Differential Neuroproteomic and Systems Biology Analysis of Spinal Cord Injury

Abstract

Acute spinal cord injury (SCI) is a devastating condition with many consequences and no known effective treatment. Although it is quite easy to diagnose traumatic SCI, the assessment of injury severity and projection of disease progression or recovery are often challenging, as no consensus biomarkers have been clearly identified. Here rats were subjected to experimental moderate or severe thoracic SCI. At 24h and 7d postinjury, spinal cord segment caudal to injury center versus sham samples was harvested and subjected to differential proteomic analysis. Cationic/anionic-exchange chromatography, followed by 1D polyacrylamide gel electrophoresis, was used to reduce protein complexity. A reverse phase liquid chromatography-tandem mass spectrometry proteomic platform was then utilized to identify proteome changes associated with SCI. Twenty-two and 22 proteins were up-regulated at 24 h and 7 day after SCI, respectively; whereas 19 and 16 proteins are down-regulated at 24 h and 7 day after SCI, respectively, when compared with sham control. A subset of 12 proteins were identified as candidate SCI biomarkers - TF (Transferrin), FASN (Fatty acid synthase), NME1 (Nucleoside diphosphate kinase 1), STMN1 (Stathmin 1), EEF2 (Eukaryotic translation elongation factor 2), CTSD (Cathepsin D), ANXA1 (Annexin A1), ANXA2 (Annexin A2), PGM1 (Phosphoglucomutase 1), PEA15 (Phosphoprotein enriched in astrocytes 15), GOT2 (Glutamic-oxaloacetic transaminase 2), and TPI-1 (Triosephosphate isomerase 1), data are available via ProteomeXchange with identifier PXD003473. In addition, Transferrin, Cathepsin D, and TPI-1 and PEA15 were further verified in rat spinal cord tissue and/or CSF samples after SCI and in human CSF samples from moderate/severe SCI patients. Lastly, a systems biology approach was utilized to determine the critical biochemical pathways and interactome in the pathogenesis of SCI. Thus, SCI candidate biomarkers identified can be used to correlate with disease progression or to identify potential SCI therapeutic targets.

Fig. 1.

CAX-PAGE differential separation of sham and SCI rat lysate (A) Ion exchange (CAX) separation chromatograms overlay of pooled rat SCI (at 24 h and 7 day postsevere injury) and sham at 24 h and 7 day lysate (pooled from n = 5 each) with the same 280 nm absorbance scale: sham in purple and SCI in green, solvent B % in orange, and conductivity in red. B, 1D-PAGE of selected CAX fractions out of 28 collected, run side by side for comparison, selected bands are boxed and labeled according to 2D position. C, Quantification of selected differential gel band intensities to derive the relative -fold increase or decrease using ImageJ densitometry software.

Fig. 2.

Comparison of differential SCI protein identification for 24 h and 7 day post-SCI versus sham. Shown are the number of identified proteins using CAX-PAGE-LC-MS/MS platform along with common proteins between 24 h and 7 days (A) up-regulated (B) down-regulated.

Fig. 3.

SCI-tissue immunoblotting validation (epi-center segment). A, Time course of post-SCI biomarkers and new candidates validation illustrated by Western blot of rat spinal cord tissue (rostral) lysate of rostral section (sham, and two SCI severity) collected at three time points (4 h, 24 h, and 7 days), candidate markers probed include transferrin, cathepsin D, TPI-1,and PEA-15. Western blot of carbonic anhydrase II served as a loading control. B, Immunoblotting quantification of spinal cord tissue lysate samples (sham and two SCI severity levels at three time points) for transferrin, CathD, TPI-1,and PEA-15 biomarkers. Mean + S.E. are shown. (* p < 0.05, compared with corresponding sham, unpaired t test).

Fig. 4.

