Persistent increase of Nogo-A-positive cells and dynamic reduction in oligodendroglia lineage cells in white matter regions following experimental and clinical traumatic brain injury

Abstract

White matter (WM) disruption and atrophy is a consequence of traumatic brain injury (TBI) and contributes to persisting cognitive impairment. An increased expression of the myelin-associated axonal outgrowth inhibitor Nogo-A and oligodendrocyte pathology might be negatively associated with postinjury WM changes. Here, we analyzed brain tissue from severe TBI patients, obtained by surgical decompression in the early postinjury phase and postmortem brain tissue of long-term TBI survivors and observed an increased number of Nogo-A⁺ cells in WM tracts such as the corpus callosum (CC). Likewise, the number of Nogo-A⁺ cells in the CC was increased from day 7 postinjury to 6 months postinjury (mpi) following central fluid percussion injury (cFPI) in mice. In addition, the number of Olig2⁺ cells in the CC and capsula externa remained constant, while the numbers of Olig2⁺/CC1⁺ and GST-π⁺ mature oligodendrocytes declined throughout the observation time of 18 months. A significantly lower number of Olig2⁺/CC1⁺ cells was found in cFPI mice compared to controls at 18 mpi. Persistent vulnerability of oligodendrocytes in combination with dynamic alterations of Nogo-A expression may have implications for the WM atrophy and insufficient recovery observed after TBI.

INTRODUCTION

Traumatic brain injury (TBI) is a complex and markedly heterogeneous disease causing long-term neurological complications that can be observed in patients for several months and years after the initial injury.^1–3 Traumatic brain injury is a risk factor for, and may share several pathological pathways with, neurodegenerative disorders including dementias. Different forms of TBI can result in axonal pathology in white matter (WM) tracts and subsequently, degeneration of neurons also in remote brain regions. A single moderate-severe TBI can lead to persistent axonal injury and chronic brain atrophy, known as posttraumatic neurodegeneration, which is associated with poor patient outcomes.^4–8 Indeed, WM injury is one of the major contributors to persistent sensorimotor and cognitive dysfunction following TBI.^9–12

Thus, TBI can lead to atrophy in various regions of the brain, including key WM tracts such as the corpus callosum (CC) and capsula externa (CE). Remyelination can be triggered following WM injury resulting from both diffuse and focal experimental TBI, as well as human TBI. This process requires oligodendrocyte progenitor cell (OPC) proliferation and differentiation.^13–15 Our previous studies showed that OPC numbers increase in acute human TBI, and in WM tracts shortly after diffuse and cortical TBI modeled in rodents.^16,17 However, it is unclear whether this increase persists at later stages after the injury.

Different factors, including increased expression of endogenous myelin-associated inhibitors to axonal regeneration such as Nogo-A, myelin-associated glycoprotein, and oligodendrocyte-myelin glycoprotein, restrict the potential endogenous recovery processes post-TBI. Accordingly, we previously observed increased levels of myelin-associated inhibitors in vulnerable brain regions in different experimental TBI models.^18,19 Nogo-A, a member of the reticulon family, is a molecule that inhibits neuronal outgrowth and dendritic spine formation. It is detected in neurons, but is also found on myelin sheets and in the cytoplasm of oligodendrocytes.²⁰ Nogo-A, in particular, plays a significant role in neurite outgrowth, which is essential to establish neuronal circuits in the process of neuronal plasticity and recovery after TBI. Inhibition of Nogo-A or its main receptor Nogo receptor-1 (NgR1), either by administering NEP1-40, NgR1(310)ecto-Fc, or monoclonal antibodies have consistently shown axonal regrowth and sprouting of uninjured axons and improved functional recovery in rodent models of traumatic spinal cord and ischemic injuries as well as TBI.²¹ Nogo-A expression has not previously been evaluated in chronic experimental and clinical TBI. Hence, the dynamic expression of Nogo-A might be instrumental for remyelination and directed neurite outgrowth.

In the present study, we determined whether oligodendroglia dysfunction in the WM (CC and CE) is a time-dependent processes that starts with the initial injury or whether this becomes prominent over time. Moreover, this is the first investigation of the long-term dynamics of Nogo-A expression in WM tracts in acute and chronic clinical and experimental TBI to understand its contribution to the degenerative and plasticity processes following TBI.

