Long pentraxin PTX3 is upregulated systemically and centrally after experimental neurotrauma, but its depletion leaves unaltered sensorimotor deficits or histopathology

Long pentraxin PTX3, a pattern recognition molecule involved in innate immune responses, is upregulated by pro-inflammatory stimuli, contributors to secondary damage in traumatic brain injury (TBI). We analyzed PTX3 involvement in mice subjected to controlled cortical impact, a clinically relevant TBI mouse model. We measured PTX3 mRNA and protein in the brain and its circulating levels at different time point post-injury, and assessed behavioral deficits and brain damage progression in PTX3 KO mice. PTX3 circulating levels significantly increased 1–3 weeks after injury. In the brain, PTX3 mRNA was upregulated in different brain areas starting from 24 h and up to 5 weeks post-injury. PTX3 protein significantly increased in the brain cortex up to 3 weeks post-injury. Immunohistochemical analysis showed that, 48 h after TBI, PTX3 was localized in proximity of neutrophils, likely on neutrophils extracellular traps (NETs), while 1- and 2- weeks post-injury PTX3 co-localized with fibrin deposits. Genetic depletion of PTX3 did not affect sensorimotor deficits up to 5 weeks post-injury. At this time-point lesion volume and neuronal count, axonal damage, collagen deposition, astrogliosis, microglia activation and phagocytosis were not different in KO compared to WT mice. Members of the long pentraxin family, neuronal pentraxin 1 (nPTX1) and pentraxin 4 (PTX4) were also over-expressed in the traumatized brain, but not neuronal pentraxin 2 (nPTX2) or short pentraxins C-reactive protein (CRP) and serum amyloid P-component (SAP). The long-lasting pattern of activation of PTX3 in brain and blood supports its specific involvement in TBI. The lack of a clear-cut phenotype in PTX3 KO mice may depend on the different roles of this protein, possibly involved in inflammation early after injury and in repair processes later on, suggesting distinct functions in acute phases versus sub-acute or chronic phases. Brain long pentraxins, such as PTX4—shown here to be overexpressed in the brain after TBI—may compensate for PTX3 absence.

Traumatic brain injury (TBI) is the leading cause of mortality in young adults and a major cause of death and disability across all ages in all countries, with no specific agent proved to be effective in dampening the neurologic sequelae post-injury 1 . Inflammatory processes contribute to outcome following TBI. Modulation of these processes represents a possibility to limit secondary neuronal injury and to improve patient outcome 2 although a detailed knowledge of the processes required to identify therapeutical targets is still largely lacking.
Pentraxins are a superfamily of phylogenetically conserved humoral mediators of innate immunity 3 that share conserved sequences. They include short (such as C-reactive protein, CRP, serum amyloid P-component, SAP) and long (such as pentraxin 3 PTX3, neuronal pentraxin 1, nPTX1, neuronal pentraxin 2, nPTX2 and pentraxin 4, PTX4) members. Among long pentraxins, PTX3 is an essential mediator of innate resistance to selected pathogens of fungal, bacterial and viral origin, and is involved in regulation of inflammation, tissue remodeling and cancer 4 . PTX3 acts as fine tuner of inflammation in several diseases, with detrimental or beneficial effects that depend on the nature of the insult 4-6 . PTX3 regulatory role for inflammatory responses may be played in a time-dependent manner, i.e. limiting neutrophil recruitment at acute phases and orchestrating tissue repair processes at chronic phases [7][8][9] . In brain disorders, available evidence indicates that PTX3 may play a protective role preserving neuron survival after seizures or stroke 10,11 . PTX3 was also proposed to dampen neutrophil recruitment to injury site after cerebral ischemia, maintain blood-brain barrier integrity, regulate glial scar formation and promote injury resolution thus highlighting its involvement in brain repair and recovery [12][13][14] . PTX3 is also know to modulate inflammation mediated by the complement system 15,16 , a key contributor to TBI pathogenesis in experimental models [17][18][19] and in TBI patients [20][21][22] .
The potential involvement of PTX3 in TBI is directly suggested by the observation that serum levels of PTX3 increase significantly after severe TBI in patients and are independently associated with hospital mortality 23 . However, the role of PTX3 in TBI remains poorly understood 24,25 . We therefore investigated PTX3 in mice subjected to controlled cortical impact (CCI), a clinically relevant model of TBI [26][27][28][29] . We analyzed PTX3 plasma and brain presence in acute (as early as 30 min), sub-acute (1-3 weeks) and chronic (4-5 weeks) phases after TBI. We identified PTX3 localization in association with different cell populations present in the injured brain, namely neutrophils, astrocytes, neurons, microglia and endothelial cells, as well as with fibrin(ogen) deposits. Then, to further investigate PTX3 relevance in TBI pathophysiology, we compared damage progression, assessed by sensorimotor and histopathological analysis in wild-type (WT) and PTX3 depleted (PTX3 KO) mice. Our data reports a long-lasting pattern of activation of PTX3 in brain and blood following TBI, and supports the hypothesis that PTX3 contributes to the progression of the lesion with effects varying over time, enhancing inflammation early after injury and fostering repair processes later on.