SCI-tissue immunoblotting validation (caudal segment). A, Time course of post-SCI biomarkers and new candidate's validation illustrated by Western blot of rat spinal cord tissue (rostral) lysate of rostral section (sham, and two SCI severity) collected at three time points (4 h, 24 h, and 7 days), candidate markers probed include transferrin, cathepsin D, TPI-1, and PEA-15. Western blot of carbonic anhydrase II served as a loading control. B, Immunoblotting quantification of spinal cord tissue lysate samples (sham and two SCI severity levels at 3 time points) for transferrin, CathD, TPI-1, and PEA-15 biomarkers. Mean + S.E. are shown. (* p < 0.05, compared with corresponding sham, unpaired t test).

Fig. 5.

SCI-tissue immunoblotting validation (rostral segment). A, Time course of post-SCI biomarkers and new candidates validation illustrated by Western blot of rat spinal cord tissue (rostral) lysate of rostral section (sham, and two SCI severity) collected at three time points (4 h, 24 h, and 7 days), candidate markers probed include transferrin, cathepsin D, TPI-1 and PEA-15. Western blot of carbonic anhydrase II served as a loading control. B, Immunoblotting quantification of spinal cord tissue lysate samples (sham and two SCI severity levels at 3 time points) for transferrin, CathD, TPI-1, and PEA-15 biomarkers. Mean + S.E. are shown. (* p < 0.05, compared with corresponding sham, unpaired t test).

Fig. 6.

Rat SCI-CSF immunoblotting validation. A, Time course of post-SCI biomarkers release and new candidate marker validation illustrated by Western blot of rat spinal cord CSF samples (sham, and two SCI severity) collected at three time points (4 h, 24 h, and 7 days), the markers probed are alphaII-spectrin, transferrin, GFAP, TPI-1, UCHL-1, and PEA-15, B, Immunoblotting quantification of CSF samples (sham and two SCI severity levels at three time points) for Alpha II-spectrin breakdown products SBDP145 & SBDP120, GFAP and GFAP-BDP (38K), TPI-1, and PEA-15 biomarkers. Mean + S.E. are shown. (* p < 0.05, compared with corresponding sham, unpaired t test).

Fig. 7.

Representative time course of SCI biomarker candidate proteins released into human CSF after SCI, detected by immunoblotting. AlphaII-spectrin, SBDP150/145 and SBDP120, GFAP, GFAP-BDP, and Transferrin, CathD, TPI-1 and PEA-15 for two SCI patients A &B at different time points postinjury and for two patients and 4 controls.

Fig. 8.

Quantification of biomarker release human CSF after SCI. (A) Biomarker levels at initial time point (First sample taken- usually within 36 postinjury) and all time points (all samples from day 0 to 6 post-SCI). Transferrin (signals × ¼ is shown), CathD, TPI-1, and PEA-15 (signals × 25 is shown). (* p < 0.05, **, p < 0.001, as compared with control CSF, unpaired t test). B, Time course of biomarker levels in CSF for all patients (n = 14). Mean + S.E. are shown. Trend lines shown are moving averages of biomarker levels from two adjacent time points.

Fig. 9.

Candidate SCI biomarker input for Systems Biology and pathway network analysis. Top 12 candidate SCI biomarkers (shown in (A)) are input into Pathway Studio to identify cellular process, regulation and common disease pathway analysis (B).

Fig. 10.

Three pathways identified as altered based on the 12 SCI biomarker candidate proteins as input. Based on 1 Biology pathway analysis, (Fig. 9), three pathways were found to have the most interactions between these 12 candidate biomarkers: A, Cell growth and Aging; B, Metabolic Dysfunction; and C, Neural Death. Original SCI biomarkers are in red ovals with blue aura. Additional interacting partner proteins are in red ovals. Cell processes are in blue rectangles, disease pathway in purple rectangles, while regulation are annotated by gray arrows.

Fig. 11.

Systems Biology analysis identifying further interactome and interplay among the top 12 new SCI biomarkers with additional CNS injury biomarkers αII-spectrin (SPTAN1; SPAN2(mouse)), UCHL1 and MBP, GFAP, and S100b. Systems Biology analysis was based on STRING program (http://string-db.org/) with action views using three species: human (A), mouse (B), and rat (C). Blue lines indicate interactions, arrows indicate regulations, whereas gray lines identify possible associations. Green circles identified key interaction clusters identified.