|Home | About | Journals | Submit | Contact Us | Français|
Outcome following traumatic brain injury (TBI) is in large part determined by the combined action of multiple processes. In order to better understand the response of the central nervous system to injury, we utilized an antibody array to simultaneously screen 507 proteins for altered expression in the injured hippocampus, a structure critical for memory formation. Array analysis indicated 41 candidate proteins have altered expression levels 24 hours after TBI. Of particular interest was catechol-O-methyl transferase (COMT), an enzyme involved in metabolizing catecholamines released following neuronal activity. Altered catecholamine signaling has been observed after brain injury, and may contribute to the cognitive dysfunctions and behavioral deficits often experienced after TBI. Our data shows that COMT expression in the injured ipsilateral hippocampus was elevated for at least 14 days after controlled cortical impact injury. We found strong co-localization of COMT immunoreactivity with the microglia marker Iba1 near the injury site. Since dopamine transporter expression has been reported to be down-regulated after brain injury, COMT-mediated catecholamine metabolism may play a more prominent role in terminating catecholamine signaling in injured areas.
The central nervous system (CNS) response to trauma is complex, and involves axonal injury, cell death, neuronal dysfunction, impaired plasticity and altered neurotransmission. Since outcome after traumatic brain injury (TBI) is dependent on the combined action of a diverse assortment of gene products, global analysis of alterations in mRNA and protein expression is likely to provide a better overall description of TBI pathophysiology [5,6,12]. In this study, we used an antibody array as a first pass screen to identify potential candidate proteins whose expression levels change in the rat hippocampus 24 h after TBI. After validating altered expression of representative candidates by western blot analysis, we subsequently targeted catechol-O-methyl transferase (COMT), an enzyme involved in the metabolism of the catecholamines dopamine and norepinephrine, for a more extensive characterization.
Numerous studies have indicated that catecholamine signaling (e.g. synthesis, release, reuptake/inactivation) can be disturbed after TBI, suggesting that altered catecholamine function may play a role in some of the behavioral deficits observed after TBI [15,17,18,30]. For example, TBI in rats results in increased expression of tyrosine hydroxylase (TH) protein, the rate-limiting enzyme in catecholamine synthesis, as well as norepinephrine and dopamine content [16,28]. In addition, Yan et al. demonstrated decreased expression of the dopamine transporter after TBI, which may affect reuptake and signaling termination . In humans, some cognitive functions correlate with a genetic polymorphism that determines COMT enzyme activity, while COMT-deficient mice exhibit altered behavior and regional disturbances in dopamine level [1,7,10]. Since TBI alters pathways involved in catecholamine synthesis and extracellular clearance, a better understanding of changes in the pathways involved in catecholamine metabolism would aid in the understanding of TBI pathophysiology and in designing potential treatment paradigms.
Harlan (Indianapolis, IN) provided male Sprague-Dawley rats (250–300g). Cy3/Cy5 monofunctional NHS-ester labeling kits were from Amersham Biosciences (Piscataway, NJ). Antibody arrays and mouse anti-COMT, -AF6, -CA150, -Ngn3, and -glutamine synthetase were obtained from BD Biosciences (San Jose, CA). Rabbit anti-GAD65/67, -GFAP, -GalC, -parvalbumin, and -NF-H were from Chemicon (Temecula, CA), and -Iba1 was from WAKO USA (Richmond, VA). Fluorescent secondaries were from Molecular Probes (Eugene, OR), and Vector Labs (Burlingame, CA) provided AP-conjugated secondaries.
All protocols involving animals are in compliance with NIH's Guide for the Care and Use of Laboratory Animals and approved by the Institutional Animal Care and Use Committee. A controlled cortical impact device was used to administer unilateral brain injury at 1.9 mm deformation essentially as described in . Sham rats received a midline incision but were not otherwise injured.
