Traumatic brain injury (TBI) is a major public health concern that affects 12% of the general population and results in high rates of death and disability worldwide [
Due to the complexity and heterogeneity of TBI-induced cognitive impairment, it is likely that multiple candidate proteins present in networks are perturbed, leading to the spectrum of cognitive symptoms. Currently, with the advent of quantitative proteomic technologies using an isobaric labelling strategy, it has become possible to quantify several proteins in a single experiment for the comparative study of global protein regulation across various biological samples, and this method has been widely applied to elucidate disease mechanisms [
However, none of the previous studies applied a quantitative proteomic technology to identify global protein changes and pathways perturbed in post-TBI cognitive impairment using clinical samples. In this study, an iTRAQ-based quantitative proteomics approach was adopted to identify and quantity the differentially expressed proteins (DEPs) in serum samples from TBI patients with cognitive impairment. In addition, DEPs were further analysed by bioinformatic platforms and validated by enzyme-linked immunosorbent assays (ELISA). These findings will further the understanding of the pathophysiological mechanisms underlying post-TBI cognitive impairments.
All protocols involving the use of human subjects were reviewed and approved by the Ethics Committee of Central South University, Changsha, China (Grant no. 201404366), and all experiments were performed in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants (or their legal guardians) enrolled in this study.
All participants presented at the Brain Trauma Specialist Department, Department of Encephalopathy of the National Key Specialty, and the Health Centre of the Xiangya Hospital, Central South University, Changsha, China, between February 2014 and December 2014. The subjects were divided into three groups: healthy controls (HC group), TBI patients without cognitive deficits (negative group), and TBI patients with cognitive deficits (positive group).
As described in our previous study [
Individuals were excluded if they met any of the following criteria: (i) serious conditions causing mental disability prior to the TBI, such as a developmental handicap (Down’s syndrome), residual disability after previous TBI, confirmed dementia, or serious chronic mental illness (schizophrenia, psychosis, or well-confirmed bipolar disorder); (ii) severe renal or hepatic impairment; (iii) uncontrolled cardiovascular disease; (iv) a current history of severe abuse of drugs or alcohol; and (v) being pregnant or lactating.
When TBI patients were recruited into study, the overall level of cognitive and behavioural functioning of them was assessed using the Rancho Los Amigos Scale (RLAS, also referred to as “Rancho” or the “Levels of Cognitive Functioning Scale”) by two doctors [
In addition, the healthy control group was composed of healthy volunteers with no current or previous lifetime history of neurological diseases or systemic medical illness. Healthy controls were matched with TBI patients for age and gender. The demographic and clinical chemistry characteristics of enrolled subjects are shown in Table
Demographic characteristics of the enrolled participants.
Positive | Negative | HC | |
---|---|---|---|
Number | 51 | 51 | 51 |
Female/male | 20/31 | 22/29 | 24/27 |
Age | | | |
Disease duration (days) | | | / |
Cause of TBI | |||
Transport accidents | 35 | 40 | / |
Fall | 10 | 8 | / |
Assaults | 2 | 1 | / |
Others | 4 | 2 | / |
GCS score at admission | 7 (4–9) | 8 (4–10) | / |
Moderate/severe① | 31/20 | 26/25 | / |
Neurosurgery② | |||
No/yes | 28/23 | 31/20 | / |
Multiple ICD-10 diagnosis (S06) | 43 | 44 | / |
Note: IQR, interquartile range. ①Moderate/severe indicates the classification of TBI according to GCS score at admission. ②Neurosurgery. “Yes” indicates the patients who underwent neurosurgical operative intervention at admission, whereas “No” indicates those who did not.
The subjects fasted for at least 12 hours before blood was drawn. The blood samples were obtained specifically for the purpose of this study and were coded to maintain anonymity. Relevant medical data were recorded and coded to match the extracted blood samples. A 3 mL blood serum sample was collected from each enrolled subject. The serum samples were placed in Eppendorf tubes without anticoagulant at 4°C and allowed to stand for 1 h. The sample was centrifuged at 3000
Pooled serum samples were generated by combining equal volumes of the 15 individual plasma samples from each group (
As described in our previous study [
Briefly, trypsin digestion and iTRAQ labelling were performed according to the manufacturer’s protocol (Applied Biosystems). First, 50
The mixed peptides were fractionated by SCX chromatography on an ultimate high-performance liquid chromatography (HPLC) system (Shimadzu, Kyoto, Japan) with an SCX column (Luna SCX 100A, Phenomenex). Based on the SCX chromatograms, 10 SCX fractions were collected along the gradient. Each SCX fraction was dried, dissolved, and then analysed on a reverse-phase liquid chromatography column (Strata-X C18 column, 5
Mass spectrometry (MS) analysis of the iTRAQ-labelled samples was performed on a Q Exactive LC-MS/MS (ThermoFisher Scientific, Waltham, MA, USA) mass spectrometer. Sequences for the peptide and reporter ions were generated to identify the protein from which the peptide originated. To minimize the effect of experimental variation, three independent MS/MS runs were performed for each sample.
