alexa Comparison of Two Solid-Phase Extraction (SPE) Methods for the Identification and Quantification of Porcine Retinal Protein Markers by LC?MS/MS

ISSN: 2379-1764

Advanced Techniques in Biology & Medicine

Reach Us +44-1764-910199

Comparison of Two Solid-Phase Extraction (SPE) Methods for the Identification and Quantification of Porcine Retinal Protein Markers by LC?MS/MS

Carsten Schmelter#, Sebastian Funke#, Jana Treml, Anja Beschnitt, Natarajan Perumal, Caroline Manicam, Norbert Pfeiffer and Franz H Grus*
Department of Experimental and Translational Ophthalmology, University Medical Center, Johannes Gutenberg University, Mainz, Germany
#Contributed equally to this work
*Corresponding Author: Franz H Grus, Department of Experimental and Translational Ophthalmology, University Medical Center, Johannes Gutenberg University, Mainz, Germany, Tel: +49 6131/173328 , Fax: +(49) 6131/175509, Email: [email protected]

Received Date: Jul 19, 2018 / Accepted Date: Aug 03, 2018 / Published Date: Aug 10, 2018

Abstract

Proper sample preparation protocols represent a critical step for liquid chromatography-mass spectrometry (LC-MS)-based proteomic study designs and influence the speed, performance and automation of high-throughput data acquisition. Main objective of this study was to compare two commercial Solid-Phase Extraction (SPE)-based sample preparation protocols (comprising SOLAμTM HRP SPE spin plates from Thermo Fisher Scientific and ZIPTIP® C18 pipette tips from Merck Millipore) for analytical performance, reproducibility and analysis speed. The house swine (Sus scrofa domestica) represents a promising animal model for studying human eye diseases including glaucoma and provides excellent requirements for the qualitative and quantitative MS-based comparison in terms of ocular proteomics. In total 6 technical replicates of two protein fractions (extracted with 0.1% dodecyl-ß-maltoside (DDM) or 1% trifluoroacetic acid (TFA)) of porcine retinal tissues were subjected to in-gel trypsin digestion and purified with both SPE-based workflows (N=3) prior LC-MS/MS analysis. On average both protein fractions (DDM and TFA) provided the identification of 550 ± 70 and 305 ± 48 proteins after ZIPTIP® purification protocol and SOLAμTM workflow resulted in the detection of 513 ± 55 and 300 ± 33 proteins (FDR<1 %). Furthermore, Venn diagram analysis revealed an average overlap of 65 ± 2% (DDM fraction) and 69 ± 4% (TFA fraction) in protein identifications between both SPE-based methods. Quantitative analysis of 24 glaucoma-related protein markers also showed no significant differences (P>0.05) regarding protein recovery between both SPE methods. However, only glaucoma protein marker methyl-CpG-binding protein 2 (MECP2) showed a significant (P=0.02) higher abundance in ZIPTIP®-purified replicates in comparison to SOLAμTM-treated study samples. Nevertheless, this result was not confirmed by in-gel trypsin digestion of recombinant MECP2 (P=0.24). In conclusion, both SPE-based purification methods worked equally well in terms of analytical performance and reproducibility, whereas the analysis speed and the semi‑automation of the SOLAμTM spin plates workflow is much more convenient in comparison to the manual ZIPTIP® C18 pipette tip protocol.

Keywords: Mass spectrometry; Glaucoma animal model; Biomarkers; Sample clean-up; ZIPTIP® C18 pipette tips; SOLAμTM HRP SPE spin plates

