par Santacruz, Carlos
Président du jury Shiffman, S. N.
Promoteur Taccone, Fabio
Co-Promoteur Debacker, Daniel
Publication Non publié, 2024-09-11
Thèse de doctorat
Résumé : Acute Brain Injury (ABI) is defined as a direct injury to the brain parenchyma that is neither hereditary, congenital, nor acquired at birth. It may be classified into two main types: traumatic and non-traumatic. Non-traumatic ABI can be further divided into infectious and non-infectious categories, while traumatic ABI can be classified as acute or chronic. In this presentation, we will focus on severe acute traumatic and non-infectious ABI. Both types are major contributors to mortality and long-term disability worldwide, encompassing a range of conditions including traumatic brain injury (TBI), subarachnoid hemorrhage (SAH), stroke, and intracerebral hemorrhage (IPH). Globally, traumatic brain injury (TBI) represents a significant burden, with nearly 30 million new cases annually and more than 7 million days lost due to disability. Additionally, a stroke occurs every 3 seconds, meaning that by the end of this meeting, there will be approximately 400 new cases of ischemic or hemorrhagic strokes worldwide. This equates to 12.2 million new cases each year, imposing an economic burden amounting to billions of dollars.The pathophysiology of ABI involves both primary and secondary injury mechanisms. Specifically, the proteomic response to primary injury occurs immediately after the acute insult, resulting in direct neuronal damage, and axonal injury. Secondary injury is characterized by a series of deleterious processes such as excitotoxicity, oxidative stress, inflammation, and apoptosis, which can exacerbate both local and distant damage. During the chronic phase, these proteomic changes have been associated with the development of neurodegenerative diseases such as Alzheimer’s and Parkinson’s disease. Understanding these molecular mechanisms is crucial for elucidating ABI pathophysiology and predicting patient outcomes.So, it makes sense to use Proteomics, the large-scale study of proteins, as a powerful tool in the research of ABI biomarkers. It allows for the identification and quantification of proteins expressed in response to brain injury, providing insights into the underlying mechanisms and potential biomarkers for diagnosis and prognosis. By analyzing the proteomic profiles of ABI patients, we can gain a deeper understanding of the molecular pathways involved after injury and during recovery. Other platforms such as lipidomic or genomics are not well studied in critically ill patients. Previous studies have focus on the role of single grouped proteomic biomarkers (the so-called DAMPs) for the evaluation of prognosis of patients after ABI. Cerebrospinal fluid (CSF) glial fibrillary acidic protein (GFAP) has emerged as a promising biomarker for the prognostication of patients with severe acute brain injury. GFAP is an intermediate filament protein expressed by astrocytes in the central nervous system, and its levels in the CSF can indicate astrocytic activation or injury. In the patients with TBI and IS, elevated CSF GFAP levels were associated with the severity of brain injury and survival and can provide valuable insights into the extent of neuronal damage. In this study, we employed a state-of-the-art MS/MS SWATH technology to analyze cerebrospinal fluid (CSF) samples collected from ABI patients at various time points post-injury. In comparison between micro-LC/MS SWATH proteomic analysis and other proteomic technologies (e.g., shotgun proteomics) reveals significant advantages inherent to the SWATH-MS approach. Notably, SWATH-MS excels in supporting quantitative analyses of peptides that span thousands of proteins with remarkable quantitative accuracy which results in the generation of highly accurate and reproducible quantitative proteomes. Also, the SWATH MS approach offers higher specificity for the identification of peptides as compared to ather proteomic approaches.Initially, we performed a systematic review of the literature to evaluate to current knowledge on the use of CSF biomarkers for the prognostication of patients with severe ABI. The results of our analysis concluded 3 main important findings:1. Many studies focus on single biomarkers, which is highly un-probable that a single protein explains the complete pathophysiological response after ABI.2. Most studies focused on TBI, and a lack of data is scarce on other types of ABI.3. A common pathophysiological response involving pathways related to inflammation, oxidative stress, apoptosis seems to be present in heterogenous cohorts of patients with traumatic and non-traumatic ABI.This first study, published in the NCC journal, included 50 ABI patients and 13 non-ABI patients. Inclusion criteria and objectives of the study are shown in the slide. Importantly, we collected vCSF in a protocolized fashion to reduce the risk of selection bias, followed each patient for 5 consecutive days during the acute period in the ICU and performed a scheduled phone all 3 months after ICU discharge. These are the basal characteristics of the included patients. As you can see, it was a heterogeneous cohort of patients with a vary of different causes of ABI. This is important because it depicts a fundamental part