Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

RIPK1 or RIPK3 deletion prevents progressive neuronal cell death and improves memory function after traumatic brain injury

RIPK1 or RIPK3 deletion prevents progressive neuronal cell death and improves memory function... Traumatic brain injury ( TBI) causes acute and subacute tissue damage, but is also associated with chronic inflam- mation and progressive loss of brain tissue months and years after the initial event. The trigger and the subsequent molecular mechanisms causing chronic brain injury after TBI are not well understood. The aim of the current study was therefore to investigate the hypothesis that necroptosis, a form a programmed cell death mediated by the interaction of Receptor Interacting Protein Kinases (RIPK) 1 and 3, is involved in this process. Neuron-specific RIPK1- or RIPK3-deficient mice and their wild-type littermates were subjected to experimental TBI by controlled cortical impact. Posttraumatic brain damage and functional outcome were assessed longitudinally by repetitive magnetic resonance imaging (MRI) and behavioral tests (beam walk, Barnes maze, and tail suspension), respectively, for up to three months after injury. Thereafter, brains were investigated by immunohistochemistry for the necroptotic marker phosphorylated mixed lineage kinase like protein(pMLKL) and activation of astrocytes and microglia. WT mice showed progressive chronic brain damage in cortex and hippocampus and increased levels of pMLKL after TBI. Chronic brain damage occurred almost exclusively in areas with iron deposits and was significantly reduced in RIPK1- or RIPK3-deficient mice by up to 80%. Neuroprotection was accompanied by a reduction of astrocyte and microglia activation and improved memory function. The data of the current study suggest that progressive chronic brain damage and cognitive decline after TBI depend on the expression of RIPK1/3 in neurons. Hence, inhibition of necroptosis signaling may represent a novel therapeutic target for the prevention of chronic post-traumatic brain damage. Keywords: Traumatic brain injury, Chronic posttraumatic brain damage, Magnetic resonance imaging, Necroptosis, Ferroptosis, Neuroprotection Introduction of death and disability in all age groups worldwide, espe- With an estimated case load of 69 million per year [1], cially in children and young adults. The incidence of TBI traumatic brain injury (TBI) represents a leading cause is expected to increase in the coming decades, as the number of the two main etiologies—motor vehicle acci- dents and falls—are expected to rise due to an increase in motorization and an aging population, respectively [2]. *Correspondence: nikolaus.plesnila@med.uni-muenchen.de; nicole. terpolilli@med.uni-muenchen.de The socio-economic impact of TBI is vast, with estimated Nikolaus Plesnila and Nicole Angela Terpolilli authors are equally costs of approximately 400 billion US$ annually [3]. This contributed to this work. number does not only include the direct costs due to Institute for Stroke and Dementia Research (ISD), LMU Klinikum, Ludwig- Maximilians University Munich, Feodor-Lynen-Str. 17, 81377 Munich, acute primary care, but also long-term follow-up costs Germany since many TBI survivors suffer from mood changes, Full list of author information is available at the end of the article © The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Wehn et al. acta neuropathol commun (2021) 9:138 Page 2 of 18 memory deficits, and loss of fine motor skills, have dif - Material and methods ficulties returning to their previous occupation, and therefore require lifelong support [4–6]. Consequently, Ethical statement TBI is increasingly recognized as a chronic neurological All animal experiments were reviewed and approved by disorder with socio-economic implications comparable the Ethical Review Board of the Government of Upper to conditions like Alzheimer’s disease or other neurode- Bavaria. The results of the study are reported in accord - generative disorders [7]. ance with the ARRIVE guidelines [24]. Animal hus- While the pathophysiology of acute brain damage has bandry, health screens, and hygiene management checks been investigated in detail in experimental animals and were performed in accordance with Federation of Euro- in humans over the past decades, relatively little is known pean Laboratory Animal Science Associations (FELASA) about the mechanisms determining long-term outcome guidelines and recommendations [25]. Only male mice after TBI. Chronic functional deficits in TBI patients between 6 and 8  weeks old mice were used. All surgical may be caused by progressive brain atrophy in cortex procedures, behavioral testing, imaging, and data analysis and hippocampus [8, 9] and hydrocephalus formation were performed in a randomized fashion by a researcher [10–13]. So far, clinical and experimental studies suggest blinded to the genotype and group allocation of the ani- that inflammation plays an important role for the devel - mals. Group allocation was obtained by drawing lots by a opment of chronic posttraumatic brain damage [14–18], third party not involved in the study or data analysis. however, the cellular and molecular mechanisms down- stream of this process are not fully understood. Particu- Animals larly, the trigger and the intracellular signaling cascades Inducible neuronal Ripk1 cKO mice were generated as causing neuronal cell death weeks and months after TBI previously described [26] and bred as Ripk1fl/fl (WT) are still unknown. and Ripk1fl/fl::Camk2a-Cre + /Cre (cKO) in our facility. Necroptosis is a form of necrotic regulated cell death, For induction of neuronal RIPK1 deletion, Ripk1 cKO which involves the upstream assembly of the necropto- mice (starting at 4  weeks of age) received three intra- some complex formed by the interaction of receptor peritoneal injections of 2 mg Tamoxifen (Sigma-Aldrich, interacting protein kinase 1 and 3 (RIPK1 and 3) [19] and Taufkirchen, Germany, # T5648) in a 100 µl Miglyol sus- downstream RIPK3-mediated phosphorylation of mixed pension (Caelo, Hilden, Germany, #3274) every 48 h (day lineage kinase like protein (MLKL). Necroptosis can be 1, 3, 5). RIPK3-deficient mice were generated and kindly initiated by activation of Toll-like receptors (TLR) 3 and provided by V. M. Dixit, Genentech Inc., San Francisco, −/− 4, TNF-alpha receptor 1 or, as more recently shown by CA, [27] and bred heterozygously to obtain Ripk3 +/+ us and others, by cylindromatosis (CYLD)-mediated deu- (KO) and Ripk3 (WT) cohorts. The genotype of each biquitination of RIPK1 [20]. CYLD is prone to activation RIPK1 and RIPK3 deficient mouse was proven by geno - by reactive oxygen species (ROS) [21]. Hence, necropto- typing (Additional file  1: Figure S1a and b). PCR was per- sis may be activated by several events, TNF release, TLR formed using the AccuStartTM II Mouse Genotyping Kit activation, inflammation and ROS production, which (Quanta Biosciences, Beverly, MA, #95,135–500) accord- are all believed to occur in the brain following TBI [16, ing to the manufacturer’s instructions. Primers were 22, 23]. Based on these findings, we hypothesize, that obtained by Metabion (metabion GmbH Planegg, Ger- necroptosis may be a relevant intracellular mechanism many). Neuronal specific conditional knock-out of Ripk1 which triggers chronic neurodegeneration after TBI. To was further proven by immunohistochemistry for RIPK1 address this issue, we used mice deficient for RIPK1 in (Additional file  1: Figure S1c). Induction of neuronal spe- neurons or RIPK3 and investigated lesion progression by cific Cre recombinase resulted in an 80% reduction of longitudinal magnetic resonance imaging (MRI), behav- RIPK1 expression in cortical neurons (Additional file  1: ioral outcome, and necroptotic signaling up to three Figure S1d). Neither neuronal RIPK1 nor global RIPK3 months after TBI in a clinically relevant mouse model deficient mice had any obvious phenotype and were born of TBI. Our results demonstrate that necroptosis is an at normal Mendelian distributions. important novel mediator of chronic neurodegeneration after TBI. Controlled Cortical Impact model of traumatic brain injury Animals were subjected to experimental traumatic brain injury using the previously described Con- trolled Cortical Impact (CCI) method [23, 28–30]. CCI induces a highly reproducible focal lesion and causes progressive brain damage and cognitive decline thereby replicating many acute and chronic characteristics of W ehn et al. acta neuropathol commun (2021) 9:138 Page 3 of 18 human TBI [23]. In short, after induction of anesthesia until they regained full motor activity in order to pre- with buprenorphine (0.1 mg/kg Bw) and isoflurane (4%, vent hypothermia. Carprofen (4 mg/kg every 24 h) was 30 s), animals were sedated with 1.5–2.5% isoflurane in administered i.p. for the following 72 h for analgesia. 30% oxygen and 70% nitrogen under continuous moni- toring of body temperature and heart rate. After right Experimental time‑line parietal craniotomy, the impact was directly applied to Motor function, depression-like behavior, and memory the intact dura with a pressure-driven steel piston with function were evaluated three days before trauma to a diameter of 3  mm (L. Kopacz, University of Mainz, obtain baseline values and up to three months thereafter. Germany; 8  m/s impact velocity, 1  mm penetration Lesion volume and tissue iron was evaluated by repeti- depth, 150 ms contact time). For sham-surgery, the pis- tive MRI up to three months after injury and validated by ton was placed on the dura, but no impact was applied. histology. At the end of the observation time brains were The craniotomy was resealed with tissue glue (VetBond, removed for immunohistochemical analysis of lesion 3  M animal care products, St. Paul, MN) and animals were kept in an incubator at 34 °C and 60% air humidity TBI MRI (T1, T2) Histology Reactive astrogliosis(GFAP) Lesion volume (Nissl) velocity = 8 mm/s depth = 1 mm contact = 150 ms Lesion volume (T2) Activated microglia (Iba-1) Iron deposits (Prussian blue) Iron deposits (T1) Fractal Sholl analysis analysis Behavior RIP1 Beam Walk fl/flCre Days Weeks RIP3 -3 0 123456 7 28 34567 91011 12 KO Tail Suspension RIP1 fl/fl RIP3 WT Barnes Maze RIP3 KO Naïve (no TBI) RIP1 fl/flCre Fig. 1 Experimental groups, methods, and time line Wehn et al. acta neuropathol commun (2021) 9:138 Page 4 of 18 Behavioral testing volume, tissue iron, astrocyte activation, and microglia To exclude age-related factors as a cause for behavioral morphology (Fig. 1). changes during the observation period, all behavioral tests were performed with an additional control group Body weight and general condition of non-traumatized (naive) RIPK3 or RIPK1 deficient Animals were weighed daily from three days before mice (Fig. 1). CCI until day 7 after trauma, then weekly. General con- dition, surgery wounds, behavior, nutrition, and fluid Motor function—Beam Walk balance were checked daily in the early postoperative The Beam Walk Test was performed as previously period, then weekly. described [23, 28, 30] on a 1 m long and 1 cm wide sus- pended wooden rod. Time to cross the beam and the number of missteps was recorded. Animals with more Magnetic resonance imaging and analysis than two missteps in pre-trauma testing were excluded For longitudinal determination of lesion volume, MRI from randomization. measurements were performed 15  min, 24  h, 7  days, one, two, and three months after TBI. For all animals, T1 weighted, T2 weighted, and diffusion weighted Memory and learning behavior—Barnes Maze imaging (DWI) sequences were collected as previously The Barnes Maze test, a well-established paradigm for described [23] under isoflurane anesthesia (1–1.5% in assessing memory function [31, 32], was performed 1, 30% oxygen/70% nitrogen) and multimodal monitor- 2, and 3  months after CCI as previously described [23]. ing of physiological parameters using a 3 T nanoScan In short, the animal was placed on a brightly lit round PET/MR (Mediso, Münster Germany). Sequences were platform with 20 identical holes along its outer rim and collected in the following order: T2-weighted imag- trained to locate a box affixed below one of the apertures ing (2D fast-spin echo (FSE), TR/TE = 3000/57.1  ms , (home cage) as fast as possible. Time to reach the home averages 14, matrix size = 96 × 96; field of cage (latency) as well as distance travelled and walking view = 16 mm × 16 mm; slice thickness = 500 µm, inter- speed were recorded and analyzed using a video track- slice gap = 60  µm), T1-weighted imaging (2D fast-spin ing software (EthoVison XT©, Version 11, 2014 Noldus echo (FSE), TR/TE = 610/28,6  ms, averages 14, matrix Information Technology). Animals were trained for four size = 96 × 96; field of view = 16  mm × 16  mm; slice consecutive days and memory function was evaluated on thickness = 500  µm, interslice gap = 60  µm). Total the sixth day. imaging time was approximately 35 min per mouse and time-point. Lesion volume was measured using ImageJ Tail Suspension test software (Rasband, W.S., ImageJ, U. S. National Insti- The Tail Suspension test is a paradigm to assess depres - tutes of Health, Bethesda, Maryland, USA, https:// sion-like behavior in rodents and was performed as previ- imagej. nih. gov/ ij/, 1997–2018) in T2 sequences. 14 ously described [23]. Briefly, animals were suspended by slices surrounding the lesion were chosen for each data- the tail for three minutes and their movements recorded set and the area segmented using the polygon tool. Vol- and analyzed using a video tracking software (EthoVison ume was then calculated using the following equation: XT©, Version 11, 2014, Noldus Information Technology). The time of inactivity was used as a proxy for depression- V = d ∗ (A1/2 + A2 + A3 . . . + An/2) like behavior [33]. with d being the distance between slices in mm (slice Histological assessment thickness + interslice gap), and A being the measured Lesion volume/ hippocampus volume area in mm . Three months after TBI, animals were fixed with 4% Hippocampal atrophy was assessed in T2-weighted PFA in deep anesthesia by transcardial perfusion. Four- images as previously described [23]. One section teen sequential 50  µm thick coronal sections were cut located in the center of the lesion containing the hip- at 500  µm intervals on a vibratome (Leica, Germany) in pocampus was chosen at the same position for each order to match the tissue volume investigated by MRI, animal. Areas of both hippocampi were segmented stained with cresyl violet according to Nissl, and evalu- using the polygon tool and the ipsilesional area of the ated by histomorphometry for lesion volume using the hippocampus was expressed as % of the area of the following formula as previously described [23]: uninjured contralateral hippocampus. V = d ∗ (A1/2 + A2 + A3 . . . + An/2) W ehn et al. acta neuropathol commun (2021) 9:138 Page 5 of 18 The volume of the hippocampus in the traumatized 100 μm away from the lesion in the hippocampus and at hemisphere was determined in six slices one mm anterior 300 μm in the cortex. For p-MLKL staining, a 5 × objec- until four mm posterior to bregma and then normalized tive was used (EC Plan-Neofluar 5x/0.16 Pol M27) with to the contralesional side. an image matrix of 1434 × 1434 pixel, a pixel scaling of 3.321 × 3.321  μm and a depth of 8 bit. Whole brain Iron deposits images were collected in z-stacks as tile scans with One section per animal was stained with Prussian blue a slice-distance of 5  μm and a total range of 25  μm. To (Iron Stain Kit, Sigma-Aldrich, # HT20) at -1.5 mm from demonstrate the intracellular localization of p-MLKL, bregma to visualize iron deposits using a light micro- a 100 × objective (Epiplan-Neofluar 100x/1.3 Oil Pol scope (Axioscope, Carl Zeiss Microscopy GmbH, Jena M27) was used with an image matrix of 512 × 512 pixel, Germany). Tile scans were then processed in ImageJ a pixel scaling of 0.166 × 0.166  μm and a depth of 8 bit. by using the color deconvolution tool to separate color Images were collected in z-stacks with a slice-distance channels. The blue channel was binarized and the inte - of 0,280 μm and a total range of 5.6 μm. After obtaining grated density was measured using the particle analy- a maximum intensity projection, images were imported sis plugin. Values of the traumatized hemisphere were into ImageJ [36] and intensity of p-MLKL measured expressed as percentage of the contralateral side. in the rim of the lesion and normalized to the signal of DAPI to correct for differences in staining. The corrected signal was then normalized to the same sized region of Immunohistochemistry interest on the contralateral hemisphere. Fifty micrometer thick floating coronal sections were prepared as previously described [34]. Blocking and Analysis of astrocyte coverage incubation with the primary antibody was performed in Assessment of astrocyte coverage was performed using 1% bovine serum albumin, 0.1% gelatin from cold water ImageJ in sections stained for GFAP (see above). Z fish skin, 0.5% Triton X-100 in 0.01 M PBS at pH 7.2–7.4 -stacks were imported into Fiji and split into individual for 72  h at 4  °C. The following primary antibodies were channels. GFAP intensity in five ROI (250 × 250  μm; at used: IBA-1 (rabbit, Wako, #019–19,741, 1:200), GFAP- 0, 250, 500, 750, and 1000  μm distance from the lesion Cy3 (mouse, Sigma Aldrich, #2905, 1:200), RIPK1 (rabbit, in the striatum) was then measured using the mean grey Novusbio, #NBP1-77077SS, 1:100), NeuN (guinea pig, value and normalized to the measurements of the con- Synaptic Systems, #266 004), and phosphorylated MLKL tralesional hemisphere to adjust for possible differences (rabbit, Cell signaling technology, # 91689S, 