Introduction

Solid Phase Extraction (SPE) is a crucial technique in Mass Spectrometry (MS)‑based protein biomarker recovery, and therefore, the choice of appropriate SPE techniques has to be planned carefully. Robustness, reproducibility, sensitivity and economic parameters encompassing time and costs have to be addressed along with the selection of proper SPE protocols. There is an increasing number of SPE sorbent materials feasible for MS based proteomics in the current market, which compromise various formats e.g. pipette tips, cartridges, discs, multi‑well‑plates, magnetic beads and sorbent materials, as reviewed elsewhere [1-7]. Since the implementation of SPE in MS workflows should reduce the laboratory expenses and work time while enhancing analysis speed, reproducibility and minimizing processing errors, an important feature of a SPE technique is its ability for automation in compliance with high-throughput proteomics methodology. Moreover, the SPE technique should also be compatible with the biological material of interest, in particular retinal tissue in the present study. In the field of glaucoma proteomics there is a growing demand to study the retina, which represents the primary ocular site affected in this neurodegenerative disorder that is indicated by aberrant proteomic alterations [8]. To study these proteomic alterations “bottom up” (BU)‑MS workflows represent the state-ofthe- art strategy providing sensitive and semi‑quantitative information of the identified protein species [9-11]. Especially for Electrospray Ionization (ESI) MS BU workflows, SPE is recommended for desalting, ionic detergent-removal (e.g. SDS) and peptide enrichment prior LCMS/ MS analysis and is mandatory for proper peptide ionization. So far, several different kinds of devices and sorbents have been developed and commercialized during the last years in order to provide proper sample clean-up [12]. BU high performance liquid chromatography ESI tandem mass spectrometry (BULCMS) was successfully used for the in‑depth proteomic characterization of retina samples from rat [13,14], pig [15] and human [16]. Also, the BULCMS workflow has been used for the proteomic analysis of a recently established porcine retina organ culture [17] in our laboratory, which allows the monitoring of retinal proteins under experimental glaucomatous condition. In the former mentioned retina proteomic studies, ZIPTIP® C18 pipette tip SPE was used as standard operation protocol (SOP) for sample clean-up prior LC‑MS/MS analysis. The purpose of the present work was to evaluate the suitability of a less laborious SPE method to be implemented in the established BULCMS workflow. Due to that reason, it was aimed forward to compare the performance of SOLAμTM HRP SPE spin plates with the ZIPTIP® C18 pipette tips regarding the MS-based analysis of the porcine retinal proteome. Furthermore, main focus was on specifically selected retinal proteins, which were recently found to be associated with the pathophysiology of glaucoma [16] and represent interesting marker candidates for quantitative protein recovery. Both SPE techniques will be evaluated for specific protein/peptide recovery performance as well as analysis speed of the retinal porcine proteome and possibly considering an (semi-)automated integration of both SPEbased methods.

Methods & Materials

Sample preparation and protein extraction protocols

Retina tissues were prepared from freshly enucleated eye bulbs (N=45) from domestic swine Sus scrofa domestica Linnaeus, 1758 individuals (sacrificed at 3-6 mon, female: male=3:2) provided by local slaughterhouses (Landmetzgerei Harth, Stadecken-Elsheim, Germany; Metzgerei Köppel, Mainz, Germany). Eye bulbs were equatorially opened with a scalpel to remove lens, vitreous body, iris and ciliary body. Retina tissue was carefully removed from pigment epithelium with a PBS coated paintbrush. After this, the whole retina was cut off from the optic nerve head and stored at -80°C until further protein extraction protocols. For the first protein extraction protocol 30 isolated retina tissues were pooled and homogenized with an Ultra- Turrax T25 sonicator (Janke & Kunkel IKA Labortechnik, Staufen im Breisgau, Germany) as described in detail in an earlier study [15]. After homogenization, 0.1% dodecyl-ß-maltoside (DDM) buffer was added to the aliquots and sonicated for 10 min on ice. Afterwards the samples were gently mixed and incubated for 30 min at RT followed by centrifugation at 10.000 g for 12 min at 4°C. The supernatant was collected and stored at ‑20°C. For the second protein extraction protocol 15 isolated retina tissues were pooled and subjected for further homogenization using a Precellys® 24 homogenizer (VWR International GmbH, Darmstadt, Germany) combined with a 1.4/2.8 mm Precellys® Ceramic kit (VWR International GmbH, Darmstadt, Germany). Prior homogenization, frozen retina tissues were added to the tubes containing the 1.4/2.8 mm ceramic balls and filled with 1.5 ml phosphate-buffered saline (PBS). Retina samples were homogenized three times for 45 s at 5,000 rpm and centrifuged afterwards at 10.000 g for 12 min at 4°C. The supernatant containing mostly cytoplasm-derived proteins was stored at -20°C and the remaining pellet was resuspended in 500 μl of 1 % TFA. Funke et al. (2016) [15] have shown previously that 1% TFA is a suitable buffer for the extraction of many nucleus-derived retinal proteins. Retina samples were homogenized three times for 45 s at 5.000 rpm followed by centrifugation at 10.000 g for 12 min at 4°C. The supernatant containing mostly nucleus-derived proteins was stored at ‑20°C. Protein measurements of both protein fractions were performed using BCA protein Assay Kit (Thermo Fisher Scientific, Rockford, USA) according to manufacturer’s introduction and measured three times with a Multiscan Ascent photometer (Thermo Fisher Scientific, Rockford, USA) at a wavelength of 570 nm. Human Recombinant protein MECP2 (Methyl-CpG-binding protein 2, cat. no. 14-1067) for validation experiment was purchased from Millipore (Billerica, USA).