Proteome Discoverer Software (Thermo Scientific version 1.3) was used for the data acquisition and quantification. The data sifted by Proteome Discoverer were used to identify proteins using Mascot (version 2.3.0, Matrix Science, London, UK) software and the Uniprot-rat database (
As described in a previous study [
As described in our previous study [
All data are expressed as the means ± SE. One-way ANOVA was used to compare the differences between the groups. All statistical analyses were conducted using the SPSS (version 22.0, Chicago, IL).
Using iTRAQ-based quantitative proteomics, a total of 331,259 MS/MS spectra were obtained, of which 40,273 were matched. Then, 48,350 PSMs were assigned to 3079 peptides after 1% FDR was applied. Through this strategy, 359 proteins were identified for further study. Of the 359 proteins, we identified 50 DEPs in the positive and negative subjects, including 23 upregulated and 27 downregulated proteins. Meanwhile, 108 DEPs were identified in the positive and control subjects, including 54 upregulated and 54 downregulated proteins. Additionally, 87 DEPs were identified in the negative and control subjects, including 37 upregulated and 40 downregulated proteins. As shown in Figure
Differentially expressed proteins (DEPs) identified using iTRAQ coupled with LC-MS/MS.
Number | Accession | Description | Mascot score | Coverage (%) | MW (kDa) | Fold change | Regulation | |
---|---|---|---|---|---|---|---|---|
Positive/negative | Positive/HC | |||||||
1 | A8MT79 | Putative zinc-alpha-2-glycoprotein-like 1 | 26.26 | 4.90 | 23 | 0.767 | / | Down |
2 | O76076 | WNT1-inducible-signaling pathway protein 2 | 72.31 | 11.60 | 26.8 | 0.818 | 0.768 | Down |
3 | P01598 | Ig kappa chain V–I region EU | 119.29 | 26.85 | 11.8 | / | 0.827 | Down |
4 | P01616 | Ig kappa chain V–II region MIL | 83.47 | 33.04 | 12 | / | 1.279 | Up |
5 | P01620 | Ig kappa chain V–III region SIE | 318.55 | 59.63 | 11.8 | / | 0.824 | Down |
6 | P01625 | Ig kappa chain V-IV region Len | 150.17 | 36.84 | 12.6 | / | 0.791 | Down |
7 | P01701 | Ig lambda chain V–I region NEW | 48 | 15.32 | 11.4 | 1.39 | / | Up |
8 | P01717 | Ig lambda chain V-IV region Hil | 75.53 | 17.76 | 11.5 | 1.319 | 1.297 | Up |
9 | P01742 | Ig heavy chain V–I region EU | 58.23 | 10.26 | 12.5 | 1.291 | 1.449 | Up |
10 | P01764 | Ig heavy chain V–III region VH26 | 109.91 | 35.04 | 12.6 | / | 0.773 | Down |
11 | P01766 | Ig heavy chain V–III region BRO | 134.83 | 25.00 | 13.2 | / | 0.812 | Down |
12 | P01768 | Ig heavy chain V–III region CAM | 106.7 | 23.77 | 13.7 | 0.674 | 0.595 | Down |
13 | P01779 | Ig heavy chain V–III region TUR | 138 | 26.72 | 12.4 | 0.8 | 0.749 | Down |
14 | P02655 | Apolipoprotein C-II | 344.57 | 57.43 | 11.3 | / | 1.353 | Up |
15 | P02751 | Fibronectin | 4895.14 | 52.68 | 262.5 | / | 0.827 | Down |
16 | P02788 | Lactotransferrin | 670.88 | 34.51 | 78.1 | / | 1.234 | Up |
17 | P04180 | Phosphatidylcholine-sterol acyltransferase | 498.47 | 29.77 | 49.5 | / | 0.791 | Down |
18 | P04278 | Sex hormone-binding globulin | 51.27 | 5.47 | 43.8 | 1.275 | 1.333 | Up |
19 | P04406 | Glyceraldehyde-3-phosphate dehydrogenase | 170.08 | 18.21 | 36 | 0.495 | 0.444 | Down |
