Heterogeneous Nuclear Ribonucleoprotein K Autoantibodies in Patients who Suffered Severe Traumatic Brain Injury

Heterogeneous Nuclear Ribonucleoprotein K Autoantibodies in Patients who Suffered Severe Traumatic Brain Injury María J. Acevedo-Calado1,4, Marco A. Montes-Cano1,4*, Francisco Murillo-Cabezas1,3, Ainara Madrazo-Atutxa2,4, Salvador ToroLlamas1,4, Ángel Luis Castaño Núñez1,4, José L. Caro Oleas1,4, Alfonso Leal-Cerro2,4 and Antonio Núñez-Roldán1,4 1Servicio de Inmunología, Hospital Universitario Virgen del Rocío, Sevilla, Spain 2Unidad de Gestión de Endocrinología, Hospital Universitario Virgen del Rocío, Sevilla, Spain 3División de Cuidados Críticos, Hospital Universitario Virgen del Rocío, Sevilla, Spain 4Instituto de Biomedicina de Sevilla (IBiS), Sevilla, Spain SOJ Immunology Open Access Research article


Introduction
In recent year's the knowledge about the alterations occurred after Traumatic Brain Injury (TBI) has undergone significant developments [1].These advances have contributed to clarify the relationship between post-traumatic edema and neuropathological sequelae that are largely responsible for involvement of this protein in apoptosis, tumors development, angiogenesis, cell invasion [10][11][12], and also altered gene expression patterns of hnRNPK have been found in many human cancers [13,14].Additionally it has been described, as a high expression of this protein during smooth muscle cell proliferation in both, aortas from animal models of atherosclerosis and in human occluded veins [15].We cannot rule out that hnRNPK auto antibodies could be also over expressed after endothelial damage and inflammation produced after TBI.However, at present there is no information on the involvement of anti-hnRNPK antibodies generation and the outcome of TBI.Therefore, the aim of this work was to analyze the presence of hnRNPK auto antibodies in patients who suffered different severity of TBI and its possible role in the outcome.