METHODS

Ethics statement

The Regional Research Ethics Committee at Uppsala University granted permission for all experimental and clinical research (decision numbers 2005/103, 2008/303, 2009/89, and 2010/379) included in the study and the research were conducted in accordance with the ethical standards given in the Helsinki Declaration of 1975, as revised in 2008. All animal experiments were approved by the Uppsala County Animal Ethics committee and followed the rules and regulations of the Swedish Agricultural Board. For the acute TBI cohort, informed consent was obtained from the patient’s closest relative to allow inclusion into the tissue bank (Uppsala Brain Bank-Trauma). At approximately 6 months postinjury, the patients who had survived and recovered sufficiently were contacted for informed consent. For the chronic TBI cohort, brain donors in this study were part of the Pacific Northwest Brain Donor Network, which collects tissue from military veterans and civilians, both with and without a history of TBI in collaboration with local Seattle area medical examiners. Consent for brain donation was obtained from the legal next of kin who agreed to be contacted and were interested in the UW research program. All studies at the University of Washington (UW) utilizing brain donation are approved by the UW School of Medicine Compliance office and Institutional Review Board and collected with informed consent.

For comparisons, frontal lobe control samples (n = 5) from the Uppsala Biobank/archived sample collection at the Department of Clinical Pathology and Cytology were obtained from uninjured patients without previous history of TBI or neurodegenerative disorders and who had died from systemic, unrelated causes. Written consent was obtained from each patient prior to inclusion in this biobank.¹⁵

Patient cohorts

TBI groups

Eleven male TBI patients were included. Demographics and clinical characteristics are shown in Table 1. In the acute cohort, requiring surgical treatment in the first postinjury week, all 5 patients had severe TBI, defined as a postresuscitation Glasgow Coma Scale score ≤8. Patients were endotracheally ventilated and sedated and continuous measurements of intracranial pressure and cerebral perfusion pressure were performed. Patients included in this study were subjected to surgical focal decompression due to life-threatening elevations of intracranial pressure and/or the presence of a space-occupying brain swelling or hemorrhage. The patients in the chronic cohort had a moderate-severe TBI, defined as the presence of a TBI associated with coma or hospitalization greater than 1 week or skull fractures/contusions for severe, and a TBI associated with disabling symptoms such as headaches, posttraumatic stress disorder, or seizures following injury. These definitions are not based of the Glasgow Coma Scale scores as medical records are not always available for brain donors; much of the TBI details are given by family report. All chronic TBI brain donors survived for many years postinjury (Table 1); the postmortem interval (time from death to tissue fixation) was 17 ± 21 h (range 5-59 h) and cases did not have significant neurodegenerative disease pathology in the neocortex, near regions of analysis for this study. The chronic TBI cohort brain tissue samples were obtained from the University of Washington Biorepository and Integrated Neuropathology (BRaIN) Laboratory as described above which is part of the CONNECT-TBI study (U54 NS115322) and supported by the Alzheimer’s Disease Research Center (AG066509), the Adult Changes in Thought Study (AG006781), the ENRICH study (lEveraging Nationwide Research Infrastructure to EnriCH Brain Health after TBI) (DoD W81XWH-21-S-TBIPH2), and the NEW-HOPE-TBI study (Neuroimaging and clinical Endpoints With High-dimensional analysis OF Pathologic Endophenotypes in TBI) (U01 NS137484). All patients were >16 years of age at the time of injury and no patient had any other known neurological disorder prior to their TBI.

Control group

The control patients had no known previous history of TBI or neurodegenerative disorders. Control patients were 56 ± 6 years old (range 50-61). The postmortem time of the included control patients (time from death to tissue fixation) was 40 ± 17 h (range 10-48 h). There were no differences in age among the cohorts (Kruskal-Wallis test, P = .229).

