Integrated dataset on acute phase protein response in chicken challenged with Escherichia coli lipopolysaccharide endotoxin

Data herein describe the quantitative changes in the plasma proteome in chickens challenged with lipopolysaccharide (LPS), a bacterial endotoxin known to stimulate the host innate immune system obtained by shotgun quantitative proteomic tandem mass tags approach using high-resolution Orbitrap technology. Statistical and bioinformatic analyses were performed to specify the effect of bacterial endotoxin. Plasma from chicken (N=6) challenged with Escherichia coli (LPS) (2 mg/kg body weight) was collected pre (0 h) and at 12, 24, 48, and 72 h post injection along with plasma from a control group (N=6) challenged with sterile saline. Protein identification and relative quantification were performed using Proteome Discoverer, and data were analysed using R. Gene Ontology terms were analysed by the Cytoscape application ClueGO based on Gallus gallus GO Biological Process database, and refined by REVIGO. Absolute quantification of several acute phase proteins, e.g. alpha-1-acid glycoprotein (AGP), serum amyloid A (SAA) and ovotrensferrin (OVT) was performed by immunoassays to validate the LC-MS results. The data contained within this article are directly related to our research article”Quantitative proteomics using tandem mass tags in relation to the acute phase protein response in chicken challenged with Escherichia coli lipopolysaccharide endotoxin” [1]. The raw mass spectrometric data generated in this study were deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD009399 (http://proteomecentral.proteomexchange.org/cgi/GetDataset?ID=PXD009399).

& 2018 Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Subject area
Veterinary medicine, Biomedicine More specific subject area Proteomics, statistics, bioinformatics, immunoassays Type of data Excel files, graphs, figures How data was acquired 1. LC-MS/MS analysis was performed using Ultimate 3000 RSLCnano system (Dionex, Germering, Germany) coupled to Q Exactive Plus mass spectrometer (Thermo Fisher Scientific, Bremen, Germany). 2. Acute phase proteins absolute quantification was performed using ELISA tests (for AGP, SAA) and radial immunodiffusion (for OVT).

Data format
Integration of raw and analyzed data Experimental factors Non-depleted plasma samples Experimental features Quantitative proteomic, bioinformatic and immunoassay analyses of chicken serum Data source location University of Glasgow Cochno Farm & Research Centre, Glasgow, United Kingdom Data accessibility The mass spectrometry proteomics raw data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD009399MS (http://proteomecentral.proteo mexchange.org/cgi/GetDataset?ID ¼PXD009399). All other data are available within this article.

Value of the data
This data provides information about changes in plasma proteome in chickens challenged with Escherichia coli lipopolysaccharide during 72 h with the emphasis on acute phase proteins such as alpha-1-acid glycoprotein (AGP), serum amyloid A (SAA) and ovotrensferrin (OVT).
Peptide/protein information and pathway analysis datasets might be useful as a basis for future targeted analysis of proteins deregulated during inflammation.
The data can be useful for other researchers investigating inflammation or pathophysiological mechanisms in veterinary medicine as well as in biomedical research.

