optimized pressure cycling parameters for PCT-MicroPestle, and integrated the method into the general PCT-SWATH workflow.

Materials and methods Tissue preparation Mouse tissues (liver, heart and brain) were collected from C57BL/6J mice and stored at -80°C. The practice protocol complied with the official ETH Zürich ethical guidelines. Human kidney tissue samples were collected from nephrectomy specimens at the Institute of Pathology, Kantonsspital St. Gallen. Within 1 hour after surgical removal, kidney samples were snap-frozen and stored at -80°C. The study protocol was approved by the ethics committee “Ethikkommission St. Gallen”. Before tissue collection, written informed consent was obtained from each included patient. The weight of tissues was measured with an analytic balance (METTLER TOLEDO, XS205). PCT-based lysis and digestion Sample lysis and protein digestion were performed in the Barocycler model NEP2320-Enhanced (Pressure BioSciences, Inc, South Easton, MA), which can process 12 samples concurrently. Tissue samples were prepared with a PCT-MicroPestle or by the conventional PCT-MicroCap method. Tissue pieces were placed in microTubes (Pressure BioSciences) either with PCT-MicroPestle (Pressure BioSciences) in 30 µl lysis buffer or with PCT-MicroCap (Pressure BioSciences) in 60 µl lysis buffer. The lysis buffer for all samples contained 8 M urea and 0.1 M ammonium bicarbonate, supplemented 3


Page 5 of 24

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


with COMPLETE protease inhibitor cocktail (Roche, Switzerland) and PhosSTOP phosphatase inhibitor cocktail (Roche, Switzerland). For the new PCT-MicroPestle-assisted tissue disruption protocol, the PCT-MicroPestle was used for the tissue lysis/extraction step and PCT-MicroCaps were used for the subsequent digestion steps. PCTMicroPestle-assisted tissue disruption was optimized by comparing different pressure regimens and cycling parameters. Samples were processed with 30 cycles of 20 seconds at high pressure (25,000 p.s.i., 35,000 p.s.i. and 45,000 p.s.i.) and 10 seconds at atmospheric pressure (scheme: 20s+10s) to optimize the high pressure level. Thereafter, the effect of the number of pressure cycles was examined by comparing 10, 60, and 90 cycles at the optimized pressure regimen, using the default scheme (20s+10s). Finally, 3 different cycling schemes (10s+10s, 20s+10s, 50s+10s) with optimized pressure and number of cycles were tested to further optimize the extraction conditions. The optimized PCT-MicroPestle-assisted tissue disruption protocol was then compared with the conventional PCT method described previously.13 Samples for PCT-MicroPestle lysis were sonicated in the microTubes 3 times, 10 seconds each (Branson 5510, USA). After sonication, sample lysis and protein extraction were performed with PCT-MicroPestle using 60 pressure cycles, each consisting of 20 seconds of 45,000 p.s.i. high pressure and 10 seconds of atmospheric pressure, at 33°C in the Barocycler. The supernatant was collected for BCA protein assay. Thereafter, proteins were digested right away or stored in -80°C. For protein reduction and alkylation, Tris (2-carboxyethyl) phosphine (10 mM) and iodoacetamide (40 mM) were added to the solution for 30-minute incubation in the dark at room temperature. After replacing PCT-MicroPestle with PCT-MicroCap (size 50 µl), samples were diluted with 0.1M ammonium bicarbonate to ~40 µl to reduce the urea concentration to 6 M prior to digestion with Lysyl endoproteinase Lys-C (mass spectrometry grade, Wako, Richmond, VA, USA, Lot CTQ0582) at an enzyme to substrate ratio of 1:40. Digestion was performed in the Barocycler at 33°C using 45 cycles, each consisting of 50 seconds at 20,000 p.s.i. high pressure and 10 seconds at atmospheric pressure. The urea concentration in the samples was then further diluted with 0.1M ammonium bicarbonate to 1.6 M urea by adjusting the total volume to ~150 µl per tube. Trypsin (sequencing grade modified, Promega, Madison, WI, USA, Lot 0000124422, enzyme to substrate ratio 1:20) digestion was carried out in the Barocycler with PCT-MicroCap (size 150 µl) under the condition of 90 cycles, each consisting of 50 seconds at 20,000 p.s.i. high pressure and 10 seconds at atmospheric pressure at 33°C. Thereafter, digestion was stopped by 10% trifluoroacetic acid (TFA, sequencing grade, Prod. No 28904, Thermo Fisher, MA, USA). The peptides were cleaned with SEPPAK C18 cartridges (Waters Corp., Milford, MA, USA), dried and stored in -80°C before SWATHMS analysis.13 For samples disrupted using the previously published conventional PCT method,13 the PCTMicroCaps were used in both sample lysis and protein digestion steps. The long caps (50 µl volume) used in the initial lysis step were replaced with shorter caps (100 and 150 µl) to accommodate the 4