Materials and Methods
This study included patients with severe TBI admitted to neurosurgical ICU at the Virgen Del Rocio University Hospital, Seville, Spain, between July 2004 and Jan 2006.The study was approved by the hospital ethics committee and informed consent was obtained from a next of kin, given that all eligible patients were in coma.Inclusion criteria were: (1) male or female over 14 years of age with severe TBI (GCS score ≤ 8 after resuscitation) diagnosed by history and clinical examination with at least one reactive pupil; (2) to obtain at least two serum samples from each patient, one of them done within 24 hours after accident and a second serum sample at least six months post-TBI and (3) haemodynamically stable (mean arterial pressure > 75 mmHg with no or low-dose vasoactive drugs).Patients were excluded based on the following criteria: (1) presence of two reactive pupils; (2) suffered cardiac arrest after TBI; (3) any spinal cord injury, pregnancy, or coma suspected to be primarily due to other causes (E.g: alcohol); (4) suffered multiple injury as measured through the Abbreviated Injury Score (AIS) > 2; (5) no possibility of follow-up during one year and (6) presence of chronic or autoimmune diseases.
Data collection included demographic and clinical variables (age, sex, cause of injury, GCS and pupil reaction after resuscitation, occurrence of prehospital hypotension (systolic blood pressure < 90 mmHg), hypoxia (peripheral oxygen saturation (SpO 2 )) < 90% and occurrence of sepsis (< 4 days).Sepsis was defined according to the Sepsis Consensus Conference criteria.Assessment of overall injury severity was based on the Injury Severity Score (ISS).Patients underwent an initial CT scan after resuscitation.Neuroradiological findings were classified according to the Traumatic Coma Data Bank (TCDB).This classification is divided into six groups: the first four indicate the presence and severity of the diffuse injury and the rest of the categories indicate the presence of an evacuated or non-evacuated mass lesion.A neuroradiologist, blind to the study goals and data reviewed and completed this classification.
Intra-Parenchymal Intra-Cranial Pressure (VENTRIX, INTEGRA Neuroscience, and Plainsboro, NJ), mean arterial blood pressure (obtained with a radial artery fluid-coupled system), Cerebral Perfusion Pressure (CPP), brain tissue oxygen pressure (LICOX, IMC System, GMS Kiel-Mielkendorf, and Germany), endtidal carbon dioxide and SaO 2 were continuously monitored in the ICU.All patients were managed according to Brain Trauma Foundation guidelines and local protocols.Treatment was targeted at maintaining intracranial pressure at < 25 mmHg, CPP at > 60 mmHg, and brain tissue oxygen pressure at > 15 mmHg.The outcome assessment was carried out 12 months after patient discharge using the Glasgow Outcome Score (GOS).To relate our findings with the auto antibodies we dichotomized our results according to the GOS into two groups, severe (2-3) and mild (4-5) sequelae.
Two serum samples were obtained and analyzed from each patient; one at the time of TBI (basal sample, obtained within 24 hours after TBI), and the second within six months after trauma.As healthy controls, we included a total of 124 non related individuals with a range of age 26-53 (39.9 ± 13.9 years).Samples were immediately centrifuged, frozen at -80°C and were stored for later analysis.Anti-hnRNPK IgG antibodies were detected by ELISA using hnRNPK recombinant protein previously purified in our laboratory [9].Recombinant hnRNP-K protein, once affinity-purified by polyhistidine-tag, was plated at 2 µg/ml (in 0.1 mol/liter carbonate-bicarbonate buffer, p H 9.5) onto a polystyrene flat-bottom ELISA plate (Nunc, Roskilde, Denmark) and incubated for 16 hours at 4°C.The non-specific binding of Igs was prevented by adding Tris-buffered saline with TBS-TM for 2 hours.The same serum-positive sample used in the library screening was used to construct a standard curve (dilutions of 1:100 to 1:1,600) to rule out non-specific antibody activities.For each ELISA, standards and samples (diluted 1:100 in TBS-TM) were added to duplicate wells and incubated for 1 hour at room temperature.Plates were washed, horseradish peroxidase conjugated rabbit anti-human IgG (Phadia AB, Uppsala, Sweden) was added, and the plates were for 1 additional hour at room temperature.The enzyme reaction was started by adding 100 µl of 3,3', 5,5'-tetramethylbenzidine and stopped 45 minutes later with 50 µl of 0.5 mol/liter H 2 SO 4 .Finally, the optical density was read at 450 nm in a micro titer plate reader (Bio-Tek Instruments, Winooski, VT).To calculate results, background reactivity of the reference mixture was subtracted to establish the frequency of newly identified antibodies in healthy individuals, a healthy control group of 124 individuals (72 men and 52 women; mean age, 40 ± 1.24 years) was also included.Patient serum samples with optical density values higher than the 95 th percentile for control subjects (0.5966) were considered positive.
The study of Anti-Endothelial Cell Antibodies (AECA) was performed using the Indirect Immunofluorescence (IIF) method on commercially available slides of Human Umbilical Vein Endothelial Cells (HUVEC; Euroimmun, Lübeck, Germany).Antibodies present in patients' sera were detected with a fluorescein-conjugated secondary antibody against human IgG (Euroimmun, Lübeck, Germany) and hnRNP-K was localized using a rabbit anti-human hnRNP-K antibody (AbCam, Cambridge, UK) with a fluorescein conjugated secondary antibody against rabbit IgG (Jackson Immuno Research Laboratories Inc, West Grove, PA).After washing with phosphate-buffered saline-Tween, slides were read on an epifluorescence microscope.Cutoff was set at 1:80 dilutions, at this dilution; all sera from 124 healthy individuals gave negative results.Positive samples were classified according to their IIF pattern.
Statistical analysis was performed with SPSS software version 18.0 (SPSS © , Chicago, IL).Qualitative variables were compared for statistically significant using the χ 2 test and Fisher's exact test.
The Wilcoxon signed-rank test was used to evaluate the statistical significance between the levels of hnRNPK antibodies at basal time and after TBI.The p values below 0.05 were considered statistically significant.

Results
During the present study 32 TBI patients were enrolled.Only 19 patients aged 19-64 (28.48 ± 10.3) met the inclusion criteria.None of the patients were excluded due to cardiac arrest.Twelve patients were excluded due to extra cerebral injury (AIS > 2), and 1-year follow-up could not be done on one patient due to administrative issues.Data concerning age, sex, initial GCS, GOS, mechanism of injury and other clinical information were displayed in Table 1.All patients showed pathological CT findings (none within TCDB category I).Six patients underwent neurosurgery, 4 of them at the admission and two require neurosurgery while in the ICU.No decompression craniectomy (duraplasty plus bone removal) was found in the sample.After discharge from the ICU, patients were admitted to the hospital´s Neuro-Rehabilitation Unit.In this series we did not have any death and the final outcome was evaluated one year after TBI by GOS and was performed by a Neuro radiologist not involved in the study.
AECA auto antibodies were present in 2 patients at basal time and also during post -TBI evolution, while 5 more patients developed de novo auto antibodies after trauma (Table 2).Additionally, after the screening of hnRNPK by ELISA assay, we observed that only 3 (15.8%)out of 19 patients had antibodies at the time of TBI, nevertheless 10/19 (52.6%) developed the antibodies during follow-up, showing a significant increase of anti-hnRNPK antibodies after TBI (Table 3).A statistically significant difference for the AECA factor (p = 0.1) was not reached, although in the case of hnRNPK the differences were statistically significant despite the small number of patients included (p = 0.01).Furthermore, when we analyze the concentration of anti-hnRNPK at the time and after TBI, we detected a significant increase of IgG hnRNPK, that were higher after TBI than the antibody levels found at the time of trauma p = 0.001, (Figure 1).
From our study cohort, eleven patients developed permanent injuries (ranked by GOS) after TBI and also, in all the cases it shows the presence of AECA auto antibodies in the serum samples obtained after trauma, and only one patient with sequelae was negative for hnRNPK antibodies.We found a significant correlation between the appearance of AECA and hnRNPK auto antibodies and the development of permanent injuries (p = 0.018 and p = 0.040 respectively).