Data
Protein and peptide identifications, as well as their corresponding peptide spectrum matches (PSMs), obtained by label-based proteomic approach, in plasma from chicken challenged with Escherichia coli lipopolysacharride (LPS) endotoxin (2 mg/kg body weight) pre (0 h) and at 12, 24, 48, and 72 h post injection along with plasma from a control group (N ¼6) challenged with sterile saline are reported, with the corresponding peptide spectrum matches (PSMs). Furthermore, relative quantification data after statistical analysis together with subsequent pathway analysis results and immunoassays data are also presented.
Results of analyze performed on this dataset has been represented in different figures and tables included in this Data in Brief article. Fig. 1 represent fold changes of proteins between LPS-treated and saline groups, and their associated p-values. Fig. 2 represent how time affect proteins quantities in LPS animals. Fig. 3 represent pathways up and down regulated, associated with LPS treatment. Fig. 4 represent pathways affected by time, associated with LPS treatment. Evolution of proteins fold changes (LPS vs saline groups) are represented for each time-affected proteins. Fig. 5 represent quantification of 3 proteins (α1-acid glycoprotein, SAA, ovotransferrin) performed by ELISA at 5 time points (0, 12, 24, 48 and 72 h). Fig. 6 represent differences in fold changes (LPS vs saline) between 4 times points (12, 24, 48, 72 h) and 0 h, to compare ELISA and LS-MS quantification. Table 1 list proteins significantly different between LPS and saline group, with their associated fold changes and p-values. Table 2 list proteins significantly different between LPS and saline group which are affected by time effect, with associated fold changes among time and p-values. Table 3 list GO terms associated by LPS challenge, with their associated p-values. Table 4 list GO terms associated by LPS challenge and time, with their associated p-values. Table 5 present different group and time effects for the proteins SAA, AGP and OVT, quantified by ELISA. Table 6 present results about time effect on proteins AGP, SAA and OVT, between LPS stimulated samples and controls, and inside the LPS-stimulated group.  were available ad libitum. From the second day, one group per day was handled and moved into the trial room. All chickens were confirmed to be climatized to handling by 15 days old. Room temperature was maintained within the thermal neutral zone at 18°C (range 18.0-18.3) and a 20 h:4 h light: dark cycle was implemented.

In
The experiment commenced when the chickens were 15 days old. Twenty four birds were injected subcutaneously (SC) at time point 0, with Escherichia coli lipopolysaccharide (LPS from E. coli O111:B4 purified by phenol extraction, L2630-25MG; Sigma-Aldrich, Dorset, UK) (2 mg/kg body weight) in a volume of 0.5 mL as the treatment group and another 24 birds injected SC by sterile normal saline (0.5 mL) as a control group. There were 5 blood sampling time points; pre (0 h) and post injection (PI) at 12, 24, 48, and 72 h. Plasma was collected from the same 6 chicken in the treated group and from the same 6 chicken in the untreated group, subsequently, at each time point for further analyses by proteomic and immunoassay methods. The remaining 18 birds in each group were not used in the plasma proteome investigation. Approximately 1.2 mL of blood was collected from the wing vein using heparinized tubes at each time point. The heparinised blood was centrifuged (3000g) for 15 min at 4°C and the plasma aspirated and immediately frozen at À20°C.
After the trial, all chickens were culled by over dose (1.5-2 mL/bird) i.v. injection of barbiturate (Euthatal 200 mg/mL, Merial, Woking, UK). Research was conducted under Home Office license (60/4466), and approved by ethical review of the University of Glasgow, MVLS College Ethics Committee.