Tissue collection and handling

Tissue handling followed the same protocol for both TBI and control patients. In the acute cohort using surgically resected tissue, brain tissue was immediately put in a routinely used fixative, 4% buffered formalin (HistoLab Products AB, Gothenburg, Sweden, cat. no. 02176). The samples were fixed for 24-72 h and then paraffin-embedded and processed by hardware TissueTek VIP (Sakura, CA, USA). Seven micrometer-thick sections were cut and placed on SuperFrost plus slides (Menzel-Gläser, Vienna, Austria) for immunohistochemical analysis. Brain sections (thickness 5 µm) from the cohort with chronic TBI were paraffin-embedded on SuperFrost plus slides. Intact sections containing predominantly intact WM were chosen for subsequent staining.

Immunohistochemistry

Staining for Nogo-A was performed as well as colabeling of Nogo-A with the astrocytic marker glial fibrillary acidic protein (GFAP), the neuronal marker neuronal-specific nuclear protein (NeuN), and oligodendrocyte marker CC1 (antiadenomatous polyposis coli; APC, clone CC1), respectively. Times for development were standardized, and all samples were run in parallel for each staining protocol.

Single labeling

Slides were placed in xylene and rehydrated as described previously.¹⁵ After washing in deionized water, slides were placed in citrate buffer (pH 6.0) and microwaved 2×10 min at 500 W. The slides were allowed to cool for 30 min and then washed in phosphate-buffered saline (PBS) 3×5 min. Endogenous peroxidase activity was blocked (Bloxall, Dako Real AB, Stockholm, Sweden) for 10 min and nonspecific binding was blocked using 5% normal rabbit serum in 1× PBS+0.1% Triton X-100 for 1 h in room temperature (RT). The primary antibody (rabbit anti-Nogo-A, Novus NBP2-41071, diluted at 1:200) was applied in 1× PBS+0.1% Triton over night at 4 °C. Samples were washed with 1× PBS+0.1% Triton X-100 3×5 min and then incubated with biotinylated antirabbit IgG (Vector Laboratories, Burlingame, CA) for 1 h followed by incubation with avidin-biotin-peroxidase complex (ABC; Vectastain Elite Kit, Vector Laboratories) for 30 min. After washing samples, 3,3-diaminobenzidine was used as chromogen (Vector Laboratories) for visualization and hematoxylin was used as counterstain. Finally, samples were washed in deionized water, dehydrated and placed in xylene for 2 min and mounted using Pertex. Nogo-A-positive cells of 1 eye view in the CC were quantified manually and normalized to cells per square millimeter.

Colabeling

To study the expression pattern of Nogo-A, in TBI tissue, colabeling with markers for neurons, astrocytes, and oligodendrocytes was made. Samples were placed in xylene and rehydrated. After antigen retrieval using citrate buffer, samples were blocked using background sniper (Histolab Products AB, Gothenburg, Sweden) for 10 min at RT. All washing steps used 1× TBS+0.1% Triton X-100. A mixture of antibodies against the 3 different cell types and Nogo-A, respectively, was used. Antibody incubation was performed over night at 4 °C and washed away before incubation with MACH 2 Double Stain 2 (BioCare Medical, Pacheco, CA, USA). For antibodies made in rabbit, the HRP substrate Vina Green (chromogen substrate, BioCare Medical) was used according to manufactures’ instructions giving a blue/green result and for antibodies made in mouse, the AP substrate Warp Red (chromogen kit BioCare Medical) was used according to manufactures’ instructions giving red positive cells. Samples were then quickly placed in 99% EtOH and in xylene before mounting using Pertex.