Discussion
To our knowledge, this study first demonstrates the    development of AECA and anti-hnRNPK IgG antibodies after TBI and its likely relationship with the outcome.It is well established that the persistent inflammation produced after TBI activate the immune response and culminating in the production of different auto antibodies against endothelial and central nervous system antigens.In fact, the presence of AECA after TBI could be explained by an ulterior exposition of antigens produced by endothelial lesions.Similarly, in human and experimental models, the alterations produced in the blood-brain barrier after TBI have been related to neuronal damage caused by the activation of immune system cells that also impairs neurological recovery after TBI [16].On the other hand, there are evidences of cell-mediated immune response within the brain and the systemic circulation after TBI, were antibodies and B cells are also pivotal players [1].More specifically, the pathological sequelae that accompanies to CNS trauma has characteristics of a self-directed immunological disease, and the production of specific auto antibodies such as gangliosides, phospholipids, beta-tubulin III , nuclear antigens or anti pituitary antibodies has been related with neurological/neuroendocrinological diseases [17][18][19][20][21][22].Although some of these antibodies have been found in healthy individuals, several findings suggest a pathogenic role of AECA auto antibodies in autoimmune rheumatic diseases such as systemic lupus erythematosus, scleroderma and vasculitis.These antibodies are able to induce proinflammatory and procoagulant effects on endothelium (increased expression of adhesion molecules and tissue factors, increased release of cytokines).In addition, in systemic vasculitis such as Takayasu arteritis and antineutrophil cytoplasm antibody-positive vasculitis, AECA have been reported to correlate with disease activity [23].These antibodies have been found after TBI in 15 out of 19 patients and only in five of them were preformed (Table 2).Thus, after trauma a specific immune response against endothelium cell tissue has been occurred.With respect to hnRNPK antibodies, the present study shows that 3 out of the 19 patients (15.8%) had antibodies against hnRNPK at the basal time sera, while after trauma these antibodies could be detected in serum sample from 13 of the patients (68.4%).Among the patients who were hnRNPK antibodies positive before the TBI, 2 suffered some type of trauma prior TBI (one suffered a serious stab wound and the other a motorcycle accident).These patients remained anti-hnRNPK positive antibodies during follow-up.Surprisingly, when patients were stratified by the presence/absence of any type of sequels, 100% and 91% of the patients with permanent injuries were AECA and hnRNPK antibodies positive respectively.

CT-TCDB classification at admission
In the previous study we describe that antibodies against hnRNPK were associated with a vascular endothelial damage in heart transplanted patients [9].Similarly, in the current study, TBI patients who developed autoantibodies after trauma are probably due to vascular endothelium damage.Accordingly, a significant association between the presence and, in particular, higher levels of anti-hnRNPK antibodies were found.These differences could be explained by the endothelial damage/ activation that occurs after of trauma triggering immune mechanisms.Our results can be added to other published studies that demonstrate the importance of the definition of a good marker and/or the combination of different markers in outcome prediction after TBI [5].
A major limitation of this study is the small number of patients included.However, the study included a very homogeneous cohort, with clinical follow-up of at least one year.Furthermore, this study addresses a pathway which is not sufficiently explored, and it is the role of the endothelium in the genesis of some accompanying phenomena to the pathology of TBI.

Summary
Measuring anti-hnRNPK antibodies in patients who suffered a severe TBI could be a helpful biomarker for evaluating endothelial injury as well as the pathophysiological events accompanying these traumas.
More studies including larger populations are needed to clarify whether these antibodies can predict the occurrence of secondary damage produced by the trauma.

Figure 1 :
Figure 1: Immunoglobulin (IgG) titers against human recombinant hnRNP-K in TBI patients at the time of trauma (basal) and post-TBI.The box plots show the median and 25 th ¯75 th percentile range as well as the minimum and maximum values.Lines inside the boxes represent the mean.Circles indicate outliers.Values under the boxes are the percentage of patients with IgG titers above the cut-off value (0.5966), calculated as the 95 th percentile from the control population.

Table 1 :
Demographic and clinical data for 19 patients with severe TBI.

Table 2 :
Evolution of AECA autoantibodies analyzed in basal and post-TBI sera.

Table 3 :
Evolution of hnRNPK autoantibodies analyzed in basal and post-TBI sera.