Spinal cord injury disrupts plasma extracellular vesicles cargoes leading to neuroinflammation in the brain and neurological dysfunction in aged male mice

Abstract

Aged individuals with spinal cord injury (SCI) are prevalent with increased mortality and worse outcomes. SCI can cause secondary brain neuroinflammation and neurodegeneration. However, the mechanisms contributing to SCI-induced brain dysfunction are poorly understood. Cell-to-cell signaling through extracellular vesicles (EVs) has emerged as a critical mediator of neuroinflammation, including at a distance through circulation. We have previously shown that SCI in young adult (YA) male mice leads to robust changes in plasma EV count and microRNAs (miRs) content. Here, our goal was to investigate the impact of old age on EVs and brain after SCI. At 24 h post-injury, there was no difference in particle count or size distribution between YA and aged mice. However, aged animals increased expression of EV marker CD63 with SCI. Using the Fireplex^® miRs assay, Proteomics, and mass spectrometry-based Lipidomics, circulating EVs analysis identified distinct profiles of miRs, proteins, and lipid components in old and injury animals. In vitro, plasma EVs from aged SCI mice, at a lower concentration comparable to those of YA SCI mice, induced the secretion of pro-inflammatory cytokines and neuronal apoptosis. Systemic administration of plasma EVs from SCI animals was sufficient to impair general physical function and neurological function in intact animals, which is associated with pro-inflammatory changes in the brain. Furthermore, plasma EVs from young animals had rejuvenating effects on naïve aged mice. Collectively, these studies identify the critical changes in circulating EVs cargoes after SCI and in aged animals and support a potential EV-mediated mechanism for SCI-induced brain changes.

1. Introduction

Given increased falls in the elderly, an incidence rate of traumatic spinal cord injury (SCI) has been significantly increasing in that population (Ikpeze and Mesfin, 2017). Older age with SCI not only aggravates disability and mortality but also negatively impacts outcomes (Furlan and Fehlings, 2009; Hsieh et al., 2013). Human and rodent studies have shown that SCI can cause secondary brain neuroinflammation and neurodegeneration leading to elevated dementia risk (Li et al., 2020c; Mahmoudi et al., 2021). Preclinical studies support that neuropathology and neurological behavioral deficits are exacerbated with age and worse in males (Fenn et al., 2014; Gaudet et al., 2021; Hooshmand et al., 2014; Siegenthaler et al., 2008; Takano et al., 2017; Zhang et al., 2015). Thus, understanding the pathophysiology of SCI in geriatric patients is essential to eventually improve recovery of function and life expectancy.

The focus of most SCI research relates to mechanisms of primary injury-induced tissue loss and associated sensorimotor and autonomic deficits. While the systemic complications are the most often recognized after SCI, less well acknowledged are the cognitive and affective symptoms that nearly 50% of patients report (Li et al., 2020c). SCI can lead to decreases in attention span, concentration, learning and memory (Sachdeva et al., 2018). However, our understanding of SCI-induced brain dysfunction remains largely observational, and few studies have examined this phenomenon with a mechanistic approach to date.

Extracellular vesicles (EVs) are now recognized as a critical component of intercellular communication (Welsh et al., 2024). Originally thought to be just a mechanism for cellular waste removal, EV functions can be attributed to a growing number of normal physiologic and pathophysiologic processes (Colombo et al., 2014; Valadi et al., 2007). EVs are nanometer-sized structures enclosed by a lipid bilayer that transmit proteins, lipids, and nucleic acids, such as microRNAs (miRs), from their originating cells to alter the function of recipient cells. Since EVs can be found in circulating biological fluids like blood plasma, they may contribute to long-distance signaling mechanisms between organ systems, including the spread of inflammatory molecules. However, our current knowledge of EV signaling is largely derived from in vitro studies (Clos-Sansalvador et al., 2022; Gardiner et al., 2016), and corresponding in vivo evidence for proposed functions is limited. We recently characterized plasma EVs dynamics at acute, subacute, and chronic time-points after SCI in young adult mice and showed that EVs derived from SCI mice can cause brain inflammation (Khan et al., 2021; Li et al., 2023b). However, changes in plasma-derived EVs in old age following SCI are poorly understood.

Here, we detected plasma EVs cargoes in both aged and young adult male mice following SCI. The underlying mechanisms linking SCI and brain dysfunction were examined. We report for the first time that old age and SCI induce distinct profiles of plasma EVs cargo contents including miRs, proteins, and lipid components. In vitro, plasma EVs derived from SCI mice induces glial activation and neuronal apoptosis. Importantly, we demonstrate that systemic administration of SCI-EVs into intact mice is sufficient to impair overall well-being and neurological function associated with neuroinflammation in the brain. Together, we present novel data that old age alters circulating plasma EVs pro-inflammatory response to SCI, potentially contributing to brain dysfunction after SCI.

2. Materials and Methods

2.1. Mouse spinal cord contusion and plasma EVs delivery

C57BL/6 aged (18-month-old) and young adult (3-month-old) male mice were obtained from NIA through Charles River Laboratories and housed on a 12 hours (h) light/dark cycle with food and water ad libitum. Spinal cord injury was conducted at moderate injury level (60 kDyn, T10) using the Infinite Horizon spinal cord impactor (Precision Systems and Instrumentation) as previously described (Li et al., 2021; Wu et al., 2016). Sham mice underwent the same procedure except for contusion. All surgical procedures were approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Maryland School of Medicine.

Intravenous injection of plasma EVs:

The plasma EVs were prepared at 1d post-injury from young adult and aged male C57BL/6 mice subjected to SCI (Khan et al., 2021). The Young SCI EVs (YSE) and Aged SCI EVs (ASE) pooled from 10 young or aged mice, separately, were measured for protein concentration using the Pierce BCA Assay Kit (Cat# 23225, Thermo Fisher). A separate set of 20 naïve aged male mice were randomly divided into two groups, and intravenously injected with 100 μl/injection (20 μg protein in PBS) of YSE or ASE once every other day for a total of three injections. The same procedure was applied to a different set of 20 naïve young adult male mice. The same volume of PBS was administrated into naïve young adult mice as control group. A battery of behavioral tests was conducted five days before the first injection and ten days after the last injection. To minimize stress and fatigue, each behavioral test was performed on a different day.

2.2. EVs isolation and Western blot

Blood was collected into precoated EDTA tubes (Cat# 365974, BD Biosciences) through terminal cardiac puncture from each mouse under anesthesia. The blood samples were immediately centrifuged at 500g for 15 min, 2500g for 10 min, and 2500g for 10 min to generate platelet free plasma (PFP), which was aliquoted into multiple tubes and stored at −80°C for later analysis or other applications. Total plasma EVs were isolated by centrifuging at 110,000g for 120 min at 4°C with supernatant removed as previously described (Khan et al., 2021). To increase EV purity, we washed the pellet at 110,000g for 120 min after an initial 110,000g spin. For tissue EVs isolation, about 10 mm mouse spinal cord around the lesion area was dissected after perfusing with ice-cold saline. The tissue samples were digested by collagenase (Cat# LS004176, 40U/ml, Worthington Biochemical), then mixed with protease inhibitor (Cat# 11697498001, Millipore Sigma) and phosphatase inhibitor (Cat# 4906837001, Millipore Sigma), next centrifuged at 300g for 5 minutes, 2000g for 10 minutes, 10,000g for 30 minutes, and 100,000g for 70 minutes at 4 °C with the supernatant from each step. The final supernatant was removed carefully, and the remaining tissue EVs pellet was resuspended in 50 μl of PBS. Detailed protocols including Nanoparticle Tracking Analysis (NTA) can be referred to our previous study (Khan et al., 2021; Li et al., 2023b).

According to the guidelines from the International Society for Extracellular Vesicles (Welsh et al., 2024), no recommendations were made for selection of a housekeeping protein that is stably expressed in all EVs. Thus, loading an equal volume of isolated EVs is a common practice for characterization of EVs markers with Western blot (Khan et al., 2021; Li et al., 2022). In the present study, equal volumes of resuspended plasma EVs sample pellets were loaded onto 4–15% Criterion^™ TGX Stain-Free^™ Precast gels (Cat# 5678083, Bio-Rad) and transferred onto nitrocellulose membranes (Cat# 1704159, Bio-Rad). Membranes went through blocking with 5% nonfat milk in PBS plus 0.1% Tween 20 (PBS-T) for 1 hour at room temperature (RT), incubation with primary antibodies overnight at 4°C, and species-specific, horseradish peroxidase (HRP)-conjugated secondary antibodies for 1 hour at RT, washing three times with PBS-T prior to the addition of chemiluminescence substrate (Cat# 37071, Thermo Fisher Scientific). The Western Blot image was visualized in the ChemiDoc^™ MP Imaging System (Bio-Rad), and protein bands were analyzed by Image Lab^™ software Version 6.0.1 (Bio-Rad). The following antibodies were used: anti-CD63 (Cat# D263–3, 1:500, MBL International) and HRP-conjugated goat anti-Rabbit (Cat# 111-035-003, 1:3000; Jackson ImmunoResearch Laboratories).

