Neuronal and perineuronal changes of cerebral cortex after exposure to inhaled particulate matter

The inhalation of particulate matter (PM) increases the perineuronal nets (PNNs) in the cerebral cortex; however, little is known about the related molecular changes. We explored how PM exposure impacted cognitive function and the levels of PNN-related genes. BALB/c mice (6-week-old females, n = 32) were exposed to 1–5-μm diesel-extracted particles (DEPs) (100 µg/m3, 5 hours per day, 5 days per week) and categorized into the following four groups: 1) 4-week DEP exposure (n = 8); 2) 4-week control (n = 8); 3) 8-week DEP exposure (n = 8); and 4) 8-week control (n = 8). The Y-maze test and olfactory function test were conducted after 4 and 8 weeks of DEP exposure. The prefrontal cortex, olfactory bulb and temporal cortex were harvested from the animals in each group. The expression of genes related to PNNs (Tenascin C, matrix metalloproteinase [MMP]14, MMP9) and synaptic vesicular transporters of vesicular glutamergic transporter 1 (VGLUT1), VGLUT2, vesicular GABAergic transporter (VGAT) were measured. The temporal cortex was immunostained for neurocan, VGLUT1, and VGAT. The 4-week DEP group had lower total arm entry in the Y-maze test and olfactory sensitivity. These impaired behavioral functions recovered in the 8-week DEP group. Expression of tenascin C and MMP9 were increased in the cerebral cortex in the 8-week DEP group compared with the control group. The levels of VGLUT1, VGLUT2, and VGAT were elevated in the cerebral cortex of the 8-week DEP group compared with the control group. In immunostaining of the temporal cortex, the expression of neurocan, VGLUT1, and GAD67 were increased in the 8-week DEP group compared with the control group. The 4-week DEP inhalation impaired spatial activities and olfactory sensitivities. After 8 weeks of DEP exposure, the PNN components and their proteolytic enzymes and the vesicular transporters increased in the cerebral cortex.

Breakdown of nets produces unwanted sprouting of synapses and changes in the balance of excitatory and inhibitory neuronal transmission. For instance, parallel plastic changes of excitatory and inhibitory synaptic transporters of vesicular glutamergic transporter 1 (VGLUT1), VGLUT2 and vesicular GABAergic transporter (VGAT), as well as changes in PNN densities, were observed during vestibular compensation 9 .
We postulated that PM inhalation impaired spatial cognitive function and olfactory sensitivities. These behavioral changes might accompany the dynamic changes of PNNs encompassing degradation as well as synthesis. Furthermore, these changes in PNNs might be linked to both excitatory and inhibitory changes in synaptic transporter levels. To test this hypothesis, spatial learning memory function and olfactory function tests were performed after PM inhalation. In addition, components of PNNs, their proteolytic enzymes, and excitatory and inhibitory vesicular transporters were evaluated for longitudinal changes after PM inhalation. No prior study has described synaptic vesicular transporters and molecular changes of PNNs after PM exposure.

Results
Effect of DEP exposure on olfactory function and working memory as revealed by the olfactory and Y-maze tests. Body weights, and the Y-maze and olfactory test results, were compared between the control and the 4-/8-week DEP groups (Fig. 1). Body weight did not change from commencement of the experiment to 4 weeks later. However, body weight was significantly lower in the 8-week DEP group than in the 8-week control group (Fig. 2). Four weeks of DEP exposure reduced the number of total arm entries compared to that of the control group ( Fig. 3A and Supplemental Table S1). However, the numbers of total arm entries of the 8-week DEP and control groups were comparable. The spontaneous alternation percentages were similar among the four  groups (Fig. 3B). On olfactory testing, the 4-week DEP group exhibited fewer sniffs of both attractive and aversive scents than the control group (Fig. 3C,D). changes in neurocan and tenascin c levels. The expression levels of the lectican-cleaving protease MMP 9 were higher in the prefrontal cortex, olfactory bulb, and temporal cortex of the 8-week DEP group than in those tissues of the control group ( Fig. 4A and Supplemental Table S2). The level of mRNA encoding MMP14 was higher in the temporal cortex of the 4-week DEP group than the control group (Fig. 4B). The levels of MMP9 and MMP14 proteins increased in the DEP groups ( Fig. 4D and Supplemental Fig. 1). The ADAMTS1 level tended to increase in the prefrontal cortex, olfactory bulb, and temporal cortex of the 8-week DEP group (compared to controls), but statistical significance was not attained (Fig. 4C).
The tenascin C levels were higher in the prefrontal cortex, olfactory bulb, and temporal cortex of the 8-week DEP group than in those tissues of the control group ( Fig. 5A and Supplemental Table S3). The levels in the 4-week DEP group were comparable to those of controls. In the immunofluorescence assay, the neurocan level in the temporal cortex was higher (as evidenced by greater numbers of immunopositive cells) in the 8-week DEP group compared to the control group (Fig. 5B). changes in the levels of excitatory and inhibitory vesicular transporters. The VGLUT1 levels were higher in the prefrontal and temporal cortices of the 8-week DEP group compared to the control group ( Fig. 6A and Supplemental Table S3). The VGLUT1 expression levels were comparable between the 4-week DEP and control groups. The olfactory bulb VGLUT1 expression level did not differ between the 4-and 8-week DEP groups and the control group. VGLUT2 levels were elevated in the prefrontal cortex of the 8-week DEP group and in the olfactory bulb of the 4-week DEP group (Fig. 6B). All groups exhibited comparable VGLUT2 levels in the temporal cortex. VGAT levels were increased in the prefrontal cortex and olfactory bulb of the 8-week DEP group compared to controls (Fig. 6C). Temporal cortex VGAT levels were similar in all groups.
In the immunofluorescence assay, the VGLUT1 level in the temporal cortex was elevated (as evidenced by greater numbers of immunopositive cells) in the 8-week-DEP group compared to the control group (Fig. 6D). The extent of GAD67-immunopositivity was higher in the 8-week DEP group than in the control group (Fig. 6E).