20 | P04434 | Ig kappa chain V–III region VH (fragment) | 65.35 | 23.28 | 12.7 | 0.68 | 0.781 | Down |
21 | P05155 | Plasma protease C1 inhibitor | 167.06 | 13.60 | 55.1 | / | 1.208 | Up |
22 | P06753 | Tropomyosin alpha-3 chain | 123.25 | 10.88 | 32.9 | / | 1.253 | Up |
23 | P06858 | Lipoprotein lipase | 104.22 | 7.58 | 53.1 | 0.808 | / | Down |
24 | P08185 | Corticosteroid-binding globulin | 22.27 | 1.98 | 45.1 | 0.801 | / | Down |
25 | P08519 | Apolipoprotein(a) | 130 | 25.62 | 501 | 1.436 | 1.62 | Up |
26 | P0C0L5 | Complement C4-B | 5043.1 | 65.19 | 192.6 | 1.226 | / | Up |
27 | P11021 | 78 kDa glucose-regulated protein | 123.72 | 10.55 | 72.3 | 0.762 | / | Down |
28 | P11597 | Cholesteryl ester transfer protein | 210.43 | 12.37 | 54.7 | 1.229 | / | Up |
29 | P12814 | Alpha-actinin-1 | 104.07 | 5.16 | 103 | 1.32 | 1.373 | Up |
30 | P15169 | Carboxypeptidase N catalytic chain | 69.71 | 7.21 | 52.3 | 0.827 | 0.808 | Down |
31 | P23142 | Fibulin-1 | 1123.98 | 42.11 | 77.2 | / | 0.833 | Down |
32 | P28066 | Proteasome subunit alpha type 5 | 48.13 | 4.98 | 26.4 | 0.767 | / | Down |
33 | P29122 | Proprotein convertase subtilisin/kexin type 6 | 29.19 | 1.03 | 106.4 | / | 0.777 | Down |
34 | P35443 | Thrombospondin-4 | 226.54 | 8.84 | 105.8 | 1.228 | 1.405 | Up |
35 | P49913 | Cathelicidin antimicrobial peptide | 73.93 | 12.35 | 19.3 | / | 0.686 | Down |
36 | P62158 | Calmodulin | 45.27 | 26.17 | 16.8 | 1.255 | 1.224 | Up |
37 | P68032 | Actin, alpha cardiac muscle 1 | 244.16 | 25.20 | 42 | / | 1.214 | Up |
38 | P68104 | Elongation factor 1-alpha 1 | 42.01 | 4.33 | 50.1 | 0.77 | / | Down |
39 | P68366 | Tubulin alpha-4A chain | 66.67 | 6.25 | 49.9 | / | 1.362 | Up |
40 | P68871 | Hemoglobin subunit beta | 196.67 | 52.38 | 16 | / | 1.303 | Up |
41 | Q02985 | Complement factor H-related protein 3 | 160.25 | 11.52 | 37.3 | / | 1.212 | Up |
42 | Q14515 | SPARC-like protein 1 | 219.59 | 17.77 | 75.2 | / | 0.753 | Down |
43 | Q6ZV73 | FYVE, RhoGEF, and PH domain-containing protein 6 | 36.67 | 1.12 | 160.7 | 1.29 | / | Up |
44 | Q86UX7 | Fermitin family homolog 3 | 22.07 | 2.40 | 75.9 | 0.8 | 0.798 | Down |
45 | Q8IWZ6 | Bardet-Biedl syndrome 7 protein | 39.01 | 1.12 | 80.3 | / | 1.24 | Up |
46 | Q92540 | Protein SMG7 | 34.49 | 0.70 | 127.2 | / | 1.222 | Up |
47 | Q92954 | Proteoglycan 4 | 233.65 | 7.05 | 151 | / | 0.829 | Down |
48 | Q96KK5 | Histone H2A type 1-H | 82.29 | 21.88 | 13.9 | / | 0.735 | Down |
49 | Q9H4B7 | Tubulin beta-1 chain | 57.95 | 6.21 | 50.3 | 0.74 | 0.81 | Down |
50 | Q9H6R4 | Nucleolar protein 6 | 33.17 | 0.61 | 127.5 | 0.793 | / | Down |
51 | Q9NUD7 | Uncharacterized protein C20orf96 | 27.39 | 2.48 | 42.8 | / | 1.229 | Up |
52 | Q9UBU7 | Protein DBF4 homolog A | 42.74 | 0.89 | 76.8 | / | 1.367 | Up |
53 | Q9UHG3 | Prenylcysteine oxidase 1 | 248.97 | 15.84 | 56.6 | / | 0.819 | Down |
54 | Q9UK55 | Protein Z-dependent protease inhibitor | 615.93 | 31.98 | 50.7 | 0.825 | 0.717 | Down |
55 | Q9Y490 | Talin-1 | 61 | 1.61 | 269.6 | 1.248 | 1.427 | Up |
56 | Q9Y5C1 | Angiopoietin-related protein 3 | 40.67 | 1.74 | 53.6 | / | 1.268 | Up |
Note: regulation: up or down indicates the DEPs that were upregulated or downregulated, respectively, in the positive group relative to the reference groups.