1:100). After in staining intensity. incubation, sections were washed in PBS and incubated with the following secondary antibodies: anti-rabbit cou- Analysis of microglia coverage and morphology pled to Alexa-fluor 594 (goat anti-rabbit, Thermo Fisher Microglia coverage was manually assessed in maximum Scientific, #A-11012), anti-guinea pig coupled to Alexa- intensity projections of iba-1 stained sections. One sec- fluor 488 (goat anti-guinea pig, Thermo Fisher Scientific, tion per animal was chosen at 1.5 mm from bregma and # A-11073), and anti-mouse coupled to Alexa-fluor 647 two ROIs chosen on the ipsilesional hemisphere, one in (goat anti-mouse, Thermo Fisher Scientific, #A- 32,728) layer V of the cortex at 300 µm away from the lesion and in 0.01 M PBS at pH 7.2–7.4 containing 0.05% Tween 20. one in the CA1a region of the hippocampus. The number Nuclei were stained with 4’,6-Diamidin-2-phenylindol of microglia was normalized to total DAPI positive cell (DAPI, Invitrogen, #D1306) 1:10,000 in 0.01 M PBS. count and expressed in as percentage of coverage in sham Imaging was performed using a ZEISS LSM 900 operated animals. confocal microscope (Carl Zeiss Microscopy GmbH, To assess microglia morphology, Sholl and fractal anal- Jena Germany). GFAP staining was recorded using a ysis were performed to indicate ramification, cell range, 10 × objective (EC Plan-Neofluar 10x/0.30 Pol M27) total cell size, and circularity using a modified protocol with an image matrix of 512 × 512 pixel, a pixel scal- from Young and Morrison [37]. Z-stack images were ing of 0.2 × 0.2  μm and a depth of 8 bit. Whole brain converted to a maximum intensity projection and cells images were collected in z-stacks as tile scans with a were individually cut out using the polygon selection slice-distance of 2  μm and a total range of 14  μm [35]. tool in ImageJ [35]. Only cells fully captured within the For microglia analysis, images were acquired using a z-stack were selected. After background subtraction, 40 × objective (EC Plan-Neofluar 40x/1.30 Oil DIC M27) images were binarized and resized to 600 × 600 pixels with an image matrix of 1024 × 1024 pixel, a pixel scal- keeping the original scale. Speckles or debris around ing of 0.2 × 0.2 μm and a depth of 8 bit. Specific regions the cells were removed using the paintbrush tool. Sholl of interest were collected in Z-stacks to include the analysis was performed using the Sholl analysis plugin entire slice thickness with a slice-distance of 0.4  μm at Wehn et al. acta neuropathol commun (2021) 9:138 Page 6 of 18 in ImageJ [38]. Centered on the soma, concentric cir- dots, suggesting that pMLKL is part of a protein com- cles with an increasing radius of 2  μm were drawn, the plex such as the necrosome (Fig.  2a, lower panel, white number of intersections measured at each radius. After arrowheads). pMLKL staining was almost absent in neu- converting binary images to outlines, fractal analysis ronal RIPK3 deficient mice suggesting that necroptotic was performed using the FracLac plugin for ImageJ [39]. signaling in neurons did essentially not occur in these As described previously [37], the total number of pixels animals (Fig.  2b). Quantification of pMLKL staining present in the cell image of either the filled or outlined showed a highly significant increase of activated MLKL binary image were calculated and later transformed to in wild type mice of both strains, while neuronal Ripk1 or 2 2 μm (pixel area = 0.208 μm ). Cell circularity was calcu- global Ripk3 knock-out completely blunted this response lated as Circularity = 4*π*Area/Perimeter . Maximum (Fig. 2c and d). span across the convex hull represents the maximum dis- tance between two points in the convex hull. Chronic posttraumatic brain damage is reduced in RIPK1 or RIPK3 deficient mice Statistical analysis After demonstrating neuronal necroptotic signaling three Sample size was calculated with the following param- months after TBI in wild type mice and showing that eters: alpha error = 0.05, beta error = 0.2, calculated RIPK1 or 3 deficiency prevented this process, we evalu - standard deviation ranged from 15 to 20% (depending ated lesion volume in cortex and hippocampus of wild type on the parameter investigated), and biologically relevant and RIPK deficient mice by longitudinal MR imaging. Two difference = 30%. All data is given as mean ± standard animals (one Ripk1 fl/fl::CamK2a Cre and one Ripk3 WT) deviation (SD) if not indicated otherwise. For compari- of the study cohort died for unknown reasons before TBI son between groups, Student t-test was used for nor- and were excluded from analysis. All other animals com- mally distributed data and Mann–Whitney Rank Sum pleted the study. Anesthesia and craniotomy did not have test for non-normally distributed data according to the any influence on general outcome parameters: sham-oper - result of Shapiro–Wilk normality test. Measurements ated and CCI animals recovered equally well from surgery over time were tested between groups using One-way in terms of bodyweight (Ripk1 cKO: Additional file  1: Fig- or Two-way ANOVA for Repeated Measurements, fol- ure S2a, Ripk3 KO: Additional file  1: Figure S2b) and gen- lowed by Tukey’s multiple comparisons test for normally eral health score (Ripk1 cKO: Additional file  1: Figure S2c, and Holm-Sidak’s multiple comparisons test for non-nor- Ripk3 KO: Additional file  1: Figure S2d). Exemplary three mally distributed data as post hoc test. Calculations were dimensional reconstructions of the brain showed a large performed with Sigma Plot version 14.0 (Systat Software lesion and a small hippocampus in the ipsilateral hemi- GmbH, Erkrath, Germany). sphere of traumatized wild type mice three months after TBI, while the lesion was significantly smaller and the hip - Results pocampus significantly larger in neuronal RIPK1 deficient fl/fl Cre A total of 33 male RIP1 deficient mice (naïve RIP1 mice (Fig. 3a). Longitudinal investigation of lesion size and fl/fl Cre fl/fl group: n = 4, RIP1 and RIP1 sham groups = 5 hippocampal volume by repetitive MRI showed that the fl/fl fl/fl each, CCI RIP1 group: n = 10, RIP1 Cre group: n = 9) primary damage measured 15 min after trauma was com- −/− +/+ and 32 male RIP3 deficient mice (naïve RIP3, RIP3 , parable in RIPK1 deficient mice and their respective wild −/− +/+ and RIP3 sham group: n = 5 each, CCI R IP3 group type controls (Fig.  3b, t = 15  min: Ripk1 cKO: 18.4 ± 2.6 −/− 3 3 n = 9, CCI RIP3 group n = 8) were operated, assessed, mm , wt: 18.4 ± 1.7 mm ) indicating that the initial trauma +/+ and analyzed for the present study. One RIP3 ani- was similar in all investigated animals. In agreement with mal was excluded from randomization and not used for previous results in this model, lesion size peaked 24 h after the study due to its performance in the Beam Walk Test TBI in both experimental groups (Ripk1 cKO: 27.0 ± 1.8 3 3 before TBI (more than 2 missteps at baseline). mm , + 47% vs. 15  min; wt: 29.8 ± 4.3 mm , + 61% vs. 15 min) as a representation of acute secondary brain dam- Traumatic brain injury induces long‑term necroptotic age. Lesion volume was not different between wild type signaling in neurons and neuronal RIPK1 deficient mice at this time point indi - Phosphorylated MLKL (pMLKL) was used as a spe- cating that necroptosis does not play a significant role for cific marker for necroptosis [40]. Three months after acute lesion progression. Within the first month after TBI, TBI large amounts of pMLKL were found in the rim removal of necrotic tissue and scar formation resulted in of the traumatic cavity, the presumed site of progres- an apparent shrinkage of the lesion. One to two months sive chronic post-trauma brain damage (Fig.  2a, upper after TBI progressive loss of brain tissue started to occur, panel). pMLKL was found by high resolution confocal a process we previously demonstrated to continue for at imaging in the cytoplasm of selected neurons as small least one year after experimental trauma [23]. Chronic W ehn et al. acta neuropathol commun (2021) 9:138 Page 7 of 18 Fig. 2 Phosphorylated MLKL, a marker of necroptosis is reduced in RIP knockout animals. a. Exemplary staining of pMLKL and NeuN (upper panel) and a pMLKL positive neuron at higher magnification (lower panel) in the rim of a traumatic contusion in a wild type mouse. b. pMLKL is significantly reduced in a RIPK3 deficient mouse. c and d p-MLKL signal intensity was significantly increased after TBI in wild type animals (white bars) compared to sham animals indicating significant presence of necroptosis three months after TBI; this increase was significantly blunted in RIPK1 (c) as well as in RIPK3 d mice where there was no difference between TBI and sham animals. Data are presented as mean ± SD; n = 9–10 for RIPK1, n = 8–9 for RIPK3. Two-way RM ANOVA with Tukey’s multiple comparisons test was used. ***p < 0.001, n.s. indicates no significant statistical difference between groups post-trauma tissue injury was significantly reduced in after TBI was similar in RIP deficient and wild type animals neuronal RIPK1 deficient mice as compared to their wild (Additional file  1: Figure S4), while chronic brain dam- type littermate controls (t = 1  month: Ripk1 cKO: 5.7 ± 1.6 age three month after TBI correlated well with the injury 3 3 mm , wt 8.2 ± 1.9 mm , p = 0.0355, t = 3 mon Ripk1 cKO: assessed by MRI (Additional file  1: Figure S5a and b) and 3 3 6.5 ± 1.7 mm , wt: 11.3 ± 2.0 mm , p = 0.0002). A similar was significantly reduced in both knock-out strains (Addi - dynamic was seen in RIPK3 deficient animals. Acute injury tional file 1 : Figure S5c and d). was not affected by RIP 3 knock-out (Fig.  3c t = 24 h, Ripk3 Since memory deficits are a hallmark of chronic post - 3 3 KO: 24,8 ± 3,3 mm , wt: 26,1 ± 5,3 mm ), while chronic traumatic brain damage in mice and TBI patients [23, 41, lesion progression was significantly attenuated from one to 42], we next investigated long-term hippocampal dam- three months after TBI (t = 1  month: Ripk3 KO: 8.8 ± 2.9 age by MRI. In the current TBI model, the hippocampus 3 3 mm , wt: 14.0 ± 1.9 mm , p = 0.006; t = 3  months: Ripk3 is only marginally injured acutely after TBI, but severely 3 3 KO: 7.3 ± 1.3 mm , wt: 11.6 ± 3.6 mm , p = 0.04). Individual affected by progressive chronic damage as previously traces for lesion volumes in each animal are given in Addi- shown [23]. All wild type mice showed significant loss tional file  1: Figure S3a (RIP1 KO) and Additional file  1: Fig- of hippocampal tissue in the traumatized hemisphere ure S3b (RIP3 KO). The MRI findings were corroborated by already one month after TBI (Fig.  3d and e, open bars). histopathological evaluations, i.e. acute brain damage 24 h Starting one month after trauma, hippocampal loss was Wehn et al. acta neuropathol commun (2021) 9:138 Page 8 of 18 significantly less pronounced in neuronal RIPK1 and points were significantly reduced and partly normalized global RIPK3 deficient mice (Fig.  3d, RIPK1: reduction to in neuronal specific RIPK1 and in global RIPK3 deficient 74.9 ± 18.5%, wt: reduction to 40.3 ± 18.7% of contralat- mice (Fig.  5b–m, closed symbols), suggesting that less eral hippocampus, p = 0.0004; Fig.  3e, RIPK3: reduction neuronal cell death was associated with less microglial to 78.0 ± 30.7%, wt: reduction to 48.2% ± 19.6% of con- activation. tralateral hippocampus, p = 0.01). This significant dif - ference persisted until the end of the observation period Neuronal RIPK1 and RIPK3 deficiency improves cognitive three months after TBI and was corroborated by histol- outcome three months after TBI ogy (Additional file 1: Figure S5e and f ). To investigate whether the reduction of lesion size, hip- pocampal damage, scar formation, and microglial activa- Astrogliosis and microglial activation are reduced in RIPK1 tion had an effect of functional outcome, we investigated and RIPK3 deficient mice motor function by beam walk, depression-like behav- The formation of a glial scar and the activation of micro - ior by the Tail Suspension test, and long-term memory glia are other hallmarks of chronic brain damage after using the Barnes Maze test. TBI significantly deterio - TBI [43, 44]. Indeed, we observed a marked increase in rated motor function and induced depression-like behav- glial fibrillary acid protein (GFAP), an astrocyte marker, ior compared to pretrauma performance as previously in the rim of the traumatic cavity and in perilesional tis- described [23], genetic deletion of RIPK1 or RIPK3, how- sue in wild type mice three months after TBI (Fig.  4a, ever, had no effect on these parameters (Fig.  6a–d). Long- left panel). Quantification of GFAP expression showed term memory was normal in non-traumatized Ripk1 an almost four-fold increase in the rim of the lesion or Ripk3 knock-out mice; it was, however, significantly with a decreasing intensity towards perilesional areas disturbed in traumatized animals, i.e. TBI increased the (Fig.  4b and c, open bars). GFAP expression was far less time needed to find the home cage, the latency to goal, pronounced in neuronal specific RIPK1 deficient mice by more than ten times and memory loss progressed over (Fig.  4a, right panel). Quantification of astrocyte density time (Fig.  6e and f, triangles and open circles). In neu- by pixel-based analysis corroborated these findings and ronal specific RIPK1 and in global RIPK3 deficient mice, reveled that activation of astrocytes was significantly however, long-term memory function was almost com- decreased by 25–35% in neuronal specific RIPK1 and in pletely preserved and resembled that of not traumatized global RIPK3 deficient mice (Fig. 4b and c, closed bars). animals (Fig.  6e and f, closed circles). Hence, protection Staining for the microglia marker Ionized calcium- of hippocampal neurons by genetic deletion of RIP 1 or binding adaptor molecule 1 yielded similar but more RIPK3 resulted in preserved long-term memory function. lesion associated findings three months after TBI (Iba1; Fig.  5). In wild type mice Iba-1 staining was most pro- Chronic lesion progression after TBI is associated with iron nounced within a distance of 100 µm from the rim of the deposits and reduced in RIP‑deficient mice traumatic cavity, while only subtle changes were observed Since our data suggest that necroptotic signaling is in areas 300  µm away from the lesion site (Fig.  5, WT). important for chronic brain damage after TBI, we were The density of iba-1 staining was heavily reduced in neu - interested to identify the mechanisms triggering this pro- ronal specific RIPK1 deficient mice (Fig.  5, Ripk1 cKO). cess. Traumatic contusions are associated with hemor- To quantify these changes, we assessed tissue coverage, rhage and subsequent deposition of iron in perilesional area, circularity, and maximal span of microglia in the brain parenchyma. Since free iron is well-known to trig- rim (100  µm) and in the vicinity (300  µm) of the trau- ger ferroptotic [45] and possibly necroptotic cell death matic lesion (Fig.  5b–i). In wild type mice all investi- signaling [46], we hypothesized that chronic posttrau- gated parameters pointed towards a significant activation matic brain damage may be associated with iron deposi- of microglia near the rim of the lesion, i.e. the coverage tion. Since iron can alter MRI signals, we looked for signal and circularity increased, while the area and the maxi- alterations using MRI scans. Indeed, we found hyperin- mal span of microglia decreased (Fig.  5b–i, open bars). tense signals at the border of the lesion one month after Microglia activation and the number of microglial branch (See figure on next page.) Fig. 3 RIPK1 and RIPK3 deficiency significantly reduces posttraumatic brain damage. a 3D reconstruction of lesion volume (blue) in relation to ipsi- and contralateral hippocampus (green) for a RIPK3 wild type and a RIPK3 deficient mouse 3 months after TBI. Scale bar = 5 mm. b and c Lesion volume over time quantified by repetitive T2-weighted MR imaging in RIPK1 (b) and RIPK3 (c) knockout animal. d and e. Hippocampal atrophy over time assessed in longitudinal T2-weighted MRI. RIPK1 (d) and RIPK3 (e) knockout mice show better preservation of hippocampal tissue over time. Mean ± SD; n = 9–10 for RIPK1, n = 8–9 for RIPK3. Two-way RM ANOVA with Sidak’s multiple comparisons test was used. *p < 0.05, **p < 0.005, ***p < 0.001 W ehn et al. acta neuropathol commun (2021) 9:138 Page 9 of 18 Fig. 3 (See legend on previous page.) Wehn et al. acta neuropathol commun (2021) 9:138 Page 10 of 18 Fig. 4 RIPK1 and RIPK3 deficiency reduces reactive astrogliosis three months after TBI. a Exemplary GFAP stainings in a wild type (left) and a RIPK1 knockout (right) animal at different distances from the traumatic contusion. Both images show the right hemisphere, but the left image has been mirrored along the midline for better visualization. Scale bar = 20 µm. Significant astrogliosis is present in the WT mouse, while it is heavily reduced in the neuronal RIPK1 deficient mouse. b and c Quantification of astrocyte coverage at different distances from the lesion. Astrocyte coverage was decreased in neuronal RIPK1 (b) and global RIPK3 (c) knockout animals compared to controls. Data are presented as mean ± SD; n = 9–10 for RIPK1, n = 8–9 for RIPK3. Students t-test for parametric and Man-Whitney-Rank-Sum test for non-parametric data were used. *p < 0.05, **p < 0.005, ***p < 0.001 W ehn et al. acta neuropathol commun (2021) 9:138 Page 11 of 18 TBI by T1-weighted MRI (Fig. 7a, upper panel) and could deficient mice were significantly protected from chronic demonstrate that these signals showed a close spatial cor- brain injury and showed improved neurocognitive func- relation with iron deposits as identified by Prussian blue tion up to three months after