Rats were decapitated 24 h after injury (n=5) or sham surgery (n=4) and the hippocampi removed under ice-cold artificial cerebrospinal fluid. Cytosolic extracts from individual hippocampi were prepared in 500 μl buffer containing 10 mM HEPES pH 7.5, 1 mM EDTA, 1 mM EGTA, 10 mM NaF, 2 mM sodium pyrophosphate, 25 mM β-glycerophosphate, 1 mM DTT, 1 mM diisopropyl fluorophosphate, 100 nM okadiac acid, and 1 mM sodium orthovanadate. Extracts were centrifuged 10 min, 4°C, 10,000 x g, and the supernatant recovered and quantified by Bradford assay. All subsequent steps were carried out as described by in the manufacturers’ protocols. Equal amounts of sham or injured protein extracts were pooled and labeled with Cy3- or Cy5-NHS ester dyes to generate four labeled samples: TBI-Cy3, TBI-Cy5, Sham-Cy3, and Sham-Cy5. The labeled samples were purified, and the dye/protein ratio measured. Arrays were incubated with 20 μg labeled sample mixes (forward: TBI-Cy5/Sham-Cy3; reverse: Sham-Cy5/TBI-Cy3), sequentially washed with buffers 1–7, centrifuged 25 min at 1000 x g, and imaged on an Axon Genepix 4200A slide scanner at 10 μm resolution.
Sham and 24 h injured ipsilateral hippocampal cytosolic extracts were prepared as described in . Total extracts were prepared from sham, 3 d and 14 d samples by solubilizing hippocampi in 10 volumes lysis buffer [6 M urea, 2 M thiourea, 2% CHAPS, and 65 mM DTT]. Equal amounts of protein were separated by SDS-PAGE, blotted, and probed with the following antibodies: mouse anti-COMT (1:10,000), anti-glutamine synthetase (1:10,000), anti-AF6 (1:500), anti-Ngn3 (1:500) and anti-CA150 (1:250). Blots were developed using CPD-Star, and the resulting images quantified using ImageJ.
Immunohistochemical detection of target proteins was carried out as previously described , except picric acid was omitted. Immunoreactivity was detected using anti-mouse IgG-Alexa-568 and anti-rabbit IgG-Alexa-488. Images were collected using a Bio-Rad MRC 1024 confocal microscope with an Olympus BX 50W camera.
Western blot data were compared using 2-tailed unpaired Student’s t-test, and considered significant at p < 0.05. The Ab array data was analyzed using two methods. First, we analyzed the fluorescent intensity signal using the manufacturer’s suggested protocol (internally normalized ratio, INR) by taking the square root of each protein’s averaged forward and reverse ratios [√(TBI-Cy5/Sham-Cy3)/(Sham-Cy5/TBI-Cy3)]. We then applied the Rank Product method  as outlined in Srivastava et al. . Calculated RP values were converted to approximate expected values (E, analogous to BLAST expected values) as described in . Briefly, we performed 100 simulated experiments with the same parameters as the Ab array, used a random number generator to obtain spot values, and determined the number of times a simulated RP value smaller than a given experimental value occurred. We used E <0.05 as the cutoff to determine which candidates had altered expression. To minimize potential false positives, we only considered proteins meeting both analyses threshold criterion as potential candidates.
To control for differences in protein quantification and labeling efficiency, reciprocally labeled aliquots of sham and TBI extract mixes were incubated on replicate arrays. Figure 1A shows a representative Ab array region incubated with the fluorescently-labeled extract mixes. Proteins exhibiting no change in expression between sham and TBI groups appear yellow, while proteins with altered expression appear red or green. The array had low overall background, and duplicate spots had similar signal intensities. After scanning, the average internally normalized ratio (INR) was calculated for each protein. As recommended by previous investigators, proteins with INRs >1.35 or <0.75 were considered to have different expression levels . We then reanalyzed the data set using Rank Product (RP) analysis . RP analysis has been suggested to provide a more robust and statistically accurate methodology for comparing proteomic data sets than the manufacturer recommended INR analysis . Proteins exhibiting differential expression between sham and TBI groups by both INR and RP analysis were considered potential candidates, and are listed in Table 1.