Discussion
Principal findings of the present study. Four weeks of DEP exposure reduced spatial activity and olfactory sensitivity compared to those of the control group. Although spatial recognition memory was not definitively impaired, spatial mobilization was decreased in the 4-week DEP group. Such immobility is suggestive of depressive behavior [10][11][12] . These changes became normalized after 8 weeks of DEP exposure. However, weight loss was more significant in the 8-then 4-week DEP group (compared to the control group). In molecular terms, the PNN levels, as in our previous study 6 , and the levels of neurocan and tenascin C (in our current study) were increased after 8 weeks of DEP exposure. Moreover, the temporal cortex levels of the lectican-cleaving proteases MMP9 and MMP14 increased after 8 weeks of DEP exposure. Thus, both PNN degradation and synthesis may increase after

Increases in PNN component and MMP9/14 levels after 4-week DEP exposure. The levels of
PNN components and related molecules (neurocan, tenascin C, MMP9, and MMP14) increased after 4-week DEP inhalation, consistent with our previous finding that PNN density increased in an 8-week DEP group 6 .
MMP9 and MMP14 levels were elevated in the temporal cortex. The protein level of MMP9 were increased in both 4-week and 8-week DEP exposure groups, although the mRNA level of MMP9 showed significant increase only in the 8-week DEP exposure group in the present study. These discrepancies could be caused by the potential activation and secretion of MMP9 in protein levels, which preceded the up-regulation of MMP9 transcription. MMPs are secretory endopeptidases regulating PNN formation by cleaving PNN components. MMP9 is a secreted inducible enzyme that is activated both by free radicals and other enzymes or factors 16 . For instance, sorting nexin (SNX) 10 promoted the activation and secretion of MMP9, and the inhibition of SNX 10 reduced the MMP9 activities although the protein levels of MMP9 were increased 17 . In addition, the protein level of MMP9 in the western blotting may include both pro-and active-MMP9. To discriminate the active-MMP9, the secreted MMP9 measurement using enzyme linked immunosorbent assay or zymography assay will be needed 18 . A previous study using MMP9-deficient mice reported that MMP9 played a crucial role in PNN remodeling 19 . Extracellular matrix components, including CSPGs and other surface molecules, are substrates of MMP14 20,21 . MMP14 expression is upregulated in patients with neurodegenerative diseases such as Alzheimer's disease, multiple sclerosis and stroke, as well as neuroinflammatory diseases 22 . Therefore, the observed increases in MMP9 and MMP14 levels suggest possible neurological effects of 4-week DEP exposure; both synthetic and degradative changes in PNNs were apparent.
Temporal cortex neurocan and tenascin C levels increased since 4 weeks of DEP exposure. Neurocan is a chondroitin sulfate proteoglycan (CSPG), thus composed of PNNs. Tenascin C is an ECM glycoprotein that mediates changes in neuronal plasticity via several mechanisms 19,23 . The protein interacts with several CSPGs of PNNs 19 . In the adult central nervous system, tenascin C is expressed in regions of active neurogenesis and areas exhibiting neuronal plasticity 19,23,24 . The observed rise in tenascin C levels suggests that PNN synthesis is activated after 4-week DEP exposure. Such exposure may affect synaptic functions by changing PNN levels. Neurocan modulates the level of PNN structural constituents, actively remodels the synapses of GABAergic interneurons by binding to the immunoglobulin-class cell-adhesion molecule NCAM, and competitively inhibits the tyrosine kinase EphA3 25 . Increased levels of PNN components may upregulate the presynaptic glutamergic transporter. A recent, synaptic fluorescence imaging study showed that increased PNN levels in the cerebral cortex were accompanied by VGLUT2 upregulation 26 . However, VGAT did not modulate inhibitory synapse formation in interneurons, or the PNN levels thereof 27 . This may explain why no definitive change in VGAT expression level in the temporal cortex was evident in the current study.