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 24

increased sample volume during the digestion steps. Prior to PCT, samples were sonicated as described above. After sonication, sample lysis and protein extraction were performed using 60 pressure cycles, each consisting of 50 seconds of 45,000 p.s.i. high pressure and 10 seconds of atmospheric pressure, at 33°C in Barocycler. After quantification by BCA protein assay, proteins were digested or stored in -80°C. Protein reduction and alkylation were carried out as described above for the samples generated with the MicroPestle method. Samples were then diluted with 0.1M ammonium bicarbonate to ~80 µl to bring the urea concentration down to 6 M prior to digestion with Lys-C in the Barocycler. The samples were then further diluted with 0.1M ammonium bicarbonate to ~300 µl with 2 microTubes (~150 µl per tube), to bring the urea concentration down to 1.6 M for Trypsin digestion. The conditions for Lys-C and Trypsin digestions were the same as previously described.13 Protein and peptide concentration measurement Protein concentrations were determined using a BCA protein assay kit (Thermo scientific, Wilmington, DE, USA, No 23225 23227) with Multi-Mode Microplate reader (SynergyTMHT, BioTek, USA). MSready peptides were dissolved in HPLC grade water with 0.1% formic acid and 3% acetonitrile and measured with NanoDrop 1000 spectrophotometer (Thermo scientific, Wilmington, DE, USA) at 280nm (1Ab=1mg ml-1). Triple-TOF MS analysis in SWATH mode Tissue samples of 2-4 mg were obtained from mouse liver, heart, brain, and human kidney. The samples were lysed and digested with PCT using either conventional or PCT-MicroPestle assisted homogenization, in duplicate. Clean total peptide (0.6 µg) from each sample was injected into an Eksigent 1D+ Nano liquid chromatography (LC) systems (Eksigent, Dublin, CA, USA) for SWATHMS in duplicate. The solvents for peptide separation include buffer A (2% acetonitrile and 0.1% formic acid in water) and buffer B (98% acetonitrile and 0.1% formic acid in water). A flow rate of 300 nl min-1 with 2% to 35% buffer B was set for 120 min reverse phase gradient. A 5600 Triple-TOF mass spectrometer (AB Sciex, Concord, Canada) was used for SWATH-MS analysis. The isolation window was 26 Da (containing 1 Da for the window overlap) with an ion accumulation time of 100 milliseconds per SWATH. The precursor mass range (400-1200 Da) was covered by a set of 32 overlapping SWATH windows.10, 14 SWATH data analysis Mouse and human libraries used in this study were built by Shao et al.13 and Guo et al.10 SWATH wiff files were converted into mzML files using ProteoWizard, and the data were analyzed with OpenSWATH in the iPortal workflow.15 The false discovery rate for peptide identification was set below 0.1%. Brutus Cluster, the central resource for high-performance/high-throughput scientific computation at ETH Zurich, was used for computation. High quality peptide features from different 5


Page 7 of 24

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


samples were aligned with an in-house alignment tool.13 Label-free protein-levels were quantified using the function ProteinInference, which was performed with the R package aLFQ (version 1.3.1) on output from OpenSWATH. Statistical analysis Quantile normalization was used to correct systematic differences in expression levels across the samples.16 Regression plots, heatmaps, and volcano plots were produced with R. P-values were calculated using Student’s t-test.