1-D SDS Page

Both protein fractions (50 μg per lane) were separated on 10-well NuPAGE 12% Bis-Tris minigels (Thermo Fisher Scientific, Rockford, USA) under reducing conditions. Gels were incorporated in the XCell SureLock™ Mini-Cell Electrophoresis System (Invitrogen, Carlsbad, USA) and prepared with NuPAGE™ MOPS SDS Running Buffer 20X (Thermo Fisher Scientific, Rockford, USA) according to the supplier’s protocol. In addition, as molecular weight reference 10 μl of the SeabluePlus 2 Pre-Stained Protein Standard (Thermo Fisher Scientific, Rockford, USA) was added and separated at 150 V for 1.5 h at 4°C. After separation, gels were fixed and stained using Novex Colloidal Blue Staining Kit (Thermo Fisher Scientific, Rockford, USA) according to the manufacturer’s introduction. Gels were destained for at least 16 h and scanned using an Epson Perfection V600 Photo Scanner (Seiko Epson Corporation, Suma, Nagano, Japan) at 700 dpi. Protein migration pattern were manually inspected and subjected for further In-gel trypsin digestion.

In-gel Trypsin Digestion

In total six lanes of both protein fractions were subjected for further In-gel trypsin digestion according to a modified protocol from a previous study [18]. At first, each protein lane was subdivided into 17 slices according to their characteristic protein migration profile and cut into small pieces. The gel pieces were destained with 100 mM ammonium bicarbonate (ABC) in 50 % acetonitrile (ACN) and afterwards dehydrated with pure ACN before further reduction and alkylation process. The samples were evaporated in the Speedvac (Eppendorf, Darmstadt, Germany) for 10 min at 30°C until dryness. Afterwards the gel pieces were resolved with 10 mM dithiothreitol (DTT) in 100 mM ABC and incubated for 30 min at 56°C followed by incubation with 55 mM iodoacetamide (IAA) in 100 mM ABC for 30 min at RT in the dark. Then the samples were dehydrated one more time with pure ACN and dried for 10 min in the SpeedVac at 30°C until dryness. The reduced and alkylated proteins were further digested with 10 μg/ml sequencing grade trypsin (Promega, Madison, USA) in 10 mM ABC 10% ACN and incubated overnight for at least 16 h at 37°C. Next day, the supernatant was collected and the tryptic peptides were extracted with 10% formic acid (FA) 70% ACN for 30 min at 350 rpm. Both supernatant fractions were pooled und dried until dryness in the SpeedVac at 30°C.

SPE-based peptide purification

At first, the extracted peptides were dissolved in 20 μl 0.1 % TFA. In order to reduce technical variability during execution of the sample preparation protocols, the slices (N=17) of two lanes (DDM or TFA protein fractions) were pooled to a total volume of 40 μl 0.1 % TFA. Then the peptide pools were subsequently split into two equal amounts (each 20 μl) and subjected for further SPE purification via SOLAμ™ SPE HRP plates (Thermo Fisher Scientific, Rockford, USA) or via C18 SPE pipette tips (Millipore, Billerica, USA). In total 3 technical replicates of each protein fraction (DDM or TFA) were purified with both SPE methods. Purification via ZIPTIP® C18 SPE pipette tips represents the standard operation protocol (SOP) in our lab and was performed according to previous publications [9-11]. In brief, the ZIPTIP® C18 SPE pipette tips were conditioned and equilibrated by pipetting 10 μl ACN three times followed by 10 μl 0.1% TFA three times. Then the sample was loaded on the stationary C18 phase by aspirate and dispense 20 times, washed three times with 0.1% TFA and finally eluted two times in 10 μl 50 % ACN 0.1% TFA. This procedure was repeated, the pooled eluate fractions (40 μl) were evaporated in the Speedvac at 30°C until dryness and stored at ‑20°C. The SPE-based peptide purification via SOLAμ™ SPE HRP plates was performed according to the manufacturer’s introduction. Activation and elution of the SOLAμTM SPE membranes was performed with 100 μl methanol (MeOH), whereas the washing step was performed with 100 μl 5% MeOH. After each step of the sample preparation protocol the SOLAμTM SPE plate was centrifuged at 4.000g for 3 min and the flow‑through/eluate fractions were collected in 96- Well microtiter microplates (Costar Corning Incorporated, Corning, USA). Eluates were transferred into new reaction tubes, evaporated in the Speedvac at 30°C until dryness and stored at -20°C prior further LC‑MS/MS analysis.