Methods
Mice. Procedures involving animals and their care were conducted in conformity with institutional guidelines in compliance with national and international laws and policies. The experimental protocols were approved by Ethical Committee at the Istituto di Ricerche Farmacologiche Mario Negri IRCCS and by Italian Ministry of Health (prot. 9F5F5.81 authorization nº 753/2017-PR). We used male 9 week old C57BL/6 J mice (Charles Rivers-Italy) weighting 22-28 g, either WT or with targeted depletion of PTX3 (PTX3 KO, C57BL/6 J inbred genetic background, Humanitas Clinical and Research Center) 30 . The protocols and details of this report are in Quantification of PTX3 presence in the brain cortex. The brain coronal Sects. (20 µm) were incubated overnight at 4 °C with primary polyclonal goat antibody anti-mouse PTX3 (0.2 mg/mL, 1:100; R&D Systems, Minneapolis, MN, USA) followed by a secondary biotinylated antibody against goat or rat IgG. Positive cells were stained with Tyramide Fluorescein (1:300, Perkin Elmer, Milan, Italy). Cell nuclei were stained with 40-6-diamidino-2-phenylindole (Hoechst, 1 mg/ml, Invitrogen, Carlsbad, CA). For negative control staining, the primary antibody was omitted, and no staining was observed. Three 20 μm-thick coronal sections at 0.4, 1.6, and 2.8 mm posterior to bregma were selected from each mouse brain for quantification. Confocal microscopy was done with a Nikon A1 confocal scan unit with a 20 × 0.5 numerical aperture (NA) objective, managed by NIS elements software. Tissues were imaged at laser excitation of 405 nm (for nuclei) and 488 nm (for PTX3) 37 . Image acquisition was done at 12-bit, keeping the fluorescent signal in a non-saturated range (0-1000 greyscale values). The acquisition was done over an area sized 2 × 2.5 mm, positioned in the ipsilateral hemisphere along the cortical region proximal to the lesion, with a pixel size of 0.62 µm. Acquisition was done over 8.3 µm thick stacks, with a step size of 2.075 µm. The different focal planes were merged into a single stack by maximum intensity projection to ensure consistent focus throughout the sample. Immunostaining for PTX3 was quantified by assessing fluorescence intensity using Fiji software. To subtract the background signal, a minimum threshold was applied based in the highest grayscale value of background 18 . PTX3 signal was reported as integrated density (ID). Dako, Santa Clara, CA, USA). Primary antibodies were followed by appropriate Alexa 488-, Alexa-546 or Alexa 555-conjugated secondary antibodies raised in goat (4 µg/mL, Life Sciences, Hercules, CA, USA). Cell nuclei were stained with Hoechst (1 mg/ml, Invitrogen, Carlsbad, CA, USA). For negative control staining, the primary antibodies were omitted, and no staining was observed. Immunofluorescence was acquired using a scanning sequential mode to avoid bleed-through effects by an IX81 microscope equipped with a confocal scan unit FV500 with 4 laser lines: Ar-Kr (488 nm), He-Ne red (646 nm), and He-Ne green (532 nm) (Olympus, Tokyo, Japan) and a UV diode. Three-dimensional volumes were acquired over 10 µm stacks, with 0.23 µm step size. Images were managed and elaborated to obtain three-dimensional renderings with Imaris v.6 (Bitplane) and GNU Image Manipulation Program (GIMP).
Sensorimotor deficits. Composite neuroscore. Mice were scored from 4 (normal) to 0 (severely impaired) for each of the following: (1) forelimb function during walking on the grid and flexion response during suspension by the tail; (2) hindlimb function during walking on the grid and extension during suspension by the tail;(3) resistance to lateral right and left push. The best total score is 12 18 . Quantification of histological stainings. Contusion volume. Eight coronal section from bregma + 0.6 to -4.0 mm were acquired with an Olympus BX-61 Virtual Stage microscope using a 2 × objective lens, with a pixel size of 3.49 µm. Contusion volume 1w and 5w after TBI was analysed as previously described 40 .
Image acquisition for histopathological analysis. Neuronal cell count, contralateral white matter quantification, collagen deposition and immunohistochemistry assays were performed 5w after TBI. Three 20 µm thick coronal sections at 0.4, 1.6, and 2.8 mm posterior to bregma were selected from each mouse brain. The entire sections were acquired with an Olympus BX-61 Virtual Stage microscope using a 20 × objective lens, with a pixel size of 0.346 µm. Acquisition was done over 10 µm thick stacks, with a step size of 2 µm. The different focal planes were merged into a single stack by mean intensity projection to ensure consistent focus throughout the sample 33 .
Neuronal density. Neuronal density was performed at 1w and 5w after TBI by segmenting the cells over the entire cortex and in the corresponding contralateral hemisphere and excluding the round-shaped signal sized below the area threshold of 25 mm 2 that is known to be associated with glial cells as reported previously 17 . Quantification was performed by Fiji software. Data were expressed as the total number of neurons quantified in the selected cortical region.
Contralateral white matter. White matter areas of corpus callosum (CC) and external capsula (EC) were disrupted by the focal trauma pathology in the ipsilateral hemisphere, so only the contralateral hemisphere was quantified. The whole contralateral CC was selected, and the contralateral EC was analysed up to the lower limit of the primary somatosensory cortex. The whole selected region of interest (ROI) were analysed.
Collagen deposition. Collagen deposition was quantified over an area included within a 300 µm radius from the contusion edge. Quantification was performed by Fiji software by segmenting the positive red signal. Data were expressed as the percentage of collagen area within the ROI. Biotinylated secondary antibodies (7.5 µg/ml, Vector Laboratories, Burlingame, CA, USA) were used. GFAP, CD11b and CD68 immunopositive cells were identified by reaction with 3,3 diaminobenzidine-tetrahydrochloride (DAB, Vector Laboratories, Burlingame, CA, USA) as previously described 41 . Negative control studies, without the primary antibody, were performed in parallel. www.nature.com/scientificreports/ The ipsilateral cortex was analyzed over an area included within a 350 µm radius from the contusion edge. Images were analyzed using Fiji software by segmenting the positive signal. GFAP, CD11b and CD68 immunostained area were expressed as positive pixels/total assessed pixels and reported as the percentage of total stained area 18 .