Results Design of PCT-MicroPestle In our previously published method,10, 13 protein extraction from tissue samples was carried out in microTubes by alternating cycles of ultra-high pressure, without additional mechanical homogenization. These microTubes were used with PCT-MicroCaps (Figure 1), which are available in three different lengths to accommodate three sample volume increments (50 µl, 100 µl, and 150 µl). In the current study we tested a new device called PCT-MicroPestle (Figure 1A, B) to further improve the efficiency of the tissue homogenization and protein extraction steps of the overall protocol. The PCT-MicroPestle is made from Polytetrafluoroethylene (PTFE), a synthetic fluoropolymer of tetrafluoroethylene, also known as Teflon® (a brand of DuPont Co). This material is low binding, retains its shape well during pressure cycling, and provides an essentially incompressible surface against which tissue sample can be homogenized during pressure cycling. The PCT-MicroPestle is designed to form a tight fit between the pestle, the inner walls, and bottom of the tube. This homogenizer-like microTube/MicroPestle arrangement can accommodate up to 30µl of extraction reagent. As shown in Figure 1A, the tight fit of the MicroPestle traps the small tissue sample at the bottom of the tube. Hydrostatic pressure compresses the tube walls against the insert, causing the tube to collapse and shorten (Figure 1C). Repeated pressure cycles lead to efficient homogenization, as the tissue sample is repeatedly forced through the narrow gap between the tip of the MicroPestle and the tube walls. Optimization of lysis conditions with PCT-MicroPestle We initially optimized the pressure cycling parameters to integrate the PCT-MicroPestle into the PCTSWATH protocol. Optimization was performed using 2-3mg pieces of mouse liver tissue. First, we tested 30 cycles of 20 seconds of high pressure and 10 seconds of atmospheric pressure (20s+10s) at 3 different high pressure levels (25,000 p.s.i., 35,000 p.s.i., 45,000 p.s.i.) (Figure S1A). 45,000 p.s.i. is the maximum pressure that can be generated in the barocycler model used here. After PCT-assisted lysis, the solubilized proteins were measured by BCA assay. The specific protein yield was calculated 6



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 24

as protein amounts in µg per mg of wet liver tissue. The results suggested that the protein yield at 45,000 p.s.i. was significantly higher compared with either 25,000 p.s.i. (P10 mg), the yields converged nicely at lower scale (< 2mg) to about 80 µg protein per mg tissue. Integration of the PCT-MicroPestle with protein digestion workflow We next evaluated whether the improved protein yields can be translated into higher peptide yields after integration with protein digestion, using mouse heart, liver or brain tissues and human kidney tissue. Two to four mg of tissue were used because this sample size could be directly and accurately determined by weighing. We applied the optimized lysis condition established for the PCTMicroPestle, i.e. 60 cycles of 20s at 45,000 p.s.i. and 10s at atmospheric pressure in 30ul of lysis buffer. The conditions for the conventional PCT method were 60 cycles of 50s at 45,000 p.s.i. followed by 10s at atmospheric pressure in 60 µl of lysis buffer. The digestion conditions for these two methods were essentially the same,10, 13 except for the difference in volume as described in Materials and Methods. The workflow for the PCT-MicroPestle method is shown in Figure S2. The MS-ready peptide mass was quantified by Nanodrop. The specific peptide yield was calculated as peptide amount per mg of tissue (µg peptide/mg tissue). The data showed that the specific yield ranged from 24 to 61 µg of peptide per mg tissue when samples were lysed using PCT-MicroPestle method, and from 19 to 45 µg per mg tissue in samples lysed by the conventional PCT method (Figure 3A). The PCT-MicroPestle method significantly increased the yield by around 19-36% when compared to conventional PCT method (Figure 3B). Furthermore, based on the data in Figure 3A, we calculated the coefficient of variation (CV) values of specific peptide yield to evaluate the reproducibility of PCT-assisted sample preparation. As shown in Figure 3C, the median CV for the samples generated by the PCT-MicroPestle method (mouse liver, heart, brain, and human kidney, each in triplicate, n=12) was 2.7%, lower than the median CV for the samples generated by conventional PCT method (median CV: 6.8%, n=12). We did not observe statistical difference (P=0.63, n=12) between these two methods, indicating that the reproducibility of the PCT-MicroPestle method was comparable to the conventional PCT method. Additionally, the value of the CV for conventional PCT method in this study was markedly lower than that in our previous paper,13 which is probably attributable to the improved estimation of peptide yields. Here, we calculated the CV according to the specific peptide yield normalized to tissue starting mass (µg 8