TBI. Of note, no protection staining (Fig.  7a, lower panel and Fig.  7b). Comparison was observed within the first days after TBI, suggest - of the volume of iron deposits between wild type and ing that necroptosis is not involved in acute injury, but RIP deficient mice revealed equal amounts of iron in all specifically mediates chronic traumatic brain damage. animals, suggesting that the amount of hemorrhage was Hence, our data further suggest that mechanistically equal in all experimental groups (Fig. 7c and d). In a next acute and chronic neuronal cell death seem to be medi- step, we investigated the spatial and temporal relation- ated by different processes. ship between chronic lesion expansion and iron deposits. Longitudinal MRI and histological assessment revealed For this purpose, we recorded iron deposits one month that lesion progression was associated with parenchy- and lesion area three months after TBI, a time point when mal iron deposition and that neuronal or global deletion the lesion already expanded. In wild type mice the rim of of RIPK1 or RIPK3 prevented lesion progression. These the lesion area, the site of lesion progression, colocal- findings suggest that chronic traumatic brain damage ized with iron deposits, while in neuronal specific RIPK1 may be triggered by iron and mediated by necroptotic deficient mice colocalization was minimal (Fig.  7e). The signaling. The association of lesion progression with iron quantification of lesion area and iron deposition in wild deposits is intriguing and may indicate that iron plays an type, RIPK1 and RIPK3 deficient mice, demonstrated important role in this process, however, further experi- that tissue loss colocalizing with histopathological detec- ments addressing this issue in more detail will need to tion of iron was significantly reduced in RIP deficient further evaluate whether there is a causal or just a correl- animals, suggesting that free iron may be involved in the ative relationship between iron deposition and neuronal pathophysiology of chronic neuronal necroptosis follow- necroptosis. ing TBI (Fig. 7f and g). Hemorrhage and the subsequent degradation of red blood cells releases large amounts of hemoglobin, heme, Discussion and free iron, i.e. molecules with high cytotoxic activity, It is increasingly recognized that next to its acute seque- into brain tissue [55]. From numerous studies investigat- lae, traumatic brain injury is a chronic disease [14]. ing intraparenchymal hemorrhage, a subtype of hemor- Chronic post-trauma brain damage is associated with rhagic stroke, it is well known that specifically free iron inflammation, persists for years after the initial insult, generates reactive oxygen species thereby damaging cell and may spread to areas initially not affected by the initial membranes and causing tissue damage and neurological impact [47]. Affected patients often suffer from neuro- dysfunction [36, 56–59]. Cerebral macro- and microhe- cognitive and mood disorders, personality changes, morrhages are common after TBI [60–62]. Specifically, neurocognitive dysfunction, or even dementia [48–54]. microbleeds have been shown to exert toxic effects on So far, no therapeutic concepts targeting the long-term endothelial cells, astrocytes, neurons, oligodendrocytes, sequelae of TBI exist as the pathophysiology of chronic and microglia and may thus lead to blood–brain barrier traumatic brain injury is still poorly understood. damage, neuronal cell death, demyelination, and chronic Here, we propose, to our knowledge for the first time, inflammation [63]. The importance of blood degrada - that programmed cell death signaling mediates chronic tion products for the pathophysiology of TBI is further neuronal injury after TBI. More specifically, we identified demonstrated by the fact that the presence and extent of the necroptosis signaling molecules RIPK1 and RIPK3 hemorrhage show a close correlation with injury severity to be major players in this process. Neurons affected by and long-term clinical outcome in TBI patients [63, 64]. chronic traumatic damage showed necroptotic signaling In line with these clinical studies, we previously demon- as evidenced by enhanced levels of pMLKL. Moreover, strated that the TBI model used in current study shows Ripk3 global knockout animals as well as neuronal RIPK1 acute macro-hemorrhage, which, however, resolves (See figure on next page.) Fig. 5 RIPK1 and RIPK3 deficiency reduces microglia activation. a. Exemplary stainings for the microglia marker iba1 in WT (upper inserts) and neuronal RIPK1 deficient mice (lower inserts) 100 µm (left inserts) and 300 µm (right inserts) from the rim of the lesion. b ‑i. Coverage and fractal analysis of microglia. In areas closer to the lesion site (100 µm, left side of each panel), knockout animals of both lines showed a decrease in microglia coverage (b. RIP 1, d. RIPK3) compared to their wild type littermates. Fractal analysis revealed that microglia of knockout animals in proximity to the lesion have less processes (c. RIP 1, e. RIPK3) are less circular (f. RIP 1, h. RIPK3), and overall smaller (g. RIP 1, i. RIPK3). In the more distal region, cells resembled those in sham operated animals, with no differences between genotypes. j‑m. Sholl analysis also shows increased ramification, i. e. more active cells, close to the lesion (j. RIP 1, l. RIPK3), but not further away from the lesion site (k. RIP 1, m. RIPK3). Data are presented as mean ± SD; n = 9–10 for RIPK1, n = 8–9 for RIPK3. Student t-test for normalized and Man-Whitney-Rank-Sum-test for non-normalized data was used. *p < 0.05, **p < 0.005, ***p < 0.001. n.s. indicates no significant statistical difference between groups Wehn et al. acta neuropathol commun (2021) 9:138 Page 12 of 18 Fig. 5 (See legend on previous page.) W ehn et al. acta neuropathol commun (2021) 9:138 Page 13 of 18 Fig. 6 RIPK1 and RIPK3 deficiency improves neurocognitive performance three months after TBI. a and b Motor impairment after TBI. Beam Walk Test revealed long term impaired motor function of the left hind limb in TBI animals (missteps compared to respective baseline, # for WT, * for KO), but there were no differences between a RIPK1 or b RIPK3 knockout animals and their respective controls. c and d Depression-like behavior after TBI. Mice of both strains showed an increase of total immobility time throughout the time course of three months, however no significant differences could be detected between RIPK1 (c) and RIPK3 (d) knockout animals compared to wild type. e and f Learning and memory dysfunction after TBI. CCI induces severe long-term memory deficits in WT mice (open circles) while RIPK1 or RIPK3 knockout animals (grey circles) show similar long-term memory function as uninjured littermate controls. Data are presented as mean ± SD; n = 9–10 for RIPK1, n = 8–9 for RIPK3. Two-way RM ANOVA with Tukey’s multiple comparisons test was used. *p < 0.05, **p < 0.005, ***p < 0.001, ****p < 0.0001 Wehn et al. acta neuropathol commun (2021) 9:138 Page 14 of 18 within the first week after injury [23]. In the current occurred in neuronal cells exposed to ferroptosis activa- study we now show by MRI and Prussian blue staining tors and knockdown of the necroptosis-mediators CYLD, that iron persists in pericontusional tissue for up to three RIPK1 or RIPK3 attenuated cell death. Interestingly, the months after the initial impact. Iron deposits identified in role for CYLD in ferroptosis also translated into neuro- still viable pericontusional brain tissue one month after protective effects in  vivo, since CYLD knockout mice TBI co-localized with damaged tissue three months after showed reduced secondary brain damage after TBI com- injury, suggesting that chronic lesion progression pref- pared to controls [21]. These findings are corroborated by erably occurred in areas with previous hemorrhage and results in models of hemorrhagic stroke suggesting that subsequent iron deposition. Since this process was sig- necroptosis and ferroptosis are indeed interconnected nificantly attenuated in neuronal Ripk1 and global Ripk3 under conditions of blood-induced tissue damage [68]. knockout animals, our findings suggest that pericontu - Despite its obvious strengths, the current study also sional iron may be involved in chronic posttraumatic has some notable limitations. We studied only young lesion expansion and that this process is mediated by male animals and are therefore not able to make any necroptotic signaling in neurons. statements on the role of necroptosis in the aged or So far, neuroinflammation was believed to be the main female brain. Further, due to technical limitations, such cause of chronic brain damage after TBI [44]. However, as the lack of specific antibodies for the study of necrop - the mechanisms by which microglial activation promotes tosis in brain tissue, we were only able to demonstrate the neuronal injury and death remain elusive. Further, micro- involvement of a single signaling molecule downstream glial activation after TBI may exert beneficial as well as of RIPK activation, namely pMLKL. Thus, future stud - detrimental effects, i.e. tissue regeneration versus acceler - ies using novel experimental tools will need to define ated damage, respectively. Therefore, disentangling these necroptotic signaling after TBI in more detail. Another opposite functions of microglia may have important shortcoming of the current study is that we demonstrate therapeutic consequences. Our current data suggest that only a spatial correlation between iron deposition and the final steps causing neuronal cell death during chronic necroptosis. Thus, further studies are needed to further post-trauma brain damage depend on RIPK1 and RIPK3 clarify whether there is a causal relationship between activity. Based on these findings, we suggest a hypotheti - iron deposition and neuronal necroptosis. Finally, we cal scenario in which chronic post-trauma brain damage want to point out that memory tests in mice are some- is initiated by the ongoing production of reactive oxy- times hard to interpret since results may be influenced by gen species (ROS) by inflammatory cells. Physiological differences in motor function or the level of disinhibition concentrations of ROS are usually well tolerated by cells which are well known to occur after TBI. We controlled since they are detoxified to water and oxygen by the glu - for differences in motor function and the intensity of tathione system and catalases [65]. However, in the pres- exploratory behavior between groups and are confident ence of iron, hydrogen peroxide is converted to highly that the presented data indeed reflect memory function, reactive hydroxyl radicals by the Fenton reaction and may however, the results need nevertheless to be interpreted initiate a form of programmed cell death called ferropto- with caution. sis [66, 67]. The link between ferroptosis and RIPK1/3- In conclusion, the current study provides evidence that mediated necroptosis in neurons is not fully established, RIPK1 and RIPK3 are critically involved in chronic post- but may be mediated by cylindromatosis (Cyld), a deu- trauma brain damage. Further, our findings suggest that biquitinase able to activate RIPK1 and downstream free iron may be involved in this process. Our results necroptosome formation under conditions of oxidative therefore help to better understand the mechanisms stress as we recently demonstrated [21]. We showed that of chronic post-trauma brain damage and suggest that CYLD-dependent RIPK1/RIPK3 necrosome-formation RIPK1- and RIPK3-mediated necroptosis may represent (See figure on next page.) Fig. 7 Lesion progression occurs in areas with iron deposits. a T1-weighted MRI (upper panel) and Prussian blue staining (lower panel) three months after injury. There is a close spatial correlation between the T1-hyperintense signal and iron staining (arrowheads). b There is high spatial correlation between the area of T1 hyperintensities and iron deposits as assessed by Prussian blue staining. Pearson product-moment correlation analysis. c and d Iron deposits in pericontusional brain tissue assessed by longitudinal MRI. Extent of hemorrhage is comparable in RIPK1 (c) and RIPK3 (d) knockout animals and controls, indicating no differences in hemorrhage size after TBI between groups. T1 hyperintensities decreased over time in all groups, suggesting a very slow resorption of iron over time. e. Co-localization of iron deposits (red) observed at 1 month after TBI (upper panels) and lesion size assessed at the end of the observation period (3 months, middle row, green) suggests a progressive expansion of the lesion towards the regions with iron deposits. f and g Quantification of overlap between iron deposits and lesion. The higher the overlap of iron deposits and lesion size at three months, the higher the rate of tissue loss/ cell death in iron containing tissue. Co-localization is significantly less pronounced in RIPK1 (f) or RIPK3 (g) deficient mice, suggesting a reduced lesion growth in RIP knockouts due to toxic iron residues. Data are presented as mean ± SD; n = 9–10 for RIPK1, n = 8–9 for RIPK3. Two-way RM ANOVA with Tukey’s multiple comparisons test was used. **p < 0.005, ***p < 0.001 W ehn et al. acta neuropathol commun (2021) 9:138 Page 15 of 18 Fig. 7 (See legend on previous page.) Wehn et al. acta neuropathol commun (2021) 9:138 Page 16 of 18 a novel therapeutic target for the treatment of patients MET_V/007), iBOF20/IBF/039 ATLANTIS, Foundation against Cancer (FAF- suffering from the long-term sequels of TBI. F/2016/865, F/2020/1505), CRIG and GIGG consortia, and VIB. The Plesnila group was funded by Munich University´s Förderprogramm für Forschung und Lehre (FöFoLe), by the BMBF-funded research consortium TRAINS (Project Supplementary Information ID: 01EW1709) and by the Munich Cluster of Systems Neurology (SyNergy; The online version contains supplementary material available at https:// doi. Project ID EXC 2145 / ID 390857198). org/ 10. 1186/ s40478- 021- 01236-0. Author contributions Conception and study design: ACW, NP, NAT. Surgery, genotyping, neuro- Additional File 1 Supplementary Fig. S1. Genotyping of RIPK1 and logical testing, histology: ACW, IK. MR imaging: ACW, MD. Data analysis & RIPK3 deficient mice and proof of neuronal specific RIPK1 knock-out in interpretation: ACW, IK, MD, NP, NAT. Statistical analysis: ACW, NAT. Manuscript flox/flox RIPK1 Camk2CreERT2 mice. a. Neuron specific RIPK1 deficient mice preparation: ACW, NP, NAT. Critical revision of the manuscript: all authors. All used for experiments were heterozygous for Camk2CreERT2 and homozy- authors read and approved the final manuscript. gous for the floxed RIPK1 allele. Littermate controls were also homozy- gous for the floxed RIPK1 allele, but did not express the Cre recombinase. Availability of data and materials b. Global RIPK3 deficient mice were homozygous for disrupted allele, The datasets used and/or analyzed during the current study available from the while control mice expressed only the wild type gene. c. and d. To corresponding author on reasonable request. flox/ demonstrate specific neuron specific RIPK1 deficiency in induced RIPK1 flox Camk2CreERT2 mice, we performed immunohistochemistry for RIPK1 and NeuN, a neuronal marker. In the cortex of control mice RIPK1 was Declarations almost exclusively expressed in neurons (upper panels), while in induced flox/flox RIPK1 Camk2CreERT2 mice RIPK1 staining was significantly reduced Ethics approval (lower panels) to 20% of baseline (d). All procedures were reviewed and approved by the respective institutional and governmental authorities and performed according to all regulations. Consent for publication Additional File 1 Supplementary Fig. 2 Body weight and physical con- Not applicable. dition after experimental TBI. a. and b. Weight after TBI. Animals recovered from weight loss directly after trauma within one week after injury; in the Competing interests following observation period weight constantly increased. No differences The authors declare that they have no competing interests. were detected between CCI and sham-operated animals in the RIPK1 (a) and RIPK3 (b) groups. c. and d. General health score to assess recovery. Conflict of interest All animals’ general condition transiently worsened in the perioperative There is no conflict of interest for any of the authors. phase with a peak at day 1 after TBI, but returned to baseline within one week. There was no difference between groups, c. RIPK1, d. RIPK3. Data Author details are presented as mean ± SD; n = 5 for sham, n = 8–10 for TBI. 1 Institute for Stroke and Dementia Research (ISD), LMU Klinikum, Ludwig- Maximilians University Munich, Feodor-Lynen-Str. 17, 81377 Munich, Germany. 2 3 Munich Cluster of Systems Neurology (SyNergy), Munich, Germany. I nstitute for Pharmacology and Clinical Pharmacy, Biochemical-Pharmacological Center Additional File 1 Supplementary Fig. 3. Individual lesion volume pro- Marburg, University of Marburg, Karl-von-Frisch Straße 2 K03, 35032 Marburg, gression for a. RIP1 and b. RIP3 deficient mice. 4 5 Germany. University of Marburg, Marburg, Germany. Molecular Signaling and Cell Death Unit, VIB-UGent Center for Inflammation Research, UGent-VIB Research Building FSVM, Technologiepark 71, 9052 Ghent, Belgium. Depar t- Additional File 1 Supplementary Fig. 4. RIPK3 deficiency does not affect ment of Biomedical Molecular Biology, Ghent University, Ghent, Belgium. acute brain injury after TBI. No differences in lesion volume as assessed by Graduate School of Systemic Neurosciences (GSN), Ludwig-Maximilians histology was detected between RIPK3 knockout mice and C57BL/6 wild University Munich, Munich, Germany. Department of Neurosurgery, LMU type controls at 24 h after TBI. Data are presented as mean ± SD; n = 10. Klinikum, Ludwig-Maximilians University Munich, Munich, Germany. Present Address: Medical Image Analysis Center (MIAC AG) and Qbig, Department of Biomedical Engineering, University of Basel, Basel, Switzerland. Additional File 1 Supplementary Fig. 5. Lesion volumes by MRI and his- Received: 28 May 2021 Accepted: 27 July 2021 tology. a. T2-weighted MRI and Nissl stained coronal section three months after injury. Lesion volume was quantified in T2-weighted MRI images as well as Nissl stained sections obtained in the same animals three months post injury. b. Correlation analysis of lesion volumes assessed by histomorphometry or T2-weighted MRI revealed a strong positive cor- References relation between both methods. Differences in measured lesion size are 1. Dewan MC, Rattani A, Gupta S, Baticulon RE, Hung YC, Punchak M, due to shrinking of tissue in the fixation process, with a shrinking factor Agrawal A, Adeleye AO, Shrime MG, Rubiano AM et al (2018) Estimating of 0.77 ± 0.23. c.