2.3. EVs microRNA assay and analysis

The total plasma and tissue EVs of all experimental groups were used for miR assay including 65 microRNAs through FirePlex^® technology (Abcam) which detected microRNAs and measured their abundance by fluorescence intensity. The data was analyzed and visualized as previously described (Khan et al., 2021).

2.4. Proteomics and data analysis

EVs from various cohorts were subjected to lysis using RIPA lysis buffer. Subsequently, lysed samples, each with an equal concentration of 0.5 μg/μl to a total volume of 50 μl, were introduced in a randomized fashion into a 96-well PCR plate. This plate was then dispatched to Olink Proteomics (Boston, USA) for comprehensive analysis. The cargo of these EVs was scrutinized utilizing the Olink Mouse exploratory panels. The Olink platform, based on Proximity Extension Assay (PEA), employs a multiplex DNA-coupled immunoassay-based targeted proteomic approach. This method entails the detection of each target protein through a pair of distinct oligonucleotide-labeled antibodies. Upon binding to the protein, these oligonucleotide probes undergo polymerization, forming a target sequence that is subsequently quantified via RT-PCR. To ensure robustness and reliability, samples were randomly arranged on the PCR plate, and quality control was upheld through the incorporation of four internal controls. Samples failing to meet the technical criteria of the assay, as per the company’s guidance, were systematically excluded from the analysis.

2.5. Lipidomics and data analysis

Lipid Extraction:

Total lipid extracts from EVs were prepared using a modified HPLC grade tert-Butyl methyl ether (MTBE, Sigma Aldrich) lipid extraction protocol (Matyash et al., 2008). Briefly, 400 μl of cold methanol and 10 μl of internal standard mixture (EquiSPLASH lipidomix, Avanti Polar Lipids, Inc.) were added to each sample. Each sample contained 100 μg of protein. The sample was incubated at 4°C, 650 RPM shaking for 15 minutes. Next, 500 μl of cold MTBE was added followed by incubation at 4°C for 1 hour with 650 RPM shaking. 500 μl of cold water was added slowly and resulting extract was maintained 4°C, 650 rpm shaking for 15 minutes. Phase separation was completed by centrifugation at 8,000 RPM for 8 min at 4 °C. The upper, organic phase was removed and set aside on ice. The bottom, aqueous phase was re-extracted with 200 μl of MTBE followed by 15 minutes of incubation at 4°C with 650 RPM shaking. Phase separation was completed by centrifugation at 8,000 RPM for 8 min at 4 °C. The upper, organic phase was removed and combined with previous organic extract. The organic extract was dried under a steady stream of nitrogen at 30 °C. The recovered lipids were reconstituted in 100 μl of acetonitrile:isopropanol:water (1:2:1, v/v/v, Fisher Scientific).

Lipid Analysis:

Total lipid extracts were analyzed by liquid chromatography coupled to targeted tandem mass spectrometry (LC-MS/MS). The LC-MS/MS analyses were performed on an Ultimate 3000 Ultra High-Performance Liquid Chromatograph (UHPLC) coupled to a Thermo TSQ Altis Tandem Quadrupole Mass Spectrometer (Thermo Scientific). LC-MS/MS methodology was adopted from literature (Medina et al., 2023). The separation was achieved using an ACQUITY Amide BEH column (1.7 μm; 2.1 × 100 mm) column (Waters) maintained at 45 °C. Mobile phase compositions for solvents A and B consisted of ACN/H₂O (95:5, v/v) and (50:50, v/v) respectively, with 10 mM ammonium acetate The gradient profile had a flow rate of 0.6 ml min⁻¹ and ramped from 0.1 to 20% B in 2 min, from 20 to 80% B in 3 min, dropped from 80 to 0.1% B in 0.1 min, and held 0.1% B for 2.9 min. Total chromatographic run time was 8.0 min. The injection volume was 2 μl. The auto-sampler was kept at 8 °C. Gas-phase ionization was achieved using either negative or positive electrospray ionization (ESI). Mass spectrometry detection was done using selective reaction monitoring (SRM) where predetermined precursor to product ion transitions were used. ESI source parameters were set as follows: voltage 3500 V in positive mode and −2500 V in negative mode, sheath gas (Arb) = 60, aux gas (Arb) = 15, sweep gas (Arb) = 1 and ion transfer tube temperature of 380 °C. Nitrogen was used as the nebulizer and argon as collision gas (1.5 mTor). The vaporizer temperature was set to 350 °C. Collision energies and RF lens voltage were optimized for each lipid class reference standard and supported by literature (Medina et al., 2023). LC-MS/MS data was acquired using Thermo’s Xcalibur software and data processing was achieved using Xcalibur and TraceFinder. Addditional data analysis and processing was done using Prism 6, MetaboAnalyst (Xia, 2011), and LION lipid ontology (Molenaar, 2019).

2.6. Primary mouse glial cell culture, EVs stimulation, and cytokines ELISA assay

Neonatal C57BL/6 mouse cerebral cortex was used to culture microglia and astrocytes as previously described (Li et al., 2021; Sabirzhanov et al., 2019). Primary mixed glial cells were grown in Dulbecco’s Modified Eagle’s Medium/F12 supplemented with 10% fetal bovine serum (FBS) and 1% Pen/Strep (complete media) under 5% CO2 at 37°C. After 10–14 days of incubation, microglia or astrocytes were plated in 96-well plates and incubated with the complete media for 24 h or 2 d, correspondingly. The media was then replaced by adding 2% FBS containing media for serum starvation, which lasted for 1 h for the microglia or 24 h for the astrocytes, followed by 24 h incubation with 2 – 6 × 10⁹ EVs/ml from Young SCI or Sham groups, or 4 × 10⁹ EVs/ml from Aged SCI or Sham groups, or PBS. The resulting supernatants were subjected to CXCL2/MIP-2 (Cat# MM200, R&D Systems) or IL-6 ELISA (Cat# M6000B, R&D Systems) according to the manufacturer’s instructions.

2.7. Primary mouse neuronal culture and Western Blot

Mouse cerebral cortex was dissected from 16–19 days embryo as described previously (Li et al., 2021; Sabirzhanov et al., 2019). Dissociated cells were seeded in 12-well plates pre-coated with poly-D-lysine (50 μg/ml, 70–150 kDa, Cat# P6407, Sigma-Aldrich) and cultured with Neurobasal^™ media (Cat# 21103049, Gibco) supplemented with 2% B-27 (Cat# 17504044, Gibco), 0.5mM Glutamax (Cat# 35050061, Gibco), and 2% penicillin-steptomycin (Cat# 15140122, Gibco) for 7 days. After 24 h incubation with 6 × 10⁹ EVs/ml from Young SCI or Sham groups, 4 × 10⁹ EVs/ml from Aged SCI or Sham groups, the neurons were harvested and processed with RIPA lysis buffer (Cat# R0278, Sigma-Aldrich) supplemented with 1× Protease Inhibitor Cocktail, Phosphatase Inhibitor Cocktail II & III (Cat# P8340, P5726, P0044, Sigma-Aldrich), Follwoing protein determination with the Pierce BCA Assay Kit (Cat# 23225, Thermo Fisher), Western Blotting was performed as described above. The following primary antibodies were used: cleaved caspase 3 (Cat# 9661S, 1:500, Cell Signaling Technology), actin/β-actin (Cat# A1978, 1:10,000, Sigma-Aldrich).

2.8. Assessment of mouse general physical behaviors and neurological function

Marble burying test:

In the testing room with dim ambient light, an empty clean cage was filled with regular bedding of approximately 5 cm depth to make a flat surface. Twenty black glass marbles of 14 mm diameter were placed evenly on the surface. One mouse was gently placed in the corner and left in the cage with the top closed for a 30-minute period of free exploration. In the end, the animal was returned to its home cage and the number of marbles left unburied (more than 1/3 their depth) with bedding was recorded (Angoa-Perez et al., 2013). The cage and the marbles were cleaned thoroughly with 70% ethanol for the next run.