Proteomic investigation of chicken plasma
Proteomic analysis of chicken plasma samples was performed by applying TMT-based quantitative gel-free approach as described previously [2]. In brief, after total protein concentration determination using BCA assay (Thermo Scientific, Rockford, USA), 35 mg of total plasma proteins from samples and internal standard (pool of all samples) were diluted to a volume of 50 mL using 0.1 M triethyl ammonium bicarbonate (TEAB, Thermo Scientific, Rockford, USA), reduced by adding 2.5 mL of Gene ontology analysed pathways and proteins over-represented in LPS compared with saline samples. This analyse have been done with the Cytoscape application ClueGO and the REVIGO tool for GO terms selection. GO terms and proteins over-expressed in LPS are in green, lower-expressed in LPS are in red. GO terms in grey could not be attributed specifically to over or lower expressed terms/proteins. GO terms in bold represent GO terms selected to be the most representative of their GO group defined by the REVIGO tool. The yFiles radial layout algorithm was applied.
200 mM DTT (60 min, 55°C) (Sigma Aldrich, St. Louis, MO, USA), alkylated by adding 2.5 mL of 375 mM IAA (30 min, room temperature in the dark) (Sigma Aldrich, St. Lois, MO, USA) and acetoneprecipitated (addition of 300 mL, overnight, À 20°C). Protein pellets were collected subsequently by centrifugation (8000g, 4°C), dissolved in 50 mL of 0.1 M TEAB and digested using 1 mL of trypsin (1 mg/mL, Promega; trypsin-to-protein ratio 1:35, at 37°C overnight). TMT sixplex reagents (Thermo Scientific, Rockford, IL, USA) were prepared according manufacturer's procedure and an amount of 19 mL of the appropriate TMT label was added to each sample used for the labelling reaction (60 min, room temperature) which was quenched using 5% hydroxylamine (Sigma-Aldrich, St. Louis, MO, USA). Five TMT-modified peptide samples were combined with the internal standard (labelled with TMT m/z 126) into the new tube, aliquoted, dried and stored at À 20°C for further analysis. A total of 30 samples (6 chicken at 5 time points) from treated and 30 samples from control chicken led to 12 individual TMT experiments with the inclusion of internal standards in each experiment.
High resolution LC-MS/MS analysis of TMT-labelled peptides was carried out using an Ultimate 3000 RSLCnano system (Dionex, Germering, Germany) coupled to a Q Exactive Plus mass spectrometer (Thermo Fisher Scientific, Bremen, Germany). Peptides were loaded onto the trap column (C18 PepMap100, 5 mm, 100 A, 300 mm Â 5 mm), desalted for 12 min at the flow rate of 15 uL/min and separated on the analytical column (PepMap™ RSLC C18, 50 cm Â 75 μm) using linear gradient 5-45% mobile phase B (0.1% formic acid in 80% ACN) over 120 min, 45% to 90% for 2 min, held at 80% for 2 min and re-equilibrated at 5% B for 20 min at the flow rate of 300 nL/min. Loading solvent consisted of 0.1% formic acid and 2% ACN in water, while mobile phase A contained 0.1% formic acid in water. Ionisation was achieved using nanospray Flex ion source (Thermo Fisher Scientific, Bremen, Germany) containing a 10 μm-inner diameter SilicaTip emitter (New Objective, USA). The MS operated in positive ion mode using DDA Top8 method. The lock mass feature was not in use in this experiment. Full scan MS spectra were acquired in range from m/z 350.0 to m/z 1800.0 with a resolution of 70,000, 120 ms injection time, AGC target 1E6, a72.0 Da isolation window and the dynamic exclusion 30 s. HCD fragmentation was performed at step collision energy (29% and 35% NCE) with a resolution of 17,500 and AGC target of 2E5. Precursor ions with unassigned charge state, as well as charge states of þ1 and more than þ7 were excluded from fragmentation. MS2 was operated in centroid mode.
For peptide identification and relative quantification the SEQUEST algorithm implemented into Proteome Discoverer (version 2.0., Thermo Fisher Scientific) was used. Database search against Gallus gallus FASTA files downloaded from NCBI database (7/12/2017, 46105 entries, NCBI Gallus gallus Annotation Release ID 103) was performed according to the following parameters: two trypsin missed cleavage sites, precursor and fragment mass tolerances of 10 ppm and 0.02 Da, respectively; carbamidomethyl (C) fixed peptide modification, oxidation (M), deamidation (N,Q) and TMT sixplex (K, peptide N-terminus) dynamic modifications. The false discovery rate (FDR) for peptide identification was calculated using the Percolator algorithm in the Proteome Discoverer workflow based on the search results against a decoy database and was set at 1% FDR. Only proteins with at least two unique peptides and less than 5% FDR were reported as reliable identification. Protein quantification was accomplished by correlating the relative intensities of reporter ions extracted from tandem mass spectra to that of the peptides selected for MS/MS fragmentation The internal standard was used to compare relative quantification results for each protein between the experiments (sixplexes). Peptide and protein identification data are shown in Supplementary file 1.
The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository [3] with the dataset identifier PXD009399.   A two-way ANOVA was performed to model the effect of treatment and time on the quantity of the peptides, using a linear regression model. Distribution of residuals generated by the ANOVA was accessed by a Shapiro-Wilk test. A Kruskal-Wallis test was performed to access the effect of treatment and time on peptides quantity using the R package "PMCMRplus" [6]. Due to multiple comparisons performed, a local False Discovery Rate was applied using the R package "qvalue" [7]. Each p-value was transformed by the function -log10(x). Obtained data are presented in Figs. 1 and 2, as well as in Tables 1 and 2 of Ref [1].
Fold change between the 2 groups has been calculated by the function log2(Mean(Group2)/Mean (Group1)). A volcano plot was designed using the R package "plotly" [8]. Plots were generated with the "ggplot2" package [9]. Spearman's correlation were calculated to estimate the relationship between ELISA and LC-MS quantifications for the proteins AGP, SAA and OVT (Fig. 6 of Ref [1]).
All operations were scripted in R to assure the automatization of the statistics pipeline to all peptides.

