Low‐level laser therapy 810‐nm up‐regulates macrophage secretion of neurotrophic factors via PKA‐CREB and promotes neuronal axon regeneration in vitro

Abstract Macrophages play key roles in the secondary injury stage of spinal cord injury (SCI). M1 macrophages occupy the lesion area and secrete high levels of inflammatory factors that hinder lesion repair, and M2 macrophages can secrete neurotrophic factors and promote axonal regeneration. The regulation of macrophage secretion after SCI is critical for injury repair. Low‐level laser therapy (810‐nm) (LLLT) can boost functional rehabilitation in rats after SCI; however, the mechanisms remain unclear. To explore this issue, we established an in vitro model of low‐level laser irradiation of M1 macrophages, and the effects of LLLT on M1 macrophage polarization and neurotrophic factor secretion and the related mechanisms were investigated. The results showed that LLLT irradiation decreased the expression of M1 macrophage‐specific markers, and increased the expression of M2 macrophage‐specific markers. Through forward and reverse experiments, we verified that LLLT can promote the secretion of various neurotrophic factors by activating the PKA‐CREB pathway in macrophages and finally promote the regeneration of axons. Accordingly, LLLT may be an effective therapeutic approach for SCI with clinical application prospects.

stage differs from patient to patient, greatly affecting the individual prognosis. 3 After SCI, both microglia and macrophages involved in the inflammatory response. 4 However, microglia cells only play a role in the initial stage when they are involved in the clearance of damaged nerve fragments. While the peripheral macrophages, which were differentiated from blood-derived monocytes massively recruited in the damage area, are the main causes of the further aggravation of the damage in the secondary injury stage. 5 Macrophages are the most important components of the physiological microenvironment at the injury site during the second stage and consequently play an important role in the progression and prognosis of SCI. 6 Recent studies have found that different local microenvironments can elicit different macrophage polarization states and functional phenotypes.
These phenotypes are typically understood to include the cytotoxic classically activated macrophages (CAMs), termed M1 macrophages, and the anti-inflammatory, nerve growth-promoting, alternatively activated macrophages, termed M2 macrophages. 7 Large numbers of macrophages enter the lesion in the immediate period after SCI and are activated to become the M1 type. Consequently, they play a major role, secreting high levels of proinflammatory factors and promoting necrosis and apoptosis. The appearance of the M2 macrophages is comparatively late and short-lived. 8 The specific polarization of macrophages after SCI is therefore not conducive to the repair and regeneration of axons.
Neurotrophic factors such as brain-derived neurotrophic factor (BDNF), nerve growth factor (NGF), ciliary neurotrophic factor (CNTF), and leukaemia inhibitory factor (LIF) comprise a family of proteins that promote the survival and growth of neurons. 9 Many studies have shown that stimulation of the ERK-cAMP response element-binding (CREB) or protein kinase A (PKA)-CREB pathway can promote the secretion of various neurotrophic factors. In contrast to the proinflammatory M1 macrophages, the M2 type is capable of secreting many types of neurotrophic factors. 10,11 This finding implies that modulation of the polarization state of the macrophages in the SCI lesion can be used to promote nerve repair, which has become a focus of research into SCI recovery. A number of approaches can be used to modify the polarization of macrophages and thereby increase the relative abundance of the M2 cell type, which would in turn increase the levels of secreted neurotrophic factors and effectively increase the growth of neuronal axons, ultimately promoting neuronal growth and repair.
Low-level laser therapy (LLLT) denotes the biological use of lowradiation laser sources and is applied to promote tissue repair as a noninvasive growth-promoting treatment. Furthermore, LLLT has been broadly investigated in the field of SCI treatment and recovery.
Relevant studies have shown that 810-nm LLLT can promote the recovery of SCI in rats, effectively boosting the recovery of the locomotor ability, reducing scar formation, significantly improving neuronal survival, modulating the inflammatory lesion microenvironment, suppressing the secretion of proinflammatory factors, and promoting the release of inflammation-resolving factors. 12,13 In addition, LLLT improves the polarization state of the macrophages within the lesion site, reducing the relative abundance of the cytotoxic M1 macrophages while also increasing the proportion of nerve-repairing M2 macrophages. 12 However, whether LLLT can improve the secretion of neurotrophic factors by macrophages or promote the regrowth of neuronal axons is presently unknown, as well as the mechanism underlying such effects. Starting from these questions, we developed the experiments described in this manuscript to investigate the modulatory effects of LLLT on the secretion of neurotrophic factors by M1 macrophages as well as the underlying molecular mechanisms.