Morphological analysis.
Morphological analysis was carried out on CD11b-stained Sects. 1w after TBI.
Image processing was performed using Fiji software. An algorithm was created to segmentate and analyze stained cells. Briefly, images were first scaled into microns (pixel size = 0.172 × 0.172 µm). Background was subtracted and a math operation was applied so that all the gray values greater than a specified constant were replaced by the constant. The constant was defined by an operator on the basis of the best segmentation performance on pilot images and did not change throughout the experimental groups. Images were then binarized and smoothed to best fit cell shape and get rid of single positive pixels still present in the background. A further step of pixel erosion helped to achieve satisfactory cell shape fitting. To be sure to select only cells entirely present in the acquired field, cells with area > 25 µm 2 were considered for analysis. Once segmented, the objects were measured for the following parameters: area, perimeter, Feret's diameter (max caliper), circularity, and solidity 29,42 . Mean single cell values for each parameter were used for statistics.
Experimental design and statistics. Mice subjected to surgery, performed by the same investigator in order to reduce variability, were randomly allocated across cages and days. Different blinded investigators evaluated mice with behavioural, histological, immunohistological and biochemical tests. Group size is of 7 defined by the formula: n = 2σ 2f. (α, β)/Δ 2 (santard deviation, SD, in groups = σ, type 1 error α = 0.05, type II error β = 0.2, percentage difference between groups Δ = 30). Standard deviation to be used in the formula was calculated based on a pilot experiment to assess PTX3 staining in cortex 5w after TBI, resulting in σ = 21, thus yielding n = 7.742. Groups were compared by analysis of variance and post hoc testing as indicated in each figure legend. A parametric or nonparametric test was selected after the Kolmogorov-Smirnov test for normality to assess whether the data for the groups were normally distributed. The constancy of the variances was checked by the Bartlett test and, if not satisfied, a Welch correction applied to the test. Statistical analysis was performed with the standard software package GraphPad Prism (GraphPad Software Inc., San Diego, CA, USA, version 7.0); p values lower than 0.05 were considered significant.