LC-MS/MS analysis

LC-MS measurements were performed by a Rheos Allegro pump (Thermo Fisher Scientific, Rockford, USA) downscaled to a capillary HPLC system (flow rate: 6.7 ± 0.3 μl/min) online coupled to a hybrid linear ion trap- Orbitrap MS (LTQ Orbitrap XL; Thermo Fisher Scientific, Rockford, USA). Purified tryptic peptides were resolved in 10 μl 0.1 TFA and 6 μl were injected into a BioBasic® C18 column system (30 x 0.5 mm pre-column+150 x 0.5 mm analytical column; Thermo Fisher Scientific, Rockford, USA). Solvent A consists of 1.94 % ACN, 0.06 % MeOH, and 0.05% FA in water and solvent B consists of 95% ACN, 3 % MeOH and 0.05% FA in water. Peptides were eluted within 50 min using following gradient program: 15–20% B (0‑2 min), 20–60% B (2–35 min), 60–100% B (35–40 min), 100–0% B (40–45 min) and 0% B (45–50 min). The LTQ Orbitrap operated in positive ionization mode and data-dependent acquisition (DDA) mode: High-resolution survey full scan (from m/z 300 to 2000) was performed in the Orbitrap with a resolution of 30.000 at 400 m/z and the target automatic gain control was set to 1 x 106 ions. For internal calibration the lock mass was set to 445.120025 m/z (polydimethylcyclosiloxane). Dynamic exclusion mode was enabled with the following settings: repeat count=1, repeat, duration=30 s, exclusion list size=100, exclusion duration=90 s and exclusion mass width = ± 20 ppm. Based on the high‑resolution MS scan the 5 most intense ions were selected for further collision induced dissociation (CID) fragmentation in the ion trap employing normalized collision energy of 35%. Manual inspection of the total ion current (TIC) chromatogram was performed using Qual Browser v. 2.0.7 SP1 (Thermo Fisher Scientific, Rockford, USA).

Peptide identification and quantification

For protein identification acquired LC-MS profiles were analyzed with software package Proteome Discoverer (Version 1.1; Thermo Fisher Scientific, Rockford, USA) using the Mascot search engine (version 2.2.07) to obtain peptide scoring information. Tandem MS spectra were searched against SwissProt database (SwissProt_150301) with a combination of Homo sapiens and Sus scrofa as taxonomies with following settings: peptide mass tolerance of ± 30 ppm in the range of 150–2000 Da, fragment mass tolerance of ± 0.5 Da, tryptic cleavage, a maximum of one missed cleavages, carbamidomethylation as fixed modification, acetylation (Protein N‑terminal) and oxidation as variable modification. Output data were filtered considering stringent conditions (mascot significance threshold filter p<0.01 and false discovery rate (FDR)<1 %). Label-free quantification (LFQ) of the proteins was performed with MAXQuant computational proteomics platform version 1.5.2.8 (Max Planck Institute of Biochemistry, Martinsried). Tandem MS spectra were searched against a user‑defined database, containing glaucoma-associated proteins (SwissProt_170531), with the previously described database search settings. In addition, MAXQuant specific feature “match between run” was enabled and proteins were identified considering FDR<1%. Glaucoma-associated protein database contains a selection of retinal protein sequences which showed at least a tendency (P<0.1) to be differentially expressed between glaucoma patients and healthy controls according to a recent publication [16].

Data analysis

Graphical presentation and t-test statistics for the qualitative analysis of the MS output data was performed with software package Statistica version 13 (Statsoft, Tulsa, USA). Normal distribution of the data-sets was verified by Shapiro-Wilk test. Student’s t-test was applied for parametric data and Mann-Whitney U-test for non-parametric data. Venn diagram analysis was performed with statistics program R version 3.2.0 with VennDiagram package version 1.6.17 (https://www.r-project. org/). Combined protein lists from the SOLAμTM-and ZipTip-purified fractions were subjected for Gene Ontology (GO) analysis using software program Cytoscape version 2.8.3 with BINGO 2.44 plugin (www. cytoscape.org). Thereby, the GO category “cellular component” was screened for potential annotations. Statistical analysis of the MaxQuant generated output data was performed using software package Perseus version 1.5.5.0 (Max Planck Institute of Biochemistry, Martinsried). At first, LFQ intensities of the detected proteins were log2 transformed for further analysis [19]. Prior statistical analysis the output data were filtered for contaminants, reversed hits, “only identified by site” and for a minimum number of 3 valid values in at least in one group. Finally, two-sided t-test statistics with p values <0.05 was applied in order to identify significant level changes in protein abundances between both SPE-based purification methods (SOLAμTM and ZIPTIP®).