To independently validate some of the antibody array results, we randomly chose two candidate proteins (AF6, CA150) for western blot analyses. To test selection criteria robustness, glutamine synthetase and neurogenin-3 were also tested because, unlike CA150 and AF6, they met only one analysis methodology criteria and therefore were not considered potential candidates. Figure 1B-D shows representative western blots and summary results for the expression levels of AF6, CA150, and glutamine synthetase. Neurogenin-3 expression was not quantifiable, suggesting its array signal may have resulted from non-specific binding. Also consistent with exclusion as a candidate, glutamine synthetase expression level was not significantly different between sham and 24 h TBI groups (Fig. 1B). However, western blot analysis found significantly decreased expression of both AF6 (Fig. 1C) and CA150 (Fig. 1D) 24 h after injury, consistent with the Ab array data. These results suggest the proteins included in Table 1 are viable candidates with altered expression levels after TBI.
Since a number of studies have indicated that altered catecholamine signaling may contribute to TBI-associated cognitive dysfunction, COMT was selected for further examination [15–17,20,30]. As shown in figure 2A, analysis of 24 h cytosolic extracts showed a mean increase in soluble COMT expression after injury, similar to the results obtained from the Ab array.
COMT is expressed as soluble (S) and membrane-bound (MB) isoforms that are derived from a single gene, though expression is regulated by independent promoters . Therefore, we generated additional total protein extracts to assess possible differential regulation of COMT isoforms. We selected 3 d and 14 d post-injury since these times have been previously demonstrated to exhibit TBI-associated behavioral effects. As shown by the representative blots and summary data (Fig. 2), analysis of 3 and 14 d hippocampal total protein extracts showed significant alterations in both S- and MB-COMT isoform expression. Summary results of S-COMT showed significantly elevated ipsilateral expression at both 3 d and 14 d post-TBI. Interestingly, contralateral S-COMT was differentially regulated, with increased expression at 3 d but decreased expression at 14 d (Fig. 2C). Ipsilateral MB- COMT was significantly elevated at both 3 and 14 days post-TBI, while contralateral expression was increased only at 3 d (Fig. 2D).
COMT expression has been reported in a wide variety of CNS cell types, including neurons, astrocytes, microglia, vascular endothelium, and ependymal cells. Immunohistochemical staining of uninjured sham sections indicated broadly distributed low-level COMT immunoreactivity throughout the hippocampus (not shown). The most striking change in COMT localization following TBI was a large increase in intensely labeled COMT-positive cells throughout the ipsilateral injured hippocampus observed at both 3 and 14 days post-injury. In order to identify this cell population, we performed double immunohistochemical staining (Fig. 3A) using 14 d post-injury tissue sections. The intense COMT immunoreactivity did not co-localize with cells expressing NF-H (neuron), GFAP (astrocyte), GalC (oligodendrocyte, not shown), GAD65/67 or parvalbumin (inhibitory interneuron, not shown). However, we detected co-localization between COMT immunoreactivity and Iba1, which detects both quiescent and activated microglia.
To further investigate the changes in COMT expression, we co-labeled sections from various time points post-TBI with anti-COMT and -Iba1 antibodies. Control sections incubated without primary antibodies had no detectable signal (not shown). Sham sections showed evenly distributed ramified Iba1-positive microglia with low level COMT expression (Fig. 3B). Within 24 h after injury, there is an apparent increase in Iba1+/COMT+ activated (amoeboid) microglia throughout the ipsilateral hippocampus. By 3 d post-TBI, when western blots detected the largest increase in COMT expression, there is a substantial increase in ispilateral hippocampal Iba1+ amoeboid microglia also expressing intense COMT immunoreactivity. In contrast, there was no noticeable increase in either the number or activation state of contralateral hippocampal microglia, indicating the changes in COMT expression are largely confined to the ipsilateral hemisphere. Fourteen days after injury, the Iba1+ microglial population in the ipsilateral hippocampus remained significantly elevated, though the co-localized COMT staining intensity appeared reduced, consistent with the western blot data.
Protein discovery using 2-D gel and mass spectrometry proteomic approaches have aided in deciphering the complex biochemical changes following TBI. [6,11,12]. These approaches, although informative, require specialized expertise and equipment. Antibody arrays, however, provide a target-directed methodology that facilitates high-throughput screening of large numbers of known or suspected candidate proteins. The antibody array analysis presented in this study identified 41 candidates whose hippocampal expression levels were significantly altered 24 h after injury, indicating the complexity of the brain’s response to TBI.