Bioinformatics
Proteins ID (Gallus gallus) were converted into Gene ID (Gallus gallus) by the platform DAVID (david-d.ncifcrf.gov/conversion.jsp) conversion tool. Gene Ontology enrichment analysis was performed using the Cytoscape (v3.6.0) [10] plugin ClueGO (v2.5.0) [11] on GO-Biological Processes (08/03/2018). For treatment effect (LPS versus saline), two clusters of proteins differentially expressed between the 2 groups were set: one cluster for over-expressed proteins following LPS treatment, the other for proteins exhibiting lower-expression following LPS. The analysis was performed using the following parameters: evidence code ¼All, GO levels 3 to 15, minimal number of gene¼ 3, minimal percentage of gene ¼3, Kappa score threshold¼0.4, p-values corrected by Bonferroni step down.
For time effect, differentially expressed proteins with time were analyzed at once using the following parameters: evidence code¼ All, GO levels 3 to 8, minimal number of gene¼3, minimal percentage of gene¼3, Kappa score threshold ¼0.4, p-values corrected by Bonferroni step down.
The two lists of GO terms over-expressed in the context of group and time effects were submitted to an analysis by REVIGO (revigo.irb.hr) [12] to remove redundant GO terms and group similar terms based on their description. For both analyses, the database used was Gallus gallus, with the SimReal semantic similarity measure.
Pathways of relationship between GO terms filtered according to REVIGO with their proteins/ genes were designed on Cytoscape. Fold change data was included for the time effect analysis on samples from the LPS treated group. Pathway analysis data are shown in Figs. 3 and 4, as well as in Tables 3 and 4 of Ref [1].

Immunoassays
The concentrations of AGP, SAA and OVT were determined in the plasma according to previously described procedures [13]. The ELISA assays for chicken AGP and SAA were obtained from Life Diagnostics Inc (West Chester, USA). They were performed according to the manufacturer's instructions with a dilution factor for the plasma samples of 1:10,000 for AGP and 1:20 for SAA. Each individual sample was run in duplicate. The plasma concentration of OVT was assessed by radial immunodiffusion (RID) using specific antibody for chickens OVT as described previously [50]. Data are presented in Fig. 5 of Ref [1].

Statistics for immunoassays
Statistics on immunoassay were performed by non-parametric tests due to group size and distribution. Group effect was assessed by a Wilcoxon-test (2-sided), and a Kruskal-Wallis test was used to assess mixed effect Group x Time on all groups and Time effect on LPS and saline groups separately. For each time point (0 h/12 h/24 h/48 h/72 h), difference between LPS and saline was assessed by a Wilcoxon-test (2-sided) and fold change of expression calculated between times 12 h/24 h/48 h/72 h versus 0 h in LPS group. Correlation between these proteins was assessed on LPS group by a Spearman rank test. Immunoassays-related statistical data are shown in Tables 5 and 6 of Ref. [1].