Grip strength test:

The grip strength meter (Cat# 47200, Ugo Basile) was set up according to the instructions in the manual and switched to “Peak Force” mode. A mouse was taken out of its home cage with the base of its tail gripped by the thumb and forefinger and lowered over the grid until its forepaws grabbed the grid firmly. The mouse was gently pulled back by its tail horizontally at a constant velocity until the animal let go of the device. The maximal grip strength value was recorded automatically. This procedure was repeated 2–3 times to obtain 3 measurements of forelimb grip strength. The recording is voided if the mouse does not hold the grid actively (Takeshita et al., 2017). For the grip strength of all limbs, both forelimbs and hindlimbs were allowed to grab the grid during the measurements. After each run, the mouse was weighed and returned to its home cage. The device was cleaned with 70% ethanol.

Nesting test:

Mouse was singly housed in a new cage one hour before the Dark Cycle. The nestlet (5cm × 5cm × 5mm, ~2.5g) was weighted and placed in the same corner of the cage. After overnight, each nestlet was graded on a rating scale of 1–5 as previously defined (Deacon, 2006).

Y-maze test:

The Y-maze (Stoelting, Wood Dale, IL) consists of three identical arms at an angle of 120° from each other (Ritzel et al., 2022). The distal 30 cm of each arm was labelled as A, B, C, whereas the remaining area was defined as the center. At the start of the test, the mouse was placed at the end of a random arm, facing the center and allowed to explore the maze freely for 6 minutes. An arm entry was defined as all four paws of the mouse entering the arm, while an alternation was designated when the mouse entered the three arms consecutively. The percentage of alternation was calculated with the following equation: total alternations ×100 / (total arm entries - 2). An ‘arm return’ was defined as the mouse returning to a previously entered arm after entry to a different one, with the same equation being used for percentage.

Novel object recognition (NOR) test:

An open field (40 × 40 cm²) arena with plexiglass black walls was placed in a dark room with red lights for illumination (Ritzel et al., 2022). Mice were placed in the center and allowed to move freely in a 5-minute habituation period on the first day. Next day, two identical objects were placed on the diagonal line symmetrical to the center, which were placed 12 cm from the walls. On the third day, one of the familiar objects was replaced with a novel object, with the position being switched for each testing mouse. The time mice spent with either object was recorded with ANY-maze tracking software (Stoelting) until a total exploration time of 30 s was reached.

2.9. NanoString transcriptomic analysis

Following intracardiac perfusion with ice cold normal saline, the mouse brain was harvested. The somatosensory cortex and hippocampus from both hemispheres were dissected and flash-frozen on dry ice. Total RNA was extracted with the RNeasy mini kit (Cat# 74104, Qiagen) and sent to the Institute for Genome Sciences at University of Maryland School of Medicine for NanoString analysis with Mouse Neuroinflammation panel. The raw data of mRNA read counts was analyzed with “Advanced Analysis” in NanoString nSolver software (version 4.0), including transcript normalization, group comparisons to obtain differentially expressed genes (DEGs) and pathway score assessment. Samples that clustered outside the boundary of score distance in Orthogonal Partial Least Squares (OPLS) (R, ropls package, 1.22.0) were taken as outliers and excluded from further analysis. Heatmaps were generated based on the ComplexHeatmap package (version 2.13.2) (Gu et al., 2016). Partial least squares discriminant analysis (PLS-DA) graphs were produced with the mixOmics package (http://mixomics.org/), while Volcano graphs were created via EnhancedVolcano (version 1.11.3) (Blighe, 2018) and Venn diagrams were plotted by VennDiagram package (version 1.73).

2.10. Statistical analysis

All data are presented as mean ± SEM from the indicated number of independent experiments. All behavioral and ex vivo studies were performed by investigators blinded to group designations. Statistical analysis was performed using GraphPad Prism 8.4.2 (GraphPad Software, LLC) for most bar graphs, or R for transcriptomic data. Normal distribution of data was assessed with the Shapiro-Wilk test. For multiple comparisons, one-way or two-way ANOVA was performed followed by Tukey’s or Dunnett’s multiple comparisons post-hoc test for parametric (normality and equal variance passed) data. Nonparametric data was analyzed by Mann-Whitney test. For analysis of miR assay data, a two-way ANOVA test was used and was followed by Tukey’s multiple comparisons test to compare the main effects of age or injury. Significance was set at p ≤ 0.05 and detailed in figure legends.

3. Results

3.1. Age alters the plasma extracellular vesicles (EVs) response after SCI

To examine changes in plasma EVs, we subjected aged (18-month-old) and young adult (3-month-old) male C57BL/6 mice to a moderate contusion. At 24 h post-injury, total plasma EVs were isolated by ultracentrifugation (Khan et al., 2021). Nanoparticle Tracking Analysis (NTA) was used to measure plasma EVs particle count and size as described previously (Khan et al., 2021). No significant differential changes in particle concentration, particle size, and coefficient of variation (COV) were detected between aged and young adult animals (Fig. 1A–D). Injury did not significantly alter these parameters in either group except for reduced particle concentration in SCI mice compared to sham ones. We then analyzed the plasma EV isolates by western blot for expression of EVs marker CD63. We observed increased CD63 expression levels in aged mice compared to young adult mice (Fig. 1E). Statistical analysis showed two-way ANOVA main effect of age, but not related to injury (Fig. 1F). These data suggest that despite modest differences in particle count and size, old age alters tetraspanin protein CD63 expression level in the plasma EVs.

Figure 1. Old age alters the plasma EV response after SCI.

(A-D) EV particle parameters were analyzed by Nanoparticle Tracking Analysis (NTA) of one-day post-injury in aged (18-month-old mice) and young adult (3-month-old) male animals. (E-F) Western Blot images (unedited gels) and analysis showed significantly upregulated CD63 expression in aged mice compared to young groups. n=6 mice/group (2 sets of samples, 3 mice/set). **p<0.01, Two-way ANOVA with Tukey’s post hoc test was performed.

3.2. Old age and SCI alter EVs miRs content associated with inflammation and autophagy signaling

Next, we examined acute changes in EV cargo in young and aged animals. EVs were isolated from both the plasma and spinal cord (SC) injury site at 1d post-SCI to compare the overall miRs profile using a CNS-enriched Fireplex^® assay. Of the 65 miRs tested, main effects analysis revealed 12 miRs (9 up, 3 down) in plasma EVs and 18 miRs (10 up, 8 down) in SC EVs that were differentially expressed after injury as well as 6 miRs (2 up, 4 down) in plasma and 11 miRs (7 up, 4 down) in SC with age (Fig. 2A–C). Corroborating prior reports (Choi et al., 2018; Hatse et al., 2014), these age-modified miRs were mainly associated with inflammatory activation (“inflammaging”), including increases in miR-146a-5p and miR-155–5p, and decreases in miR-214–3p, miR-93–5p, and miR-20a-5p.

Figure 2. Old age and SCI alter microRNA (miRs) cargo profiles in both plasma and tissue EVs.

(A-C) Main effect analysis indicated the numbers of EVs-contained miRs altered by 1 d injury or age (A), which were listed as from plasma (B) or from spinal cord (SC) tissue (C). (D-G) Detailed information showed that different miRs were up/down-regulated in plasma or tissue with age or after injury. n=6 mice/group.

To address the relationship between miRs changes in plasma EVs and SC tissue EVs, we compared the data of differentially expressed miRs by injury, age, and the interaction (Fig. 2D–G). The upregulation of miR-15b-5p and downregulation of miR-150–5p were consistent in both samples after injury (Fig. 2D). All three miRs (miR-125b-5p, miR-206, miR-145–5p) decreased after injury in SC EVs and increased in plasma EVs. Further analyses indicated that only injury effect was significant for miRs cargo alterations in SC tissue; however, group effects of both age and injury, as well as their interaction, were significant for miRs changes in plasma (Fig. S1A–C). Another miR (miR-23a-3p) in the plasma EVs and two miRs (let-7d-5p, miR-103a-3p) in the SC EVs were also found to have a significant interaction between the group effects of age and injury (Fig. S1D–F). miR-23a-3p, similar to miR-206 and miR-145–5p, increased in the plasma of young mice following SCI with more modest changes in aged mice. In the injured tissue, let-7d-5p increased and miR-103a-3p decreased in Aged SCI mice compared to Young SCI group.