—f. Quantification of lesion volume (c. RIPK1, d. RIPK3) the global incidence of traumatic brain injury. J Neurosurg. https:// doi. as well as hippocampal damage (e. RIPK1, f. RIPK3) using both methods. org/ 10. 3171/ 2017. 10. JNS17 352 Data are presented as mean ± SD; n = 8–10 for TBI. Pearson correlation, 2. Injury GBDTB, Spinal Cord Injury C (2019) Global, regional, and national Students t-test for parametric and Man-Whitney-Rank-Sum test for non- burden of traumatic brain injury and spinal cord injury, 1990-2016: a sys- parametric data were used. *p < 0.05, ***p < 0,001. tematic analysis for the Global Burden of Disease Study 2016. The Lancet Neurology 18: 56-87https:// doi. org/ 10. 1016/ S1474- 4422(18) 30415-0 3. Maas AIR, Menon DK, Adelson PD, Andelic N, Bell MJ, Belli A, Bragge Acknowledgements P, Brazinova A, Buki A, Chesnut RM et al (2017) Traumatic brain injury: We want to thank Dr. Ulrike Schillinger for her support during the analysis integrated approaches to improve prevention, clinical care, and research. of the MRI data. Research in the Vandenabeele group is supported by Flem- Lancet Neurol 16:987–1048. https:// doi. org/ 10. 1016/ S1474- 4422(17) ish grants (EOS MODEL-IDI, FWO Grant 30826052), FWO research grants 30371-X (G.0E04.16N, G.0C76.18N, G.0B71.18N, G.0B96.20N), Methusalem (BOF16/ W ehn et al. acta neuropathol commun (2021) 9:138 Page 17 of 18 4. Rabinowitz AR, Levin HS (2014) Cognitive sequelae of traumatic brain 23. Mao X, Terpolilli NA, Wehn A, Chen S, Hellal F, Liu B, Seker B, Plesnila N injury. Psychiatr Clin North Am 37:1–11. https:// doi. org/ 10. 1016/j. psc. (2019) Progressive histopathological damage occurring up to one year 2013. 11. 004 after experimental traumatic brain injury is associated with cognitive 5. Selassie AW, Zaloshnja E, Langlois JA, Miller T, Jones P, Steiner C (2008) decline and depression-like behavior. J Neurotrauma: Doi https:// doi. org/ Incidence of long-term disability following traumatic brain injury 10. 1089/ neu. 2019. 6510 hospitalization, United States, 2003. J Head Trauma Rehabil 23:123–131. 24. Kilkenny C, Browne WJ, Cuthill IC, Emerson M, Altman DG (2010) Improv- https:// doi. org/ 10. 1097/ 01. HTR. 00003 14531. 30401. 39 ing bioscience research reporting: the ARRIVE guidelines for reporting 6. Silver JM, McAllister TW, Arciniegas DB (2009) Depression and cogni- animal research. PLoS Biol 8:e1000412. https:// doi. org/ 10. 1371/ journ al. tive complaints following mild traumatic brain injury. Am J Psychiatry pbio. 10004 12 166:653–661. https:// doi. org/ 10. 1176/ appi. ajp. 2009. 08111 676 25. Guillen J (2012) FELASA guidelines and recommendations. J Am Assoc 7. Ramos-Cejudo J, Wisniewski T, Marmar C, Zetterberg H, Blennow K, de Lab Anim Sci 51:311–321 Leon MJ, Fossati S (2018) Traumatic brain injury and Alzheimer’s disease: 26. Takahashi N, Vereecke L, Bertrand MJ, Duprez L, Berger SB, Divert T, Gon- the cerebrovascular link. EBioMedicine 28:21–30. https:// doi. org/ 10. calves A, Sze M, Gilbert B, Kourula S et al (2014) RIPK1 ensures intestinal 1016/j. ebiom. 2018. 01. 021 homeostasis by protecting the epithelium against apoptosis. Nature 8. MacKenzie JD, Siddiqi F, Babb JS, Bagley LJ, Mannon LJ, Sinson GP, Gross- 513:95–99. https:// doi. org/ 10. 1038/ natur e13706 man RI (2002) Brain atrophy in mild or moderate traumatic brain injury: a 27. Newton K, Sun X, Dixit VM (2004) Kinase RIP3 is dispensable for normal longitudinal quantitative analysis. AJNR Am J Neuror 23:1509–1515 NF-kappa Bs, signaling by the B-cell and T-cell receptors, tumor necrosis 9. Reider-Groswasser I, Cohen M, Costeff H, Groswasser Z (1993) Late factor receptor 1, and Toll-like receptors 2 and 4. Mol Cell Biol 24:1464– CT findings in brain trauma: relationship to cognitive and behavioral 1469. https:// doi. org/ 10. 1128/ mcb. 24.4. 1464- 1469. 2004 sequelae and to vocational outcome. AJR Am J Roentgenol 160:147–152. 28. Terpolilli NA, Zweckberger K, Trabold R, Schilling L, Schinzel R, Tegtmeier https:// doi. org/ 10. 2214/ ajr. 160.1. 84166 13 F, Plesnila N (2009) The novel nitric oxide synthase inhibitor 4-amino- 10. Cardoso ER, Galbraith S (1985) Posttraumatic hydrocephalus–a retrospec- tetrahydro-L-biopterine prevents brain edema formation and intracranial tive review. Surg Neurol 23:261–264. https:// doi. org/ 10. 0090/ 3019(85) hypertension following traumatic brain injury in mice. J Neurotrauma 90092- 8[pii] 26:1963–1975. https:// doi. org/ 10. 1089/ neu. 2008- 0853 11. Guyot LL, Michael DB (2000) Post-traumatic hydrocephalus. Neurol Res 29. Zweckberger K, Eros C, Zimmermann R, Kim SW, Engel D, Plesnila N 22:25–28 (2006) Eec ff t of early and delayed decompressive craniectomy on 12. Honeybul S, Ho KM (2012) Incidence and risk factors for post-traumatic secondary brain damage after controlled cortical impact in mice. J Neu- hydrocephalus following decompressive craniectomy for intractable rotrauma 23:1083–1093. https:// doi. org/ 10. 1089/ neu. 2006. 23. 1083 intracranial hypertension and evacuation of mass lesions. J Neurotrauma 30. Zweckberger K, Stoffel M, Baethmann A, Plesnila N (2003) Eec ff t of 29:1872–1878. https:// doi. org/ 10. 1089/ neu. 2012. 2356 decompression craniotomy on increase of contusion volume and func- 13. Mazzini L, Campini R, Angelino E, Rognone F, Pastore I, Oliveri G (2003) tional outcome after controlled cortical impact in mice. J Neurotrauma Posttraumatic hydrocephalus: a clinical, neuroradiologic, and neuropsy- 20:1307–1314. https:// doi. org/ 10. 1089/ 08977 15033 22686 102 chologic assessment of long-term outcome. Arch Phys Med Rehabil 31. Barnes CA (1979) Memory deficits associated with senescence: a neu- 84:1637–1641 rophysiological and behavioral study in the rat. J Comp Physiol Psychol 14. Faden AI, Loane DJ (2015) Chronic neurodegeneration after traumatic 93:74–104. https:// doi. org/ 10. 1037/ h0077 579 brain injury: alzheimer disease, chronic traumatic encephalopathy, or 32. Gawel K, Gibula E, Marszalek-Grabska M, Filarowska J, Kotlinska JH (2019) persistent neuroinflammation? Neurotherapeutics 12:143–150. https:// Assessment of spatial learning and memory in the Barnes maze task in doi. org/ 10. 1007/ s13311- 014- 0319-5 rodents-methodological consideration. Naunyn Schmiedebergs Arch 15. Johnson VE, Stewart JE, Begbie FD, Trojanowski JQ, Smith DH, Stewart Pharmacol 392:1–18. https:// doi. org/ 10. 1007/ s00210- 018- 1589-y W (2013) Inflammation and white matter degeneration persist for years 33. Cryan JF, Mombereau C, Vassout A (2005) The tail suspension test as a after a single traumatic brain injury. Brain 136:28–42. https:// doi. org/ 10. model for assessing antidepressant activity: review of pharmacological 1093/ brain/ aws322 and genetic studies in mice. Neurosci Biobehav Rev 29:571–625. https:// 16. Loane DJ, Kumar A, Stoica BA, Cabatbat R, Faden AI (2014) Progressive doi. org/ 10. 1016/j. neubi orev. 2005. 03. 009 neurodegeneration after experimental brain trauma: association with 34. Ghosh M, Balbi M, Hellal F, Dichgans M, Lindauer U, Plesnila N (2015) Peri- chronic microglial activation. J Neuropathol Exp Neurol 73:14–29 cytes are involved in the pathogenesis of cerebral autosomal dominant 17. Ramlackhansingh AF, Brooks DJ, Greenwood RJ, Bose SK, Turkheimer FE, arteriopathy with subcortical infarcts and leukoencephalopathy. Ann Kinnunen KM, Gentleman S, Heckemann RA, Gunanayagam K, Gelosa Neurol 78:887–900. https:// doi. org/ 10. 1002/ ana. 24512 G et al (2011) Inflammation after trauma: microglial activation and trau- 35. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, matic brain injury. Ann Neurol 70:374–383. https:// doi. org/ 10. 1002/ ana. Preibisch S, Rueden C, Saalfeld S, Schmid B et al (2012) Fiji: an open- 22455 source platform for biological-image analysis. Nat Methods 9:676–682. 18. Ritzel RM, Doran SJ, Barrett JP, Henry RJ, Ma EL, Faden AI, Loane DJ (2018) https:// doi. org/ 10. 1038/ nmeth. 2019 Chronic alterations in systemic immune function after traumatic brain 36. Hua Y, Xi G, Keep RF, Hoff JT (2000) Complement activation in the brain injury. J Neurotrauma 35:1419–1436. https:// doi. org/ 10. 1089/ neu. 2017. after experimental intracerebral hemorrhage. J Neurosurg 92:1016–1022. 5399https:// doi. org/ 10. 3171/ jns. 2000. 92.6. 1016 19. Hitomi J, Christofferson DE, Ng A, Yao J, Degterev A, Xavier RJ, Yuan J 37. Young K, Morrison H (2018) Quantifying microglia morphology from pho- (2008) Identification of a molecular signaling network that regulates a tomicrographs of immunohistochemistry prepared tissue using imageJ. J cellular necrotic cell death pathway. Cell 135:1311–1323. https:// doi. org/ Vis Exp: Doi https:// doi. org/ 10. 3791/ 57648 10. 1016/j. cell. 2008. 10. 044 38. Ferreira TA, Blackman AV, Oyrer J, Jayabal S, Chung AJ, Watt AJ, Sjostrom 20. Moquin DM, McQuade T, Chan FK (2013) CYLD deubiquitinates RIP1 in PJ, van Meyel DJ (2014) Neuronal morphometry directly from bitmap the TNFα-induced necrosome to facilitate kinase activation and pro- images. Nat Methods 11:982–984. https:// doi. org/ 10. 1038/ nmeth. 3125 grammed necrosis. PLoS ONE 8:e76841. https:// doi. org/ 10. 1371/ journ al. 39. Karperien A (1999–2013) FracLac for ImageJ. pone. 00768 41 40. Sun L, Wang H, Wang Z, He S, Chen S, Liao D, Wang L, Yan J, Liu W, Lei X 21. Ganjam GK, Terpolilli NA, Diemert S, Eisenbach I, Hoffmann L, Reuther C, et al (2012) Mixed lineage kinase domain-like protein mediates necrosis Herden C, Roth J, Plesnila N, Culmsee C (2018) Cylindromatosis mediates signaling downstream of RIP3 kinase. Cell 148:213–227. https:// doi. org/ neuronal cell death in vitro and in vivo. Cell Death Differ 25:1394–1407. 10. 1016/j. cell. 2011. 11. 031 https:// doi. org/ 10. 1038/ s41418- 017- 0046-7 41. Ariza M, Serra-Grabulosa JM, Junque C, Ramirez B, Mataro M, Poca A, Bar- 22. Krieg SM, Voigt F, Knuefermann P, Kirschning CJ, Plesnila N, Ringel F (2017) gallo N, Sahuquillo J (2006) Hippocampal head atrophy after traumatic Decreased secondary lesion growth and attenuated immune response brain injury. Neuropsychologia 44:1956–1961. https:// doi. org/ 10. 1016/j. after traumatic brain injury in Tlr2/4(-/-) mice. Front Neurol 8:455. https:// neuro psych ologia. 2005. 11. 007 doi. org/ 10. 3389/ fneur. 2017. 00455 Wehn et al. acta neuropathol commun (2021) 9:138 Page 18 of 18 42. Jorge RE, Acion L, Starkstein SE, Magnotta V (2007) Hippocampal 57. Derry PJ, Vo ATT, Gnanansekaran A, Mitra J, Liopo AV, Hegde ML, Tsai AL, volume and mood disorders after traumatic brain injury. Biol Psychiatry Tour JM, Kent TA (2020) The chemical basis of intracerebral hemorrhage 62:332–338. https:// doi. org/ 10. 1016/j. biops ych. 2006. 07. 024 and cell toxicity with contributions from eryptosis and ferroptosis. Front 43. Jassam YN, Izzy S, Whalen M, McGavern DB, El Khoury J (2017) Neuroim- Cell Neurosci 14:603043. https:// doi. org/ 10. 3389/ fncel. 2020. 603043 munology of traumatic brain injury: time for a paradigm shift. Neuron 58. Letarte PB, Lieberman K, Nagatani K, Haworth RA, Odell GB, Duff TA (1993) 95:1246–1265. https:// doi. org/ 10. 1016/j. neuron. 2017. 07. 010 Hemin: levels in experimental subarachnoid hematoma and effects 44. Morganti-Kossmann MC, Semple BD, Hellewell SC, Bye N, Ziebell JM on dissociated vascular smooth-muscle cells. J Neurosurg 79:252–255. (2019) The complexity of neuroinflammation consequent to traumatic https:// doi. org/ 10. 3171/ jns. 1993. 79.2. 0252 brain injury: from research evidence to potential treatments. Acta Neuro- 59. Wan J, Ren H, Wang J (2019) Iron toxicity, lipid peroxidation and ferropto- pathol 137:731–755. https:// doi. org/ 10. 1007/ s00401- 018- 1944-6 sis after intracerebral haemorrhage. Stroke Vasc Neurol 4:93–95. https:// 45. Dixon SJ, Lemberg KM, Lamprecht MR, Skouta R, Zaitsev EM, Gleason CE, doi. org/ 10. 1136/ svn- 2018- 000205 Patel DN, Bauer AJ, Cantley AM, Yang WS et al (2012) Ferroptosis: an iron- 60. Hasiloglu ZI, Albayram S, Selcuk H, Ceyhan E, Delil S, Arkan B, Baskoy L dependent form of nonapoptotic cell death. Cell 149:1060–1072. https:// (2011) Cerebral microhemorrhages detected by susceptibility-weighted doi. org/ 10. 1016/j. cell. 2012. 03. 042 imaging in amateur boxers. AJNR Am J Neuroradiol 32:99–102. https:// 46. Pasparakis M, Vandenabeele P (2015) Necroptosis and its role in inflam-doi. org/ 10. 3174/ ajnr. A2250 mation. Nature 517:311–320. https:// doi. org/ 10. 1038/ natur e14191 61. Park JH, Park SW, Kang SH, Nam TK, Min BK, Hwang SN (2009) Detection 47. Plesnila N (2016) The immune system in traumatic brain injury. Curr Opin of traumatic cerebral microbleeds by susceptibility-weighted image of Pharmacol 26:110–117. https:// doi. org/ 10. 1016/j. coph. 2015. 10. 008 MRI. J Korean Neurosurg Soc 46:365–369. https:// doi. org/ 10. 3340/ jkns. 48. Al-Khindi T, Macdonald RL, Schweizer TA (2010) Cognitive and functional 2009. 46.4. 365 outcome after aneurysmal subarachnoid hemorrhage. Stroke 41:e519– 62. van der Horn HJ, de Haan S, Spikman JM, de Groot JC, van der Naalt J e536. https:// doi. org/ 10. 1161/ STROK EAHA. 110. 581975 (2018) Clinical relevance of microhemorrhagic lesions in subacute mild 49. Alghnam S, AlSayyari A, Albabtain I, Aldebasi B, Alkelya M (2017) Long- traumatic brain injury. Brain Imaging Behav 12:912–916. https:// doi. org/ term disabilities after traumatic head injury ( THI): a retrospective analysis 10. 1007/ s11682- 017- 9743-6 from a large level-I trauma center in Saudi Arabia. Inj Epidemiol 4:29. 63. Glushakova OY, Johnson D, Hayes RL (2014) Delayed increases in micro- https:// doi. org/ 10. 1186/ s40621- 017- 0126- 7[pii] vascular pathology after experimental traumatic brain injury are associ- 50. Andelic N, Howe EI, Hellstrom T, Sanchez MF, Lu J, Lovstad M, Roe C ated with prolonged inflammation, blood-brain barrier disruption, and (2018) Disability and quality of life 20 years after traumatic brain injury. progressive white matter damage. J Neurotrauma 31:1180–1193. https:// Brain Behav. https:// doi. org/ 10. 1002/ brb3. 1018doi. org/ 10. 1089/ neu. 2013. 3080 51. Draper K, Ponsford J (2008) Cognitive functioning ten years following 64. Lawrence TP, Pretorius PM, Ezra M, Cadoux-Hudson T, Voets NL (2017) traumatic brain injury and rehabilitation. Neuropsychology 22:618–625. Early detection of cerebral microbleeds following traumatic brain injury https:// doi. org/ 10. 1037/ 0894- 4105. 22.5. 618 using MRI in the hyper-acute phase. Neurosci Lett 655:143–150. https:// 52. Hackett ML, Anderson CS (2000) Health outcomes 1 year after subarach-doi. org/ 10. 1016/j. neulet. 2017. 06. 046 noid hemorrhage: an international population-based study. The austral- 65. Conrad M, Pratt DA (2019) The chemical basis of ferroptosis. Nat Chem ian cooperative research on subarachnoid hemorrhage study group. Biol 15:1137–1147. https:// doi. org/ 10. 1038/ s41589- 019- 0408-1 Neurology 55:658–662 66. Li Q, Han X, Lan X, Gao Y, Wan J, Durham F, Cheng T, Yang J, Wang Z, Jiang 53. Hart T, Whyte J, Polansky M, Millis S, Hammond FM, Sherer M, Bushnik T, C et al (2017) Inhibition of neuronal ferroptosis protects hemorrhagic Hanks R, Kreutzer J (2003) Concordance of patient and family report of brain. JCI insight 2:e90777. https:// doi. org/ 10. 1172/ jci. insig ht. 90777 neurobehavioral symptoms at 1 year after traumatic brain injury. Arch 67. Ratan RR (2020) The chemical biology of ferroptosis in the central nerv- Phys Med Rehabil 84:204–213. https:// doi. org/ 10. 1053/ apmr. 2003. 50019 ous system. Cell Chem Biol 27:479–498. https:// doi. org/ 10. 1016/j. chemb 54. Himanen L, Portin R, Isoniemi H, Helenius H, Kurki T, Tenovuo O (2005) iol. 2020. 03. 007 Cognitive functions in relation to MRI findings 30 years after traumatic 68. Zille M, Karuppagounder SS, Chen Y, Gough PJ, Bertin J, Finger J, Milner brain injury. Brain Inj 19:93–100 TA, Jonas EA, Ratan RR (2017) Neuronal death after hemorrhagic stroke 55. Regan RF, Panter SS (1996) Hemoglobin potentiates excitotoxic injury in in vitro and in vivo shares features of ferroptosis and necroptosis. Stroke cortical cell culture. J Neurotrauma 13:223–231. https:// doi. org/ 10. 1089/ 48:1033–1043. https:// doi. org/ 10. 1161/ strok eaha. 116. 015609 neu. 1996. 13. 223 56. Bhasin RR, Xi G, Hua Y, Keep RF, Hoff JT (2002) Experimental intracerebral Publisher’s Note hemorrhage: effect of lysed erythrocytes on brain edema and blood- Springer Nature remains neutral with regard to jurisdictional claims in pub- brain barrier permeability. Acta Neurochir Suppl 81:249–251. https:// doi. lished maps and institutional affiliations. org/ 10. 1007/ 978-3- 7091- 6738-0_ 65 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Acta Neuropathologica Communications Springer Journals

RIPK1 or RIPK3 deletion prevents progressive neuronal cell death and improves memory function after traumatic brain injury

Loading next page...
