Human tear fluid proteome dataset for usage as a spectral library and for protein modeling

This article provides a detailed dataset of human tear fluid proteins. Samples were fractionated by sodium dodecyl sulfate (SDS) gel electrophoresis resulting in 48 fractions that were spiked with an indexed retention time (iRT) peptide standard. These data are based on a data-dependent acquisition (DDA) mass spectrometric approach and can be used for example as a spectral library for tear fluid proteome analysis by data-independent acquisition (DIA). Moreover, the provided data set can be used with optimized HPLC and mass spectrometric settings for proteins/peptides of interest. Besides these aspects, this dataset can serve as a protein overview for gene ontology enrichment analysis and for modeling and benchmarking of multiple signaling pathways associated with the ocular surface in healthy or disease stages. The mass spectrometry proteomics data from the described workflow have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD011075.


a b s t r a c t
This article provides a detailed dataset of human tear fluid proteins. Samples were fractionated by sodium dodecyl sulfate (SDS) gel electrophoresis resulting in 48 fractions that were spiked with an indexed retention time (iRT) peptide standard. These data are based on a data-dependent acquisition (DDA) mass spectrometric approach and can be used for example as a spectral library for tear fluid proteome analysis by data-independent acquisition (DIA). Moreover, the provided data set can be used with optimized HPLC and mass spectrometric settings for proteins/peptides of interest. Besides these aspects, this dataset can serve as a protein overview for gene ontology enrichment analysis and for modeling and benchmarking of multiple signaling pathways associated with the ocular surface in healthy or disease stages. The mass spectrometry proteomics data from the described workflow have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD011075.
© 2019 The Authors. Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons. org/licenses/by/4.0/).

Data
The here presented proteomic dataset offers mass spectrometry data files generated from human tear fluid (see Fig. 1). The tear fluid was collected with Schirmer test strips. Afterwards, proteins were eluted, the protein concentration determined and proteins separated by SDS gel electrophoresis. The resulting lanes were cut into single bands followed by an in-gel digestion with trypsin. After peptide extraction iRT peptides were added to each sample and the samples were measured with a datadependent based mass spectrometric approach.

Experimental design, materials and methods
For a detailed workflow overview see Fig. 1.

Sample collection
Tear fluid was collected via Schirmer test (Haag-Streit UK Ltd, United Kingdom) without anesthesia from 20 healthy individuals. During this process, a small filter strip was placed inside the lower lid of the right and left eye, respectively. Then individuals closed their eyes and tears were collected for 5 min. After that, strips were removed. Samples were pooled and tear fluid was eluted from the strips with 10 mL solution buffer containing phosphate buffered saline (Thermo Fisher Scientific Inc., USA), 0.1% (v/v) Triton-X-100 (AppliChem GmbH, Germany), and an EDTA-free protease inhibitor tablet Specifications table

Subject area
Proteomics More specific subject area DIA based mass spectrometric analysis of human tear fluid Type of data Raw files, msf.files How data was acquired Mass spectrometry (Q Exactive HF mass spectrometer operated in data-dependent acquisition (DDA) mode performing HCD fragmentation)

Data format
Raw files, unfiltered Experimental factors Data were obtained by mass spectrometric DDA measurements of human tear fluid spiked with iRT peptides for use as a spectral library.

Experimental features
Tear fluid was fractionated via gel electrophoresis and in-gel digested resulting in 48 bands.

Data source location
Bochum, Germany (51 26 0 43.4 00 N 7 15 0 27.9 00 E) Data accessibility The data files are hosted in the public repository ProteomeXchange with identifier PXD011075. Value of the data The dataset supplies standard proteomic data based on human tear fluid measured in DDA mode. The samples were spiked with iRT peptides.
This spiked data set can be used as a spectral library for human tear fluid analysis in DIA mode for example in the context of ocular surface-related diseases. Data can serve as an overview with detailed information for further gene ontology enrichment analysis. It may help to find optimized parameters for identification of proteins/peptides of interest. It can be applied for modeling and benchmarking of multiple signaling pathways associated with the eye. (Roche Diagnostics GmbH, Germany). The incubation was performed at 4 C continuous shaking overnight. After this, the supernatant was transferred, aliquoted into 1.5 mL reaction tubes and frozen at À80 C until further usage. The remaining filter strips without solution were discarded.