Venn diagram showing the number of differentially expressed proteins (DEPs) and their overlap. The results indicated 108 proteins showed differential expression in the positive versus healthy control (HC) groups (green cycle), 50 proteins in the positive versus negative groups (blue cycle), and 87 proteins in the negative versus HC groups (red cycle). A total of 56 DEPs which included 28 DEPs in the positive versus negative comparison, 11 DEPs in the positive versus negative comparison, and 17 DEPs in the overlapping regions between both comparisons (positive versus negative and positive versus negative) were specific to the positive group.
To gain insight into the biological changes in TBI patients with cognitive impairment, the DEPs were categorized according to the following Gene Ontology (GO) classes: biological process, molecular function, and cellular components (Figure
GO analysis of the differentially expressed proteins (DEPs). All identified proteins were functionally annotated in GO database according to their biological process (a), molecular function (b), and cellular component (c). In addition, the GO term enrichment analysis was conducted, and the significantly enriched categories (
GO biological process
GO molecular function
GO cellular component
To determine whether the DEPs were enriched in certain groups, we employed the PANTHER Overrepresentation Test and used the Bonferroni correction for multiple comparisons. As shown in Figure
Because we were interested in the signaling pathways enriched in DEPs, a KEGG pathway analysis was performed. As shown in Table
KEGG pathways associated with the differentially expressed proteins (DEPs) identified by iTRAQ analysis.
Number | Pathway ID | Pathway | DEPs with pathway annotation | Mapped DEPs | |
---|---|---|---|---|---|
1 | ko05322 | Systemic lupus erythematosus | 3 | P0C0L5, P12814, Q96KK5 | 0.001338 |
2 | ko05133 | Pertussis | 3 | P05155, P0C0L5, P62158 | 0.003080 |
3 | ko04510 | Focal adhesion | 4 | P02751, P12814, P35443, Q9Y490 | 0.010071 |
4 | ko05144 | Malaria | 2 | P35443, P68871 | 0.011312 |
5 | ko05010 | Alzheimer's disease | 3 | P04406, P06858, P62158 | 0.015594 |
6 | ko05130 | Pathogenic | 2 | P68366, Q9H4B7 | 0.021948 |
7 | ko04260 | Cardiac muscle contraction | 2 | P06753, P68032 | 0.026523 |
8 | ko04261 | Adrenergic signaling in cardiomyocytes | 3 | P06753, P62158, P68032 | 0.027736 |
9 | ko04970 | Salivary secretion | 2 | P49913, P62158 | 0.032064 |
10 | ko04610 | Complement and coagulation cascades | 2 | P05155, P0C0L5 | 0.032890 |
11 | ko04512 | ECM-receptor interaction | 2 | P02751, P35443 | 0.037802 |
12 | ko05410 | Hypertrophic cardiomyopathy (HCM) | 2 | P06753, P68032 | 0.038242 |
13 | ko05414 | Dilated cardiomyopathy (DCM) | 2 | P06753, P68032 | 0.039573 |
14 | ko04145 | Phagosome | 3 | P35443, P68366, Q9H4B7 | 0.041505 |
15 | ko05146 | Amoebiasis | 2 | P02751, P12814 | 0.043672 |
16 | ko04540 | Gap junction | 2 | P68366, Q9H4B7 | 0.045073 |
The 56 DEPs identified in the current study were submitted to STRING to assess the PPI networks (Figure
Protein-protein interactions for the differentially expressed proteins identified using iTRAQ-based proteomics were analysed with STRING V10.0. In the network, the proteins are represented as nodes. The colors of the lines connecting the nodes represent different evidence types for the protein linkage.
Based on the results of the bioinformatic analysis and the correlations with disease pathogenesis, five candidate DEPs, namely, GAPDH, CaM, apolipoprotein(a) (APO(a)), thrombospondin-4 (THBS4), and Talin-1 (TLN1), were selected for validation in an additional 105 cases using ELISA. These cases included 35 TBI patients with cognitive impairment, 35 TBI patients without cognitive impairment, and 35 healthy controls.