Moreover, miR-146a-5p, known to modulate neuroinflammation during the progression of Alzheimer’s disease or neuropathic pain (Lei et al., 2021; Wang et al., 2018b), increased in both compartments with age in our study (Fig. 2E). miR-93–5p was upregulated in SC EVs with both age and injury but downregulated in plasma EVs with age (Fig. 2E). Two miRs (miR-24–3p and miR-15b-5p) increased in SC EVs with age as well as in plasma EVs after injury, contrasting with miR-107 and miR-150–5p (Fig. 2F). Meanwhile, miR-132–3p was upregulated in plasma after injury and downregulated in SC EVs with age (Fig. 2F). Two miRs (miR-20a-5p and miR-486–5p) were found increased in SC EVs after injury while decreased in plasma EVs with age (Fig. 2G). Among these miRs, several (miR-93–5p, miR-107, and miR-486–5p) were reported to regulate apoptosis and cellular autophagy (Li et al., 2020a; Li et al., 2018) and some (miR-24–3p, miR-132–3p, and miR-20a-5p) were reported to regulate inflammation response (Gong et al., 2022; Oladejo et al., 2022). Collectively, these data show that EVs miRs contents are dysregulated by age and SCI, which are associated with neuroinflammation and disrupted autophagy-lysosome pathways.

3.3. Plasma EVs proteomics altered by age and SCI associated with proinflammatory pathways

Protein is another important category of EVs content responsible for facilitating signal transduction to recipient cells. The Olink platform that uses multi-plex DNA-coupled immunoassay-based targeted proteomic approach was used to identify protein cargo in EVs by employing the mouse exploratory panel. The abundance of a total of 92 proteins was measured and normalized by Olink. Generally, most proteins were upregulated in aged plasma EVs following SCI compared to other groups (Fig. 3A). The proteomic samples clustered into different pools by group in PLS-DA with 20.8% variations in 1st component and 7.1% variations in 2nd component (Fig. 3B). Further analysis by Two-way ANOVA showed the group effects of age, SCI and the interaction (Table S1). Il10, Fli1, and Sez6l2 were the top 3 proteins with significant interaction effects of age and injury. Qdpr and Eno2 were significantly affected by both age and SCI.

Figure 3. Circulating EVs protein contents are altered by age or SCI at 1 d post-injury.

(A) The protein expression levels in Olink mouse exploratory panel from the plasma EVs of four groups are displayed as z-score heatmap, exhibiting upregulations in large clusters and downregulations in small clusters with Aged SCI mice compared to other groups. (B) Based on the same dataset, four groups clustered into different pools by PLS-DA with 1^st component of 20% and 2^nd component of 7.3% protein variations. (C-D) Differential expression (DE) analysis demonstrated the proteins regulated significantly in the plasma EVs of Aged Sham mice compared to Young Sham group. Protein expression levels of the DE proteins from (C) was visualized as z-score heatmap (D) by group, showing general upregulations in Aged Sham group. (E-F) The upregulated DE proteins in the plasma EVs of Aged SCI mice further potentiated pathways relating to inflammatory responses in recipient cells. (G-H) The upregulated DE proteins were introduced into protein clustering enrichment analysis based on three databases (GO, KEGG, and WikiPathway), indicating the top 3 pathways promoted potentially respectively. (I-N) Similar analyses suggested SCI induced proteomic responses in the plasma EVs of young mice (I-J) and aged animals (L-M), leading to pathways affected predominantly (K, N). n= 6 mic (Young Sham group), 7 mice (Young SCI group), 5 mice (Aged Sham group), and 5 (Aged SCI group).

To clarify the proteome response caused by age or SCI, differential expression analysis of four group combinations was performed to extract the protein cargos enriched or deprived in plasma EVs with aging or following injury. Specifically, 13 proteins were significantly upregulated in plasma EVs between two sham groups (Fig. 3C–D). A total of 15 upregulated and 3 down-regulated proteins were detected between two SCI groups (Fig. 3E–F). Considering the homogeneity of receiving cells, we analyzed the functional pathways stimulated potentially by these differentially expressed proteins (DEPs) assuming their complete absorption. Clustering enrichment analysis based on three databases (GO, KEGG, and WikiPathways) indicated that similar top pathways were upregulated by the DEPs with aging in Fig. 3G–H, which implies chemokine/cytokine activity and chronic pro-inflammatory responses. On the other hand, injury effects from SCI caused 8 DEPs in plasma EVs between Young SCI and Young Sham (5 up- and 3 down-regulated (Fig. 3I–J), which also suggests the promotion of chemokine/cytokine activity and receptor binding. (Fig. 3K). In contrast, SCI in aged mice resulted in 7 DEPs (5 up- and 2 down-regulated, Fig. 3L–M) mainly affecting immune and inflammatory responses, along with microglia functions (Fig. 3N).

Of the 92 proteins tested, there were 22 cytokines and chemokines contained in the plasma EVs (Fig. 4A). The heatmap of differences in these cytokines and chemokines highlighted alterations in proinflammatory markers. Individual plots (Fig. 4B–D) showed that aged SCI EVs were noted to be significantly enriched in the chemokine monocyte chemoattractant protein-1 (Ccl2) and a key pro-inflammatory cytokine interleukin 23 receptor (Il23r) compared to the young SCI EVs. Moreover, the expression of the chemokine (C-X-C motif) ligand 1 (Cxcl1) was significantly elevated in young SCI EVs compared to the Young Sham EVs.

Figure 4. Cytokines and chemokines in plasma EVs are dysregulated by old age after SCI.

(A) Heat map showing the expression of 22 cytokines and chemokines in EV protein cargo across the groups. (B-D) Individual plots indicate normalized protein expression (NPX) expressed in a log2 scale of the differentially expressed cytokines and chemokines. Aged SCI EVs showed significantly enriched in the chemokine monocyte chemoattractant protein-1 (Ccl2) and a key pro-inflammatory cytokine interleukin 23 receptor (Il23r) compared to the young SCI EVs. The expression of the chemokine (C-X-C motif) ligand 1 (Cxcl1) was significantly elevated in young SCI EVs compared to the Young Sham EVs. n=6 mic (Young Sham group), 7 mice (Young SCI group), 5 mice (Aged Sham group), and 5 (Aged SCI group). #p<0.05, **p<0.01, Two-way ANOVA with Tukey’s post hoc test was performed.

Together, we demonstrate that the plasma EVs protein contents are altered by SCI, which is associated with pro-inflammatory responses to injury and is exacerbated in aged mice.

3.4. Old age and SCI lead to distinct changes of lipid composition in plasma EVs

Plasma EVs are enriched with various lipid classes, which play essential structural and functional roles in exosome biogenesis and cell-to-cell communication. The total lipid extracts of EVs samples from the four groups were subjected to liquid chromatography-tandem mass spectrometry (LC-MS/MS) for category and abundance assessment. The lipidomic variations separated two sham groups into well-separated pools in PCA in both positive and negative mode (Fig. 5A–B), demonstrating robust up- or down-regulations in different lipid categories (Fig. 5C–D). Further differential expression analysis indicated multiple lipid clusters changed significantly with age and the lipids with biggest variations were shown in the volcano graphs (Fig. 5E–F). Triacylglycerols (TG, 30 in total) and phosphatidylcholines (PC, 36 in total) were the top two lipid categories altered by aging. Next, the lipidomic data was analyzed using Lipid Ontology (LION) enrichment analysis for functionality assessment. Fig. 5G exhibited the top 10 LION-terms of significant enrichment, including “diacylglycerophosphocholines” (DG), multi-terms of “fatty acid” (FA), “glycerophospholipids” (GP), “membrane component”, “endoplasmic reticulum” (ER), “average lateral diffusion”, and “neutral intrinsic curvature”. Similarly, distinct lipidomic profiles of two SCI groups were displayed as PCA plots (Fig. 5H–I) and heatmaps (Fig. 5J–K) with DE results shown in Fig. 5L–M. Noticeably, 87 out 96 DE TGs and 26 out of 27 DE PCs increased significantly in aged mice compared to young ones following SCI. Accordingly, the top enriched LION-items cover multiple classes of “triacylglycerols”, “lipid storage”, “lipid droplet”, “glycerolipids” (GL), and “headgroup with neutral charge” (Fig. 5N).