Results
This study was carried out according to the plans in Fig. 1. First, PTX3 was analyzed at different time points after TBI or sham surgery for its circulating levels or brain gene expression and protein presence (Fig. 1A). Next, in order to define PTX3 role in TBI pathophysiology we investigated behavioral deficits and brain damage progression over 5 weeks after TBI comparing WT to PTX3 KO mice (Fig. 1B).
PTX3 plasma levels and brain presence over 5 weeks after TBI. At Fig. 3A), implying an effect of the surgical procedure-not selectively associated with TBI induction. Analyzing further time points from surgery, PTX3 levels were significantly higher than naïve (7.51 ± 1.37) from week 1 to week 3 (13.63 ± 1.25, 17.63 ± 2.51 and 14.5 ± 2.18, respectively). Brain PTX3 levels were investigated in cortex after TBI by immunofluorescence and quantified under a ROI placed in the ipsilateral cortex within the first 350 μm from the edge of the contusion (left panel, Fig. 3B). PTX3 increased starting from 48 h (52.02 × 106 ± 79.12 × 105, ID ± SEM), reached its maximum at 2w (10.49 × 107 ± 14.41 × 106) and up to 5w after TBI (97.63 × 106 ± 12.89 × 106) compared to naive mice (9.05 × 106 ± 24.87 × 105, Fig. 3C). In contrast with the circulating levels, brain PTX3 was not affected by the surgical procedure, i.e. at 24 h after the sham procedure PTX3 was not seen in the brain (Fig. 3D) www.nature.com/scientificreports/ specifically caused by TBI. It appeared in cellular structures (arrows in Fig. 3E) up to 1 week after TBI, while it was located mainly extracellularly at longer time points (from 2 to 5w after TBI, Fig. 3E).

Confocal analysis of PTX3 presence in the contused cortex at different time points after TBI.
Starting from the notion that PTX3 has been reported in neutrophil granules 46 , we assessed the possible co-localization between neutrophil marker elastase and PTX3 by immunofluorescence. PTX3 was localized in  www.nature.com/scientificreports/ proximity of neutrophils 48 h after TBI (Fig. 4A). By SIM, we observed PTX3 with branches surrounding a neutrophil, likely indicating its presence on extracellular NETs (Fig. 4B). Immunofluorescent controls showed no signal in PTX3 KO mice and in WT stained without the primary antibody, confirming the staining specificity ( Supplementary Fig. 2). We next evaluated PTX3 localization in relation to different cell populations present in the injured tissue including astrocytes, neurons, microglial and endothelial cells, 1 week after TBI, a time point at which PTX3 signal appears next to cell-like structures (Fig. 3C). PTX3 signal was present next to but did not co-localize with any of these cell populations in the contused cortex (Fig. 5).
We then addressed extracellular presence of PTX3 at 1w and 2w after TBI and investigated possible connections with fibrin deposits. At both time points we found a strong co-localization between PTX3 and fibrin(ogen) ( Fig. 6).
A summary of the results of PTX3 expression and presence after brain injury is reported in Table 1.

Effects of PTX3 genetic depletion on TBI long-term outcomes. Sensorimotor function was
assessed over 4 weeks after sham injury or TBI using Neuroscore (Fig. 7A) and Beam Walk (Fig. 7B) tests in WT and PTX3 KO mice. No difference was observed in post-traumatic sensorimotor deficits between the two strains ( Fig. 7C-D). We then assessed lesion volume and neuronal density at 5 weeks after TBI by cresyl violet staining. We observed an extensive macroscopic area of cortical tissue loss, extending rostrocaudally from bregma + 0.4 to -3.6 mm, both in WT and PTX3 KO animals, without differences in the lesion volume between the two genotypes (Fig. 7E). Also, the neuronal density in a cortical region, traced at a distance of 350 μm from the contusion edge, did not differ between WT and PTX3 KO mice (Fig. 7F).