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 24

peptide per mg tissue), rather than the raw peptide mass, thus minimizing the variation caused by any differences in the amount of tissue mass per sample. To quantitatively analyze the peptides and proteins by mass spectrometry, 0.6 µg of the peptide material from each sample was subjected to SWATH-MS in duplicate. Sixteen SWATH maps were obtained from the three types of mouse tissue and human kidney tissue prepared by PCT-MicroPestle and conventional PCT method, in duplicates (Table S2-17). After OpenSWATH analysis, about 10,000 proteotypic peptides from about 1,500-2,000 Swiss-Prot proteins were detected in the tissues (Figure 4 A, B). The data from samples prepared by the conventional PCT method were consistent with the data from our previous study which further demonstrated the reproducibility of SWATH-MS analysis.13 The number of peptides/proteins was slightly higher in the samples extracted using the PCT-MicroPestle compared with the PCT-MicroCap method, suggesting that improved tissue homogenization by the MicroPestle may lead to better proteomic identification coverage. We found that most proteins (90.39%, CV = 5.15%) identified by SWATH-MS were shared in samples prepared by both methods (Figure 4C). Based on Gene Ontology (GO) annotation, we further analyzed the cellular components and function of the nonoverlapping proteins identified in samples processed by the two methods, and found no significant difference (data not shown). Comparative analysis of the quantified proteome in samples generated by PCT-MicroPestle and the conventional PCT method We evaluated the reproducibility of the SWATH-based peptide and protein quantification based on the proteins commonly identified in samples prepared by different methods. The correlation coefficients (R2) of quantified proteins between runs across different sample types were calculated (Figure S3). We found that the average R2 was 0.98, indicating a high degree of reproducibility of the workflow. To determine whether the proteins quantified by these two PCT methods are comparable, the proteins generated by the PCT-MicroPestle and conventional PCT methods were compared. We found that proteins were highly correlated and the R2 between the two methods ranging from 0.91 to 0.96 with an average R2 of 0.94±2.24% (Figure S3). Our results illustrate that both methods led to comparable quantification for the proteins commonly identified. To determine the similarity of the protein populations identified from the SWATH maps generated by PCT-MicroPestle and conventional PCT methods, we performed an unsupervised hierarchical clustering analysis (Figure S4). We observed that the overall proteomes generated from PCTMicroPestle method were consistent with those from conventional PCT method across different tissues. Additionally, proteins across different tissue samples were analyzed by volcano plots (Figure S5). Proteins showing a notable difference in abundance in samples extracted using the two different methods were defined as fold change higher than 2 (PCT-MicroPestle/PCT-MicroCap) with P value lower than 0.05 (Student’s t-test between PCT-MicroPestle and PCT-MicroCap). Such difference 9


Page 11 of 24

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


meant that they were extracted more or less efficiently by the PCT-MicroPestle compared to the conventional PCT method. We found that few proteins (less than 5% of whole measured proteome within each sample type) showed significant difference between the PCT-MicroPestle and the conventional PCT method, indicating that the measured proteomes, identified from the same amount of peptide injected into the mass spectrometer, were quantitatively comparable in samples extracted using the two methods.