Figure 5. The lipid composition of plasma EVs are altered by aging with SCI.

(A-B) The lipidomics data of plasma EVs acquired under positive or negative mode LC-MS/MS was analyzed via PCA, indicating Aged Sham and Young Sham groups clustered separately. (C-D) Heatmap of differentially expressed (DE) lipids from (A-B) demonstrated downregulations and upregulations of different lipid categories between the two sham groups. (E-F) DE lipids were displayed as volcano plots highlighting fold change and significance. (G) LION enrichment analysis suggested top functional terms regulated by the total DE lipid categories with Aged Sham mice compared to Young Sham group. (H-M) Aging with SCI caused distinct changes of lipid compositions in plasma EVs as visualized by multivariate analysis (PCA, H, I), clustering (J-K) and univariate analysis (L-M). (N) Top 10 functional terms driven by aging with SCI in LION enrichment analysis. n=6 mice (Young Sham group), 6 mice (Young SCI group), 5 mice (Aged Sham group), and 5 mice (Aged SCI group).

On the other hand, both young and aged mice exhibited distinct lipid profiles in plasma EV following SCI, which were visualized in PCA plots (Fig. S2A–N). DE analysis showed that a big part of lipid clusters was downregulated in young and aged mice post-injury, especially the TG. SCI resulted in a significant reduction of 107 out of 119 DE TGs in young mice and 70 out of 82 in aged mice. PC is another lipid category altered by injury, as 8 out of 16 in young mice and 18 out 19 in aged mice decreased significantly. As PCs are major components of EVs membranes, their downregulation may reflect reduction of EVs particles in plasma, which is consistent with our data in Fig. 4a. Moreover, LION enrichment analysis between two young groups indicated that multiple fatty-acid-related terms changed significantly along with “membrane component”, “headgroup with positive charge/zwitter-ion”, “sphingolipids”, “glycerophospholipids”, “endoplasmic reticulum”, and “sphingomyelins”. In aged mice plasma EVs, the DE lipid clusters were significantly enriched in “plasma membrane”, “sphingomyelins”, “endosome/lysosome”, “sphingolipids”, etc. from Aged SCI versus Aged Sham.

Interestingly, several classes of lysophospholipids, including lysophosphatidylcholine (LPC), ether-linked lysophosphatidylcholine (LPC-ether), and lysophosphatidylethanolamine (LPE) significantly decreased by age with SCI. Specifically, 5 out of 8 LPC-ether and 2 LPC decreased significantly with Aged Sham to Young Sham (Fig. 5C–D). All the 9 LPC-ether, 2 LPC, and 2 LPE were reduced significantly in Aged SCI compared to Young SCI (Fig. 5J–K). 4 LPC-ether but not 1 LPC nor 1 LPE decreased in Young SCI to Young Sham (Fig. S2C–D). All 8 DE LPC-ether and 1 LPE were reduced significantly in Aged SCI compared to Aged Sham (Fig. S2J–K). Together, these results indicate that old age and SCI alter lipid compositions of circulating EVs, reflecting dysregulated lipid pathways including lysosome function in the host cells.

3.5. Circulating EVs from SCI mice induce the secretion of pro-inflammatory cytokines and neuronal apoptosis in vitro

To determine the inflammatory response of plasma EVs derived from SCI mice, primary microglia, astrocytes, and neurons were cultured from mouse neonatal or embryonic cortices. Plasma EVs were isolated from young adult or aged mice at 1 d post-injury or sham mice. NTA was measured to obtain EVs concentration (EVs particles/ml). Based on pilot data, primary microglia were exposed to three different dosages (2 × 10⁹, 4 × 10⁹, and 6 × 10⁹ particles/ml) of EVs for 24 h, separately and the resulting supernatants were collected for ELISA assays. We initially assessed the secretion of cytokines CXCL2, IL-6, and TNFα in cultured microglia and found that CXCL2 release was increased in response to the EVs stimulation. In contrast, TNFα and IL-6 were barely detected in the microglial supernatants. EVs from young SCI-mice increased CXCL2 secretion in a dose-dependent manner compared to EVs stimulation from sham animals or PBS treatment (Fig. 6A). The highest concentration of EVs (6 × 10⁹ particles/ml) derived from young SCI mice significantly increased secretion of CXCL2 compared to Sham EVs or PBS (Fig. 6A–B). As SCI in aged mice causes profound neuroinflammation and neurodegeneration in the brain, we hypothesized that plasma EVs derived from these animals had more proinflammatory features contributing to the process. Therefore, we selected average EVs concentration (4 × 10⁹ particles/ml) from aged animals for microglial stimulation. EVs from aged SCI mice at a middle concentration level induced significant CXCL2 secretion (Fig. 6C, Fig. S3A–F). To ascertain the microglial responses were due to the EVs stimulation, we tested PBS, plasma supernatant collected after the ultracentrifugation spin. None of these were able to induce CXCL2 secretion in cultured microglia, indicating the specific effects of EVs derived from SCI animals.

Figure 6. Circulating EVs from SCI mice induce the secretion of pro-inflammatory cytokines and neuronal apoptosis in vitro.

Plasma EVs were isolated from young adult (12 male mice) or aged mice (12 male mice) at 1 d post-injury or sham mice (12 male mice). Nanoparticle Tracking Analysis (NTA) was measured to obtain EVs concentration (EVs particles/ml). (A-B) Primary microglia were exposed to three different dosages (2 × 10⁹, 4 × 10⁹, and 6 × 10⁹ particles/ml) of EVs for 24 h, separately and the resulting supernatants were collected for CXCL2 ELISA assays. Different dosages of Young SCI EVs were tested for the optimal condition with PBS or Young Sham EVs (A) as control. The fold changes of CXCL2 levels (B) were compared among groups after the stimulation of Young SCI EVs (6 × 10⁹ particles/ml, n=11 mice × 2 independent microglia cultures). Two batches of CXCL2 ELISA assays are presented in Supplemental Figure S4a–c. (C) The fold changes of CXCL2 levels were compared among groups after the stimulation of Aged SCI EVs (4 × 10⁹ particles/ml, n=12 mice × 2 independent microglia cultures). Two batches of CXCL2 ELISA assays are presented in Supplemental Figure S4d–f. (D-E) The fold changes of IL-6 levels released from cultured astrocytes were compared among groups after the stimulation of Young SCI EVs (D, 6 × 10⁹ particles/ml, n=12 mice × 2 independent astrocytes cultures) or Aged SCI EVs (E, 4 × 10⁹ particles/ml, n=12 mice × 2 independent astrocytes cultures) compared to Young Sham EVs (n=12 mice × 2 independent astrocytes cultures) or Aged Sham EVs (n=11 mice × 2 independent astrocytes cultures) or PBS. Two batches of IL-6 ELISA assays are presented in Supplemental Figure S4g-l. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001. One-way ANOVA with Tukey’s post hoc test was performed (A-E). (F-G) Expression levels of cleaved caspase 3 protein in cultured neurons treated with various plasmas EVs isolated from young adult mice (F, 6 × 10⁹ particles/ml) and aged animals (G, 4 × 10⁹ particles/ml). Each blot lane represents an individual animal EVs. After 1d treatment, the harvested neuronal samples were blotted with indicated antibodies. In each case loading control (actin) is from the same blot as the experimental sample. *p < 0.05; **p < 0.01. n=4–5 mice/group using Mann-Whitney test (F-G).

Among these cytokines tested in cultured microglia, we found that IL-6 response to the EVs stimulation was marked in cultured astrocytes. Both EVs derived from young SCI mice at 6 × 10⁹ particles/ml or aged SCI animals at 4 × 10⁹ particles/ml significantly increased IL-6 secretion compared to their Sham EVs or PBS (Fig. 6D–E). We also tested PBS, plasma supernatant collected after the ultracentrifugation spin (Fig. S3G–L). None of these induced IL-6 secretion in cultured astrocytes, indicating the specific effects of EVs derived from SCI animals.

Primary cortical neurons were incubated with EVs derived from young or aged SCI mice. After 24 h treatment, the cells were harvested for protein extraction and neuronal apoptosis assay. Western blotting analysis showed that protein expression level of cleaved caspase 3 was significantly elevated in EV groups derived from young SCI mice at 6 × 10⁹ particles/ml or aged SCI animals at 4 × 10⁹ particles/ml (Fig. 6F–G). Fig. S4 showed additional experiments in which the EVs were derived from a different set of animals and the treatment in a different culture. Bafilomycin A1 (BFA, 100 nM) was used as a positive control for inducing neuronal apoptosis. Fig. S5 showed unedited western blots of Fig. S4 and Figure 6F, G. Together, these data indicate the pro-inflammatory features of circulating EVs from SCI mice, with the comparable effects by a lower concentration of EVs-derived from aged SCI animals.