| Ethical statement
The present study was approved by the Institutional Animal Care

| Animals
BALB/c mice (6-to 8-week-old females and 1-to 3-day-old male and female juveniles) of the same strain were obtained from the animal facility of the Fourth Military Medical University (Xi'an, China).

| Cultivation of dorsal root ganglion (DRG) neurons
Five juvenile BALB/c mice (1-3 days old) were individually sacrificed in 75% ethanol at −20°C. The DRGs were extracted and placed in Dulbecco's modified Eagle's medium DMEM/F12 culture medium cooled on ice. The DRGs were completely shredded and digested with 0.25% trypsin and 0.1% type-4 collagenase. followed by mixing and incubation at 37°C in a humidified atmosphere comprising 5% CO 2 for 30 minutes to dissolve the tissue and release single cells. The process was assessed every 5 minutes, with gentle shaking to improve the mixing of the tissue pieces and the digestion solution. A fivefold volume of DMEM/F12 medium with 20% FBS was added to the centrifuge tube and allowed to equilibrate for 5 minutes without agitation to stop the digestion. Then, the cells were centrifuged at 300 × g for 5 minutes and resuspended in neurobasal serum-free medium with B27 and 1% penicillin/streptomycin solution added. The cells were plated in 6-well plates at 1 × 10 4 . After 24 hours, the medium was exchanged again, and a solution comprising 5 µmol/L cytarabine (MedChemExpress, Monmouth Junction, NJ), B27, 1% penicillin/streptomycin solution, and Neurobasal serum-free medium was added. This medium was exchanged for normal culture medium after another 48 h.

| Macrophage cultivation
Six-to eight-week-old female BALB/c mice were killed by administration of an anaesthetic overdose (0.5 mL/10 g 5% chloral hydrate) and placed in a 70% ethanol solution for 10 min. The complete femur and tibia were removed and the metaphyses of these bones were opened. The filtrate was repeatedly agitated to obtain a well-mixed suspension. A 200 screen mesh (YA0949; Solarbio, China) was used to filter the obtained liquid, which was collected in a 15-mL centrifuge tube, to which a threefold volume of red-blood-cell lysis solution was added and incubated at 37°C for 10 min to obtain complete lysis of erythrocytes. The resulting mixture in the 15-mL centrifuge tube was centrifuged at 300 × g for 5 minutes to discard the supernatant. The pellet containing the cells was gently resuspended in preconditioned DMEM medium with 10% FBS, 1% penicillin/streptomycin solution and 10 ng/mL macrophage colony-stimulating factor (M-CSF; Sino Biological Inc, Beijing, China). The cell suspension was seeded into 35-mm-diameter culture plates at 1 × 10 6 , and the medium was changed every 2 days for 7 days until the cell cultures matured.

| Flow cytometry
Macrophages received medium with the addition of 100 ng/mL LPS and 20 ng/mL IFN-γ for 48 h. After cell maturation, 0.25% trypsin with 0.02% EDTA (Gibco, Grand Island, USA) was added, and 12well were plates into the incubators for 10 minutes during digestion.
DMEM with 10% FBS was added to terminate digestion, the cells were placed into a centrifuge at 300g × 5 minutes, and the super-

| Cultivation of DRG neurons in macrophageconditioned medium
Forty-eight hours after irradiation, the culture medium was removed from all groups and centrifuged at 10,000 × g for 10 minutes to remove cells and debris, yielding the conditioned medium, which was added to the culture dishes with purified DRG neurons. After another 48 hours, cell staining was conducted.