Discussion High throughput molecular profiling of clinical tissues with high accuracy and reproducibility is essential for personalized medicine and biomarker identification.17 We have recently developed an integrated protocol, namely the PCT-SWATH pipeline (the conventional PCT method, aka “microCap method”), which performs protein extraction and digestion under hydrostatic pressure in sample tubes capped by a MicroCap.10, 13 With the Barocycler NEP2320-Enhanced instrument, up to 6 or 8 tissue samples, depending on the sample cartridge used, can be processed and digested into peptides in a single batch in 6-8 hours. Thereby the technical variation of sample preparation is minimized.13 Additionally, such extraction speed and simplified procedures make it possible to process a large number of samples in parallel, which is essential to clinical research. In this study, to further improve the performance of the PCT-SWATH method, we used a miniaturized homogenization device, the PCT-MicroPestle, and incorporated it into the workflow. The PCTMicroPestle is a disposable insert which produces a homogenizer-like effect under repeated pulses of high hydrostatic pressure in the Barocycler. We identified the optimized condition for tissue homogenization by the PCT-MicroPestle as a cycling scheme consisting of 60 cycles at 45,000 p.s.i. using a 20s+10s cycling profile. This lysis step can be completed in 30 minutes, reducing by 50% the time for tissue lysis compared to the conventional method (Figure S2).10, 13 Additionally, we used 60 µl of lysis buffer in the conventional protocol, but only 30 µl with PCT-MicroPestle. Thereby, the final reaction volume in the trypsin digestion step is reduced from 300 µl to only ~150 µl, which fits easily into a single MicroTube. Therefore, since the samples no longer need to be split into 2 microTubes, sixteen samples can be run concurrently in the NEP2320-Enhanced Barocycler, doubling the throughput of the earlier workflow. Therefore, using the PCT-MicroPestle method, the whole lysis and digestion workflow can be accelerated compared to the previous MicroCap method. As shown in Figure S2, using the new workflow, it is now practical to batch-process 12 or 16 samples, depending on the sample cartridge used, within 6 hours. In addition to the improved speed and throughput of the protocol, the yield of peptide is increased as well. The peptide yield generated by the PCT-MicroPestle method was increased by around 19%-36% compared to conventional PCT method. This confirmed that the design of MicroPestle significantly 10



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 24

improves tissue lysis and homogenization. Furthermore, our previous data showed that the conventional PCT method was more advantageous in handling minute tissue specimens but was less efficient when processing relatively larger tissue samples.13 Here, our data showed that PCTMicroPestle method substantially improved the extraction efficiency in relatively large pieces of heart and liver tissues, indicating that larger tissues, especially for fibrous and hard tissues, require more extensive homogenization compared to smaller pieces. Although relatively larger samples benefited more strongly from the PCT-MicroPestle method due to the design of the homogenizer, the specific protein yield was still lower than when smaller samples were used. This suggests that for relatively larger (~10mg range) samples, splitting the tissue into smaller pieces before PCT homogenization and lysis is likely to be beneficial. Furthermore, to evaluate the overall reproducibility of the PCT-SWATH pipeline, we first calculated the CV of peptide yield extracted with these two PCT methods. The median CV was about 3% for the MicroPestle method and about 6% for the MicroCap method. These data demonstrate that the PCT method can reproducibly generate peptides from various types of samples and that the PCTMicroPestle method further improves the results. Subsequently, 0.6 µg of MS-ready peptides was injected to SWATH-MS to generate a comprehensive and permanent digital record in the form of fragment ion spectra. We then evaluated the reproducibility of the SWATH-based protein quantification. The correlation coefficient of quantified proteins between samples was around 0.98 on average, indicating a high degree of reproducibility of the workflow. OpenSWATH analysis revealed that most of the proteins appearing in the spectral libraries can be identified. Interestingly, the optimized PCT-MicroPestle method produced a higher number of peptides and yielded a more complete set of analytes. This may be attributed, at least in part, to more uniform and full disruption of the sample matrix. Proteomes generated from these two PCT methods were further analyzed by heatmap and volcano plots. We found that proteins quantified from samples extracted by the PCT-MicroPestle method were highly correlated with proteins quantified from conventional PCT method as shown by heatmap. The proteome patterns between these two methods were almost identical, although some clusters exhibited small variation in signal intensity. Furthermore, proteins showing notably different abundance between MicroPestle and MicroCap methods were analyzed with volcano plots. These proteins that demonstrated significant differences accounted for less than 5% of the whole measured proteome across the samples, further demonstrating the comparability of these two methods. Cellular components and functions of these differentially extracted proteins were analyzed by Gene Ontology (GO) annotation and no significant difference was identified. This novel method, while very effective, still has some limitations. The lysis buffer volume cannot exceed 30 µl, which limits the sample size to less than 10mg tissue. Larger tissues would need to be divided into smaller pieces and split into several tubes when applying this workflow. In this study, 11