Results

1-D SDS page of porcine retinal proteins

Both protein fractions (extracted with 0.1% DDM or 1% TFA) showed a high degree of mass migration pattern congruency regarding SDS-PAGE (Figures 1A & 1B). Each lane was subdivided into 17 slices labeled with the most abundant proteins per slice. As already observable, both protein fractions showed an specific and reproducible protein pattern and particularly the DDM fraction contains predominantly cytoplasm-derived or membrane-associated proteins such as clathrin heavy chain 1 (≈ 150 kDa, highest protein score: 547, CLTC), retinol-binding protein 3 (≈98 kDa, highest protein score: 325, RBP3), glutathione S‑transferase P (≈ 30-20 kDa, highest protein score: 1948, GSTP1) or reticulon-4 (≈ 17 kDa, highest protein score: 558, RTN4). In contrast, the TFA fraction exhibits mostly nucleus‑derived proteins such as high mobility group protein B1 (≈28 kDa highest protein score: 818, HMGB1), different kinds of histones (e.g. Histone H1.4 ≈ 28-17 kDa, highest protein score: 3421, HIST1H1E), but also many mitochondrial (e.g. ATP synthase subunit beta, mitochondrial, ≈ 12 kDa, highest protein score: 1352, ATP5B) as well as ribosomal proteins (e.g. 60S ribosomal protein L12; ≈ 12 kDa, highest protein score: 589, RPL12).

advanced-techniqes-biology-medicine-proteins

Figure 1: 1-D SDS PAGE of porcine retinal proteins after extraction with 0.1% DDM or 1% TFA buffer solution labeled with the most abundant proteins per spot. Each sample lane contains a total protein amount of 50 μg. (A) Protein migration pattern of the 0.1% DDM extract mostly containing cytoplasm-derived or membraneassociated proteins. (B) Protein migration pattern of the 1% TFA extract predominantly consisting out of nucleus-derived proteins.

Qualitative comparison between both SPE-based peptide purification methods

In order to evaluate qualitative differences regarding protein identification rates, in total six technical replicates of two protein fractions (DDM and TFA) were purified with two different SPEprotocols (ZIPTIP® and SOLAμTM) prior LC-MS/MS analysis (as described in detail in 2.4). On average the DDM fractions provided the identification of 550 ± 70 proteins after ZIPTIP® purification protocol, whereas SOLAμTM purification protocol resulted in the identification of 513 ± 55 proteins using high-confident filtering criteria (FDR<1%). In contrast, the ZIPTIP® purification protocol leads to the detection of 305 ± 48 proteins in the TFA fraction, whereas SOLAμTM purification protocol identified 300 ± 33 proteins. No significant differences (P>0.05) regarding protein identifications were observed between both SPE-based peptide purification methods using Student’s t-test.

In the next step of the analysis, the identified proteins were classified according to their protein ion scores in order to evaluate if there are significant differences in the identification of group-specific proteins species (Figure 2). On average 7 ± 3 proteins of the SOLAμTM ‑purified DDM fraction showed a protein score >3000. Up to 318 ± 70 proteins indicated a protein score <3000 >70 and 186 ± 25 proteins were detected with a protein score <70. On the other hand, the ZIPTIP® procedure resulted in the detection 9 ± 3 proteins with a protein score >3000 in the DDM fraction. On average 359 ± 78 proteins showed a protein score <3000 >70 and up to 183 ± 48 proteins were identified with a protein score<70. Nevertheless, no significant group-specific differences (Mann-Whitney U-test, P>0.05) regarding the protein ion score were observed between both SPE methods. The proteins of the TFA fraction showed a similar distribution pattern regarding the protein ion score, but also no significant changes (Mann-Whitney U-test, P>0.05) were observed between both methods. Furthermore, we also clustered the identified peptides according to their characteristic molecular weights (MW), in order to check if there are significant differences in the detection of group-specific peptide sequences between both purification protocols (Figure 3). On average both SPE workflows resulted in comparable peptide identification rates (SOLAμTM: 1115 ± 149 peptides; ZIPTIP®: 1254 ± 156 peptides) within the mass range of 900 to 1800 Da in DDM fraction. With both SPE methods the number of identified peptides (SOLAμTM: 232 ± 31 peptides; ZIPTIP®: 258 ± 45 peptides) decreased at the edges of the mass range between 600-900 Da and >1800 Da. Nevertheless, none of the groups showed significant differences (Student’s t-test, P>0.05) between both purification methods in the DDM fraction. In addition, the MW-classified peptides of the TFA fraction showed a similar distribution pattern, but also no significant differences (Student’s t-test, P>0.05) were found between SOLAμTM - and ZIPTIP® -purified study samples.