As shown in Figure
Serum proteins levels among the positive, negative, and HC group. A
In our study, using an iTRAQ-based quantitative proteomic approach, a total of 56 DEPs were found which displayed quantitative changes unique to TBI patients with cognitive deficits relative to healthy controls and TBI patients without cognitive deficits. Of these DEPs, 30 were downregulated, including LPL and GAPDH, and 26 were upregulated, including APO(a), THBS4, and CaM.
According to the bioinformatic analysis, the 56 DEPs were suggested to be involved in a wide variety of cellular and metabolic processes, including immunity and inflammation, transportation of important regulatory biomolecules, blood coagulation, and other cell processes, and a sizeable group of significantly differentiated pathways with important biological functions. Moreover, the results of PPI analysis indicated that TBI-induced cognitive impairment is a multifactorial process of pathological progress that involves various proteins that interact with each other including the SPARCL1-CALM-ACTC1-TUBA4A-GAPDH-LPL-SHBG-SERPINA6 network. In addition, the ELISA validation results confirmed the proteomics analysis findings to some extent.
Interestingly, our results revealed that AD signaling pathways (including LPL, GAPDH, and CaM) might play an important role in the pathophysiology of post-TBI cognitive impairments. Numerous epidemiological studies have indicated that TBI can increase the risk of developing AD, which is the most common form of dementia [
As a classical glycolytic enzyme, GAPDH was validated by ELISA as being significantly downregulated in the TBI patients with cognitive impairment. In addition, GAPDH has been suggested to have high affinity for AD-associated proteins, including
Additionally, we found that lipid metabolism-associated DEPs, including ApoC-II, Apo(a), cholesteryl ester transfer protein (CETP), and LPL, were significantly altered in the TBI patients with cognitive impairment. It has been suggested that dysregulated lipid metabolism may play an important role in the pathophysiology of post-TBI cognitive impairments. As the main constituent of lipoprotein(a), Apo(a) was validated using ELISA as being significantly upregulated in the TBI patients with cognitive impairment. It has been demonstrated that Apo(a) may be involved in the pathology of dementia by participating in amyloidogenesis and playing a role in neuronal maintenance [
THBS4 and Talin-1 were also both validated using ELISA as being significantly upregulated in the TBI patients with cognitive impairment. In contrast, accumulated data indicate that THBS4 could not only regulate synapse formation but also play an important role in neurite and axon outgrowths [
Although the altered serum DEPs were identified and their possible underlying mechanisms were investigated, the present study has several limitations. First, only the serum of TBI patients with cognitive impairments was analysed. To more accurately reflect the pathophysiology of cognitive deficits following TBI, the plasma and CSF from the same individuals with post-TBI cognitive impairments should be analysed using iTRAQ-based quantitative proteomic approaches in the future. Second, the 5 DEPs selected for ELISA validation are not brain specific. Moreover, a correlation between the 5 DEPs and the degree of cognitive impairment has not been established. Third, the present study included only a small number of patients; therefore, additional studies using a larger patient population should be conducted to fully confirm/validate the current findings. Fourth, similar to previous studies [
To the best of our knowledge, the present study was the first to use an iTRAQ-based quantitative proteomic approach to identify DEPs in serum samples obtained from TBI patients with cognitive deficits to better understand the pathophysiology of cognitive impairments following TBI. Using an iTRAQ-based quantitative proteomic analysis, serum proteome alterations in patients with cognitive impairment after TBI were identified, and 56 DEPs were found to be specifically related to TBI-induced cognitive impairment. Moreover, bioinformatic analysis revealed that AD signaling pathways and lipid metabolism are involved in the pathophysiology of cognitive deficits following TBI. However, the limitations of the present study require further investigation and large-scale validation.
The authors declare that they have no competing interests.
The work presented here was conducted in collaboration between all authors. Weijun Peng conceived and designed the work. Weijun Peng, Zhe Wang, and Xin-gui Xiong analysed the data, interpreted the results, and drafted the manuscript. Qinghua Liang, Wei Huang, Yang Wang, and Zi-an Xia carried out the clinical research and participated in the acquisition, analysis, or interpretation of the data. Chuhu Zhang contributed to the data collection, analysed the data, and interpreted the results. All authors reviewed and approved the final manuscript. Weijun Peng, Zhe Wang, and Zi-an Xia contributed equally to this study.
This work was supported by the Young Scientists Fund of the National Natural Science Foundation of China (Grant nos. 81102564, 81603670) and the National Natural Science Foundation of China (Grant no. 81373705). The authors also thank Gene Denovo Co. (Guangzhou, People’s Republic of China) for technical support.