Effects of PTX3 depletion on short and long-term brain inflammatory processes.
We also analyzed brains obtained from WT and PTX3 KO TBI mice sacrificed at 1 week after TBI, in order to evaluate acute effects of PTX3 depletion. Neither the lesion volume (12.82 ± 1.055 vs 14.44 ± 1.164, mm 3 ± SEM) nor the neuronal density (3118 ± 117.9 vs 3163 ± 247.4, number of neurons/mm 2 ± SEM) showed differences between the two strains ( Supplementary Fig. 3A,B). Since microglial morphology may be associated with their function, we next analyzed few shape descriptors 42 , observing no differences between WT and PTX3 KO mice (Supplementary Fig. 3C). In the chronic phase (5 weeks after TBI) we measured astrogliosis and microglial activation, quantifying GFAP, CD11b and CD68 immunopositive area at the edge of the contusion area. No differences were present between WT and PTX3 KO mice for GFAP (Fig. 8A), CD11b (Fig. 8B) and CD68 (Fig. 8C) stainings. We also assessed the axonal loss by luxol fast blue staining. The stained hemisphere did not show any difference between WT and PTX3 KO mice in either the corpus callosum or the external capsule (Fig. 8D). In different models of tissue damage, PTX3 deficiency has been associated with enhanced collagen deposition 47 . Therefore, we measured collagen deposition within the first 350 μm from the edge of the contusion by sirius red staining.

Discussion
The present study originally shows that, following TBI: (1) PTX3 circulating levels are increased in the subacute phase, starting 1 week and up to 3 weeks post-injury; (2) in the brain, PTX3 gene expression is increased in response to TBI, as early as 24 h and up to 5 weeks after the primary insult; (3) PTX3 protein is present in the injured cortex, specifically in association with NETs in the acute phase, and bound to fibrin(ogen) deposits in the sub-acute phase; (4) the brain gene expression of long pentraxin nPTX1 and PTX4 is also increased www.nature.com/scientificreports/ after TBI. Despite the consistent presence of PTX3 at different time points after TBI, its genetic depletion did not affect TBI outcome, as assessed up to 5 weeks post-injury, in terms of sensorimotor deficits, lesion size or inflammatory markers. Although weakly expressed in basal conditions in the central nervous system (CNS), ptx3 gene transcription may be induced in different brain cells-among which glial cells, mononuclear phagocytes and endothelial cells-in response to a wide range of pro-inflammatory stimuli, including IL-1β, TNF-α, toll-like receptor (TLR) Figure 6. PTX3 co-localization with fibrin(ogen) in the contused cortex 1 and 2 weeks after TBI. PTX3 (red) co-localized with fibrin(ogen) (green) at 1w (left) and 2w (right) after TBI. Nuclei are in blue. Images are representative of at least two independent experiments. Scale bar = 20 µm. Table 1. Summary table of brain PTX3 mRNA expression, protein presence and localization over time after TBI. The long-lasting pattern of activation of PTX3 in brain and blood following TBI supports the hypothesis that PTX3 contributes to the progression of the lesion with effects varying over time. Increases versus respective controls are calculated according to quartiles and correspond to: = control group; + low increase; + + intermediate increase; + + + high increase.

Protein presence
Plasma Control group + + + (24 h) = (5w) + + + + + + + + + + + + + -= Brain cortex Control group = (24 h) = (5w) = = + + + + + + + + + + + + + + + www.nature.com/scientificreports/ ligands in addition to microbial components 43,48 . We actually found a strong activation of PTX3 mRNA expression compared to control mice in brain areas adjacent to TBI lesion core, namely cortex and striatum, occurring at early time points (24 h) and lasting up to 5w after insult. Additional brain areas next to the lesion core, such as hippocampus and thalamus, also showed PTX3 overexpression after TBI, although to a lower extent. These observations indicate that PTX3 participates in brain acute phase responses elicited by TBI in both lesion core and surrounding areas. PTX3 mRNA is still present in sub-acute and chronic responses, but exclusively in areas adjacent to the lesion core, supporting the hypothesis that, at this stage, PTX3 may have a role in the formation of the gliotic scar and/or in healing processes, as previously shown in other models or tissues 7,12 .