advanced-techniqes-biology-medicine-classification

Figure 2: Distribution plot shows the classification of the identified proteins according to their specific Protein Ion Scores. Proteins were clustered to the Protein Ion Scores in the range of ≥ 3000, <3000 to ≥ 1000, <1000 to ≥ 800, <800 to ≥ 600, <400 to ≥ 200, <200 to ≥ 70 and <70. (A) Ion Score distribution plot of the identified proteins in the DDM fraction after enrichment with to different SPE-based purification methods (SOLAμTM and ZIPTIP®). (B) Distribution profile of the identified proteins in the TFA fraction according to their Ion Score after enrichment with to different purification methods.

advanced-techniqes-biology-medicine-peptides

Figure 3: Distribution plot shows the classification of the identified peptides according to their characteristic Molecular Weight (MW). Peptides were clustered in the molecular mass range of ≥ 1800 Da, <1800 to ≥ 1500 Da, <1500 to ≥ 1200 Da, <1200 to ≥ 900 Da and <900 to ≥ 600 Da. (A) Bar plot shows the MW distribution of the identified peptides in the DDM fraction after enrichment with to different SPE-based purification methods (SOLAμTM and ZIPTIP®). (B) Distribution profiles of the MW of the identified peptides in the TFA fraction after enrichment with to different purification methods.

Venn diagram analysis revealed that on average 65 ± 2 % of all identified proteins were detected with both SPE-based purification methods (SOLAμTM and ZIPTIP®) in the DDM fraction (Figure 4). Also the TFA fraction provided an average overlap of 69 ± 4 % of all identified proteins between both purification methods (Figure 5). Even on the peptide level both methods resulted in an overlap of 60 ± 1% in the DDM fraction and an overlap of 62 ± 5% in the TFA fraction (Figures 1 and 2). Furthermore, around 98% of all identified proteins in both protein fractions (DDM and TFA) could be annotated to cellular components regarding Gene Ontology (GO) analysis. As a further result, the identified proteins of the DDM fraction as well as TFA fraction did not show any conspicuous differences regarding the cellular localization between both SPE-based purification methods (Figure 6).

advanced-techniqes-biology-medicine-microtiter

Figure 4: Venn diagram showing the overlap (percentage distribution) of identified proteins in the DDM fraction between six technical replicates. Three technical replicates were either purified by ZIPTIP® pipette tips or SOLAμTM microtiter plates. On average 65 ± 2 % of all identified proteins were detected with both SPE-based purification methods.

advanced-techniqes-biology-medicine-purified

Figure 5: Venn diagram showing the overlap (percentage distribution) of identified proteins in the TFA fraction between six technical replicates. Three technical replicates were either purified by ZIPTIP® pipette tips or SOLAμTM microtiter plates. On average 69 ± 4 % of all identified proteins were detected with both SPE-based purification methods.

advanced-techniqes-biology-medicine-clustered

Figure 6: Gene Ontology (GO) annotation analysis of all identified porcine retinal proteins. Proteins were clustered according to their cellular components such as nucleus, mitochondrion, endoplasmatic reticulum, golgi apparatus, ribosome, lysosome or peroxisome. DDM (A) as well as TFA fraction (B) did not show any conspicuous differences regarding the cellular localization of the identified proteins between both SPE-based purification methods (SOLAμTM and ZIPTIP®).