 
/lp/springer-journals/ripk1-or-ripk3-deletion-prevents-progressive-neuronal-cell-death-and-3FD1BEdS0z
Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2021
eISSN
2051-5960
DOI
10.1186/s40478-021-01236-0
Publisher site
See Article on Publisher Site

Abstract

Traumatic brain injury ( TBI) causes acute and subacute tissue damage, but is also associated with chronic inflam- mation and progressive loss of brain tissue months and years after the initial event. The trigger and the subsequent molecular mechanisms causing chronic brain injury after TBI are not well understood. The aim of the current study was therefore to investigate the hypothesis that necroptosis, a form a programmed cell death mediated by the interaction of Receptor Interacting Protein Kinases (RIPK) 1 and 3, is involved in this process. Neuron-specific RIPK1- or RIPK3-deficient mice and their wild-type littermates were subjected to experimental TBI by controlled cortical impact. Posttraumatic brain damage and functional outcome were assessed longitudinally by repetitive magnetic resonance imaging (MRI) and behavioral tests (beam walk, Barnes maze, and tail suspension), respectively, for up to three months after injury. Thereafter, brains were investigated by immunohistochemistry for the necroptotic marker phosphorylated mixed lineage kinase like protein(pMLKL) and activation of astrocytes and microglia. WT mice showed progressive chronic brain damage in cortex and hippocampus and increased levels of pMLKL after TBI. Chronic brain damage occurred almost exclusively in areas with iron deposits and was significantly reduced in RIPK1- or RIPK3-deficient mice by up to 80%. Neuroprotection was accompanied by a reduction of astrocyte and microglia activation and improved memory function. The data of the current study suggest that progressive chronic brain damage and cognitive decline after TBI depend on the expression of RIPK1/3 in neurons. Hence, inhibition of necroptosis signaling may represent a novel therapeutic target for the prevention of chronic post-traumatic brain damage. Keywords: Traumatic brain injury, Chronic posttraumatic brain damage, Magnetic resonance imaging, Necroptosis, Ferroptosis, Neuroprotection Introduction of death and disability in all age groups worldwide, espe- With an estimated case load of 69 million per year [1], cially in children and young adults. The incidence of TBI traumatic brain injury (TBI) represents a leading cause is expected to increase in the coming decades, as the number of the two main etiologies—motor vehicle acci- dents and falls—are expected to rise due to an increase in motorization and an aging population, respectively [2]. *Correspondence: nikolaus.plesnila@med.uni-muenchen.de; nicole. terpolilli@med.uni-muenchen.de The socio-economic impact of TBI is vast, with estimated Nikolaus Plesnila and Nicole Angela Terpolilli authors are equally costs of approximately 400 billion US$ annually [3]. This contributed to this work. number does not only include the direct costs due to Institute for Stroke and Dementia Research (ISD), LMU Klinikum, Ludwig- Maximilians University Munich, Feodor-Lynen-Str. 17, 81377 Munich, acute primary care, but also long-term follow-up costs Germany since many TBI survivors suffer from mood changes, Full list of author information is available at the end of the article © The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/. The Creative Commons Public Domain Dedication waiver (http:// creat iveco mmons. org/ publi cdoma in/ zero/1. 0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Wehn et al. acta neuropathol commun (2021) 9:138 Page 2 of 18 memory deficits, and loss of fine motor skills, have dif - Material and methods ficulties returning to their previous occupation, and therefore require lifelong support [4–6]. Consequently, Ethical statement TBI is increasingly recognized as a chronic neurological All animal experiments were reviewed and approved by disorder with socio-economic implications comparable the Ethical Review Board of the Government of Upper to conditions like Alzheimer’s disease or other neurode- Bavaria. The results of the study are reported in accord - generative disorders [7]. ance with the ARRIVE guidelines [24]. Animal hus- While the pathophysiology of acute brain damage has bandry, health screens, and hygiene management checks been investigated in detail in experimental animals and were performed in accordance with Federation of Euro- in humans over the past decades, relatively little is known pean Laboratory Animal Science Associations (FELASA) about the mechanisms determining long-term outcome guidelines and recommendations [25]. Only male mice after TBI. Chronic functional deficits in TBI patients between 6 and 8  weeks old mice were used. All surgical may be caused by progressive brain atrophy in cortex procedures, behavioral testing, imaging, and data analysis and hippocampus [8, 9] and hydrocephalus formation were performed in a randomized fashion by a researcher [10–13]. So far, clinical and experimental studies suggest blinded to the genotype and group allocation of the ani- that inflammation plays an important role for the devel - mals. Group allocation was obtained by drawing lots by a opment of chronic posttraumatic brain damage [14–18], third party not involved in the study or data analysis. however, the cellular and molecular mechanisms down- stream of this process are not fully understood. Particu- Animals larly, the trigger and the intracellular signaling cascades Inducible neuronal Ripk1 cKO mice were generated as causing neuronal cell death weeks and months after TBI previously described [26] and bred as Ripk1fl/fl (WT) are still unknown. and Ripk1fl/fl::Camk2a-Cre + /Cre (cKO) in our facility. Necroptosis is a form of necrotic regulated cell death, For induction of neuronal RIPK1 deletion, Ripk1 cKO which involves the upstream assembly of the necropto- mice (starting at 4  weeks of age) received three intra- some complex formed by the interaction of receptor peritoneal injections of 2 mg Tamoxifen (Sigma-Aldrich, interacting protein kinase 1 and 3 (RIPK1 and 3) [19] and Taufkirchen, Germany, # T5648) in a 100 µl Miglyol sus- downstream RIPK3-mediated phosphorylation of mixed pension (Caelo, Hilden, Germany, #3274) every 48 h (day lineage kinase like protein (MLKL). Necroptosis can be 1, 3, 5). RIPK3-deficient mice were generated and kindly initiated by activation of Toll-like receptors (TLR) 3 and provided by V. M. Dixit, Genentech Inc., San Francisco, −/− 4, TNF-alpha receptor 1 or, as more recently shown by CA, [27] and bred heterozygously to obtain Ripk3 +/+ us and others, by cylindromatosis (CYLD)-mediated deu- (KO) and Ripk3 (WT) cohorts. The genotype of each biquitination of RIPK1 [20]. CYLD is prone to activation RIPK1 and RIPK3 deficient mouse was proven by geno - by reactive oxygen species (ROS) [21]. Hence, necropto- typing (Additional file  1: Figure S1a and b). PCR was per- sis may be activated by several events, TNF release, TLR formed using the AccuStartTM II Mouse Genotyping Kit activation, inflammation and ROS production, which (Quanta Biosciences, Beverly, MA, #95,135–500) accord- are all believed to occur in the brain following TBI [16, ing to the manufacturer’s instructions. Primers were 22, 23]. Based on these findings, we hypothesize, that obtained by Metabion (metabion GmbH Planegg, Ger- necroptosis may be a relevant intracellular mechanism many). Neuronal specific conditional knock-out of Ripk1 which triggers chronic neurodegeneration after TBI. To was further proven by immunohistochemistry for RIPK1 address this issue, we used mice deficient for RIPK1 in (Additional file  1: Figure S1c). Induction of neuronal spe- neurons or RIPK3 and investigated lesion progression by cific Cre recombinase resulted in an 80% reduction of longitudinal magnetic resonance imaging (MRI), behav- RIPK1 expression in cortical neurons (Additional file  1: ioral outcome, and necroptotic signaling up to three Figure S1d). Neither neuronal RIPK1 nor global RIPK3 months after TBI in a clinically relevant mouse model deficient mice had any obvious phenotype and were born of TBI. Our results demonstrate that necroptosis is an at normal Mendelian distributions. important novel mediator of chronic neurodegeneration after TBI. Controlled Cortical Impact model of traumatic brain injury Animals were subjected to experimental traumatic brain injury using the previously described Con- trolled Cortical Impact (CCI) method [23, 28–30]. CCI induces a highly reproducible focal lesion and causes progressive brain damage and cognitive decline thereby replicating many acute and chronic characteristics of W ehn et al. acta neuropathol commun (2021) 9:138 Page 3 of 18 human TBI [23]. In short, after induction of anesthesia until they regained full motor activity in order to pre- with buprenorphine (0.1 mg/kg Bw) and isoflurane (4%, vent hypothermia. Carprofen (4 mg/kg every 24 h) was 30 s), animals were sedated with 1.5–2.5% isoflurane in administered i.p. for the following 72 h for analgesia. 30% oxygen and 70% nitrogen under continuous moni- toring of body temperature and heart rate. After right Experimental time‑line parietal craniotomy, the impact was directly applied to Motor function, depression-like behavior, and memory the intact dura with a pressure-driven steel piston with function were evaluated three days before trauma to a diameter of 3  mm (L. Kopacz, University of Mainz, obtain baseline values and up to three months thereafter. Germany; 8  m/s impact velocity, 1  mm penetration Lesion volume and tissue iron was evaluated by repeti- depth, 150 ms contact time). For sham-surgery, the pis- tive MRI up to three months after injury and validated by ton was placed on the dura, but no impact was applied. histology. At the end of the observation time brains were The craniotomy was resealed with tissue glue (VetBond, removed for immunohistochemical analysis of lesion 3  M animal care products, St. Paul, MN) and animals were kept in an incubator at 34 °C and 60% air humidity TBI MRI (T1, T2) Histology Reactive astrogliosis(GFAP) Lesion volume (Nissl) velocity = 8 mm/s depth = 1 mm contact = 150 ms Lesion volume (T2) Activated microglia (Iba-1) Iron deposits (Prussian blue) Iron deposits (T1) Fractal Sholl analysis analysis Behavior RIP1 Beam Walk fl/flCre Days Weeks RIP3 -3 0 123456 7 28 34567 91011 12 KO Tail Suspension RIP1 fl/fl RIP3 WT Barnes Maze RIP3 KO Naïve (no TBI) RIP1 fl/flCre Fig. 1 Experimental groups, methods, and time line Wehn et al. acta neuropathol commun (2021) 9:138 Page 4 of 18 Behavioral testing volume, tissue iron, astrocyte activation, and microglia To exclude age-related factors as a cause for behavioral morphology (Fig. 1). changes during the observation period, all behavioral tests were performed with an additional control group Body weight and general condition of non-traumatized (naive) RIPK3 or RIPK1 deficient Animals were weighed daily from three days before mice (Fig. 1). CCI until day 7 after trauma, then weekly. General con- dition, surgery wounds, behavior, nutrition, and fluid Motor function—Beam Walk balance were checked daily in the early postoperative The Beam Walk Test was performed as previously period, then weekly. described [23, 28, 30] on a 1 m long and 1 cm wide sus- pended wooden rod. Time to cross the beam and the number of missteps was recorded. Animals with more Magnetic resonance imaging and analysis than two missteps in pre-trauma testing were excluded For longitudinal determination of lesion volume, MRI from randomization. measurements were performed 15  min, 24  h, 7  days, one, two, and three months after TBI. For all animals, T1 weighted, T2 weighted, and diffusion weighted Memory and learning behavior—Barnes Maze imaging (DWI) sequences were collected as previously The Barnes Maze test, a well-established paradigm for described [23] under isoflurane anesthesia (1–1.5% in assessing memory function [31, 32], was performed 1, 30% oxygen/70% nitrogen) and multimodal monitor- 2, and 3  months after CCI as previously described [23]. ing of physiological parameters using a 3 T nanoScan In short, the animal was placed on a brightly lit round PET/MR (Mediso, Münster Germany). Sequences were platform with 20 identical holes along its outer rim and collected in the following order: T2-weighted imag- trained to locate a box affixed below one of the apertures ing (2D fast-spin echo (FSE), TR/TE = 3000/57.1  ms , (home cage) as fast as possible. Time to reach the home averages 14, matrix size = 96 × 96; field of cage (latency) as well as distance travelled and walking view = 16 mm × 16 mm; slice thickness = 500 µm, inter- speed were recorded and analyzed using a video track- slice gap = 60  µm), T1-weighted imaging (2D fast-spin ing software (EthoVison XT©, Version 11, 2014 Noldus echo (FSE), TR/TE = 610/28,6  ms, averages 14, matrix Information Technology). Animals were trained for four size = 96 × 96; field of view = 16  mm × 16  mm; slice consecutive days and memory function was evaluated on thickness = 500  µm, interslice gap = 60  µm). Total the sixth day. imaging time was approximately 35 min per mouse and time-point. Lesion volume was measured using ImageJ Tail Suspension test software (Rasband, W.S., ImageJ, U. S. National Insti- The Tail Suspension test is a paradigm to assess depres - tutes of Health, Bethesda, Maryland, USA, https:// sion-like behavior in rodents and was performed as previ- imagej. nih. gov/ ij/, 1997–2018) in T2 sequences. 14 ously described [23]. Briefly, animals were suspended by slices surrounding the lesion were chosen for each data- the tail for three minutes and their movements recorded set and the area segmented using the polygon tool. Vol- and analyzed using a video tracking software (EthoVison ume was then calculated using the following equation: XT©, Version 11, 2014, Noldus Information Technology). The time of inactivity was used as a proxy for depression- V = d ∗ (A1/2 + A2 + A3 . . . + An/2) like behavior [33]. with d being the distance between slices in mm (slice Histological assessment thickness + interslice gap), and A being the measured Lesion volume/ hippocampus volume area in mm . Three months after TBI, animals were fixed with 4% Hippocampal atrophy was assessed in T2-weighted PFA in deep anesthesia by transcardial perfusion. Four- images as previously described [23]. One section teen sequential 50  µm thick coronal sections were cut located in the center of the lesion containing the hip- at 500  µm intervals on a vibratome (Leica, Germany) in pocampus was chosen at the same position for each order to match the tissue volume investigated by MRI, animal. Areas of both hippocampi were segmented stained with cresyl violet according to Nissl, and evalu- using the polygon tool and the ipsilesional area of the ated by histomorphometry for lesion volume using the hippocampus was expressed as % of the area of the following formula as previously described [23]: uninjured contralateral hippocampus. V = d ∗ (A1/2 + A2 + A3 . . . + An/2) W ehn et al. acta neuropathol commun (2021) 9:138 Page 5 of 18 The volume of the hippocampus in the traumatized 100 μm away from the lesion in the hippocampus and at hemisphere was determined in six slices one mm anterior 300 μm in the cortex. For p-MLKL staining, a 5 × objec- until four mm posterior to bregma and then normalized tive was used (EC Plan-Neofluar 5x/0.16 Pol M27) with to the contralesional side. an image matrix of 1434 × 1434 pixel, a pixel scaling of 3.321 × 3.321  μm and a depth of 8 bit. Whole brain Iron deposits images were collected in z-stacks as tile scans with One section per animal was stained with Prussian blue a slice-distance of 5  μm and a total range of 25  μm. To (Iron Stain Kit, Sigma-Aldrich, # HT20) at -1.5 mm from demonstrate the intracellular localization of p-MLKL, bregma to visualize iron deposits using a light micro- a 100 × objective (Epiplan-Neofluar 100x/1.3 Oil Pol scope (Axioscope, Carl Zeiss Microscopy GmbH, Jena M27) was used with an image matrix of 512 × 512 pixel, Germany). Tile scans were then processed in ImageJ a pixel scaling of 0.166 × 0.166  μm and a depth of 8 bit. by using the color deconvolution tool to separate color Images were collected in z-stacks with a slice-distance channels. The blue channel was binarized and the inte - of 0,280 μm and a total range of 5.6 μm. After obtaining grated density was measured using the particle analy- a maximum intensity projection, images were imported sis plugin. Values of the traumatized hemisphere were into ImageJ [36] and intensity of p-MLKL measured expressed as percentage of the contralateral side. in the rim of the lesion and normalized to the signal of DAPI to correct for differences in staining. The corrected signal was then normalized to the same sized region of Immunohistochemistry interest on the contralateral hemisphere. Fifty micrometer thick floating coronal sections were prepared as previously described [34]. Blocking and Analysis of astrocyte coverage incubation with the primary antibody was performed in Assessment of astrocyte coverage was performed using 1% bovine serum albumin, 0.1% gelatin from cold water ImageJ in sections stained for GFAP (see above). Z fish skin, 0.5% Triton X-100 in 0.01 M PBS at pH 7.2–7.4 -stacks were imported into Fiji and split into individual for 72  h at 4  °C. The following primary antibodies were channels. GFAP intensity in five ROI (250 × 250  μm; at used: IBA-1 (rabbit, Wako, #019–19,741, 1:200), GFAP- 0, 250, 500, 750, and 1000  μm distance from the lesion Cy3 (mouse, Sigma Aldrich, #2905, 1:200), RIPK1 (rabbit, in the striatum) was then measured using the mean grey Novusbio, #NBP1-77077SS, 1:100), NeuN (guinea pig, value and normalized to the measurements of the con- Synaptic Systems, #266 004), and phosphorylated