Protein localization
Other members of the pentraxin family, namely the long pentraxins nPTX1 and PTX4, but not nPTX2 (nor the short pentraxins CRP and SAP), were also upregulated, although to a lower extent compared to PTX3, in the cortex at the early phases after TBI, suggesting the involvement of nPTX1 and PTX4 in the acute brain responses to TBI. nPTX1 was reported to have a role in neuronal death after ischemia in vitro 49   www.nature.com/scientificreports/ its involvement in the traumatic pericore tissue, an area subjected to post-injury hypoxia 22 . At variance, scanty information is available on PTX4 45 which however shows a clearcut upregulation after TBI. Available data show that circulating PTX3 levels increase rapidly in response to infections and cardiovascular diseases 3,6 . Elevated PTX3 plasma levels are recognized as independent predictors of mortality at three months after acute myocardial infarction 51,52 . Circulating PTX3 is also associated with plaque vulnerability/rupture [53][54][55] , incidence of heart failure, cardiac arrest 6,56-58 and hypoxic respiratory failure 59 , suggesting a role as biomarker of cardiovascular risk 48 . In the CNS, PTX3 role as biomarker was proposed also in acute brain injury. Specifically, PTX3 was been identified as a novel and independent prognostic marker in ischemic stroke in both human 60 and mice 12 . Finally, PTX3 protein levels increased in human cerebrospinal fluid (CSF) early (48 h) after subarachnoid hemorrhage (SAH), with a second peak of expression in the following 48-96 h associated with increased occurrence of vasospasm 61 . In TBI patients, serum PTX3 levels, measured within 72 h of hospital admission, were reported to increase significantly after severe injury and to be independently associated with hospital mortality 23 . In our TBI model we show a persistent increase in PTX3 circulating levels. We observed a sharp increase in plasma PTX3 at 24 h post-injury which however appeared to be non-specific and possibly due to PTX3 release following skin injury, as previously demonstrated at the same time-point by Doni et al. 7 . At later time points we observed a clear-cut and long-lasting increase in plasma PTX3 compared to non-TBI controls, starting from 1w after TBI. At least part of PTX3 systemic increase can be due to its release by brain cells into cerebral blood flow through the injured blood brain barrier (BBB). Leukocytes may be an additional source of PTX3, thus potentially contributing to PTX3 circulating levels by protein release in blood 62 .
TBI induces neutrophil recruitment to the site of damage, making them an important source of PTX3 in the traumatic brain 63 . In particular, neutrophils act as a reservoir of ready-made PTX3, stored as glycosylated form in specific granules 48 , and rapidly released to sites where tissue damage (or microbial stimulation) occurs 64 . PTX3 is complexed with components of NETs, a structure involved in inflammation, phagocytosis and coagulation 46,48,65,66 . PTX3 function in NETs is known in non-sterile inflammation, where it brings the NET component proteins into close proximity with the pathogens to enhance pathogen clearance 65 . We report for the first time in TBI that PTX3 is present in the mouse brain parenchyma in acute phases after injury and is associated with NETs. The exact functions associated with NET-bound PTX3 in sterile inflammation, like that associated with TBI, still need clarification. PTX3 was reported to reduce neuroinflammation at early time points after intrastriatal lipopolysaccharide (LPS) administration or after stroke by dampening neutrophil recruitment into the brain 14 . Other potential roles of PTX3 in the context of acute brain injury comprise NET-mediated thrombosis [67][68][69] . We show here that PTX3 was present in cortex next to astrocytes, neurons, microglia and endothelial cells in the sub-acute phase, suggesting a possible release from these cells with a paracrine function. The expression of PTX3 by each of these brain cell populations has been associated with a specific function in response to brain damage. Namely, PTX3 in astrocytes has been related to BBB integrity in the acute phase of stroke 13,43 ; in neurons it conferred resistance to neuronal damage at sub-acute phases after seizures 10 and mediated neurogenesis and angiogenesis after stroke 11,70 . At variance with the protective functions described above, in cultured endothelial cells PTX3 was released in response to inflammatory stimuli 71-74 acting as critical determinant of the endothelial dysfunction 48 . In addition, it promoted glial proliferation and glial scar formation after stroke, thus possibly limiting the recovery of synaptic plasticity after acute brain injury 12 .