of the hypothesis, which is that a common proteomic response is present after an acute injury to the brain. Also, we focused on severely ill patients, as represented by the high mortality and poor functional outcomeVolcano plots clearly showed significantly downregulated and upregulated vCSFprotein expression between ABI patients and non-ABI patients from day 1 to day 5. According to the PCA results, protein expression in patients with and without ABI clustered into two distinct groups in the first dimension, and this clustering was present across the 5 days.Heatmap analysis revealed up- and downregulation of vCSF proteins in patients with ABI compared to those without ABI and in patients with different types of ABI, with the height of the dendrogram showing links separating ABI patients from non-ABI patients (intergroup variability) but also suggesting different groups among ABI patients (intragroup variability).A Venn diagram illustrated the distinct expression patterns of 42 vCSF proteins expressed in all 34 areas from 5 overlapping ovals corresponding to Day 1 to Day 5 (Figure 8). Considering only individual areas (e.g., individual Days 1 to 5 without any overlapping area), an average of 2-3 proteins per day were significantly expressed, corresponding individually to ~4% of all documented proteins. The proteins that were significantly differentiated but participate from day 1 to day 5 represent most of the total. These proteins seem to be involved in diverse biological processes, such as proteolysis inhibition (e.g., metalloproteinase inhibitor 1), blood coagulation (e.g., fibrinogen alpha chain), inflammation (e.g., protein S100-A8) and metabolic activities (e.g., carbonic anhydrase 1), as well as molecular functions, such as cellular binding and endopeptidase inhibition.We also found significant differences between patients with traumatic vs non-traumatic, high vs normal ICP, but not between patients with good vs bad GOS outcome.To progress to the second phase of the study, we concentrated in the type and the confidence of interactions between proteins that were significant. For this, we used public servers that qualify the interaction between the proteins based on a large database of The way to analyze the interactions between the proteins is by using using graph theory. The average node degree and clustering coefficient along with a significant PPI p-value indicates that interactions among these proteins are not by chance alone and that they are connected biologically as a group. The DAMP ICP 30 shows several interesting characteristics. In the center of the graph, so proteins with a high connectivity, we found the interactions between de Ankerin and Spectrin proteins. These proteins are structural proteins in charge of maintaining the cellular structures in place. After severe proteolysis, these proteins de-gradate and cellular structures are displaced from their original spot. This triggers the activation of current dependent sodium channels that finally produces inward current of water into the cell. Subsequently, increase in intra-cellular water activates caspase-related apoptosis and programed cellular death. Other interesting interaction is the ubiquitin-lead pos-traslational modification of proteins, which means that proteins can pass to unprogrammed activities increasing the risk of secondary injury. Other proteins found in the ICP 30 proteome were related to increase inflammation and cerebrovascular motor tone dysregulation. Briefly, we know that CSF levels of Amyloid 1-42 in AD are hypothesised to be caused by its aggregation and sequestration in cerebral plaques, with less amyloid being available to diffuse into the CSF. Also, Altered apolipoprotein E reduces the cell capacity to transport amyloid beta-plaques reducing inflammation. So, altered proteomic expression of these two proteins could be a risk factor for the development of chronic neuro-degenerative diseases.Compared to non- ABI (control) patients, day 1 to day 5 ApoEswath protein expression was significantly lower after nontraumatic ABI (p<0.0001). APLP1, but not APP or APLP2, was also significantly lower in patients with nontraumatic ABI (p<0.0001) over the study period. No differences in ApoE or Aβ protein levels were found between patients with unfavorable and favorable neurological outcomes.Several critical considerations need to be addressed to establish the validity and reliability of proteomic results as prognostic indicators.First, there is inherent uncertainty associated with proteomic analyses, stemming from technical variability, sample preparation methods, and instrument sensitivity. These factors can contribute to inconsistencies and false positives/negatives in biomarker identification.Second, within-patient variability in proteomic profiles over time adds complexity to the interpretation of results. The dynamic nature of ABI pathology and the influence of various factors, such as treatment interventions and patient-specific variables, may confound the relationship between proteomic signatures and clinical outcomes. Third, while the study identified DAMP patterns associated with ABI subtypes and outcomes, further evidence is needed to establish the causal relationships and mechanistic pathways involved.

Documents en relation

DI-fusion