MLKL tralesional hemisphere to adjust for possible differences (rabbit, Cell signaling technology, # 91689S, 1:100). After in staining intensity. incubation, sections were washed in PBS and incubated with the following secondary antibodies: anti-rabbit cou- Analysis of microglia coverage and morphology pled to Alexa-fluor 594 (goat anti-rabbit, Thermo Fisher Microglia coverage was manually assessed in maximum Scientific, #A-11012), anti-guinea pig coupled to Alexa- intensity projections of iba-1 stained sections. One sec- fluor 488 (goat anti-guinea pig, Thermo Fisher Scientific, tion per animal was chosen at 1.5 mm from bregma and # A-11073), and anti-mouse coupled to Alexa-fluor 647 two ROIs chosen on the ipsilesional hemisphere, one in (goat anti-mouse, Thermo Fisher Scientific, #A- 32,728) layer V of the cortex at 300 µm away from the lesion and in 0.01 M PBS at pH 7.2–7.4 containing 0.05% Tween 20. one in the CA1a region of the hippocampus. The number Nuclei were stained with 4’,6-Diamidin-2-phenylindol of microglia was normalized to total DAPI positive cell (DAPI, Invitrogen, #D1306) 1:10,000 in 0.01 M PBS. count and expressed in as percentage of coverage in sham Imaging was performed using a ZEISS LSM 900 operated animals. confocal microscope (Carl Zeiss Microscopy GmbH, To assess microglia morphology, Sholl and fractal anal- Jena Germany). GFAP staining was recorded using a ysis were performed to indicate ramification, cell range, 10 × objective (EC Plan-Neofluar 10x/0.30 Pol M27) total cell size, and circularity using a modified protocol with an image matrix of 512 × 512 pixel, a pixel scal- from Young and Morrison [37]. Z-stack images were ing of 0.2 × 0.2  μm and a depth of 8 bit. Whole brain converted to a maximum intensity projection and cells images were collected in z-stacks as tile scans with a were individually cut out using the polygon selection slice-distance of 2  μm and a total range of 14  μm [35]. tool in ImageJ [35]. Only cells fully captured within the For microglia analysis, images were acquired using a z-stack were selected. After background subtraction, 40 × objective (EC Plan-Neofluar 40x/1.30 Oil DIC M27) images were binarized and resized to 600 × 600 pixels with an image matrix of 1024 × 1024 pixel, a pixel scal- keeping the original scale. Speckles or debris around ing of 0.2 × 0.2 μm and a depth of 8 bit. Specific regions the cells were removed using the paintbrush tool. Sholl of interest were collected in Z-stacks to include the analysis was performed using the Sholl analysis plugin entire slice thickness with a slice-distance of 0.4  μm at Wehn et al. acta neuropathol commun (2021) 9:138 Page 6 of 18 in ImageJ [38]. Centered on the soma, concentric cir- dots, suggesting that pMLKL is part of a protein com- cles with an increasing radius of 2  μm were drawn, the plex such as the necrosome (Fig.  2a, lower panel, white number of intersections measured at each radius. After arrowheads). pMLKL staining was almost absent in neu- converting binary images to outlines, fractal analysis ronal RIPK3 deficient mice suggesting that necroptotic was performed using the FracLac plugin for ImageJ [39]. signaling in neurons did essentially not occur in these As described previously [37], the total number of pixels animals (Fig.  2b). Quantification of pMLKL staining present in the cell image of either the filled or outlined showed a highly significant increase of activated MLKL binary image were calculated and later transformed to in wild type mice of both strains, while neuronal Ripk1 or 2 2 μm (pixel area = 0.208 μm ). Cell circularity was calcu- global Ripk3 knock-out completely blunted this response lated as Circularity = 4*π*Area/Perimeter . Maximum (Fig. 2c and d). span across the convex hull represents the maximum dis- tance between two points in the convex hull. Chronic posttraumatic brain damage is reduced in RIPK1 or RIPK3 deficient mice Statistical analysis After demonstrating neuronal necroptotic signaling three Sample size was calculated with the following param- months after TBI in wild type mice and showing that eters: alpha error = 0.05, beta error = 0.2, calculated RIPK1 or 3 deficiency prevented this process, we evalu - standard deviation ranged from 15 to 20% (depending ated lesion volume in cortex and hippocampus of wild type on the parameter investigated), and biologically relevant and RIPK deficient mice by longitudinal MR imaging. Two difference = 30%. All data is given as mean ± standard animals (one Ripk1 fl/fl::CamK2a Cre and one Ripk3 WT) deviation (SD) if not indicated otherwise. For compari- of the study cohort died for unknown reasons before TBI son between groups, Student t-test was used for nor- and were excluded from analysis. All other animals com- mally distributed data and Mann–Whitney Rank Sum pleted the study. Anesthesia and craniotomy did not have test for non-normally distributed data according to the any influence on general outcome parameters: sham-oper - result of Shapiro–Wilk normality test. Measurements ated and CCI animals recovered equally well from surgery over time were tested between groups using One-way in terms of bodyweight (Ripk1 cKO: Additional file  1: Fig- or Two-way ANOVA for Repeated Measurements, fol- ure S2a, Ripk3 KO: Additional file  1: Figure S2b) and gen- lowed by Tukey’s multiple comparisons test for normally eral health score (Ripk1 cKO: Additional file  1: Figure S2c, and Holm-Sidak’s multiple comparisons test for non-nor- Ripk3 KO: Additional file  1: Figure S2d). Exemplary three mally distributed data as post hoc test. Calculations were dimensional reconstructions of the brain showed a large performed with Sigma Plot version 14.0 (Systat Software lesion and a small hippocampus in the ipsilateral hemi- GmbH, Erkrath, Germany). sphere of traumatized wild type mice three months after TBI, while the lesion was significantly smaller and the hip - Results pocampus significantly larger in neuronal RIPK1 deficient fl/fl Cre A total of 33 male RIP1 deficient mice (naïve RIP1 mice (Fig. 3a). Longitudinal investigation of lesion size and fl/fl Cre fl/fl group: n = 4, RIP1 and RIP1 sham groups = 5 hippocampal volume by repetitive MRI showed that the fl/fl fl/fl each, CCI RIP1 group: n = 10, RIP1 Cre group: n = 9) primary damage measured 15 min after trauma was com- −/− +/+ and 32 male RIP3 deficient mice (naïve RIP3, RIP3 , parable in RIPK1 deficient mice and their respective wild −/− +/+ and RIP3 sham group: n = 5 each, CCI R IP3 group type controls (Fig.  3b, t = 15  min: Ripk1 cKO: 18.4 ± 2.6 −/− 3 3 n = 9, CCI RIP3 group n = 8) were operated, assessed, mm , wt: 18.4 ± 1.7 mm ) indicating that the initial trauma +/+ and analyzed for the present study. One RIP3 ani- was similar in all investigated animals. In agreement with mal was excluded from randomization and not used for previous results in this model, lesion size peaked 24 h after the study due to its performance in the Beam Walk Test TBI in both experimental groups (Ripk1 cKO: 27.0 ± 1.8 3 3 before TBI (more than 2 missteps at baseline). mm , + 47% vs. 15  min; wt: 29.8 ± 4.3 mm , + 61% vs. 15 min) as a representation of acute secondary brain dam- Traumatic brain injury induces long‑term necroptotic age. Lesion volume was not different between wild type signaling in neurons and neuronal RIPK1 deficient mice at this time point indi - Phosphorylated MLKL (pMLKL) was used as a spe- cating that necroptosis does not play a significant role for cific marker for necroptosis [40]. Three months after acute lesion progression. Within the first month after TBI, TBI large amounts of pMLKL were found in the rim removal of necrotic tissue and scar formation resulted in of the traumatic cavity, the presumed site of progres- an apparent shrinkage of the lesion. One to two months sive chronic post-trauma brain damage (Fig.  2a, upper after TBI progressive loss of brain tissue started to occur, panel). pMLKL was found by high resolution confocal a process we previously demonstrated to continue for at imaging in the cytoplasm of selected neurons as small least one year after experimental trauma [23]. Chronic W ehn et al. acta neuropathol commun (2021) 9:138 Page 7 of 18 Fig. 2 Phosphorylated MLKL, a marker of necroptosis is reduced in RIP knockout animals. a. Exemplary staining of pMLKL and NeuN (upper panel) and a pMLKL positive neuron at higher magnification (lower panel) in the rim of a traumatic contusion in a wild type mouse. b. pMLKL is significantly reduced in a RIPK3 deficient mouse. c and d p-MLKL signal intensity was significantly increased after TBI in wild type animals (white bars) compared to sham animals indicating significant presence of necroptosis three months after TBI; this increase was significantly blunted in RIPK1 (c) as well as in RIPK3 d mice where there was no difference between TBI and sham animals. Data are presented as mean ± SD; n = 9–10 for RIPK1, n = 8–9 for RIPK3. Two-way RM ANOVA with Tukey’s multiple comparisons test was used. ***p < 0.001, n.s. indicates no significant statistical difference between groups post-trauma tissue injury was significantly reduced in after TBI was similar in RIP deficient and wild type animals neuronal RIPK1 deficient mice as compared to their wild (Additional file  1: Figure S4), while chronic brain dam- type littermate controls (t = 1  month: Ripk1 cKO: 5.7 ± 1.6 age three month after TBI correlated well with the injury 3 3 mm , wt 8.2 ± 1.9 mm , p = 0.0355, t = 3 mon Ripk1 cKO: assessed by MRI (Additional file  1: Figure S5a and b) and 3 3 6.5 ± 1.7 mm , wt: 11.3 ± 2.0 mm , p = 0.0002). A similar was significantly reduced in both knock-out strains (Addi - dynamic was seen in RIPK3 deficient animals. Acute injury tional file 1 : Figure S5c and d). was not affected by RIP 3 knock-out (Fig.  3c t = 24 h, Ripk3 Since memory deficits are a hallmark of chronic post - 3 3 KO: 24,8 ± 3,3 mm , wt: 26,1 ± 5,3 mm ), while chronic traumatic brain damage in mice and TBI patients [23, 41, lesion progression was significantly attenuated from one to 42], we next investigated long-term hippocampal dam- three months after TBI (t = 1  month: Ripk3 KO: 8.8 ± 2.9 age by MRI. In the current TBI model, the hippocampus 3 3 mm , wt: 14.0 ± 1.9 mm , p = 0.006; t = 3  months: Ripk3 is only marginally injured acutely after TBI, but severely 3 3 KO: 7.3 ± 1.3 mm , wt: 11.6 ± 3.6 mm , p = 0.04). Individual affected by progressive chronic damage as previously traces for lesion volumes in each animal are given in Addi- shown [23]. All wild type mice showed significant loss tional file  1: Figure S3a (RIP1 KO) and Additional file  1: Fig- of hippocampal tissue in the traumatized hemisphere ure S3b (RIP3 KO). The MRI findings were corroborated by already one month after TBI (Fig.  3d and e, open bars). histopathological evaluations, i.e. acute brain damage 24 h Starting one month after trauma, hippocampal loss was Wehn et al. acta neuropathol commun (2021) 9:138 Page 8 of 18 significantly less pronounced in neuronal RIPK1 and points were significantly reduced and partly normalized global RIPK3 deficient mice (Fig.  3d, RIPK1: reduction to in neuronal specific RIPK1 and in global RIPK3 deficient 74.9 ± 18.5%, wt: reduction to 40.3 ± 18.7% of contralat- mice (Fig.  5b–m, closed symbols), suggesting that less eral hippocampus, p = 0.0004; Fig.  3e, RIPK3: reduction neuronal cell death was associated with less microglial to 78.0 ± 30.7%, wt: reduction to 48.2% ± 19.6% of con- activation. tralateral hippocampus, p = 0.01). This significant dif - ference persisted until the end of the observation period Neuronal RIPK1 and RIPK3 deficiency improves cognitive three months after TBI and was corroborated by histol- outcome three months after TBI ogy (Additional file 1: Figure S5e and f ). To investigate whether the reduction of lesion size, hip- pocampal damage, scar formation, and microglial activa- Astrogliosis and microglial activation are reduced in RIPK1 tion had an effect of functional outcome, we investigated and RIPK3 deficient mice motor function by beam walk, depression-like behav- The formation of a glial scar and the activation of micro - ior by the Tail Suspension test, and long-term memory glia are other hallmarks of chronic brain damage after using the Barnes Maze test. TBI significantly deterio - TBI [43, 44]. Indeed, we observed a marked increase in rated motor function and induced depression-like behav- glial fibrillary acid protein (GFAP), an astrocyte marker, ior compared to pretrauma performance as previously in the rim of the traumatic cavity and in perilesional tis- described [23], genetic deletion of RIPK1 or RIPK3, how- sue in wild type mice three months after TBI (Fig.  4a, ever, had no effect on these parameters (Fig.  6a–d). Long- left panel). Quantification of GFAP expression showed term memory was normal in non-traumatized Ripk1 an almost four-fold increase in the rim of the lesion or Ripk3 knock-out mice; it was, however, significantly with a decreasing intensity towards perilesional areas disturbed in traumatized animals, i.e. TBI increased the (Fig.  4b and c, open bars). GFAP expression was far less time needed to find the home cage, the latency to goal, pronounced in neuronal specific RIPK1 deficient mice by more than ten times and memory loss progressed over (Fig.  4a, right panel). Quantification of astrocyte density time (Fig.  6e and f, triangles and open circles). In neu- by pixel-based analysis corroborated these findings and ronal specific RIPK1 and in global RIPK3 deficient mice, reveled that activation of astrocytes was significantly however, long-term memory function was almost com- decreased by 25–35% in neuronal specific RIPK1 and in pletely preserved and resembled that of not traumatized global RIPK3 deficient mice (Fig. 4b and c, closed bars). animals (Fig.  6e and f, closed circles). Hence, protection Staining for the microglia marker Ionized calcium- of hippocampal neurons by genetic deletion of RIP 1 or binding adaptor molecule 1 yielded similar but more RIPK3 resulted in preserved long-term memory function. lesion associated findings three months after TBI (Iba1; Fig.  5). In wild type mice Iba-1 staining was most pro- Chronic lesion progression after TBI is associated with iron nounced within a distance of 100 µm from the rim of the deposits and reduced in RIP‑deficient mice traumatic cavity, while only subtle changes were observed Since our data suggest that necroptotic signaling is in areas 300  µm away from the lesion site (Fig.  5, WT). important for chronic brain damage after TBI, we were The density of iba-1 staining was heavily reduced in neu - interested to identify the mechanisms triggering this pro- ronal specific RIPK1 deficient mice (Fig.  5, Ripk1 cKO). cess. Traumatic contusions are associated with hemor- To quantify these changes, we assessed tissue coverage, rhage and subsequent deposition of iron in perilesional area, circularity, and maximal span of microglia in the brain parenchyma. Since free iron is well-known to trig- rim (100  µm) and in the vicinity (300  µm) of the trau- ger ferroptotic [45] and possibly necroptotic cell death matic lesion (Fig.  5b–i). In wild type mice all investi- signaling [46], we hypothesized that chronic posttrau- gated parameters pointed towards a significant activation matic brain damage may be associated with iron deposi- of microglia near the rim of the lesion, i.e. the coverage tion. Since iron can alter MRI signals, we looked for signal and circularity increased, while the area and the maxi- alterations using MRI scans. Indeed, we found hyperin- mal span of microglia decreased (Fig.  5b–i, open bars). tense signals at the border of the lesion one month after Microglia activation and the number of microglial branch (See figure on next page.) Fig. 3 RIPK1 and RIPK3 deficiency significantly reduces posttraumatic brain damage. a 3D reconstruction of lesion volume (blue) in relation to ipsi- and contralateral hippocampus (green) for a RIPK3 wild type and a RIPK3 deficient mouse 3 months after TBI. Scale bar = 5 mm. b and c Lesion volume over time quantified by repetitive T2-weighted MR imaging in RIPK1 (b) and RIPK3 (c) knockout animal. d and e. Hippocampal atrophy over time assessed in longitudinal T2-weighted MRI. RIPK1 (d) and RIPK3 (e) knockout mice show better preservation of hippocampal tissue over time. Mean ± SD; n = 9–10 for RIPK1, n = 8–9 for RIPK3. Two-way RM ANOVA with Sidak’s multiple comparisons test was used. *p < 0.05, **p < 0.005, ***p < 0.001 W ehn et al. acta neuropathol commun (2021) 9:138 Page 9 of 18 Fig. 3 (See legend on previous page.) Wehn et al. acta neuropathol commun (2021) 9:138 Page 10 of 18 Fig. 4 RIPK1 and RIPK3 deficiency reduces reactive astrogliosis three months after TBI. a Exemplary GFAP stainings in a wild type (left) and a RIPK1 knockout (right) animal at different distances from the traumatic contusion. Both images show the right hemisphere, but the left image has been mirrored along the midline for better visualization. Scale bar = 20 µm. Significant astrogliosis is present in the WT mouse, while it is heavily reduced in the neuronal RIPK1 deficient mouse. b and c Quantification of astrocyte coverage at different distances from the lesion. Astrocyte coverage was decreased in neuronal RIPK1 (b) and global RIPK3 (c) knockout animals compared to controls. Data are presented as mean ± SD; n = 9–10 for RIPK1, n = 8–9 for RIPK3. Students t-test for parametric and Man-Whitney-Rank-Sum test for non-parametric data were used. *p < 0.05, **p < 0.005, ***p < 0.001 W ehn et al. acta neuropathol commun (2021) 9:138 Page 11 of 18 TBI by T1-weighted MRI (Fig. 7a, upper panel) and could deficient mice were significantly protected from chronic demonstrate that these signals showed a close spatial cor- brain injury and showed improved neurocognitive func- relation with