| RT-qPCR
Twenty-four hours after the LLLT, the OMEGA RNA extraction kit (BioLegend, San Diego, CA) was used according to the manufacturer's protocol to extract the total RNA. The RNA obtained was reverse-transcribed using PrimeScript RT Master Mix (TaKaRa Bio Inc, Kusatsu, Japan) according to the manufacturer's protocol. The beta-actin housekeeping gene was used for normalization of gene expression by parallel amplification. The relative expression levels of the target mRNA were calculated. PCR primer sequences are listed in Table 1.

| Western blot analysis
Forty-eight hours after LLLT irradiation, the medium was removed from all the experimental groups, and the cells were washed with PBS, followed by the addition of 100-150 µL RIPA buffer with phosphatase inhibitor and digestion for 5 minutes at 4°C. A cell scraper was used to completely harvest the protein from the bottom of the culture plate, which was transferred to a 1.5-mL centrifuge tube and centrifuged at 4°C and 12 000 rpm for 5 min.
The DNA in the centrifuge tube was discarded.

| Immunofluorescence staining
The cells were transferred into immunofluorescence plates (SB-Shifix20; Shikhar Biotech, Lalitpur, Nepal); washed with PBS three times, 5 minutes each; permeabilized using 0.3% Triton X-100 for a total of 15 minutes; and blocked with 1% FBS at room temperature

| ELISA
Forty-eight hours after LLLT irradiation, the culture supernatants were collected, centrifuged to remove any floating cells and prepared as samples for ELISA, together with BDNF, NGF, CNTF and LIF standards, according to the instructions of the ELISA kits (BOSTER, Wuhan, China). After completion of the reaction, the absorbance of the samples and standards was measured at 450 nm using a TA B L E 1 Primers used in this study Target gene/ primer name Primer sequence

| Data analysis
Bar charts were rendered using GraphPad Prism 7.0 (GraphPad Software). Axon length analysis was conducted using ImageJ software. Statistical analysis was conducted using SPSS 19.0 (IBM Corp.). Each experiment was independently repeated five times, with the mean values used as the results. The data are expressed as the mean ± standard error of the mean. The results were initially tested for normality using the F test. Parametric data were compared using one-way analysis of variance (ANOVA).
Nonparametric data were compared using Tukey's test. Sidak's test was used to compare each group with the corresponding group that received the inhibitor. P < .05 was considered to indicate statistical significance.

| Identification of macrophages in vitro
The results from flow cytometry showed that the positive ratio of F4/80 was 99.8% (region B), indicating that macrophages were cultured successfully ( Figure 1A,B).

| Identification of M1-type macrophages in vitro
The results from flow cytometry showed that after addition of LPS and IFN-γ, the ratio of CD86 + , F4/80 + M1-type macrophages increased from 1.8% to 91.7% (C4 quadrant), indicating successful polarization of macrophages to M1-type macrophages ( Figure 1C,D).

| The influence of LLLT on M1 macrophages in vitro
Compared with that of the M1 group, the iNOS expression was extremely low in the CON group and markedly reduced in the M1+LLLT group (both P < .001) ( Figure 3A,B).
The ARG-1 content of the CON group was higher than that of the M1 group, but the difference was not significant. Further comparison revealed that ARG-1 expression was significantly increased in the M1+LLLT group compared with the M1 group (P < .01) ( Figure 3C,D).

| Influence of conditioned medium from M1 macrophages irradiated by LLLT on the axonal growth of DRG neurons
The M1+LLLT group had significantly better neuronal growth than the M1 group (P < .05). Compared with that of the CON group, the length of the DRG axons was increased in the M1+LLLT group, while the length was shorter in the M1 group, but the differences were not significant (Table 2, Figure 4A,B).