3.6. Systemic administration of plasma EVs from SCI mice impairs neurological outcomes and general physical function in intact mice

Given that post-injury derived plasma EVs are capable of inducing proinflammatory responses in cultured glia and neurons, we next assessed their pathophysiological effects in vivo through intravenous injection in intact aged mice. The experimental scheme is shown in Fig. 7A. Both the groups of aged mice receiving Young SCI EVs (YSE) or Aged SCI EVs (ASE) exhibited significantly lower scores in the Nesting test compared to their baselines (Fig. 7B, Statistical details in Table S2), suggesting changes in general health of these mice. Representative examples are demonstrated in Fig. 7B, showing a YSE mouse and an ASE one, both of whom performed worse in post-injection nesting behavior. In the Marble burying test, ASE mice but not YSE animals showed significantly more unburied marbles than baseline in the testing cage (Fig. 7C, Table S2), reflecting the reduction of their exploration activities in a novel environment. Representative examples exhibit the changes between pre- and post- injections as well as between the two groups. Grip strength analysis showed no difference of the forelimbs strength force across the groups (Fig. 7D). However, significantly higher values were detected from all limbs of YSE group compared to their baseline, implicating the rejuvenating effect of plasma EVs from young mice. In the NOR test, the two groups achieved no difference in both object bias score and novelty preference (Fig. 7E), suggesting intact cognition and recognition memory. In the Y-maze test reflecting hippocampus-dependent spatial memory, YSE mice exhibited significantly increased alteration compared to their baselines, indicating enhanced cognitive function and reinforcing rejuvenating effects of plasma EVs from young mice. In contrast, the aged mice receiving ASE showed significant reduction of spontaneous alteration compared to the mice injected with YSE, indicating impaired spatial memory (Fig. 7F, Table S2). Although there is no difference of “Arm return”, both groups showed significantly reduced locomotion activities, including total distance and total entries. Collectively, these results suggest that plasma EVs from SCI mice are able to impair general physical performance and spatial memory in naïve aged animals with worse effects from Aged SCI EVs. These findings also imply rejuvenating effects of plasma EVs from young mice.

Figure 7. Systemic administration of plasma EVs from SCI mice impairs general physical function and neurological outcomes in intact aged mice.

Naïve aged mice receiving Young SCI EVs (YSE) or Aged SCI EVs (ASE) were evaluated using a battery of neurological behavioral tests before or after intravenously EVs injection. (A) Experimental paradigm is illustrated. (B) The nesting scores were quantified and the representative images of the nestlets were indicated. Aged mice receiving YSE or ASE exhibited significant lower scores in the Nesting test compared to their baselines. (C) Unburied marbles in the novel cage were recorded in the Marble burying test. Representative images of the marbles on bedding were indicated. ASE mice but not YSE animals showed significantly more unburied marbles than baseline. (D) Grip strength (peak force) in the forelimbs and all limbs was examined and normalized by animal’s body weight (BW). Aged mice receiving YSE showed significantly increased values of strength force from all limbs compared to their baseline. (E) The Novel object recognition (NOR) analysis showed no difference of the object bias score and novelty preference across the groups. (F) Spontaneous alteration, total arm entries, arm return, and total distance were recorded in the Y-maze test. YSE mice exhibited significantly increased alteration compared to their baselines. In contrast, the aged mice receiving ASE showed significant reduction of spontaneous alteration compared to the mice injected with YSE. Both YSE and ASE groups showed significantly reduced locomotion activities including total distance and total entries. n=8–10 mice/group. Nonparametric paired t tests were performed for comparisons between pre- and post-injections. Mann-Whitney tests were performed for comparison between YSE and ASE groups. *p<0.05, **p<0.01.

Furthermore, we investigated the consequences following in vivo injection of YSE or ASE in intact young adult mice with PBS as control. Three groups of mice were applied as the experimental scheme in Fig. 8A. No significant changes were observed in the Nesting test as shown in the representative examples (Fig. 8B). However, in the Marble burying test, significantly more marbles were unburied by the mice receiving YSE or ASE compared to PBS group (Fig. 8C, Statistical details in Table S2), suggesting decreased exploration activities of the animals after receiving SCI EVs. While no significant changes were detected in grip strength between PBS group and YSE or ASE mice, ASE mice showed significantly lower scores compared to YSE groups (Fig. 8D, Table S2), suggesting impaired muscle health in these mice. In the cognitive functional assessment, we did not observe significant differences between the groups in both NOR and Y-maze tests (Fig. 8E–F). These findings indicate that plasma EVs from SCI animals are sufficient to impair overall well-being of naive young adult mice without affecting cognitive function.

Figure 8. Plasma EVs from SCI animals impair overall well-being of naive young adult mice without affecting cognitive function.

Naïve young adult mice receiving Young SCI EVs (YSE) or Aged SCI EVs (ASE) were evaluated using a battery of neurological behavioral tests after intravenously EVs injection. (A) The experimental scheme was followed by PBS, YSE, and ASE groups. (B) No difference of the nesting behaviors was observed across the groups. (C) Mice receiving YSE or ASE showed increased numbers of unburied marbles compared to PBS group. (D) Significantly lower grip strength was detected in ASE mice compared to YSE group. (E-F) The three groups did not show differences in both NOR (E) and Y-maze (F) tests. n=8–10 mice/group. One-way ANOVA with Tukey’s or Dunnett’s post hoc test was performed. *p<0.05, **p<0.01.

3.7. Circulating EVs derived from SCI animals mediate neuroinflammation in the brain in intact aged mice

To determine if the observed behavioral impairment in naive aged mice after plasma EVs administration may relate to molecular changes in the brain, we collected the cerebral cortex and hippocampus from YSE and ASE mice after the completion of the behavioral tests and applied the RNA samples in NanoString Neuroinflammation panel with naïve aged mice as control (Fig. 9–10). Through PLS-DA, the transcriptomic data of the three groups clustered into different pools in the cortex and hippocampus (Fig. 9A–10A, Fig. S6A–B). The Naïve group was away from YSE or ASE groups along the 1^st components with 22% gene variations in the cortex and 25% in the hippocampus. The 2^nd components with 6% gene variations in the cortex and 9% in the hippocampus caused opposite changes in YSE group and ASE group. In-depth pathway analysis demonstrated that there are robust downregulations and upregulations from Naïve to YSE or ASE in the cortex (Fig. 9B) as well as overall upregulations in the hippocampus (Fig. 10B). Further statistical analysis showed that changes are significant in most of the target pathways in the two brain regions by comparisons between Naïve and YSE or ASE (Table S3). In the cortex, there was no significant difference of these pathways between YSE and ASE groups. However, in the hippocampus, YSE group exhibited significant upregulations in the pathways of “Angiogenesis”, “Epigenetic Regulation”, Insulin Signaling”, and “Notch”, compared to ASE group. This may reflect the differences of YSE mice to ASE ones in some of the behavior tests observed.

Figure 9. Circulating EVs derived from SCI animals mediate transcriptomic changes in the cerebral cortex of intact aged mice.

(A) The transcriptomics data of Naïve, YSE, and ASE groups from NanoString Neuroinflammation panel were processed by PLS-DA, clustering into different pools by group. (B) Pathway scoring analysis indicated downregulations and upregulations of inflammation-related pathways with YSE and ASE groups compared to the Naïve. (C-E) Volcano graphs from differential expression analysis exhibited the between-group DEGs of YSE to Naïve (C), ASE to Naïve (D), and ASE to YSE (E). (F) Venn diagram indicated the intersections and discrepancies of DEGs with heatmap exhibiting gene variations in the exclusive subset of ASE to YSE (red-circled). n=5 mice (Naïve), 10 mice (YSE), and 9 mice (ASE).

Figure 10. Transcriptomic changes are dysregulated in the hippocampus of naive aged mice after systemic administration of plasma EVs from SCI mice.