iron deposits as identified by Prussian blue tion up to three months after TBI. Of note, no protection staining (Fig.  7a, lower panel and Fig.  7b). Comparison was observed within the first days after TBI, suggest - of the volume of iron deposits between wild type and ing that necroptosis is not involved in acute injury, but RIP deficient mice revealed equal amounts of iron in all specifically mediates chronic traumatic brain damage. animals, suggesting that the amount of hemorrhage was Hence, our data further suggest that mechanistically equal in all experimental groups (Fig. 7c and d). In a next acute and chronic neuronal cell death seem to be medi- step, we investigated the spatial and temporal relation- ated by different processes. ship between chronic lesion expansion and iron deposits. Longitudinal MRI and histological assessment revealed For this purpose, we recorded iron deposits one month that lesion progression was associated with parenchy- and lesion area three months after TBI, a time point when mal iron deposition and that neuronal or global deletion the lesion already expanded. In wild type mice the rim of of RIPK1 or RIPK3 prevented lesion progression. These the lesion area, the site of lesion progression, colocal- findings suggest that chronic traumatic brain damage ized with iron deposits, while in neuronal specific RIPK1 may be triggered by iron and mediated by necroptotic deficient mice colocalization was minimal (Fig.  7e). The signaling. The association of lesion progression with iron quantification of lesion area and iron deposition in wild deposits is intriguing and may indicate that iron plays an type, RIPK1 and RIPK3 deficient mice, demonstrated important role in this process, however, further experi- that tissue loss colocalizing with histopathological detec- ments addressing this issue in more detail will need to tion of iron was significantly reduced in RIP deficient further evaluate whether there is a causal or just a correl- animals, suggesting that free iron may be involved in the ative relationship between iron deposition and neuronal pathophysiology of chronic neuronal necroptosis follow- necroptosis. ing TBI (Fig. 7f and g). Hemorrhage and the subsequent degradation of red blood cells releases large amounts of hemoglobin, heme, Discussion and free iron, i.e. molecules with high cytotoxic activity, It is increasingly recognized that next to its acute seque- into brain tissue [55]. From numerous studies investigat- lae, traumatic brain injury is a chronic disease [14]. ing intraparenchymal hemorrhage, a subtype of hemor- Chronic post-trauma brain damage is associated with rhagic stroke, it is well known that specifically free iron inflammation, persists for years after the initial insult, generates reactive oxygen species thereby damaging cell and may spread to areas initially not affected by the initial membranes and causing tissue damage and neurological impact [47]. Affected patients often suffer from neuro- dysfunction [36, 56–59]. Cerebral macro- and microhe- cognitive and mood disorders, personality changes, morrhages are common after TBI [60–62]. Specifically, neurocognitive dysfunction, or even dementia [48–54]. microbleeds have been shown to exert toxic effects on So far, no therapeutic concepts targeting the long-term endothelial cells, astrocytes, neurons, oligodendrocytes, sequelae of TBI exist as the pathophysiology of chronic and microglia and may thus lead to blood–brain barrier traumatic brain injury is still poorly understood. damage, neuronal cell death, demyelination, and chronic Here, we propose, to our knowledge for the first time, inflammation [63]. The importance of blood degrada - that programmed cell death signaling mediates chronic tion products for the pathophysiology of TBI is further neuronal injury after TBI. More specifically, we identified demonstrated by the fact that the presence and extent of the necroptosis signaling molecules RIPK1 and RIPK3 hemorrhage show a close correlation with injury severity to be major players in this process. Neurons affected by and long-term clinical outcome in TBI patients [63, 64]. chronic traumatic damage showed necroptotic signaling In line with these clinical studies, we previously demon- as evidenced by enhanced levels of pMLKL. Moreover, strated that the TBI model used in current study shows Ripk3 global knockout animals as well as neuronal RIPK1 acute macro-hemorrhage, which, however, resolves (See figure on next page.) Fig. 5 RIPK1 and RIPK3 deficiency reduces microglia activation. a. Exemplary stainings for the microglia marker iba1 in WT (upper inserts) and neuronal RIPK1 deficient mice (lower inserts) 100 µm (left inserts) and 300 µm (right inserts) from the rim of the lesion. b ‑i. Coverage and fractal analysis of microglia. In areas closer to the lesion site (100 µm, left side of each panel), knockout animals of both lines showed a decrease in microglia coverage (b. RIP 1, d. RIPK3) compared to their wild type littermates. Fractal analysis revealed that microglia of knockout animals in proximity to the lesion have less processes (c. RIP 1, e. RIPK3) are less circular (f. RIP 1, h. RIPK3), and overall smaller (g. RIP 1, i. RIPK3). In the more distal region, cells resembled those in sham operated animals, with no differences between genotypes. j‑m. Sholl analysis also shows increased ramification, i. e. more active cells, close to the lesion (j. RIP 1, l. RIPK3), but not further away from the lesion site (k. RIP 1, m. RIPK3). Data are presented as mean ± SD; n = 9–10 for RIPK1, n = 8–9 for RIPK3. Student t-test for normalized and Man-Whitney-Rank-Sum-test for non-normalized data was used. *p < 0.05, **p < 0.005, ***p < 0.001. n.s. indicates no significant statistical difference between groups Wehn et al. acta neuropathol commun (2021) 9:138 Page 12 of 18 Fig. 5 (See legend on previous page.) W ehn et al. acta neuropathol commun (2021) 9:138 Page 13 of 18 Fig. 6 RIPK1 and RIPK3 deficiency improves neurocognitive performance three months after TBI. a and b Motor impairment after TBI. Beam Walk Test revealed long term impaired motor function of the left hind limb in TBI animals (missteps compared to respective baseline, # for WT, * for KO), but there were no differences between a RIPK1 or b RIPK3 knockout animals and their respective controls. c and d Depression-like behavior after TBI. Mice of both strains showed an increase of total immobility time throughout the time course of three months, however no significant differences could be detected between RIPK1 (c) and RIPK3 (d) knockout animals compared to wild type. e and f Learning and memory dysfunction after TBI. CCI induces severe long-term memory deficits in WT mice (open circles) while RIPK1 or RIPK3 knockout animals (grey circles) show similar long-term memory function as uninjured littermate controls. Data are presented as mean ± SD; n = 9–10 for RIPK1, n = 8–9 for RIPK3. Two-way RM ANOVA with Tukey’s multiple comparisons test was used. *p < 0.05, **p < 0.005, ***p < 0.001, ****p < 0.0001 Wehn et al. acta neuropathol commun (2021) 9:138 Page 14 of 18 within the first week after injury [23]. In the current occurred in neuronal cells exposed to ferroptosis activa- study we now show by MRI and Prussian blue staining tors and knockdown of the necroptosis-mediators CYLD, that iron persists in pericontusional tissue for up to three RIPK1 or RIPK3 attenuated cell death. Interestingly, the months after the initial impact. Iron deposits identified in role for CYLD in ferroptosis also translated into neuro- still viable pericontusional brain tissue one month after protective effects in  vivo, since CYLD knockout mice TBI co-localized with damaged tissue three months after showed reduced secondary brain damage after TBI com- injury, suggesting that chronic lesion progression pref- pared to controls [21]. These findings are corroborated by erably occurred in areas with previous hemorrhage and results in models of hemorrhagic stroke suggesting that subsequent iron deposition. Since this process was sig- necroptosis and ferroptosis are indeed interconnected nificantly attenuated in neuronal Ripk1 and global Ripk3 under conditions of blood-induced tissue damage [68]. knockout animals, our findings suggest that pericontu - Despite its obvious strengths, the current study also sional iron may be involved in chronic posttraumatic has some notable limitations. We studied only young lesion expansion and that this process is mediated by male animals and are therefore not able to make any necroptotic signaling in neurons. statements on the role of necroptosis in the aged or So far, neuroinflammation was believed to be the main female brain. Further, due to technical limitations, such cause of chronic brain damage after TBI [44]. However, as the lack of specific antibodies for the study of necrop - the mechanisms by which microglial activation promotes tosis in brain tissue, we were only able to demonstrate the neuronal injury and death remain elusive. Further, micro- involvement of a single signaling molecule downstream glial activation after TBI may exert beneficial as well as of RIPK activation, namely pMLKL. Thus, future stud - detrimental effects, i.e. tissue regeneration versus acceler - ies using novel experimental tools will need to define ated damage, respectively. Therefore, disentangling these necroptotic signaling after TBI in more detail. Another opposite functions of microglia may have important shortcoming of the current study is that we demonstrate therapeutic consequences. Our current data suggest that only a spatial correlation between iron deposition and the final steps causing neuronal cell death during chronic necroptosis. Thus, further studies are needed to further post-trauma brain damage depend on RIPK1 and RIPK3 clarify whether there is a causal relationship between activity. Based on these findings, we suggest a hypotheti - iron deposition and neuronal necroptosis. Finally, we cal scenario in which chronic post-trauma brain damage want to point out that memory tests in mice are some- is initiated by the ongoing production of reactive oxy- times hard to interpret since results may be influenced by gen species (ROS) by inflammatory cells. Physiological differences in motor function or the level of disinhibition concentrations of ROS are usually well tolerated by cells which are well known to occur after TBI. We controlled since they are detoxified to water and oxygen by the glu - for differences in motor function and the intensity of tathione system and catalases [65]. However, in the pres- exploratory behavior between groups and are confident ence of iron, hydrogen peroxide is converted to highly that the presented data indeed reflect memory function, reactive hydroxyl radicals by the Fenton reaction and may however, the results need nevertheless to be interpreted initiate a form of programmed cell death called ferropto- with caution. sis [66, 67]. The link between ferroptosis and RIPK1/3- In conclusion, the current study provides evidence that mediated necroptosis in neurons is not fully established, RIPK1 and RIPK3 are critically involved in chronic post- but may be mediated by cylindromatosis (Cyld), a deu- trauma brain damage. Further, our findings suggest that biquitinase able to activate RIPK1 and downstream free iron may be involved in this process. Our results necroptosome formation under conditions of oxidative therefore help to better understand the mechanisms stress as we recently demonstrated [21]. We showed that of chronic post-trauma brain damage and suggest that CYLD-dependent RIPK1/RIPK3 necrosome-formation RIPK1- and RIPK3-mediated necroptosis may represent (See figure on next page.) Fig. 7 Lesion progression occurs in areas with iron deposits. a T1-weighted MRI (upper panel) and Prussian blue staining (lower panel) three months after injury. There is a close spatial correlation between the T1-hyperintense signal and iron staining (arrowheads). b There is high spatial correlation between the area of T1 hyperintensities and iron deposits as assessed by Prussian blue staining. Pearson product-moment correlation analysis. c and d Iron deposits in pericontusional brain tissue assessed by longitudinal MRI. Extent of hemorrhage is comparable in RIPK1 (c) and RIPK3 (d) knockout animals and controls, indicating no differences in hemorrhage size after TBI between groups. T1 hyperintensities decreased over time in all groups, suggesting a very slow resorption of iron over time. e. Co-localization of iron deposits (red) observed at 1 month after TBI (upper panels) and lesion size assessed at the end of the observation period (3 months, middle row, green) suggests a progressive expansion of the lesion towards the regions with iron deposits. f and g Quantification of overlap between iron deposits and lesion. The higher the overlap of iron deposits and lesion size at three months, the higher the rate of tissue loss/ cell death in iron containing tissue. Co-localization is significantly less pronounced in RIPK1 (f) or RIPK3 (g) deficient mice, suggesting a reduced lesion growth in RIP knockouts due to toxic iron residues. Data are presented as mean ± SD; n = 9–10 for RIPK1, n = 8–9 for RIPK3. Two-way RM ANOVA with Tukey’s multiple comparisons test was used. **p < 0.005, ***p < 0.001 W ehn et al. acta neuropathol commun (2021) 9:138 Page 15 of 18 Fig. 7 (See legend on previous page.) Wehn et al. acta neuropathol commun (2021) 9:138 Page 16 of 18 a novel therapeutic target for the treatment of patients MET_V/007), iBOF20/IBF/039 ATLANTIS, Foundation against Cancer (FAF- suffering from the long-term sequels of TBI. F/2016/865, F/2020/1505), CRIG and GIGG consortia, and VIB. The Plesnila group was funded by Munich University´s Förderprogramm für Forschung und Lehre (FöFoLe), by the BMBF-funded research consortium TRAINS (Project Supplementary Information ID: 01EW1709) and by the Munich Cluster of Systems Neurology (SyNergy; The online version contains supplementary material available at https:// doi. Project ID EXC 2145 / ID 390857198). org/ 10. 1186/ s40478- 021- 01236-0. Author contributions Conception and study design: ACW, NP, NAT. Surgery, genotyping, neuro- Additional File 1 Supplementary Fig. S1. Genotyping of RIPK1 and logical testing, histology: ACW, IK. MR imaging: ACW, MD. Data analysis & RIPK3 deficient mice and proof of neuronal specific RIPK1 knock-out in interpretation: ACW, IK, MD, NP, NAT. Statistical analysis: ACW, NAT. Manuscript flox/flox RIPK1 Camk2CreERT2 mice. a. Neuron specific RIPK1 deficient mice preparation: ACW, NP, NAT. Critical revision of the manuscript: all authors. All used for experiments were heterozygous for Camk2CreERT2 and homozy- authors read and approved the final manuscript. gous for the floxed RIPK1 allele. Littermate controls were also homozy- gous for the floxed RIPK1 allele, but did not express the Cre recombinase. Availability of data and materials b. Global RIPK3 deficient mice were homozygous for disrupted allele, The datasets used and/or analyzed during the current study available from the while control mice expressed only the wild type gene. c. and d. To corresponding author on reasonable request. flox/ demonstrate specific neuron specific RIPK1 deficiency in induced RIPK1 flox Camk2CreERT2 mice, we performed immunohistochemistry for RIPK1 and NeuN, a neuronal marker. In the cortex of control mice RIPK1 was Declarations almost exclusively expressed in neurons (upper panels), while in induced flox/flox RIPK1 Camk2CreERT2 mice RIPK1 staining was significantly reduced Ethics approval (lower panels) to 20% of baseline (d). All procedures were reviewed and approved by the respective institutional and governmental authorities and performed according to all regulations. Consent for publication Additional File 1 Supplementary Fig. 2 Body weight and physical con- Not applicable. dition after experimental TBI. a. and b. Weight after TBI. Animals recovered from weight loss directly after trauma within one week after injury; in the Competing interests following observation period weight constantly increased. No differences The authors declare that they have no competing interests. were detected between CCI and sham-operated animals in the RIPK1 (a) and RIPK3 (b) groups. c. and d. General health score to assess recovery. Conflict of interest All animals’ general condition transiently worsened in the perioperative There is no conflict of interest for any of the authors. phase with a peak at day 1 after TBI, but returned to baseline within one week. There was no difference between groups, c. RIPK1, d. RIPK3. Data Author details are presented as mean ± SD; n = 5 for sham, n = 8–10 for TBI. 1 Institute for Stroke and Dementia Research (ISD), LMU Klinikum, Ludwig- Maximilians University Munich, Feodor-Lynen-Str. 17, 81377 Munich, Germany. 