Page 13 of 24

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


mechanical lysis and chemical lysis (high concentration urea) were combined. In future work, we plan to test the effect of other buffer system on the efficiency and quality of the PCT-MicroPestle extraction method. In addition, it would be interesting to analyze the advantage of this well-controlled PCT-MicroPestle method on macromolecular and posttranslational modification (PTM) preservation, which may extend the application of this novel method in biomedical research. It has been suggested that traditional mechanical disruption, if carried out too long or too vigorously in an attempt to obtain more thorough tissue disaggregation, may lead to undesirable impairment of PTM (e.g. phosphorylation)18 of the extracted proteins, and may harm the structural integrity of other proteinsof-interest. Concluding remarks Overall, the new PCT-MicroPestle-PCT-SWATH method is an advanced proteomic pipeline due to its reduced time consumption and highly reproducible results, at both the peptide extraction and proteomic analysis levels. With this method, comparative quantitative analysis of large number of clinical samples can be performed with higher throughput and efficiency than existing methods. We suggest that the optimal specimen mass is less than 3-5 mg for a range of tissues including soft tissues (e.g. brain tissue) and fibrous tissue (e.g. heart tissue). Additionally, peptides extracted by PCTMicroPestle method can be potentially analyzed by other protein identification/quantification methods, such as Western blot, ELISA and shotgun proteomics, etc, making it a versatile tool for homogenization of small solid tissue samples. Therefore, we expect wide applications of this method in systems biology, systems medicine and translational studies.

Acknowledgements: We thank L Gillet for help in SWATH analysis and the ETH Brutus team for computational support. The work was supported by the SystemsX.ch project PhosphoNetXPPM (to R.A.), the Swiss National Science Foundation (grant no. 3100A0-688 107679 to R.A.), the European Research Council grants no. ERC-2008-AdG 233226 and ERC-20140AdG 670821 to R.A.), and National Natural Science Foundation of China (81100581 to S.S.). The authors have declared no competing financial interest. VG and AL are employees of Pressure BioSciences, Inc.

Supporting Information Supplementary information is available on the Journal of Proteome Research website. Table S1, Table S2-17, and Figure S1-S5 12



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 24

References 1. Singleton, C., Recent advances in bioanalytical sample preparation for LC-MS analysis. Bioanalysis 2012, 4, (9), 1123-40. 2. Hansen, S. H.; Stensballe, A.; Nielsen, P. H.; Herbst, F. A., Metaproteomics: Evaluation of protein extraction from activated sludge. Proteomics 2014, 14, (21-22), 2535-9. 3. Gottlieb, B.; Beitel, L. K.; Trifiro, M., Changing genetic paradigms: creating next-generation genetic databases as tools to understand the emerging complexities of genotype/phenotype relationships. Human genomics 2014, 8, 9. 4. Dounce, A. L.; Witter, R. F.; Monty, K. J.; Pate, S.; Cottone, M. A., A method for isolating intact mitochondria and nuclei from the same homogenate, and the influence of mitochondrial destruction on the properties of cell nuclei. The Journal of biophysical and biochemical cytology 1955, 1, (2), 139-53. 5. Olszowy, P. P.; Burns, A.; Ciborowski, P. S., Pressure-assisted sample preparation for proteomic analysis. Analytical biochemistry 2013, 438, (1), 67-72. 6. Powell, B. S.; Lazarev, A. V.; Carlson, G.; Ivanov, A. R.; Rozak, D. A., Pressure cycling technology in systems biology. Methods in molecular biology 2012, 881, 27-62. 7. Freeman, E.; Ivanov, A. R., Proteomics under pressure: development of essential sample preparation techniques in proteomics using ultrahigh hydrostatic pressure. Journal of proteome research 2011, 10, (12), 5536-46. 8. Lopez-Ferrer, D.; Petritis, K.; Robinson, E. W.; Hixson, K. K.; Tian, Z.; Lee, J. H.; Lee, S. W.; Tolic, N.; Weitz, K. K.; Belov, M. E.; Smith, R. D.; Pasa-Tolic, L., Pressurized pepsin digestion in proteomics: an automatable alternative to trypsin for integrated top-down bottom-up proteomics. Molecular & cellular proteomics : MCP 2011, 10, (2), M110 001479. 9. Gross, V. S.; Greenberg, H. K.; Baranov, S. V.; Carlson, G. M.; Stavrovskaya, I. G.; Lazarev, A. V.; Kristal, B. S., Isolation of functional mitochondria from rat kidney and skeletal muscle without manual homogenization. Analytical biochemistry 2011, 418, (2), 213-23. 10. Guo, T.; Kouvonen, P.; Koh, C. C.; Gillet, L. C.; Wolski, W. E.; Rost, H. L.; Rosenberger, G.; Collins, B. C.; Blum, L. C.; Gillessen, S.; Joerger, M.; Jochum, W.; Aebersold, R., Rapid mass spectrometric conversion of tissue biopsy samples into permanent quantitative digital proteome maps. Nature medicine 2015. 11. Liu, Y.; Chen, J.; Sethi, A.; Li, Q. K.; Chen, L.; Collins, B.; Gillet, L. C.; Wollscheid, B.; Zhang, H.; Aebersold, R., Glycoproteomic analysis of prostate cancer tissues by SWATH mass spectrometry discovers N-acylethanolamine acid amidase and protein tyrosine kinase 7 as signatures for tumor aggressiveness. Molecular & cellular proteomics : MCP 2014, 13, (7), 1753-68. 12. Sajic, T.; Liu, Y.; Aebersold, R., Using data-independent, high resolution mass spectrometry in protein biomarker research: Perspectives and clinical applications. Proteomics Clin Appl 2014.