(A) The transcriptomics data of Naïve, YSE, and ASE groups from NanoString Neuroinflammation panel were processed by PLS-DA, clustering into different pools by group. (B) Pathway scoring analysis indicated downregulations and upregulations of inflammation-related pathways with YSE and ASE groups compared to the Naïve. (C-E) Volcano graphs from differential expression analysis exhibited the between-group DEGs of YSE to Naïve (C), ASE to Naïve (D), and ASE to YSE (E). (F) Venn diagram indicated the intersections and discrepancies of DEGs with heatmap exhibiting gene variations in the exclusive subset of ASE to YSE (red-circled). n=5 mice (Naïve), 10 mice (YSE), and 9 mice (ASE).

Differentially expressed genes (DEGs) analysis revealed big numbers of DEGs between Naïve and YSE or ASE groups in the cortex and hippocampus (Fig. 9C–D, 10C–D), suggesting intravenous injection of SCI plasma EVs capable of inducing systemic gene responses. Among DEGs, 49 out of 582 genes in the cortex (Fig. 9E) and 140 out of 622 genes in the hippocampus (Fig. 10E) represented the DE variations in mRNA level between YSE and ASE groups. The three DEG sets were introduced into Venn diagrams (Fig. 9F, 10F), showing the between-group similarities and discrepancies. YSE mice shared most of the DEGs with ASE animals to Naïve group, 235 in the cortex (Fig. 9F) and 224 in the hippocampus (Fig. 10F). Most of the common DEG clusters were upregulated or downregulated consistently in YSE and ASE groups for both brain regions (Fig. S6C–D). The 15 DEGs changed between YSE and ASE groups exclusively in the cortex are displayed as a heatmap of the average z-score of their expression levels (Fig. 9F) as well as the 21 DEGs exclusive to ASE versus YSE in the hippocampus (Fig. 10F). Comparing to control mice, these genes in the upper cluster were upregulated in YSE mice and downregulated in ASE group, while those in the lower cluster exhibited opposite changes. Based on the gene-pathway annotations of Neuroinflammation panel, these DEGs altered reversely in the two brain regions mostly clustered in several common pathways, including “Microglia Function”, “Apoptosis”, “NF-kB”, etc. (Fig. S6E), which could drive distinct downstream responses. Together, our data demonstrate that plasma EVs derived from SCI mice cause robust neuroinflammatory gene changes in intact aged mice brain, associated with brain dysfunction.

To examine the effect that EV injections have on the brain in neuroinflammation, we used qPCR to examine markers that reflect microglial function, cytokine secretion and inflammatory signaling. In the cerebral cortex (Fig. 11A), no differences were observed for the expression of Cd68 and Tmem119. However, the VEGF receptor Flt1 (Fms related receptor tyrosine kinase 1), a gene that contributes to microglial pathology in AD brains (Ryu et al., 2009), showed a significant increase in both groups subjected to SCI EV injection. Similarly, the gene P2ry12 was observed an increase after injection of EVs obtained from both Young (YSE) and Aged SCI (ASE). For the gene Tgm2 (Transglutaminase 2), however, we only observed a significant increase of expression in the cortex of YSE mice, but not ASE mice, with post-hoc analysis even showing significantly lower levels in ASE compared to the YSE group. In the hippocampus (Fig. 11B), the changes were even more robust after EV injection, with significant changes of all five microglial markers examined. The genes Cd68, Flt1, and P2ry12 all saw significant increases in YSE and ASE mice. Mirroring its expression profile in the cortex, the gene Tgm2 also saw a significant increase in the hippocampus of YSE mice, but not ASE. In contrast, the gene Tmem119 was significantly decreased by EV injection in YSE and ASE groups. Next, we sought to assess cytokine secretion and inflammatory signaling. In the cortex (Fig. 11C), the mRNA expression levels of IL10rb (Interleukin 10 Receptor Subunit Beta), Osmr (Oncostatin M Receptor) and Tnfrsf12a (TNF Receptor Superfamily Member 12A) were significantly elevated in YSE and ASE mice. These genes also saw significant increases in the hippocampus (Fig. 11D), with the exception of Osmr which only showed significant increases in YSE mice. In contrast, the genes Csf1r and Trem2 only showed significant changes in the hippocampus region, with the former being significantly decreased in ASE compared to naïve mice, while the latter showed marked elevation in both YSE and ASE. Finally, we examined the expression levels of Pcna (Proliferating Cell Nuclear Antigen), which saw no changes in the cortex (Fig. 11E), but it was significantly elevated in the hippocampus (Fig. 11F) of YSE and ASE mice. In addition, plasma EVs from both YSE and ASE did not change innate immune response and astrocytes function in naïve aged mice (Fig. S7). Taken together, these results suggest higher susceptibility to EV induced microglial activation and inflammatory signaling in the hippocampus, highlighting a regional difference.

Figure 11. Systemic administration of plasma EVs derived from SCI mice alters microglia function and inflammatory signaling in the brain.

(A-B) qPCR analysis was utilized to examine markers of microglia function (Cd68, Flt1, P2ry12, Tgm2, Tmem119) in the cerebral cortex (A) and hippocampus (B). (C-D) mRNA expression of cytokine and inflammatory signaling (Csf1r, IL10rb, Osmr, Tnfrsf12a, Trem2) was examined in the cortex (C) and hippocampus (D). (E-F) Cell cycle and DNA damage gene Pcna was assessed via qPCR in the cortex (E) and hippocampus (F) regions. n=5 mice (Naïve), 10 mice (YSE), and 9 mice (ASE). One-way ANOVA with Dunnett’s T3 multiple comparisons post hoc test was performed. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001.

4. Discussion

Here, we present for the first time that circulating plasma EVs and their cargo contents, including miRs, proteins, and lipids, were disrupted in aged animals and SCI mice, reflecting dysregulated neuroinflammation and lipid pathways. In vitro, the plasma EVs from SCI mice induced the secretion of pro-inflammatory cytokines in microglia and astrocytes, and neuronal apoptosis, indicating the pro-inflammatory features. In vivo, we provide novel data that plasma EVs from young donors showed rejuvenating effects in naïve aged mice, evidenced by improved grip strength and spontaneous alteration in Y-maze test. Importantly, our findings demonstrated that systemic delivery of post-traumatic plasma EVs is sufficient to impair general physical function and spatial memory in intact aged animals via a pro-inflammatory mechanism in the brain. Together, our studies identify a circulating EV-mediated mechanism for SCI-induced brain dysfunction.

Emerging data suggest that EVs may participate in the progression of secondary injury by transporting parent cell-specific signaling cargoes (e.g., signal lipids, genetic information, cytokines, receptors, etc.) that alter the function of recipient cells both within and outside the CNS (Dutta et al., 2021; Rufino-Ramos et al., 2017). However, circulating EVs-mediated crosstalk following SCI and its underlying molecular mechanism(s) has received limited attention. Recently, our lab reported that SCI in young adult mice potentiated the circulating EVs response with altered miR cargo, contributing to upregulated expression of inflammation-related genes in the brain (Khan et al., 2021). In the present study, we examined changes in plasma EVs derived from aged mice. No significant differential changes in particle concentration, particle size, and coefficient of variation were detected between aged (18-month-old male mice) and young adult (3-month-old) male animals. However, in agreement with the study from Faisal et al (Alibhai et al., 2020), increased CD63 expression levels were observed in the EVs derived from aged mice. This likely reflects age-related increase of cellular senescence which contributes to changes in particle cargo and function.

EV miR cargoes analysis revealed that old age alters the EV response in both circulation and injured tissue, as well as miR cargo, following SCI. We observed 3 miRs (miR-125b-5p, miR-206, miR-145–5p) that are increased in plasma EVs, but decreased in spinal cord EVs after injury. miR-145–5p negatively affects cell proliferation and chemokine secretion, as well as acting as a regulator of SOX2 (Ozen et al., 2015; Yan et al., 2019). miR-206 is enriched in mouse and human skeletal muscle and is critical for myogenesis (Salant et al., 2020; Sempere et al., 2004). miR-125b-5p acts as a negative co-regulator of some inflammatory genes through the TRAF6/MAPKs/NF-κB pathway (Rasheed et al., 2019). These miRs may shuttle from spinal cord to plasma after pro-inflammatory activation, serving as regulators that modify gene expression in downstream target cells. Moreover, in the injured tissue, we showed that let-7d-5p increased and miR-103a-3p decreased in aged SCI mice compared to young SCI animals. The former suppresses inflammatory responses, whereas the latter reduces apoptosis and inflammation by targeting HMGB1 (Li et al., 2020b; Sun et al., 2020). Thus, altered miR cargo after SCI impacts inflammation and autophagy related pathways, potentially contributing to subsequent neuropathology.