2 3 Munich Cluster of Systems Neurology (SyNergy), Munich, Germany. I nstitute for Pharmacology and Clinical Pharmacy, Biochemical-Pharmacological Center Additional File 1 Supplementary Fig. 3. Individual lesion volume pro- Marburg, University of Marburg, Karl-von-Frisch Straße 2 K03, 35032 Marburg, gression for a. RIP1 and b. RIP3 deficient mice. 4 5 Germany. University of Marburg, Marburg, Germany. Molecular Signaling and Cell Death Unit, VIB-UGent Center for Inflammation Research, UGent-VIB Research Building FSVM, Technologiepark 71, 9052 Ghent, Belgium. Depar t- Additional File 1 Supplementary Fig. 4. RIPK3 deficiency does not affect ment of Biomedical Molecular Biology, Ghent University, Ghent, Belgium. acute brain injury after TBI. No differences in lesion volume as assessed by Graduate School of Systemic Neurosciences (GSN), Ludwig-Maximilians histology was detected between RIPK3 knockout mice and C57BL/6 wild University Munich, Munich, Germany. Department of Neurosurgery, LMU type controls at 24 h after TBI. Data are presented as mean ± SD; n = 10. Klinikum, Ludwig-Maximilians University Munich, Munich, Germany. Present Address: Medical Image Analysis Center (MIAC AG) and Qbig, Department of Biomedical Engineering, University of Basel, Basel, Switzerland. Additional File 1 Supplementary Fig. 5. Lesion volumes by MRI and his- Received: 28 May 2021 Accepted: 27 July 2021 tology. a. T2-weighted MRI and Nissl stained coronal section three months after injury. Lesion volume was quantified in T2-weighted MRI images as well as Nissl stained sections obtained in the same animals three months post injury. b. Correlation analysis of lesion volumes assessed by histomorphometry or T2-weighted MRI revealed a strong positive cor- References relation between both methods. Differences in measured lesion size are 1. Dewan MC, Rattani A, Gupta S, Baticulon RE, Hung YC, Punchak M, due to shrinking of tissue in the fixation process, with a shrinking factor Agrawal A, Adeleye AO, Shrime MG, Rubiano AM et al (2018) Estimating of 0.77 ± 0.23. c.—f. Quantification of lesion volume (c. RIPK1, d. RIPK3) the global incidence of traumatic brain injury. J Neurosurg. https:// doi. as well as hippocampal damage (e. RIPK1, f. RIPK3) using both methods. org/ 10. 3171/ 2017. 10. JNS17 352 Data are presented as mean ± SD; n = 8–10 for TBI. Pearson correlation, 2. Injury GBDTB, Spinal Cord Injury C (2019) Global, regional, and national Students t-test for parametric and Man-Whitney-Rank-Sum test for non- burden of traumatic brain injury and spinal cord injury, 1990-2016: a sys- parametric data were used. *p < 0.05, ***p < 0,001. tematic analysis for the Global Burden of Disease Study 2016. The Lancet Neurology 18: 56-87https:// doi. org/ 10. 1016/ S1474- 4422(18) 30415-0 3. Maas AIR, Menon DK, Adelson PD, Andelic N, Bell MJ, Belli A, Bragge Acknowledgements P, Brazinova A, Buki A, Chesnut RM et al (2017) Traumatic brain injury: We want to thank Dr. Ulrike Schillinger for her support during the analysis integrated approaches to improve prevention, clinical care, and research. of the MRI data. Research in the Vandenabeele group is supported by Flem- Lancet Neurol 16:987–1048. https:// doi. org/ 10. 1016/ S1474- 4422(17) ish grants (EOS MODEL-IDI, FWO Grant 30826052), FWO research grants 30371-X (G.0E04.16N, G.0C76.18N, G.0B71.18N, G.0B96.20N), Methusalem (BOF16/ W ehn et al. acta neuropathol commun (2021) 9:138 Page 17 of 18 4. Rabinowitz AR, Levin HS (2014) Cognitive sequelae of traumatic brain 23. Mao X, Terpolilli NA, Wehn A, Chen S, Hellal F, Liu B, Seker B, Plesnila N injury. Psychiatr Clin North Am 37:1–11. https:// doi. org/ 10. 1016/j. psc. (2019) Progressive histopathological damage occurring up to one year 2013. 11. 004 after experimental traumatic brain injury is associated with cognitive 5. Selassie AW, Zaloshnja E, Langlois JA, Miller T, Jones P, Steiner C (2008) decline and depression-like behavior. J Neurotrauma: Doi https:// doi. org/ Incidence of long-term disability following traumatic brain injury 10. 1089/ neu. 2019. 6510 hospitalization, United States, 2003. J Head Trauma Rehabil 23:123–131. 24. Kilkenny C, Browne WJ, Cuthill IC, Emerson M, Altman DG (2010) Improv- https:// doi. org/ 10. 1097/ 01. HTR. 00003 14531. 30401. 39 ing bioscience research reporting: the ARRIVE guidelines for reporting 6. Silver JM, McAllister TW, Arciniegas DB (2009) Depression and cogni- animal research. PLoS Biol 8:e1000412. https:// doi. org/ 10. 1371/ journ al. tive complaints following mild traumatic brain injury. Am J Psychiatry pbio. 10004 12 166:653–661. https:// doi. org/ 10. 1176/ appi. ajp. 2009. 08111 676 25. Guillen J (2012) FELASA guidelines and recommendations. J Am Assoc 7. Ramos-Cejudo J, Wisniewski T, Marmar C, Zetterberg H, Blennow K, de Lab Anim Sci 51:311–321 Leon MJ, Fossati S (2018) Traumatic brain injury and Alzheimer’s disease: 26. Takahashi N, Vereecke L, Bertrand MJ, Duprez L, Berger SB, Divert T, Gon- the cerebrovascular link. EBioMedicine 28:21–30. https:// doi. org/ 10. calves A, Sze M, Gilbert B, Kourula S et al (2014) RIPK1 ensures intestinal 1016/j. ebiom. 2018. 01. 021 homeostasis by protecting the epithelium against apoptosis. Nature 8. MacKenzie JD, Siddiqi F, Babb JS, Bagley LJ, Mannon LJ, Sinson GP, Gross- 513:95–99. https:// doi. org/ 10. 1038/ natur e13706 man RI (2002) Brain atrophy in mild or moderate traumatic brain injury: a 27. Newton K, Sun X, Dixit VM (2004) Kinase RIP3 is dispensable for normal longitudinal quantitative analysis. AJNR Am J Neuror 23:1509–1515 NF-kappa Bs, signaling by the B-cell and T-cell receptors, tumor necrosis 9. Reider-Groswasser I, Cohen M, Costeff H, Groswasser Z (1993) Late factor receptor 1, and Toll-like receptors 2 and 4. Mol Cell Biol 24:1464– CT findings in brain trauma: relationship to cognitive and behavioral 1469. https:// doi. org/ 10. 1128/ mcb. 24.4. 1464- 1469. 2004 sequelae and to vocational outcome. AJR Am J Roentgenol 160:147–152. 28. Terpolilli NA, Zweckberger K, Trabold R, Schilling L, Schinzel R, Tegtmeier https:// doi. org/ 10. 2214/ ajr. 160.1. 84166 13 F, Plesnila N (2009) The novel nitric oxide synthase inhibitor 4-amino- 10. Cardoso ER, Galbraith S (1985) Posttraumatic hydrocephalus–a retrospec- tetrahydro-L-biopterine prevents brain edema formation and intracranial tive review. Surg Neurol 23:261–264. https:// doi. org/ 10. 0090/ 3019(85) hypertension following traumatic brain injury in mice. J Neurotrauma 90092- 8[pii] 26:1963–1975. https:// doi. org/ 10. 1089/ neu. 2008- 0853 11. Guyot LL, Michael DB (2000) Post-traumatic hydrocephalus. Neurol Res 29. Zweckberger K, Eros C, Zimmermann R, Kim SW, Engel D, Plesnila N 22:25–28 (2006) Eec ff t of early and delayed decompressive craniectomy on 12. Honeybul S, Ho KM (2012) Incidence and risk factors for post-traumatic secondary brain damage after controlled cortical impact in mice. J Neu- hydrocephalus following decompressive craniectomy for intractable rotrauma 23:1083–1093. https:// doi. org/ 10. 1089/ neu. 2006. 23. 1083 intracranial hypertension and evacuation of mass lesions. J Neurotrauma 30. Zweckberger K, Stoffel M, Baethmann A, Plesnila N (2003) Eec ff t of 29:1872–1878. https:// doi. org/ 10. 1089/ neu. 2012. 2356 decompression craniotomy on increase of contusion volume and func- 13. Mazzini L, Campini R, Angelino E, Rognone F, Pastore I, Oliveri G (2003) tional outcome after controlled cortical impact in mice. J Neurotrauma Posttraumatic hydrocephalus: a clinical, neuroradiologic, and neuropsy- 20:1307–1314. https:// doi. org/ 10. 1089/ 08977 15033 22686 102 chologic assessment of long-term outcome. Arch Phys Med Rehabil 31. Barnes CA (1979) Memory deficits associated with senescence: a neu- 84:1637–1641 rophysiological and behavioral study in the rat. J Comp Physiol Psychol 14. Faden AI, Loane DJ (2015) Chronic neurodegeneration after traumatic 93:74–104. https:// doi. org/ 10. 1037/ h0077 579 brain injury: alzheimer disease, chronic traumatic encephalopathy, or 32. Gawel K, Gibula E, Marszalek-Grabska M, Filarowska J, Kotlinska JH (2019) persistent neuroinflammation? Neurotherapeutics 12:143–150. https:// Assessment of spatial learning and memory in the Barnes maze task in doi. org/ 10. 1007/ s13311- 014- 0319-5 rodents-methodological consideration. Naunyn Schmiedebergs Arch 15. Johnson VE, Stewart JE, Begbie FD, Trojanowski JQ, Smith DH, Stewart Pharmacol 392:1–18. https:// doi. org/ 10. 1007/ s00210- 018- 1589-y W (2013) Inflammation and white matter degeneration persist for years 33. Cryan JF, Mombereau C, Vassout A (2005) The tail suspension test as a after a single traumatic brain injury. Brain 136:28–42. https:// doi. org/ 10. model for assessing antidepressant activity: review of pharmacological 1093/ brain/ aws322 and genetic studies in mice. Neurosci Biobehav Rev 29:571–625. https:// 16. Loane DJ, Kumar A, Stoica BA, Cabatbat R, Faden AI (2014) Progressive doi. org/ 10. 1016/j. neubi orev. 2005. 03. 009 neurodegeneration after experimental brain trauma: association with 34. Ghosh M, Balbi M, Hellal F, Dichgans M, Lindauer U, Plesnila N (2015) Peri- chronic microglial activation. J Neuropathol Exp Neurol 73:14–29 cytes are involved in the pathogenesis of cerebral autosomal dominant 17. Ramlackhansingh AF, Brooks DJ, Greenwood RJ, Bose SK, Turkheimer FE, arteriopathy with subcortical infarcts and leukoencephalopathy. Ann Kinnunen KM, Gentleman S, Heckemann RA, Gunanayagam K, Gelosa Neurol 78:887–900. https:// doi. org/ 10. 1002/ ana. 24512 G et al (2011) Inflammation after trauma: microglial activation and trau- 35. Schindelin J, Arganda-Carreras I, Frise E, Kaynig V, Longair M, Pietzsch T, matic brain injury. Ann Neurol 70:374–383. https:// doi. org/ 10. 1002/ ana. Preibisch S, Rueden C, Saalfeld S, Schmid B et al (2012) Fiji: an open- 22455 source platform for biological-image analysis. Nat Methods 9:676–682. 18. Ritzel RM, Doran SJ, Barrett JP, Henry RJ, Ma EL, Faden AI, Loane DJ (2018) https:// doi. org/ 10. 1038/ nmeth. 2019 Chronic alterations in systemic immune function after traumatic brain 36. Hua Y, Xi G, Keep RF, Hoff JT (2000) Complement activation in the brain injury. J Neurotrauma 35:1419–1436. https:// doi. org/ 10. 1089/ neu. 2017. after experimental intracerebral hemorrhage. J Neurosurg 92:1016–1022. 5399https:// doi. org/ 10. 3171/ jns. 2000. 92.6. 1016 19. Hitomi J, Christofferson DE, Ng A, Yao J, Degterev A, Xavier RJ, Yuan J 37. Young K, Morrison H (2018) Quantifying microglia morphology from pho- (2008) Identification of a molecular signaling network that regulates a tomicrographs of immunohistochemistry prepared tissue using imageJ. J cellular necrotic cell death pathway. Cell 135:1311–1323. https:// doi. org/ Vis Exp: Doi https:// doi. org/ 10. 3791/ 57648 10. 1016/j. cell. 2008. 10. 044 38. Ferreira TA, Blackman AV, Oyrer J, Jayabal S, Chung AJ, Watt AJ, Sjostrom 20. Moquin DM, McQuade T, Chan FK (2013) CYLD deubiquitinates RIP1 in PJ, van Meyel DJ (2014) Neuronal morphometry directly from bitmap the TNFα-induced necrosome to facilitate kinase activation and pro- images. Nat Methods 11:982–984. https:// doi. org/ 10. 1038/ nmeth. 3125 grammed necrosis. PLoS ONE 8:e76841. https:// doi. org/ 10. 1371/ journ al. 39. Karperien A (1999–2013) FracLac for ImageJ. pone. 00768 41 40. Sun L, Wang H, Wang Z, He S, Chen S, Liao D, Wang L, Yan J, Liu W, Lei X 21. Ganjam GK, Terpolilli NA, Diemert S, Eisenbach I, Hoffmann L, Reuther C, et al (2012) Mixed lineage kinase domain-like protein mediates necrosis Herden C, Roth J, Plesnila N, Culmsee C (2018) Cylindromatosis mediates signaling downstream of RIP3 kinase. Cell 148:213–227. https:// doi. org/ neuronal cell death in vitro and in vivo. Cell Death Differ 25:1394–1407. 10. 1016/j. cell. 2011. 11. 031 https:// doi. org/ 10. 1038/ s41418- 017- 0046-7 41. Ariza M, Serra-Grabulosa JM, Junque C, Ramirez B, Mataro M, Poca A, Bar- 22. Krieg SM, Voigt F, Knuefermann P, Kirschning CJ, Plesnila N, Ringel F (2017) gallo N, Sahuquillo J (2006) Hippocampal head atrophy after traumatic Decreased secondary lesion growth and attenuated immune response brain injury. Neuropsychologia 44:1956–1961. https:// doi. org/ 10. 1016/j. after traumatic brain injury in Tlr2/4(-/-) mice. Front Neurol 8:455. https:// neuro psych ologia. 2005. 11. 007 doi. org/ 10. 3389/ fneur. 2017. 00455 Wehn et al. acta neuropathol commun (2021) 9:138 Page 18 of 18 42. Jorge RE, Acion L, Starkstein SE, Magnotta V (2007) Hippocampal 57. Derry PJ, Vo ATT, Gnanansekaran A, Mitra J, Liopo AV, Hegde ML, Tsai AL, volume and mood disorders after traumatic brain injury. Biol Psychiatry Tour JM, Kent TA (2020) The chemical basis of intracerebral hemorrhage 62:332–338. https:// doi. org/ 10. 1016/j. biops ych. 2006. 07. 024 and cell toxicity with contributions from eryptosis and ferroptosis. Front 43. Jassam YN, Izzy S, Whalen M, McGavern DB, El Khoury J (2017) Neuroim- Cell Neurosci 14:603043. https:// doi. org/ 10. 3389/ fncel. 2020. 603043 munology of traumatic brain injury: time for a paradigm shift. Neuron 58. Letarte PB, Lieberman K, Nagatani K, Haworth RA, Odell GB, Duff TA (1993) 95:1246–1265. https:// doi. org/ 10. 1016/j. neuron. 2017. 07. 010 Hemin: levels in experimental subarachnoid hematoma and effects 44. Morganti-Kossmann MC, Semple BD, Hellewell SC, Bye N, Ziebell JM on dissociated vascular smooth-muscle cells. J Neurosurg 79:252–255. (2019) The complexity of neuroinflammation consequent to traumatic https:// doi. org/ 10. 3171/ jns. 1993. 79.2. 0252 brain injury: from research evidence to potential treatments. Acta Neuro- 59. Wan J, Ren H, Wang J (2019) Iron toxicity, lipid peroxidation and ferropto- pathol 137:731–755. https:// doi. org/ 10. 1007/ s00401- 018- 1944-6 sis after intracerebral haemorrhage. Stroke Vasc Neurol 4:93–95. https:// 45. Dixon SJ, Lemberg KM, Lamprecht MR, Skouta R, Zaitsev EM, Gleason CE, doi. org/ 10. 1136/ svn- 2018- 000205 Patel DN, Bauer AJ, Cantley AM, Yang WS et al (2012) Ferroptosis: an iron- 60. Hasiloglu ZI, Albayram S, Selcuk H, Ceyhan E, Delil S, Arkan B, Baskoy L dependent form of nonapoptotic cell death. Cell 149:1060–1072. https:// (2011) Cerebral microhemorrhages detected by susceptibility-weighted doi. org/ 10. 1016/j. cell. 2012. 03. 042 imaging in amateur boxers. AJNR Am J Neuroradiol 32:99–102. https:// 46. Pasparakis M, Vandenabeele P (2015) Necroptosis and its role in inflam-doi. org/ 10. 3174/ ajnr. A2250 mation. Nature 517:311–320. https:// doi. org/ 10. 1038/ natur e14191 61. Park JH, Park SW, Kang SH, Nam TK, Min BK, Hwang SN (2009) Detection 47. Plesnila N (2016) The immune system in traumatic brain injury. Curr Opin of traumatic cerebral microbleeds by susceptibility-weighted image of Pharmacol 26:110–117. https:// doi. org/ 10. 1016/j. coph. 2015. 10. 008 MRI. J Korean Neurosurg Soc 46:365–369. https:// doi. org/ 10. 3340/ jkns. 48. Al-Khindi T, Macdonald RL, Schweizer TA (2010) Cognitive and functional 2009. 46.4. 365 outcome after aneurysmal subarachnoid hemorrhage. Stroke 41:e519– 62. van der Horn HJ, de Haan S, Spikman JM, de Groot JC, van der Naalt J e536. https:// doi. org/ 10. 1161/ STROK EAHA. 110. 581975 (2018) Clinical relevance of microhemorrhagic lesions in subacute mild 49. Alghnam S, AlSayyari A, Albabtain I, Aldebasi B, Alkelya M (2017) Long- traumatic brain injury. Brain Imaging Behav 12:912–916. https:// doi. org/ term disabilities after traumatic head injury ( THI): a retrospective analysis 10. 1007/ s11682- 017- 9743-6 from a large level-I trauma center in Saudi Arabia. Inj Epidemiol 4:29. 63. Glushakova OY, Johnson D, Hayes RL (2014) Delayed increases in micro- https:// doi. org/ 10. 1186/ s40621- 017- 0126- 7[pii] vascular pathology after experimental traumatic brain injury are associ- 50. Andelic N, Howe EI, Hellstrom T, Sanchez MF, Lu J, Lovstad M, Roe C ated with prolonged inflammation, blood-brain barrier disruption, and (2018) Disability and quality of life 20 years after traumatic brain injury. progressive white matter damage. J Neurotrauma 31:1180–1193. https:// Brain Behav. https:// doi. org/ 10. 1002/ brb3. 1018doi. org/ 10. 1089/ neu. 2013. 3080 51. Draper K, Ponsford J (2008) Cognitive functioning ten years following 64. Lawrence TP, Pretorius PM, Ezra M, Cadoux-Hudson T, Voets NL (2017) traumatic brain injury and rehabilitation. Neuropsychology 22:618–625. Early detection of cerebral microbleeds following traumatic brain injury https:// doi. org/ 10. 1037/ 0894- 4105. 22.5. 618 using MRI in the hyper-acute phase. Neurosci Lett 655:143–150. https:// 52. Hackett ML, Anderson CS (2000) Health outcomes 1 year after subarach-doi. org/ 10. 1016/j. neulet. 2017. 06. 046 noid hemorrhage: an international population-based study. The austral- 65. Conrad M, Pratt DA (2019) The chemical basis of ferroptosis. Nat Chem ian cooperative research on subarachnoid hemorrhage study group. Biol 15:1137–1147. https:// doi. org/ 10. 1038/ s41589- 019- 0408-1 Neurology 55:658–662 66. Li Q, Han X, Lan X, Gao Y, Wan J, Durham F, Cheng T, Yang J, Wang Z, Jiang 53. Hart T, Whyte J, Polansky M, Millis S, Hammond FM, Sherer M, Bushnik T, C et al (2017) Inhibition of neuronal ferroptosis protects hemorrhagic Hanks R, Kreutzer J (2003) Concordance of patient and family report of brain. JCI insight 2:e90777. https:// doi. org/ 10. 1172/ jci. insig ht. 90777 neurobehavioral symptoms at 1 year after traumatic brain injury. Arch 67. Ratan RR (2020) The chemical biology of ferroptosis in the central nerv- Phys Med Rehabil 84:204–213. https:// doi. org/ 10. 1053/ apmr. 2003. 50019 ous system. Cell Chem Biol 27:479–498. https:// doi. org/ 10. 1016/j. chemb 54. Himanen L, Portin R, Isoniemi H, Helenius H, Kurki T, Tenovuo O (2005) iol. 2020. 03. 007 Cognitive functions in relation to MRI findings 30 years after traumatic 68. Zille M, Karuppagounder SS, Chen Y, Gough PJ, Bertin J, Finger J, Milner brain injury. Brain Inj 19:93–100 TA, Jonas EA, Ratan RR (2017) Neuronal death after hemorrhagic stroke 55. Regan RF, Panter SS (1996) Hemoglobin potentiates excitotoxic injury in in vitro and in vivo shares features of ferroptosis and necroptosis. Stroke cortical cell culture. J Neurotrauma 13:223–231. https:// doi. org/ 10. 1089/ 48:1033–1043. https:// doi. org/ 10. 1161/ strok eaha. 116. 015609 neu. 1996. 13. 223 56. Bhasin RR, Xi G, Hua Y, Keep RF, Hoff JT (2002) Experimental intracerebral Publisher’s Note hemorrhage: effect of lysed erythrocytes on brain edema and blood- Springer Nature remains neutral with regard to jurisdictional claims in pub- brain barrier permeability. Acta Neurochir Suppl 81:249–251. https:// doi. lished maps and institutional affiliations. org/ 10. 1007/ 978-3- 7091- 6738-0_ 65

Journal

Acta Neuropathologica CommunicationsSpringer Journals

Published: Aug 17, 2021

Keywords: Traumatic brain injury; Chronic posttraumatic brain damage; Magnetic resonance imaging; Necroptosis; Ferroptosis; Neuroprotection

References