We found that PTX3 co-localized with fibrin deposits starting from the sub-acute phase up to the chronic phase after TBI. In fact, PTX3 may bind fibrinogen and/or fibrin and plasminogen and increase plasmin-mediated pericellular fibrinolysis 7 , an action related to its role in extracellular matrix remodeling. In fact PTX3 can target the extracellular matrix component fibrinogen, via its N-terminal domain 75 , and promote wound healing by favoring fibrinolysis as shown in different models of tissue damage 7,47,76,77 , a function dependent on the acidic microenvironment generated by tissue injury 7 . Thus, PTX3 may lead to fibrinolysis and gliotic scar resolution with subsequent improvement in the injury outcome. Fibrinogen extravasation has been observed in postmortem human brain months and even years after a single moderate-severe TBI 78 . Parenchymal deposition of fibrinogen is significantly increased following acute TBI and associated with a proinflammatory, pro-phagocytic microglial/macrophage phenotype, suggesting a role in augmenting and sustaining an inflammatory state in human TBI that impacts negatively on neuronal density and, potentially, axonal survival 79 .
The analysis of PTX3 mRNA and protein expression reveals that mRNA increases 24 h after TBI likely indicating protein production later on (48 and 96 h). The subsequent gene expression increase (96 h) is likely to result in enhanced protein production at sub-acute times (1w and 2w), when it reaches its maximum. At these time points PTX3 appears next to brain cells and also in relation to extracellular matrix fibrin(ogen). The protein parenchymal localization in the extracellular matrix follows the lower but still present PTX3 gene expression up to chronic phases. The acid environment present in the lesioned brain parenchyma may support the persistent presence of PTX3, and its involvement in tissue remodelling during sub-acute and chronic phases 7 .
An obvious approach was to assess TBI outcome in PTX3 depleted mice, according to the protocol for evaluating the injury evolution in our experimental model 18,33,40 . PTX3 depleted mice lacked a clear-cut phenotype compared to WT up to 5w after TBI. We can envisage a few possible scenarios to account for this lack of phenotype. First, PTX3 may subserve multiple functions throughout different phases of injury evolution, i.e. acute inflammatory functions vs. chronic reparative actions, finally resulting in a balanced phenotype. To explore possible time-dependent roles of PTX3 in TBI physiopathology, we assessed lesion size and neuronal viability at sub-acute phase (1w) after TBI in WT and PTX3 KO, finding no differences. This result does not directly support a prevalent inflammatory role of PTX3 in the early stages post-injury, however the multiple functions of PTX3 and its different interactions with injury-related molecules in TBI brains cannot exclude that the lack of phenotype of PTX3 KO mice may be due to a balance between damaging and protective effects. As an alternative hypothesis, PTX3 may have a bystander role in TBI, thus not actively participating to injury progression, in line www.nature.com/scientificreports/ with what hypothesized for a chronic pulmonary pathology 80,81 . However, since TBI elicits acute inflammation, the bystander role seems less likely compared to other hypothesis. For instance, a possible scenario would be that brain long pentraxins could compensate for the absence of PTX3 in KO mice. Actually they share a degree of identity with PTX3, namely 21.2% with nPTX1 and 19.7% with PTX4 (sequence alignment done with Uniprot, www. unipr ot. org). It may also be possible that these long pentraxins counteract PTX3 action in TBI progression, especially interfering with PTX3-associated synaptogenesis, a process that involves the remodeling of the extracellular matrix by several mediators, i.e. tumor necrosis factor-induced protein-6 (TSG6), astrocytes-secreted thrombospondins (TSPs) and β1 integrin 82,83 .Unfortunately, the lack of information available on the function of these pentraxins in the brain does not presently allow to substantiate any of these hypothesis.

Conclusion
The long-lasting pattern of activation of PTX3 in brain and blood supports a specific involvement in TBI. The lack of a clear-cut phenotype in PTX3 KO mice may depend on the different roles of this protein, possibly involved in inflammation early after injury and in repair processes later on, and/or on the presence of brain pentraxins compensating for its absence. Available data obtained in different models of injury and inflammation, show that recombinant PTX3 administered in sub-acute phases reduces inflammation and boosts tissue reparative processes 7 . As such, our work may offer reasons to administer PTX3 to alleviate sub-acute pathological sequelae after TBI.

Data availability
The data sets generated and/or analysed during the current study are available in the Figshare repository, 10.6084/ m9.figshare.13019081.