Utilizing the Olink Mouse exploratory panels with a small set of proteins, we detected more variance of protein cargos in plasma EV caused by aging than those by injury, which are functionally capable of potentiating inflammation processes in recipient cells. The common proteins upregulated with aging include Ccl2 and Tnfrsf11b. The former is short for C-C motif ligand 2, also known as monocyte chemoattractant protein-1 (MCP-1), which may cause robust accumulation of macrophages and activation of microglia in mouse brain (Selenica et al., 2013) and contribute to neurodegenerative diseases such as Alzheimer’s Disease (Joly-Amado et al., 2020). The latter Tnfrsf11b was identified as a predictive inflammatory plasma marker in human sepsis development by a recent study (Zhang et al., 2023). While pro-inflammatory chemokine Cxcl1 (Zhou et al., 2023) was significantly elevated in young SCI EVs, Ccl2 and a key pro-inflammatory cytokine Il23r (Bloch et al., 2018) were noted to be significantly enriched in the aged SCI EVs, demonstrating that the plasma EVs protein contents are altered by SCI which is associated with pro-inflammatory responses to SCI and is exacerbated in aged mice. Moreover, the proteomic responses in plasma EV also reflect the senescence status of inflammaging in originating cells, implying that proinflammatory responses could be enhanced following injury in aged animals. A recent clinical study showed a higher levels of serum chemokine (C-C motif) ligand 21 (CCL21) in acute SCI patients (Chen et al., 2020). We and others have previously reported that CCL21, a potent microglia activator, is released by injured dorsal horn neurons after SCI and accumulates in the brain via directly axonal transportation, contributing to pro-inflammatory responses in the brain after SCI (Wu et al., 2016; Wu et al., 2014; Zhao et al., 2007). While acknowledging the limitations that a small set of proteins is detected in the Olink Mouse exploratory panels, we recognized if circulating CCL21 is carried by plasma EVs participating in the progression of brain neuroinflammation after SCI is intriguing for the future investigation.

Lipids are essential components of EVs and play structural and regulatory functions during a wide range of biological activity of the vesicles. Importantly, lipid dysregulation has been linked to several disorders, including metabolic syndrome, inflammation, and neurological dysfunction (Perpina-Clerigues et al., 2023). Here, we revealed that most of the triacylglycerols (TG) and phosphatidylcholines (PC) increased with aging but decreased following SCI. These were associated with EV particle concentration which was reduced with significant injury effect and CD63 expression increased with significant age effect. Studies in human subjects reported that serum levels of TG, PC and other lipids went up with age (Kawanishi et al., 2018), but declined prior to the onset of type I diabetes (Suvitaival, 2020). Generally, the circulating EVs lipidome could undergo significant changes with age, and may be influenced by factors such as metabolic and disease state.

We have previously shown that SCI-derived plasma EVs from young adult male mice increase glial activation with neuroinflammatory potential when injected into the lateral ventricle of intact brain (Khan et al., 2021). Here, the in vitro studies revealed that cultured mouse microglia and astrocytes exhibit pro-inflammatory responses after exposure to SCI-derived blood-borne EVs. Moreover, EVs from aged SCI mice, at a lower concentration, showed comparable pro-inflammatory and neuronal cell death responses compared to those from young adult mice, indicating the pro-inflammatory features of SCI-derived circulating EVs. In vivo, when systemic delivery of SCI-derived plasma EVs into naive aged mice, we observed impairments of locomotion activities in Nesting, Marble burying, and Y-maze tests, reflecting the deterioration of animals’ physical wellbeing. These were accompanied by the transcriptomic changes in the somatosensory cortex and hippocampus, where two groups shared similar regulations in genes and pathways in terms of neuroinflammation. Furthermore, qPCR data analysis suggests higher susceptibility to EV induced microglial activation and inflammatory signaling in the hippocampus, highlighting a regional difference. This may be associated with impaired hippocampus-dependent spatial memory observed in intact aged mice receiving SCI-EVs from aged donors. In line with our data, Nicolas et al. reported that transfer of plasma from young health mice to aged mice, resulted in improvements in hippocampal-dependent learning and memory in the aged mice (Fitz et al., 2023). After young plasma treatment, the hippocampal transcriptome showed significant upregulation of the anti-aging gene Klotho, with this effect being diminished after EV depletion. Thus, our findings demonstrate that plasma EVs are capable of transducing injury signals from damaged tissue to the brain following SCI. Of potential importance was significantly improved spatial memory and all-limb grip strength detected in naïve aged mice receiving plasma EVs from young donors. This may be due to the beneficial effects produced by young serum EVs in aged mice which is consistent with previous reports (Chen et al., 2024; Wang et al., 2018a). EVs are naturally secreted nano-vesicles by living cells, which inherit several advantages over other blood components (Chen et al., 2024). In vivo injection of EVs is biocompatible, non-cytotoxic, and exhibits low immunogenicity. EVs also possess a high cargo loading capacity, which includes several proteins, miRNA cargo, and can even cross the blood-brain barrier. Further the isolated EVs are easy to store and have an increased stability for long term experimental purposes. Thus, EVs provide a simpler and safer alternative to whole blood by potentially reducing the side effects and risks of blood transfusion. Moreover, there are limitations associated with the volume of blood transfusion, whereas EVs can be concentrated during isolation and purification process. Considering the pivotal role of EVs as multifunctional messengers across a spectrum of biological processes, we reason to speculate that plasma EVs from young health donors may have potential to alleviate SCI complications in aging population.

Our finding that SCI-derived plasma EVs also impairs overall well-being of naive young adult mice is consistent with the notion of detrimental effects of EVs cargoes derived from SCI animals. Impaired grip strength observed in young mice receiving aged SCI-derived EVs support the hypothesis that EVs-derived from aged SCI animals show greater effects of the pro-inflammatory features compared to young SCI-EVs. A limitation from the present study is that we didn’t include a loss-of-function experiment by blocking the EV release due to technical feasibilities. Welshed et al. discussed the inhibition of EV release using various genetic manipulations and drugs, including RAB27A/B knockdown, neutral sphingomyelinase inhibition, and ARRDC1 inhibition (Welsh et al., 2024). Although these treatments are often claimed to be specific for EVs of particular biogenesis pathways, they may also affect EV formation and membrane trafficking more generally (Kalluri and LeBleu, 2020). Consequently, it is difficult to exclude an impact on other EVs and/or non-EV cellular processes.

EVs can be also detected in the cerebrospinal fluid (CSF), providing an alternative route for transporting cargo that promotes brain inflammation after SCI. Elevated levels of inflammasome proteins have been detected in EVs isolated from the CSF of human SCI patients (de Rivero Vaccari et al., 2016). CSF-derived EVs from female Bama miniature pigs with SCI showed a better effect in promoting vascular regeneration compared to sham CSF-EVs (Li et al., 2023a). In our recent report with the lateral ventricle injection of SCI-derived EVs (Khan et al., 2021), we observed increased Nlrp3 and Il1b mRNA expression in the cerebral cortex, supporting an association between EVs and brain inflammation through CSF circulation. We recently reported that sexually dimorphic EVs responses after chronic SCI are associated with neuroinflammation and neurodegeneration in the aged brain (Li et al., 2023b). In the present study, only male mice were chosen to minimize variability of experimental outcomes and number of animals to reach statistical significance. It is possible that female mice following SCI show differential plasma EVs cargo contents, but further investigation is required.

In summary, we report for the first time that old age and SCI induce distinct profiles of plasma EVs cargo contents including miRs, protein, and lipid components. To our knowledge, this is also the first time pro-inflammatory features of circulating EVs derived from aged animals following SCI have been tested both in vitro and in vivo. These findings suggest that old age alters circulating plasma EVs pro-inflammatory response to SCI, potentially contributing to brain dysfunction after SCI. Thus, our studies shed light on the mechanisms by which circulating EVs cargo contents shape the pathophysiology of aging and link SCI and brain dysfunction.

Highlights.

• Old age and SCI induced distinct profiles of plasma EVs cargo contents

• SCI-disrupted plasma EVs induced glial activation and neuronal apoptosis in vitro

• In vivo, SCI-derived EVs are sufficient to impair neurological function in intact mice

• Plasma EVs from young animals had rejuvenating effects on naïve aged mice

• In vivo, disrupted circulating EVs cargoes contributed to brain dysfunction after SCI

Data and materials availability

All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials.