Flow cytometry analysis of immune and glial cells in a trigeminal neuralgia rat model

Microvascular compression of the trigeminal root entry zone (TREZ) is the main cause of most primary trigeminal neuralgia (TN), change of glial plasticity was previously studied in the TREZ of TN rat model induced by chronic compression. To better understand the role of astrocytes and immune cells in the TREZ, different cell markers including glial fibrillary acidic protein (GFAP), complement C3, S100A10, CD45, CD11b, glutamate-aspartate transporter (GLAST), Iba-1 and TMEM119 were used in the TN rat model by immunohistochemistry and flow cytometry. On the post operation day 28, GFAP/C3-positive A1 astrocytes and GFAP/S100A10-positive A2 astrocytes were activated in the TREZ after compression injury, there were no statistical differences in the ratios of A1/A2 astrocytes between the sham and TN groups. There was no significant difference in Iba-1-positive cells between the two groups. The ratios of infiltrating lymphocytes (CD45+CD11b−) (p = 0.0075) and infiltrating macrophages (CD45highCD11b+) (p = 0.0388) were significantly higher than those of the sham group. In conclusion, different subtypes A1/A2 astrocytes in the TREZ were activated after compression injury, infiltrating macrophages and lymphocytes increased, these neuroimmune cells in the TREZ may participate in the pathogenesis of TN rat model.

Our previous studies observed that chronic compression of the TREZ induced local plasticity changes in a variety of glial cells and protrusions of reactive astrocytes extending from the CNS side to the PNS side 4 , but we did not distinguish microglia from infiltrating macrophages or different subtypes of astrocytes. Astrocytic glutamate-aspartate transporter (GLAST) is a well-known marker of mature astrocytes and complement C3 and S100 calcium-binding protein A10 (S100A10) are also identified as markers of A1 astrocytes and A2 astrocytes, respectively [8][9][10] . Furthermore, microglia and infiltrating macrophages in the nervous system have some similar lineage characteristics, including origin, proliferation and the expression of myeloid cell markers; therefore, it was difficult to distinguish the microglia and infiltrating macrophages in the TREZ by Iba-1 immunostaining. It was reported that detecting the differential expression of CD11b and CD45 in microglia and macrophages by flow cytometry was a good way to distinguish them 11,12 .
Therefore, different cell markers were used in this study to further clarify the categories of microglia, infiltrating macrophages, lymphocytes and astrocytes activated in the TREZ after compression injury in the TN rat model by immunohistochemistry and flow cytometry, which provide an experimental basis for further studying the pathogenesis of TN.

Materials and methods
Animal. Healthy  TN rat model of chronic compression in TREZ. The experimental animals were randomly divided into a sham operation group and a TN group (n = 4-5 for each group) with chronic compression in the TREZ. The TN animal model was established according to our previously described procedure 13,14 . A surgical incision was made at the upper edge of the right orbit of the rat. Then, the infraorbital fissure was exposed along the operative approach of the medial wall of the orbit, and a small plastic filament was carefully inserted into the intracalvarium from the inferior orbital fissure to compress the TREZ. In the sham group, the right infraorbital nerve of rats was only surgically exposed, without TREZ compression injury.
Measurement of the orofacial mechanical pain threshold. von Frey filaments were used to measure the orofacial mechanical pain threshold of the rats. The right orofacial area was stimulated by a von Frey filament 3 times with an interval of 10 s. The filament was bent during stimulation, and the strength of the filament was recorded when positive reactions of orofacial pain behaviors were observed. The mechanical pain thresholds of rats in each group were recorded 3 days before surgery to obtain baseline data and were measured again on postoperative day (POD) 28 to evaluate the mechanical allodynia of the TN animal model. was used to stain the nuclei. The immunohistochemical sections were photographed and analyzed with a Leica laser confocal microscope (Leica TCS SP8, Heidelberger, Germany). Image-Pro Plus software (version 6.0; Media Cybernetics, MD, USA) also helped to quantify and graph positive labelling in a specific cell type (i.e. C3 expression in GFAP + astrocytes) as area fraction, which is the percentage of double positive labelling in the cell type of interest within the imaged area, as previously described 15,16 . Flow cytometric analysis. TREZ  www.nature.com/scientificreports/ Biotec, Bergisch Gladbach, Germany) were used to obtain single cells. Thereafter, 1 µL of Golgi transport protein blocker GolgiPlug Protein Trnsp Inhb (BD Pharmingen, NJ, USA) was added to the 10 6 /mL cell culture medium. Then, the cells were cultured at 37 ℃ in a 5% CO 2 incubator for 6 h. Fixable Viability Stain 780 (BD Pharmingen, NJ, USA) was used to mark dead cells.

Immunohistochemistry. Rats in both groups
In each tube, to 100 μL of liquid, 2 µL of FC receptor blocker CD32 (550270, BD Pharmingen, NJ, USA) was added and incubated at room temperature for 5 min in the dark, followed by the addition of 2 µL of flow cytometry antibodies CD45-PE-Cy7 and CD11b-V450 (561588 and 562108, BD Pharmingen, NJ, USA), GLAST (130-118-344, Miltenyi Biotec, Bergisch Gladbach, Germany) and incubating at room temperature for 15 min in the dark. After incubation, the suspended cells were centrifuged at 300 × g for 5 min, and then the supernatant was discarded.
A volume of 100 µL of fixation/permeabilization solution (BD Pharmingen, NJ, USA) was added to each tube and incubated at 4 °C for 20 min in the dark. Then, 1 ml of 1 × Perm/Wash Buffer (BD Pharmingen, NJ, USA) was added, and the cells were washed twice by centrifugation at 300 × g for 5 min. After resuspension in 100 µL of 1 × Perm/Wash buffer, the cells were incubated with 0.2 μg/mL anti-S100A10-Alexa Fluor 647 antibody (FAB2377R, R&D Systems, MN, USA) and 0.1 μg/mL anti-C3 antibody (ab17456, Abcam, Cambridge, UK) at room temperature for 30 min. Then, the cells were centrifuged at 400 × g for 5 min and resuspended in ice-cold PBS. Fluorescent dye-labeled secondary antibody (A11001, Thermo, MA, USA) was added and incubated for 20 min at room temperature. Finally, the cells were washed and transferred to Trucount™ Absolute Counting Tubes for flow cytometric analysis. Data were acquired on a BD LSR Fortessa X-20 (BD Biosciences, NJ, USA) and analyzed with FlowJo V10 (BD Pharmingen, NJ, USA).

Statistical analysis.
All data are presented as the mean ± standard error of the mean (SEM). Data analysis of the flow cytometry plots was performed using FlowJo V10 (BD Pharmingen, NJ, USA). Data were analyzed using SPSS 19.0 statistical software (IBM, NK, USA) and GraphPad Prism 8 software (GraphPad Software, CA, USA). Statistical differences between groups of rats were calculated using two-way ANOVA with Sidak's multiple comparisons tests and Student's t-tests. p < 0.05 was considered statistically significant.

Results
Chronic compression of the TREZ induced orofacial mechanical hyperalgesia in TN rats. There was no significant difference in the baseline pain threshold between the sham group and the TN group. However, the mechanical stimulation hyperalgesia threshold of the TN group was significantly lower than that of the sham group on POD 28 (Fig. 1), indicating that the TN animal model was successfully established. Different types of astrocytes were activated in the TREZ after compression injury. On POD 28 of chronic compression, GFAP-positive astrocytes on the CNS side were significantly activated, thick protrusions of astrocytes extended obviously from the CNS to the PNS side, and GFAP expression had increased (Fig. 2). There were GFAP/C3-positive A1 astrocytes and GFAP/S100A10-positive A2 astrocytes on the CNS side of the TREZ region. More interestingly, some GFAP-positive reactive astrocytes expressed both C3 and Figure 1. Orofacial mechanical stimulation threshold changes in the TN rat model. There was no significant difference in the hyperalgesia threshold of the right orofacial mechanical stimulation between the TN group and the sham group (n = 4 for each group) before surgery. On the 28th day after the operation, the mechanical stimulation threshold in the sham group returned to the preoperative baseline threshold, while the mechanical stimulation threshold of rats in the TN group was significantly lower than that of the sham group and at baseline. ****p < 0.0001 as determined by two-way ANOVA with Sidak's multiple comparisons tests. All data show the mean ± SEM. www.nature.com/scientificreports/ S100A10 ( Fig. 2 and Fig. S1). And there was significant larger expression area of the C3 (p = 0.0136) and obvious smaller expression area of the S100A10 (p = 0.0304) in GFAP positive astrocytes in TREZ after chronic compression. The ratio of GFAP/C3-positive area and GFAP/S100A10-positive area was clearly increased in TN group compared with sham group (p = 0.0174) (Fig. 3). However, there was very little positive expression of Ki67 in the TREZ region (Fig. 4). According to the result of flow cytometry, there were no statistical differences in the ratios of total astrocytes (CD45 − GLAST + ) (p = 0.05968), A1 astrocytes (CD45 − GLAST + C3 + ) (p = 0.1599) and A2 astrocytes (CD45 − GLAST + S100A10 + ) (p = 0.7788) between the two groups (Fig. 5).

Different expression of immune cells in the TREZ after compression injury.
According to the immunohistochemical staining results, there was no significant difference in Iba-1-positive cells between the    www.nature.com/scientificreports/ two groups on POD 28, and the cell marker TMEM119 was present not only in Iba-1-positive cells but also some other cells (Fig. 6).
The survival rates of all cells extracted from both groups were calculated to be above 95% on POD 28 by flow cytometry. The relative cell size and the cell granularity in the two groups were not measurably different. Quantification analysis was performed by collecting cells with a Trucount™ absolute counter, and the absolute number of cells per milligram of tissue was calculated. According to the total cell number of sham group (289,182 ± 1166) vs. TN group (227,194 ± 24,227) (p = 0.00443), and percentage of sham group (73.42 ± 1.044%) vs. TN group (49.82 ± 5.43%) of cells per milligram of tissue (p = 0.0034), the total cell number and percentage in TN group were significantly less than those in the sham group (Fig. 7). www.nature.com/scientificreports/ The proportions and numbers of CD11b + cells in the TREZ and TG in the two groups were not significantly different (p = 0.3327) (Fig. 8). The ratios of infiltrating lymphocytes (CD45 + CD11b − ) (p = 0.0075) and infiltrating macrophages (CD45 high CD11b + ) (p = 0.0388) were significantly higher than those of the sham group. There was no significant difference in the ratio and number of microglia (CD45 low CD11b + ) between the two groups (p = 0.5304) (Fig. 8).

Discussion
In this study, the orofacial mechanical pain threshold tested by von Frey filaments is to confirm the successful establishment of TN animal model, and the histological analyses have been well studied in our previous researches 4, 13,17 . We consider that it is one kind of the relatively stable and mature TN animal models induced by mechanical compression on the TREZ.
Based on the immunohistochemical staining results, chronic compression of TREZ induced an increase in GFAP expression in astrocytes, and the protrusions of astrocytes became thicker and obviously extended to the peripheral side ( Fig. 2 and Fig. S1). The co-staining of the GFAP/C3-positive A1-astrocytes obviously activated after chronic compression injury in TN group, while the expression of GFAP/S100A10-positive A2-astrocytes activation was relatively low. Interestingly, some triple-labeled GFAP/C3/S100A10-positive cells www.nature.com/scientificreports/ were also observed in the TREZ of the TN rat model, while few Ki67-positive cells were observed (Fig. 4). Flow cytometry analysis also showed that the ratios of A1 astrocytes (CD45 − GLAST + C3 − ) and A2 astrocytes (CD45 − GLAST + S100A10 + ) did not change significantly. GLAST antibody is a specific marker which is widely used in astrocytes flow cytometry and sorting, as the previous literature described 18 . Besides, we focus on the activation of astrocytes on post operation day 28 in the TN animal model, the inflammation in the rats may be largely alleviated at this time. Therefore, we consider that the isolation efficacy of astrocytes by using GLAST in our study might not affect the results obviously. We considered that morphological and functional changes, rather than the proliferation of astrocytes, may be more susceptible to the influence of compression injury on the TREZ of TN rats. Flow cytometry was used to distinguish clearly and effectively between microglia (CD45 low CD11b − ) and macrophages (CD45 high CD11b − ) in our study. The ratios of infiltrating macrophages (CD45 high CD11b + ) and lymphocytes (CD45 + CD11b − ) increased, while the ratio of microglia (CD45 low CD11b + ) remained unchanged, indicating that infiltrating immune cells were more common than proliferating innate immune cells in the TREZ and TG in the process of chronic compression injury. Infiltrating macrophages (CD45 high CD11b + ) and The ratios of CD11b + cells in the TREZ and TG of the TN group were lower than those of the sham group. (c-e) The ratios of macrophages (CD45 high CD11b + ) and lymphocytes (CD45 + CD11b − ) in the TREZ and TG of the TN group were higher than those in the sham group. The difference in the proportions and numbers of microglia (CD45 low CD11b + ) between the two groups was not statistically significant. *p < 0.05, **p < 0.01, ns: not significant, as determined by two-tailed Student's t-test. SSC side scatter. www.nature.com/scientificreports/ lymphocytes (CD45 + CD11b − ) may play important roles in the occurrence and development of TN induced by chronic compression. CD11b is commonly used to identify cells in myeloid lineage, including neutrophils, monocytes, macrophages and microglia. Although the ratio of total CD11b-positive cells in the TREZ was reduced a little in our study, while there was no significant difference between sham group and TN group (Fig. 8). We think it may be the reason that compression injury on TREZ leads to the decrease of absolute cell numbers in nerve tissue samples obtained. Furthermore, chronic mechanical compression on the TREZ in TN model could also undermine the integrity of blood-nerve-barrier inside the trigeminal nerve causing the macrophages infiltrate to the parenchyma from blood-nerve-barrier, which may increase the ratio of macrophages.
Recent studies have shown that the complex interconnections between glial cells and immune cells play a crucial role in the maintenance of CNS homeostasis, inflammation, and neuropathic pain 19 . Astrocytes reversibly regulate the demyelination and remyelination of myelinated nerve fibers during nerve injury by secreting different cytokines 20 . CCL2 and CXCL10, which are produced by reactive astrocytes, can recruit lymphocytes in the peripheral blood to the CNS, and astrocytes provide nutritional support for these infiltrating immune cells 21 . Infiltrating macrophages and lymphocytes may migrate through the broken blood-nerve barrier in chronic compression nerve injury with the action of certain chemokines and participate in the occurrence and development of TN induced by chronic compression.
Activated microglia dominates the early glial response to peripheral nerve injury and participate in the induction and development of neuropathic pain, while astrocytes contribute to the persist of chronic pain 22,23 . In our previous study, Iba1-positive microglia/macrophages increased on POD 7 and POD 14 in the TREZ of TN rats, and these activated cells decreased to POD 21 and returned to a low level on POD 28 4 . Therefore, the results of the change in cell number by flow cytometry showed no change in microglia on POD 28, which is also consistent with our previous study.
However, our experiments also had some limitations. Although TMEM119 has been identified as a surface marker of microglia that can be used to reliably distinguish microglia from infiltrating macrophages 24 , recent studies also reported that the expression of TMEM119 is not restricted to microglia in inflammation 25,26 . In our study, there were few Iba1/TMEM119-positive cells in the TREZ, so we still could not well distinguish between microglia and macrophages through the two cell markers by immunohistochemistry.
In summary, chronic compression of the TREZ induced orofacial mechanical hyperalgesia in TN rats, A1-astrocytes rather than A2-astrocytes obviously activated after chronic compression injury in TN group, while infiltrating macrophages and lymphocytes increased, which indicating the neuroimmune cells in the TREZ may participate in the pathogenesis of TN rat model.