Increases in the levels of excitatory and inhibitory vesicular transporters after 8-week DEP
exposure. The levels of both excitatory and inhibitory presynaptic transporters increased after 8 weeks of DEP exposure. VGLUT1 expression increased to a greater extent than VGLUT2 expression after such exposure. VLUGT1 is the predominant moiety of the corticocortical and corticothalamic projection systems 28 . VGLUT2 is the principal vesicular glutamate transporter of the subcortical and thalamocortical circuits 28 . Thus, the results suggest that DEP-induced changes occurred principally at the level of the cerebral cortex. Similarly, several previous studies reported elevations in the synaptic glutamate level and GABA transport after particle exposure 29,30 . One in vitro study found that PM-exposed microglia exhibited increased levels of extracellular glutamate, correlating with neuronal viability 30 . In addition, presynaptic glutamate and GABA accumulated in a dose-dependent manner following exposure to carbon dots 29 . Another study reported that the levels of mRNAs encoding the GABAB1 and GABAB2 receptors rose in the prefrontal cortex of rats exposed to cigarette smoke 31 . However, changes in synaptic transporter levels did not become significant before 8 weeks of DEP exposure in the present study, whereas depressive behavior and olfactory desensitization were apparent after 4 weeks of such exposure. Although the levels of vesicular synaptic transporters did not change, the observed neuroinflammatory alterations may have contributed to the behavioral changes evident after 4 weeks of DEP exposure 6 . In addition, the moderately high concentration of DEP (100 μg/m 3 ) used in the present study may be too low to trigger early changes in the levels of vesicular synaptic transporters; exposure to DEP for 8 weeks might be necessary to increase transporter levels. In the current study, cerebral cortex VGLUT1 and VGLUT2 levels increased after 8 weeks of DEP exposure. Such alterations in vesicular neurotransmission levels may have mediated recovery from depressive behavior triggered by such exposure. Previous studies found that cerebral cortical VGLUT1 and VGLUT2 levels www.nature.com/scientificreports www.nature.com/scientificreports/ increased as depression became relieved 13,14 . Similarly, VGAT expression was elevated in the prefrontal cortex and olfactory bulb after 8 weeks of DEP exposure. An increase in vesicular GABA transmission may have aided recovery of olfactory bulb olfaction 15 . However, weight loss increased with the duration of DEP exposure; such exposure may have been accompanied by stress-induced depressive behavior 32 .
In conclusion, 4-week DEP inhalation induced depressive behavior and decreased olfactory sensitivity, but recovery was evident after longer-term DEP exposure. Although the behavioral changes became normalized, MMP9 and MMP14 expression levels increased after 8-week DEP inhalation, perhaps explaining the prolonged degenerative and inflammatory changes noted in the cerebral cortex. The recovery of behavioral changes were related to increases in the levels of excitatory and inhibitory vesicular transporters and extracellular matrix components in the cerebral cortex.