Quantitative analysis between both SPE-based peptide purification methods

The aim of the second part this study was to identify if there are significant differences regarding label-free quantification (LFQ) of proteins between both purification methods (SOLAμTM and ZIPTIP®). For this quantitative analysis, a multitude of glaucoma-associated retinal biomarker candidates were selected, which showed distinctive level alterations (P<0.1) in retinal samples of glaucoma patients [16]. Up to 15 of these glaucoma-associated biomarker candidates were identified in the DDM fraction, whereas the TFA fraction contained up to 10 retinal biomarker candidates (Figure 7 and Table 1). The three most abundant marker candidates of the DDM fraction were elongation factor 1-alpha 1 (EEF1A1), ADP/ATP translocase 3 (SLC25A6) and ras-related protein Rab-11B (RAB11B). In contrast, the TFA fraction provided the three most abundant biomarker candidates histone H1.0 (H1F0), α‑crystallin B chain (CRYAB) and methyl-CpG-binding protein 2 (MECP2). Approximately 96 % of the identified biomarker candidates did not show any significant differences (Student’s t‑test, P>0.05) regarding LFQ analysis between both SPE-based purification methods. Only MECP2 showed a significant higher abundance (Student’s t-test, P=0.02) in ZIPTIP®‑ purified replicates in contrast to SOLAμTM -purified study samples. Verification experiment with tryptic in-gel digestion of recombinant protein MECP2 (Figure 3) resulted in a slightly higher abundance in the ZIPTIP®-purified study samples (LFQ intensity=11.5×107 ± 2.7×107, N=9) compared to SOLAμTMpurified replicates (LFQ intensity=10.1×107 ± 2.2×107, N=9), however no significant difference (Student’s t-test, p=0.24) was found.

advanced-techniqes-biology-medicine-clustered

Figure 7: Bar plot shows the label-free quantification results of 25 glaucoma-associated biomarker candidates after enrichment with two different SPE purification methods (SOLAμTM and ZIPTIP®). Marker candidates on the left were identified in the DDM fraction, whereas proteins on the right were detected in the TFA fraction (FDR<1%). Only protein MECP2 showed a significant (p<0.05) higher abundance in ZIPTIP®-purified replicates in comparison to SOLAμTM purified study samples.

Discussion

By BULCMS implementation of the two SPE-based methods for sample clean-up (ZIPTIP® C18 pipette tips or SOLAμTM HRP spin plates), the proteomic analysis provided a sensitive view into the complex porcine retina proteome in confidence with recent porcine protein catalogues [15,20-22]. The present study contributes to a better characterization of the pig eye proteome and promotes the house swine as suitable candidate for ocular disease model systems [23-27]. The retina “core proteome” was congruently illustrated with both SPE methods (see supplementary file 1), including the expression of abundant proteins such as clathrin heavy chain 1 (CLTC), pyruvate kinase isozymes M1/ M2 (PKM), creatine kinase B-type (CKB), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), rhodopsin (RHO), various tubulin chains and 14-3-3 proteins.. Moreover, subcellular compartments were comparably described with both, purification techniques, highlighting nucleus and mitochondria as the main annotated organelles. Also, both systems worked equally well concerning the reconstitution of retinal proteins from two different protein fractions: DDM- and TFA-containing solvents provided a high overlap regarding protein identifications between both purification techniques. Nevertheless, both SPE protocols also showed a certain degree of contrary indicated by ZIPTIP® C18 and SOLAμTM HRP exclusively detected protein species (e.g. ZIPTIP®: VAMP2 or SOLAμTM: HSP90AB1,). In accordance, the ZIPTIP® C18 method resulted in a higher number of protein identifications not included in the retina “core proteome” supported by tendencies of higher peptide ion scores and a higher sensitivity over the inspected mass range. However, since none of these effects was supported by statistical significance an approximately equal analytical performance and reproducibility of both SPE workflows can be concluded.

In terms of quantitative recovery of glaucoma-associated retinal target proteins, there were also no distinctive differences between both SPE methods. Whereas, for some focused retinal proteins, e.g. ADP/ ATP translocase 3 (SLC25A6), guanine nucleotide-binding protein G(i) subunit α-2 (GNAI2) or Ras-related protein Rab-11B (Rab11B), the SOLAμTM SPE method showed tendencies of quantitatively higher protein recovery, for other target proteins, e.g. high mobility group protein B1 (HMGB1) or α-crystallin B chain (CRYAB), the ZIPTIP® C18 technique achieved a trend to quantitatively higher protein abundances accompanied by higher peptide ion scores. Nevertheless, since all of these findings refer to tendencies an approximately comparable protein/peptide recovery performance of the two tested SPE methods towards the selected marker proteins can be assumed. Exceptions are the results regarding MECP2 extraction, since MECP2 was significantly (P=0.02) recovered in higher quantities from retina samples by the use of the manual ZIPTIP® C18 method. In confidence, this effect could be reproduced by tendency (P=0.24) regarding the extraction of tryptic, recombinant MECP2. The different performance in MECP2 extraction emphasizes the special retention behavior of this protein. As a member of intrinsically disordered proteins (IDPs), MECP2 lacks higher structural organization [28-30] providing high plasticity for molecular interactions [31], which is reflected by its amino acid sequence and also might influence retention behavior of MECP2-derived peptides on SPE sorbents, e.g. on HRP representing a polymeric sorbent with polar and non-polar retention properties [32].