Previous studies of individually selected proteins, as well as global gene chip analysis, have indicated the involvement of many different proteins in TBI pathophysiology (for review, see ). Although some candidates may also appear within DNA microarray data sets, literature searches revealed at least 9 of the candidates in Table 1 have been specifically implicated in injury models or injury-activated pathways. For instance, the increased hypoxia-inducible transcription factor (Hif-1α) protein after TBI is consistent with altered Hif-1α mRNA after TBI , and increased protein expression after ischemic and brain hemorrhage injury [13,22]. Gao et al. recently reported a significant decrease in H3 histone acetylation levels after TBI . This is consistent with increased SMRT expression, a central component of a co-repressor complex that recruits histone deacetylase to the promoter regions of genes . These data, in conjunction with our confirming altered AF6 and CA150 expression (Fig. 1), indicate a high-throughput characterization of protein expression approach using Ab arrays may successfully identify potential candidates involved in TBI pathophysiology.
In the present study, the selection of COMT for a more detailed analysis was guided by a number of previous studies that have implicated catecholamine signaling in the behavioral sequelae of TBI [4,20,21]. Following their release, catecholamines are either taken up by specific transporters and/or are metabolized by monoamine oxidase or COMT. Transporter-mediated termination of catecholamine signaling plays a prominent role in structures such as the striatum, whereas in structures such as the prefrontal cortex that express low transporter levels, inactivation through metabolism may play a more prominent role [14,26]. An increased role for COMT after injury is consistent with the reported decrease in dopamine transporter (DAT) as early as 7 d post-injury, and increases in the levels of TH protein and TH fibers in the frontal cortex [16,28,29]. Western blot data (Fig. 2) indicated significant changes in COMT isoform expression occurred in both hemispheres at both 3 and 14 d post-TBI. These data may reflect changes in COMT expression in a wide variety of cell types, including vascular endothelial, microglial and neuronal cell types. Our double immunohistochemical-labeling experiments (Fig. 3) indicate that at all time points post-injury, microglia in the ipsilateral injured area exhibit enhanced COMT expression. In addition, increased COMT immunoreactivity in the dentate gyrus, hilus, and fiber tracts was also observed. Interestingly, recent research has shown that microglia express dopamine receptor mRNA, and their migration can be stimulated by the application of dopamine, the D1 receptor agonist dihydrexidine, or the D2 receptor agonist quinpirole . It is possible that increased dopamine release after injury  might serve as a chemotactic signal to trigger the recruitment of microglia to distressed regions.
Since scavenging is one of the primary microglia roles in the CNS, enhanced microglia COMT expression may aid in removing and metabolizing catecholamines from damaged areas. Alternatively, the increased COMT immunoreactivity detected in the activated microglia (Fig. 3) might arise from phagocytic activity as they clean the debris from the dead and dying cells in the injured hippocampus. However, elevated COMT immunoreactivity is still associated with microglia at 14 days post-injury. This would indicate that intrinsically up-regulated COMT expression in microglia is a more likely explaination, since phagocytosed COMT would not be expected to remain stable over long periods. Additional experiments are needed to determine the exact mechanism (transcription, translation, or protein stability) through which microglia regulates COMT expression, and its consequences following TBI.
In conclusion, the present study demonstrates the potential of Ab arrays to explore the complex pathophysiology of CNS injury. In addition to corroborating previous studies, the antibody array identified many novel candidate proteins. For instance, 10 of the 41 candidates possess functions related to actin dynamics, remodeling, or cell adhesion. Since blood-brain barrier integrity is compromised following TBI, defining the changes that occur in tight junction proteins (e.g. AF6, symplekin) or signaling proteins that regulate actin dynamics (e.g. PRK2, ILK, ABR, Rho-GDI 2) may be critical to understanding the mechanisms underlying blood-brain barrier dysfunction. Some of these candidates may also be involved structural remodeling after TBI. The identification of altered COMT expression, in addition to other novel candidates, may be useful for developing therapies to improve TBI outcome.
We would like to thank Melanie Moody and Min Zhang for surgical assistance, Anthony Moore for critical comments, and Dr. Ray Grill for confocal assistance. Research supported by National Institutes of Health grants NS35457, NS049160.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.