Amino acid analysis
Protein concentration of the tear fluid was determined via amino acid analysis. First, glass vials to be used for this technique were incubated in a muffle furnace (muffle furnace, Carbolite CWF 1100, USA) at 400 C for 4 h to avoid contaminations. Each sample was analyzed in duplicate. In each clean glass vial, 4 mL of the eluted tear fluid was transferred, dried in a vacuum concentrator (RVC2-25CD plus) and placed in an evacuation vessel. After that, 400 mL 6 M hydrochloric acid and one phenol crystal were added. The evacuation was performed four times in alternation and samples were aerated with argon. The acidic gas phase hydrolysis was done at 150 C for 1 h to cleave peptides into single amino acids. Then, peptide samples were evacuated again to remove the residual hydrochloric acid. Derivatization was proceeded by adding 30 mL of AccQ-Fluor borate buffer with the internal standard Norleucine and 10 mL AccQ-Fluor reagent (10 mM 6-aminoquinolyl-N-hydroxysuccinimidylcarbamate in acetonitrile), followed by an incubation at 55 C for 10 min. Thereby, primary and secondary amines were converted to stable derivatives. These derivatives were separated on a C18 reversed-phase separation column (2.1 mm Â 100 mm in length, Waters GmbH, Germany). Next, the hydrolyzed peptides were dissolved in 10 mL of 20 mM hydrochloric acid. For separation, a gradient system consisting of two solvents was used (solvent A: AccQ-Tag Ultra Eluent A, solvent B: AccQ-Tag Ultra Eluent B). To elute the derivatives, flow rate was set to 0.7 mL per minute with a column temperature of 55C and an increase of solvent A in the gradient (see Table 1). The amino acid derivatives were detected at an emission wavelength of 260 nm utilizing UV-spectrophotometry (Waters GmbH, Germany). Quantitative analysis was performed by applying different concentrations of an internal amino acid standard. In consideration of the volume and the molar mass of each amino acid, protein concentration of the tear fluid sample was calculated.

Sample preparation
400 mL of the pooled human tear fluid (concentration: 0.6 mg/mL) was lyophilized overnight (Alpha 2e4 LDplus, Martin Christ Gefriertrocknungsanlagen GmbH, Germany) and dissolved in 130 mL distilled water. 20 mL Bolt™ sample reducing agent (Thermo Fisher Scientific Inc., USA) as well as 50 mL 4Â LDS sample buffer (pH 8.5) comprising 26.5 mM Tris HCl, 35.25 mM Tris Base, 2% LDS, 10% glycerol, 0.055 mM Coomassie blue G250, and 0.045 mM phenol red was added. The sample was mixed and centrifuged briefly. To achieve denaturation, samples were incubated for 10 min at 95 C and 350 rpm in a thermomixer (Thermomixer comfort, Eppendorf AG, Germany).  in accordance with the manufacturer's instructions. Afterwards, the protein lanes were cut into single bands (in total 12 per lane) and each of them was transferred into a glass vial. Destaining and pH adjustment were achieved by incubating the gel pieces in 50 mM ammonium bicarbonate (Sigma-Aldrich Chemie GmbH, Germany). The washing solution was discarded after 10 min and another solution comprising of 50% (v/v) 50 mM ammonium bicarbonate and 50% (v/v) 100% acetonitrile (Merck KGaA, Germany) was added again for 10 min. Wash cycles were performed in repetition alternately four times. Gel pieces were then dried in a vacuum concentrator (Concentrator plus, Eppendorf AG, Germany).

NanoHPLC settings
The extracted peptides were first concentrated and then separated with an UltiMate™ 3000 RSLCnano system (Thermo Fisher Scientific Inc., USA) equipped with a capillary precolumn (100 mm Â 2 cm, particle size 5 mm, pore size 100 Å; Thermo Fisher Scientific Inc., USA). After 7 min of desalting with 0.1% trifluoroacetic acid, peptides were eluted from the pre-column and on-line transferred to a reversed-phase analytical column (PepMap C18 75 mm Â 50 cm, particle size 2 mm, pore size 100 Å; Thermo Scientific Inc., USA). Then, peptide separation was achieved by applying a stepwise 150 min gradient of solvent A (0.1% formic acid) and solvent B (84% acetonitrile, 0.1% formic acid) at 400 nL per minute flow rate (see Table 2). The temperature of the column oven was set at 60 C. The concentration of solvent B increased from 5% to 60% within 127 min, followed by a washing step with 95% solvent B equilibrating for 5 min.

Mass spectrometry settings
The nanoHPLC system was directly connected to a nano-electrospray ionization source (Thermo Fisher Scientific Inc., USA) with 250 C capillary temperature and a 1600 V spray voltage. Thereby, peptides were ionized by electrospray ionization and injected into a Q Exactive™ HF tandem mass spectrometer (Thermo Fisher Scientific Inc., USA). The instrument was operated in a data-dependent acquisition mode with the XCalibur software (Thermo Fisher Scientific Inc., USA). Mass spectra were acquired with a scan range from 350 to 1400 m/z and a full scan mode resolution set to 60,000 (AGC 3e6, 80 ms maximum injection time, 1.6 m/z wide isolation window). Fragment ions generated from the top 10 most intensive precursor ions were selected for higher-energy collisional dissociation (HCD) fragmentation at 28% NCE. Resulting fragment ions were examined in an Orbitrap mass analyzer acquiring resolving power of 30,000 at 100 m/z (AGC 1e6, 120 ms maximum injection time). Dynamic exclusion was utilized within 30 s to avoid a repeated selection of the same peptide. For internal recalibration, the lock mass polydimethylcyclosiloxane (445.120 m/z) was employed.