Materials and Methods
Animal experiments. The study was approved by the Institutional Animal Care and Use Committee of Soonchunhyang University Medical School (SCHBC-Animal-2014-013). All methods adhered to the guidelines and regulations of the Institutional Animal Care and Use Committee of Soonchunhyang University Medical School. BALB/c mice (6-week-old females, n = 32) were divided into four groups (n = 8 for each group): (1) 4-week diesel-extracted particle (DEP) exposure, (2) 4-week control, (3) 8-week DEP exposure and (4) 8-week control. Mice were raised under standard conditions and exposed to DEP (NIST SRM 2975, Sigma-Aldrich) (5-10 μm in diameter) as described in a previous study 6,33 . DEP was autoclaved and coated with BAS to prevent aggregation. DEP was suspended in serum-free medium, nebulized, and delivered at 0.25 mL/min using a MEDI-PUMP (model #1125; Thomas Corporation, IL, USA). The 4-and 8-week DEP groups were exposed to 100 μg/m 3 DEP for 5 h per day on 5 days per week for 4 and 8 weeks respectively (Fig. 1). The control mice were exposed to saline solution for periods identical to those of the experimental groups. Fresh tissues from the prefrontal cortex, olfactory bulb and temporal cortex (n = 5 for each exposure group), as well as whole brain tissue for immunostaining (n = 3 for each exposure group), were harvested at the end of the experimental period (4 or 8 weeks). To minimize circadian effects, all tissue harvest experiments were performed at a consistent time, around 4 p.m.
Behavioral tests. Body weights were measured before (day 0) and every 7 days (i.e. days 7, 14, 21, 28, 35, 42, 49 and 56) after DEP exposures in all mice. The Y-maze test for spatial working memory and olfactory sensitivity test were conducted 4 weeks and 8 weeks after DEP exposure in the 4-week DEP/control group and 8-week DEP/ control group, respectively. www.nature.com/scientificreports www.nature.com/scientificreports/ Spatial working memory in the Y-maze. Animals were acclimated to the testing area for 30 min prior to the test. A black plastic Y-maze was used to differentiate the color of the maze and the mouse. The Y-maze consisted of three identical arms (3 cm wide, 32.5 cm long, 15 cm high, 120° apart) placed in the testing room. Each arm was named A, B and C (clockwise from the top). Individual mice were placed at the end of one of the arms facing the center and allowed to explore the maze freely for 5 min. The maze was cleaned after each session. The tests were video recorded and analyzed manually by two blinded investigators. Entries were counted when the animals passed through the midpoint of the arm. Entries into new arms were considered successful alternations, as opposed to entering the two previously visited arms. Alternation percentage was calculated as shown below. olfactory sensitivity test. Animals were habituated to the testing area for 30 min prior to the test. During the test and resting phases, animals were placed in a clean, empty cage; cages were cleaned after each test. Animals were introduced to distilled water as a control scent, 10% peanut butter solution as an attractive scent and 10% trichloroacetic acid (TCA) solution as an aversive scent. A 2 × 2 in square piece of filter paper was impregnated with each scent and placed in one side of the cage. Animals were exposed to each scent for 3 min and all sessions were video recorded. All animals went through the resting phase in another clean, empty cage for 6 min between sessions to neutralize the scent they were exposed to previously. The total number of sniffs were analyzed after the experiment.
Analysis of mRnA expression. Fresh olfactory bulb, prefrontal cortex, and temporal cortex tissues were micro-punched and rapidly frozen at −20 °C on dry ice. Using the TRI Reagent ® (Sigma-Aldrich), total RNA was extracted from mouse brain tissues according to the manufacturer's instructions. Reverse transcription was performed using TOPscript TM RT DryMIX (dT18 plus; Enzynomics Co. Ltd., Daejeon, South Korea) as in our previous study 6 . Table 1 lists the forward and reverse oligonucleotides used for PCR amplification of the markers of the PNNs (tenascin C, matrix metalloproteinase [MMP]14, MMP9 and a disintegrin and metalloproteinase with thrombospondin motifs 1 [ADAMTS1]) and the vesicular transporters of VGLUT1, VGLUT2, and VGAT. All mRNA levels were expressed as percentages of the mRNA level encoding glyceraldehyde 3-phosphate dehydrogenase (GAPDH). All RT-PCR reactions were replicated three times for each gene from five mice in each group.
Western blotting. Proteins were extracted from brain tissue and subjected to Western blotting as previously described 6 . Approximately 20 μg of protein was separated by 12% sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene difluoride (PVDF) membranes (Merck Millipore, Burlington, MA, USA). The membranes were soaked in blocking buffer (5% non-fat dry milk in Tris-buffered saline containing Tween-20 [TBS-T]) for 1 h at room temperature. They were then incubated with specific primary antibodies: anti-MMP9 (rabbit polyclonal; Abcam, Cambridge, UK), anti-MMP14 (rabbit monoclonal; Abcam) and β-actin (D6A8, rabbit mAb; Cell Signaling Technology). Immunoreactive proteins were detected with a horseradish peroxidase (HRP)-coupled secondary antibody (anti-rabbit IgG, HRP-linked antibody; Cell Signaling Technology) and visualized using an enhanced chemiluminescence (ECL) kit (Bio-Rad). The protein bands were quantitated by densitometry using ImageJ gel analysis software (National Institutes of Health, Bethesda, MD, USA).