Experimental studies

Animals and the central (midline) fluid percussion injury TBI model

Adult male C57BL/6 mice (Taconic, Silkeborg, Denmark) were housed with free access to food and water for a minimum of 7 days prior to surgery. The surgical procedure for central (midline) fluid percussion injury (cFPI) has been described in detail previously.¹⁴ Briefly, anesthesia was induced by placing the mouse in a ventilated Plexiglas container with 4% isoflurane in air, and then maintained by isoflurane 1.2%-1.4% and N₂O/O₂ 70%/30% delivered through a nosecone, after being secured to a stereotaxic frame. The scalp was anesthetized using a subcutaneous injection of bupivacaine. The scalp was cut open and the skull exposed. Centered 2.5 mm behind the bregma suture, a 3.0- mm diameter craniotomy was made over the midline by keeping the dura mater and the sagittal sinus intact. Then, a plastic cap placed over the craniotomy and securely glued to the skull around was filled with isotonic saline at RT and attached to the Luer-Lock on the fluid percussion device (VCU Biomedical Engineering Facility, Richmond, VA, USA). To induce the diffuse TBI, the fluid percussion pendulum was released to create a pressure wave subsequently transmitted into the cranial cavity (19.5 ± 1.1 PSI). The injury-induced apnea was recorded (39 ± 13 s [mean±SD]). In total, acute mortality from cFPI was 8 out of 35 mice, 1 animal was found dead in its cage at day 9. Sham-injured animals were subjected to the same anesthesia and surgery, but the pendulum was not released. The cap was removed following surgery, the bone flap replaced, and the skin sutured using resorbable sutures. In total, 77 mice were used for the experiments.

Immunohistochemistry

Coronal brain sections were quenched with 3% H₂O₂ peroxidase solution for 15 min, then blocked in PBS supplemented with 0.5% Triton X-100 and 3% normal Donkey Serum (Sigma G9023) at RT for 1 h. Afterward, sections were incubated in Nogo-A (diluted at 1:200; BD Biosciences, Franklin Lakes, NJ, USA), myelin basic protein (diluted at 1:4000; Abcam ab 40390), or GST π (diluted at 1:400; BD Biosciences) antibody diluted in blocking solution at 4 °C overnight. After washing steps in 0.5% Triton X-100 in PBS (PBS-TX), sections were incubated with an antirabbit biotinylated secondary antibody (1:500, 711-065-152, Jackson ImmunoResearch, Baltimore, MD, USA), in 3% NDS PBS-T, at RT for 2 h, and the signal was enhanced by using Vectastain ABC Elite kit (Vector Laboratories). Staining was revealed using chromogen 3,3-diaminobenzidine-tetrahydrochloride (Dabsafe, Saveen Werner AB, Limhamn, Sweden) and 3% H₂O₂. Sections were dehydrated in consecutively higher concentrations of ethanol, followed by xylene and mounted using Pertex (Histolab AB, Gothenburg, Sweden). Stained sections were analyzed using Olympus BX51 light microscope and CellSens digital imaging software. Figures were composed using Photoshop CS5 software. To analyze Nogo-A-positive cells 5 high-resolution pictures from consecutive coronal sections of the CC and external capsule (EC) per brain were obtained at magnification 20×, using brightfield Olympus BX51 microscope.

Immunofluorescence

Coronal brain sections were blocked for 1 h in PBS-TX supplemented with 3% normal Donkey Serum after 3 times washing in PBS. Sections were incubated at 4 °C in primary antibodies (Oligodendrocyte transcription factor, Olig2, R&D Systems, diluted at 1:400 and anti-APC clone CC1, CC1, Calbiochem, diluted at 1:200) diluted in blocking solution overnight. After incubation with primary antibodies, sections were washed in PBS-TX. Thereafter, sections were incubated with corresponding secondary antibodies (Life Technologies and Jackson) (diluted at 1:500 in blocking solution for 2 h at RT). Sections were mounted on slides and images captured using a Zeiss LSM 780 Microscope.

Quantification of immune-positive cells in the CC and EC was performed by using standardized frames that was overlaid on respective brain areas (bregma −1.3 to 1.9, CC, lateral 0.5 mm each hemisphere, and EC, lateral 0.5-2.25 mm). A person blinded to the experimental conditions counted immunopositive cells from 3 consecutive coronal sections per animal as described previously.²²

Statistical analyses

Data were processed with GraphPad Prism 8.1.2 software (GraphPad, San Diego, CA, USA) including preparation of figures and statistical analysis. In a first step, all data underwent a test for normal distribution. The Kolmogorov-Smirnov did not show normal distribution of Nogo-A-positive cells. Therefore, the number of Nogo-A-positive cells are presented as medians with whisker including individual number of cells per animals. In all other subsequent analyses, statistical analysis was performed by 1-way ANOVA with Bonferroni correction. The number of GST-π-, Olig2-, and CC1-positive cells are shown as means±SE of the means (SEM). Differences in the number of oligodendrocyte lineage cells were calculated using the 2-way ANOVA with Tukey’s or Sidak’s posthoc multiple comparison. Detailed information on significances is provided in the Supplementary results. Throughout all experiments, the numbers of replicates may vary due to sections excluded related to the quality of sections or staining in individual experiments.

RESULTS

Dynamics of oligodendroglia lineage cell populations after cFPI

To evaluate if cFPI TBI affected the number of oligodendroglia lineage cells, we conducted immunohistochemistry/immunofluorescence studies. In both regions of interest, the CC and EC, we quantified the number of oligodendrocyte transcription factor (Olig2)⁺, anti-APC clone CC1, (CC1)⁺, and glutathione-S-transferase pi-isoform (GST-π)⁺ cells following cFPI and sham-injured mice.

First, we used Olig2, an oligodendroglia lineage marker for immunohistochemistry of the EC and CC (Figure 1A and B). In the CC, there was no change in the number of Olig2⁺ cells in the CC at any time point. In the EC, however, there was a late and significant reduction of Olig2-positive cells in the CE of mice subjected to cFPI at 18 months postinjury (n = 4048 ± 235 cells, P < .03) when compared to the sham-injured controls (n = 5319 ± 243) (Figure 1B).

To further explore long-term TBI-induced changes to the oligodendroglia population, we used CC1, a commonly used antibody for mature oligodendrocytes. This antibody does not label myelin, and colabeled with Olig2 in the CC (Figure 2A) and EC (Figure 2B). In the CC, cells colabeled with Olig2 and CC1 did not show any difference between the sham and cFPI injury groups at any time points (Figure 2A). However, we observed a gradual decline in the number of Olig2/CC1-positive cells in both groups from 30 days that was pronounced by 18 months postinjury. Moreover, the number of Olig2/CC1-positive cells in the CC was lower in sham-injured mice at 18 months compared to all other experimental groups subjected to sham injury and analyzed at earlier time points.

In the EC, the Olig2/CC1⁺ cells did not change from 2 to 18 months postinjury in sham-injured animals (Figure 2B), in contrast to the CC. In brain-injured animals, there was a reduction in the number of Olig2/CC1⁺ cells between day 7 and 14 to 18 months postinjury (Figure 2B). At 18 months postinjury, the number of Olig2/CC1⁺ cells was significantly lower in cFPI animals when compared to sham-injured controls (Figure 2B).

To confirm changes in mature oligodendrocytes in the CC and CE, we performed additional studies and quantified the number of GST-π⁺ cells (Figure 3). GST-π is another marker for mature oligodendrocytes that is also expressed in myelin. Here, in the CC, the number of cells remained stable in sham-injured mice during the first 30 days, with a decline at 18 months postinjury. There was a decreased number of cells in mice subjected to cFPI at 7, 14, and 30 days when compared to controls. In contrast, at 6 months of recovery, an increased number of GST-π⁺ cells was found in brain-injured mice when compared to sham-injured controls. Twelve months later, at 18 months postinjury, the number of cells dropped significantly and there was now no difference between sham-injured and cFPI mice. Similar observations were made in the EC (Figure 3). Together, these results show dynamic changes in oligodendroglia both at early and delayed time points postinjury. Significantly, lower cell counts were found in the CC and EC of aged mice, in both sham and cFPI groups, a possible combined effect due to injury and age.

Delayed increase of Nogo-A-positive cells in mice subjected to cFPIs

Mice were subjected to cFPI or sham injury, and endpoint analyses were performed at different time points after the injury. Similar to TBI patients, we quantified the number of Nogo-A⁺ cells in the CC and EC for up to 18 months after the insult (Figure 4A). Compared to sham-injured mice, we found increased numbers of Nogo-A⁺ cells in the CC at 7 days after cFPI (Figure 4B). Quantification of cells revealed a constant number of Nogo-A⁺ cells in sham-injured animals throughout the observation period of 18 months, covering almost the entire life span of a mouse. In contrast, the number of Nogo-A-positive cells were significantly higher in mice subjected to cFPI (Figure 4C). While no difference between the sham-injured and cFPI mice was observed at day 2, the number of cells were significantly higher at 7, 14, and 30 days as well as 6 months postinjury. At 18 months following cFPI, the number of Nogo-A-positive cells were comparable to levels found in sham-injured mice at that time point.

Increased numbers of Nogo-A-positive cells in brain tissue acutely and chronically in human TBI tissue

Expression of the myelin-associated inhibitory protein Nogo-A was studied in TBI patients. The mean age of the patients was 56.4 ± 9.4 years (range 24-78 years), younger than controls (P = .002). The contusions were surgically evacuated at a mean of 109.2 ± 45.8 h (range 4-240 h) postinjury, with only 1 patient included beyond the initial postinjury week. Similarly, in the chronic cohort brain tissue from TBI patients deceased 3-19 years after the insult were also analyzed postmortem (Table 1).

Nogo-A⁺ cells appeared cytoplasmic and rounded (Figure 5A-C), in particular in samples obtained in the early postinjury phase. At both the acute and chronic time points, increased immunoreactivity was observed in WM tissue of acute and chronic TBI patients (P < .05; Figure 5D). There were increased number of Nogo-A+ cells in the acute TBI group with a median of 148 cells/mm² (Q1: 143 cells/mm², Q3: 158 cells/mm²) (Figure 5B) as well as in patients with chronic TBI (105 cells/mm², Q1: 75 cells/mm², Q3: 138 cells/mm²) (Figure 5C) compared to the control group (average 11 cells/mm², Q1: 9 cells/mm², Q3: 12 cells/mm²; Figure 5D).

While most Nogo-A⁺ cells showed an oligodendroglia-like morphology, it is known that Nogo-A is expressed in other cell types. Thus, in the acute brain tissue samples, we performed colabeling of Nogo-A (blue/green) with neuronal (NeuN: red; Figure 6A and C) and oligodendroglia markers (CC1: red, indicated by arrows in Figure 6B). No colabeling with reactive GFAP-positive astrocytes (red) was observed (indicated by arrows in Figure 6D).

DISCUSSION

Traumatic brain injury causes a number of long-term symptoms leading to a reduced quality of life and difficulties with reintegration into society. Impaired cognition and other higher brain functions are common postinjury; WM injury and progressive WM atrophy may contribute to these deficits. In addition, impaired brain plasticity may reduce the potential for rehabilitation and recovery.²³ To develop treatments to overcome these long-term complications, the underlying mechanisms need to be identified.

Changes of oligodendrocyte lineage cells in WM tracts

Following TBI, significant numbers of mature oligodendrocytes undergo acute or delayed cell death.^5,17 Acute mechanisms include excitotoxicity as well as other noxious effects on cells, for example, an increase of reactive oxygen species or energy depletion.²⁴ Secondary degeneration of oligodendrocytes may, on the other hand, also be caused by delayed decline of injured or functionally inactive neuronal tracts caused by, for example, a prolonged inflammatory response.²⁵ Beyond the first postinjury weeks in human TBI, progressive atrophy and demyelination of WM tracts have been observed using both histological²⁶ and radiological methods.²⁷ One oligodendrocyte can myelinate up to 50 axons and provide important trophic function for the axons.²⁸ Changes in the number of oligodendrocytes observed in our present study both in the initial and late postinjury time points, may thus markedly influence WM function and brain connectivity. The EC contains various circuits connecting different brain regions and is more susceptible to diffuse TBI due to its structural composition and mechanical force distribution.²⁹ Conversely, the CC may respond differently to injuries caused by cFPI. The differential responses observed in the CC and CE indicate that the interplay between Nogo-A and GSTpi can vary across different brain regions; differences in the impact of the injury may contribute to this. For instance, in the CE, where significant reductions in Olig2⁺ CC1⁺ cells were observed at 18 months, Nogo-A levels may exert a more pronounced negative effect on recovery. On the other hand, we cannot rule out that aging per se may exacerbate the loss of Olig2⁺ OPCs, potentially impairing recovery due to a loss of regenerative capacity.

Increased number of oligodendrocyte precursor cells has been observed as early as 2 days in mice subjected to cFPI³⁰ or CCI injury,³¹ in a model of mild TBI,³² and in surgically resected tissue of severe TBI patients.⁵ An increased number of these cells might be the immediate response to replace the loss of mature oligodendrocytes. These studies did not extend their observations beyond the initial postinjury months. Here, we found that the population exclusively expressing Olig2 remained stable throughout the observation time of 18 months. A lower number of cells was only found in the CE of mice subjected to cFPI compared to sham-injured animals at the same time. The lower number of GST-π-positive cells in the CC and CE at 18 months postinjury has been communicated previously.³¹ Delayed changes in the number of oligodendrocyte lineage cells suggest the presence of yet unknown persistent molecular cues and cascades involved in delayed cell death processes and oligodendrogenesis at later stages of senescence following TBI. Such cascades may involve the activation of developmental programs of oligodendrogenesis, increased levels of growth factors, or a downregulation of factors that inhibit oligodendrogenesis.³³

Increased number of Nogo-A-positive cells in the injured brain

When we assessed Nogo-A-positive cells in WM tracts using the cFPI model in mice with a postinjury survival time of up to 18 months, we observed increased number of Nogo-A+ cells from 7 days to 6 months postinjury in our mouse TBI model. In parallel, there were increased number of Nogo-A+ cells during the initial postinjury weeks in surgically resected tissue from severe TBI patients. These findings indicate persistent injury processes in the WM tracts post-TBI that have important implications for the insufficient recovery and the persistent brain network dysfunction commonly observed in TBI patients.³⁴ In humans, regardless of the severity of the initial traumatic insult, multiple pathophysiological mechanisms contribute including sustained injury cascades that result in prolonged motor and cognitive deficits.⁶ Indeed, widespread axonal pathology in humans is commonly observed even in predominately focal TBI, or in low-energy trauma.³⁵ Similarly, in the model of cFPI, the physical impact of the delivered fluid pressure pulse leads to alterations in brain function and disruption in the neuronal circuits as well as in the WM.^22,36 This experimental TBI model consistently produces long-term behavioral and cognitive changes, although rarely evaluated beyond the first postinjury month,^14,37,38 associated with morphological changes such as reactive sprouting observed as early as 1 day after the injury.^14,39 On the other hand, we analyzed brain tissues from TBI patients for the myelin-associated inhibitor Nogo-A and observed that the number of Nogo-A⁺ cells was increased both acutely and chronically in human TBI, with some important differences. These findings indicate that the severity of the injury and the duration of the postinjury time can lead to different patterns of Nogo-A expression in patients with acute and chronic TBI. In acute TBI, immediate inflammatory responses, oxidative stress, and glial cell activation can lead to higher levels of Nogo-A expression to inhibit axonal sprouting and maladaptive neurodegeneration.^39,40

We previously showed that early anti-inflammatory treatment resulted in less oligodendrocyte death arguing for an important role for the inflammatory process post-TBI.⁴¹ Conversely, chronic TBI may result in prolonged or dysregulated levels of Nogo-A, potentially due to persistent inflammation or limited neuroplasticity due to WM atrophy and ongoing axonal pathology.²⁶ On the other hand, the cFPI model in mice is characterized by increased blood-brain barrier (BBB) permeability, diffuse axonal injury, and significant astroglial activation, mainly in the subcortical WM.³⁹ Although Nogo-A is primarily known as an axonal growth and regeneration inhibitor, it may also modulate cellular responses to oxidative stress⁴² and inflammation.⁴³ The persistent upregulation of Nogo-A and subsequent oligodendrocyte loss following TBI can thus be driven by chronic inflammation, BBB changes, and oxidative stress. Consistently, increased levels of Nogo-A, particularly in the chronic and acute phases post-TBI, can coincide with TBI-induced oxidative stress or inflammation, as indicated by the reduction in GSTpi levels in the CC.⁴⁴ Moreover, persistent expression of Nogo-A in chronic TBI was consistent with a delayed reduction in Olig2⁺ OPCs at 6 months in CC indicating a prolonged vulnerability to these cells. The Nogo-A pathway has an important role in mediating inhibition of axonal outgrowth and plasticity in the injured CNS after binding to, for example, the Nogo-66 receptor 1 (NgR1).⁴⁵ This indicates that successful axonal regeneration can create an microenvironment conducive to remyelination and activate OPCs to differentiate into mature oligodendrocytes. In addition, the Nogo-A/NgR1 pathway may be of importance also under normal physiological circumstances,⁴⁶ with involvement in circuit consolidation and to fine tune the CNS.⁴⁷ Inhibition of neurite outgrowth and structural plasticity by Nogo-A has been described in different CNS injury models such a spinal cord injury, TBI and stroke.^30,48–50 Thus, a prolonged inhibition of neurite outgrowth and plasticity by the reactive upregulation of Nogo-A is likely, a hypothesis that needs confirmation in subsequent studies. Pharmacological inhibition of Nogo-A as well as experimental approaches using genetically modified animals have shown that reduced levels or deficiency of Nogo-A is beneficial for the recovery of lost neurological functions in most studies in different spinal cord or brain injury models.⁵¹ Thus, these previous reports and the results of our present study suggest that observed changes in Nogo-A may contribute to an insufficient recovery post-TBI. However, surprisingly, an impaired recovery in Nogo-A/B deficient mice was observed following TBI, suggesting a complex role for this myelin-associated inhibitor in TBI.^52,53 In fact, prolonged upregulation of Nogo-A might be beneficial for neuronal survival by inactivating stress signaling cascades.⁵⁴

This study has several limitations. One key limitation is the variability in postmortem intervals, which can range from months to several years and are affected by numerous factors in humans. However, the survival times of several years indicated that the findings are chronic. There are obvious caveats when comparing rodent to human TBI brain tissue and important differences in brain structure, gray matter-WM volumes and biomechanical factors as well as gyrencephalic humans versus lissencephalic rodents, are merely a few of the potential reasons for a different response to TBI. However, the present study evaluates changes up to 18 months postinjury in mice, that is, almost across the entire lifespan of a mouse. While the duration cannot be directly compared to the human samples, we argue that both represent the chronic phase post-TBI. Furthermore, while the prolonged disruption of oligodendrocytes following TBI is significantly influenced by Nogo-A expression in cases of human chronic TBI, a detailed analysis of these dynamics in WM was not performed due to the small cohort size.

In conclusion, we found short- and long-term upregulations of Nogo-A in brain tissue of patients suffering from prior TBI. Likewise, we found similar results in mice subjected to diffuse experimental TBI induced by cFPI. Interestingly, the insult was accompanied by decline of mature oligodendrocytes in the CC and CE which was followed up for 18 months postinjury. Long-term preservation of mature oligodendrocytes might be a potential target in future therapeutic approaches to maintain WM function after TBI.

ACKNOWLEDGMENTS

The authors wish to thank Carin Sjölund for excellent technical assistance and are grateful to Prof. Irina Alafuzoff and Prof. Martin Ingelsson for establishing the Uppsala Brain Bank-Trauma. Tissue samples from acute TBI patients were obtained from Uppsala University Brain Bank-Trauma. Tissue samples from the chronic cohort of moderate-severe TBI patients were obtained from the University of Washington Biorepository and Integrated Neuropathology (BRaIN) Laboratory which is part of the CONNECT-TBI and TRANSFORM-TBI studies (U54 NS115322, U01NS137500) and supported by the Alzheimer’s Disease Research Center (AG066509), the Adult Changes in Thought Study (AG006781), the ENRICH study (lEveraging Nationwide Research Infrastructure to EnriCH Brain Health after TBI) (DoD W81XWH-21-S-TBIPH2), and the NEW-HOPE-TBI study (Neuroimaging and clinical Endpoints With High-dimensional analysis OF Pathologic Endophenotypes in TBI) (U01NS137484).

SUPPLEMENTARY MATERIAL

Supplementary material is available at academic.oup.com/jnen.

FUNDING

This work was supported by Swedish Research Council (toN.M.), Alborada Trust (to N.M.), Swedish Brain Foundation (to N.M. and K.R.), Crafoord Foundation (to I.Ö.), Skåne University Hospital ALF funds (to N.M. and K.R.), and Hans-Gabriel and Trolle Wachtmeister Foundation (to N.M. and K.R.).

CONFLICTS OF INTEREST

The authors declare no conflict of interest.

REFERENCES