13


Page 15 of 24

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


13. Shao, S.; Guo, T.; Koh, C. C.; Gillessen, S.; Joerger, M.; Jochum, W.; Aebersold, R., Minimal sample requirement for highly multiplexed protein quantification in cell lines and tissues by PCTSWATH mass spectrometry. Proteomics 2015. 14. Gillet, L. C.; Navarro, P.; Tate, S.; Rost, H.; Selevsek, N.; Reiter, L.; Bonner, R.; Aebersold, R., Targeted data extraction of the MS/MS spectra generated by data-independent acquisition: a new concept for consistent and accurate proteome analysis. Molecular & cellular proteomics : MCP 2012, 11, (6), O111 016717. 15. Rost, H. L.; Rosenberger, G.; Navarro, P.; Gillet, L.; Miladinovic, S. M.; Schubert, O. T.; Wolski, W.; Collins, B. C.; Malmstrom, J.; Malmstrom, L.; Aebersold, R., OpenSWATH enables automated, targeted analysis of data-independent acquisition MS data. Nature biotechnology 2014, 32, (3), 21923. 16. Bolstad, B. M.; Irizarry, R. A.; Astrand, M.; Speed, T. P., A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics 2003, 19, (2), 185-93. 17. Shao, S.; Guo, T.; Aebersold, R., Mass spectrometry-based proteomic quest for diabetes biomarkers. Biochim Biophys Acta 2014. 18. Larance, M.; Lamond, A. I., Multidimensional proteomics for cell biology. Nature reviews. Molecular cell biology 2015, 16, (5), 269-80.

14



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 24

Figure Legends Figure 1. Design of PCT-MicroPestle. (A) Appearance of PCT-MicroPestle and microTube at ambient pressure (MicroPestle shown in black, microTube shown in gray). White area represents the compressible volume (air, buffer and tissue). (B) Under high hydrostatic pressure, the microTube is compressed, causing it to narrow and shorten. (C) Appearance of PCT-MicroPestle (right) and long PCT-MicroCap (left, shorter caps were used when sample was diluted for enzyme digestions) inserted into microTubes.

Figure 2. Protein extraction comparison between PCT-MicroPestle and conventional (MicroCap) PCT method as a function of sample size. Aliquots of mouse liver, heart, and brain tissues were lysed using the PCT-MicroPestle or the conventional PCT method. All samples were processed with 3 biological replicates. The extracted protein amount was measured with BCA. The specific yield was calculated as µg protein per mg of tissue wet weight (µg protein/mg tissue) for liver (A), heart (B), and brain (C). The percentage of increase in specific protein yield by the PCT-MicroPestle over the conventional PCT method was calculated (D, E, F). Data are presented as the mean ± SD of three biological replicates. *P

Reproducible Tissue Homogenization and Protein ... - ACS Publications

Recommend Documents