With regard to effort, costs and analysis time the SOLAμTM HRP spin plate workflow is clearly a superior method compared the manual ZIPTIP® C18 protocol. Although the ZIPTIP® C18 method can be automated on robotic stations [33], the speed and robustness of such fully automated SPE pipette tip systems is hampered by the lack of resistance towards errors such as tip resin trapped air‑bubbles or sample debris blocked pipette tips. This error-prone procedure can be avoided by using SOLAμTM HRP spin plates due to the macro-porous structure of the solid phase material providing continuous solvent flow-through without any sample loss [34]. Furthermore, the factor time is crucial in fully automated SPE pipette tip platforms with regard to solvent diffusion, alterations in solvent composition, robot movement distances and robot working cycles. In accordance, the constant need for trouble shooting in such SPE pipette tip automation systems is an important limitation compared to centrifugal SPE spin plates. Indeed, the execution of the SOLAμTM HRP spin plate workflow still requires a certain degree of operator handling and represents only a semi‑automatic platform, nevertheless it provides much more standardization and reproducibility than the manual ZIPTIP® C18 pipette tip workflow. However, due to the semi-automatic system properties and the consequent compatibility to MS-based high-throughput screening strategies, centrifugal SPE spin plates have recently been “expected to become the mainstream method of sample processing in the future” [35] and have been successfully used for the BULCMS proteomic analysis of human cerebrospinal fluid [36], human serum [37], human lung tissues [38] and HeLa cells [39].

Conclusion

In conclusion, the SOLAμTM HRP spin plates represent an attractive alternative to the manual ZIPTIP® C18 pipette tip workflow and are particularly recommended for high‑throughput discovery proteomic approaches saving time, costs and operator effort. Nevertheless, considering the sensitive performance of the manually applied ZIPTIP® C18 tips, a targeted use of these pipette tips for the analysis of specific marker proteins such as SLC25A6, GNAI2, Rab11B or MECP2 should be addressed. In addition, the development of a mixed‑mode spin plate, benefiting from the properties of both stationary materials (HRP and C18), could be a future innovation for improved recovery of retinal proteins. Beyond that, fine-tuning adjustments of the SOLAμTM HRP spin plate workflow such as centrifugal speed, time, temperature, elution volume, elution buffer and number of repeat cycles could significantly improve the sensitivity and accuracy of the current proteomic measurements.

Author Contributions

C. Schmelter and S. Funke developed the study design, organized the experimental set-up and wrote the manuscript. J. Treml performed the experimental work as well as the mass spectrometric measurements and assisted in the data analysis. A. Beschnitt and N. Perumal participated in study design, proofread the manuscript and contributed important intellectual content. C. Manicam supported the manuscript revision and provided important intellectual input. N. Pfeiffer critically reviewed the manuscript and provided important intellectual input. F. Grus performed study coordination and study design, participated in review and approval of the manuscript.

Additional information

A list of all identified proteins is provided in supplementary file 1.

References

Citation: Schmelter C, Funke S, Treml J, Beschnitt A, Perumal N, et al. (2018) Comparison of Two Solid-Phase Extraction (SPE) Methods for the Identification and Quantification of Porcine Retinal Protein Markers by LC‑MS/MS. Adv Tech Biol Med 6: 262. DOI: 10.4172/2379-1764.1000262

Copyright: © 2018 Schmelter C, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Select your language of interest to view the total content in your interested language

Post Your Comment Citation
Share This Article
Relevant Topics
Recommended Conferences
Viewmore
Article Usage
  • Total views: 117
  • [From(publication date): 0-2018 - Oct 18, 2018]
  • Breakdown by view type
  • HTML page views: 98
  • PDF downloads: 19

Post your comment

captcha   Reload  Can't read the image? click here to refresh