| The influence of LLLT on the expression of neurotrophic factors by M1 macrophages
The mRNA levels of BDNF, NGF and CNTF were all significantly reduced in the M1 group compared with the CON group (P < .01). LIF F I G U R E 2 A, Immunofluorescence was used to identify the polarization of M1-type macrophages. F4/80 is a macrophage marker, and iNOS is an M1 macrophage marker. B, The ratio of F4/80 + iNOS + cells in the M1 group was significantly improved compared with the CON group. **P < .01

| The influence of LLLT on neurotrophic secretion signalling by M1 macrophages
The expression of PKA was significantly higher in the CON group than in the M1 group (P < .05). The expression of PKA was also significantly higher in the M1+LLLT group than in the M1 group (P < .05) ( Figure 7A,B). The levels of CREB were somewhat higher in both the CON and M1+LLLT groups than the M1 group, but the differences were not significant ( Figure 7C,D). Further analysis of the expression level of p-CREB revealed that CREB phosphorylation was significantly higher in the CON group than in the M1 group (P < .05).
Furthermore, the p-CREB levels were higher in the M1+LLLT group than in the M1 group (P < .01). There was no significant difference in the p-CREB levels between the M1+LLLT and CON groups ( Figure 7E,F).
Additionally, the expression levels of ERK and p-ERK were investigated. The results did not indicate any significant difference in the levels of ERK between the two groups ( Figure S1).

| Effect of the PKA inhibitor H-89 on the LLLTinduced secretion of neurotrophic factors
The secretion of neurotrophic factors was significantly higher in the CON group than in the CON+in group (BDNF, P < .001; NGF, P < .01; CNTF, P < .001; LIF, P < .05). Moreover, the secretion of all four neurotrophic factors was significantly higher in the M1+LLLT group than in the M1+LLLT+in group (BDNF, P < .01; NGF, P < .05; CNTF, P < .001; LIF, P < .001). However, there were no significant differences in the secretion of BDNF, NGF, CNTF or LIF between the M1 and M1+in groups ( Table 3, Figure 8A-D).
The p-CREB levels of each group were compared with those of the corresponding group that received inhibitor. There were no significant differences between the M1 and M1+in groups. However, the p-CREB levels were significantly higher in the CON group than in the CON+in group (P < .01), whereas those of the M1+LLLT group were higher than those of the M1+LLLT+in group (P < .05) ( Figure 8E,F).

| Effect of the PKA inhibitor H-89 on the LLLTinduced secretion of neurotrophic factors
The secretion of neurotrophic factors was significantly higher in the CON group than in the CON+in group (BDNF, P < .001; NGF, P < .01; CNTF, P < .001; LIF, P < .05). Moreover, the secretion of all four neurotrophic factors was significantly higher in the M1+LLLT group than in the M1+LLLT+in group (BDNF, P < .01; NGF, P < .05; CNTF, P < .001; LIF, P < .001). However, there were no significant differences in the secretion of BDNF, NGF, CNTF or LIF between the M1 and M1+in groups ( Table 3, Figure 8A-D).
The p-CREB levels of each group were compared with those of the corresponding group that received inhibitor. There were no significant differences between the M1 and M1+in groups.
However, the p-CREB levels were significantly higher in the CON group than in the CON+in group (P < .01), whereas those of the

| Effect of the PKA inhibitor H-89 on the axonal growth effect elicited by LLLT
Axonal growth was significantly greater in the CON group than in the CON+in group (P < .01), whereas that of the M1+LLLT group was also greater than that of the M1+LLLT+in group (P < .01) (Table 3).
However, there were no significant differences between the M1 and M1+in groups (Figure 9).

| D ISCUSS I ON
After SCI, the formation of the lesion produces a physiological microenvironment that inhibits neuronal survival, blocks axon regrowth and leads to irreversible damage to the spinal cord. 14 Macrophages play key roles in both time and space in the physiological microenvironment. 15  In contrast, this study used LLLT, which is a widely clinically used noninvasive therapy that can suppress inflammation and promote recovery. 20 Early studies have already discovered that LLLT can increase the production of IL-4 and IL-10 within the lesion site, down-regulate the relative abundance of M1 macrophages and increase the number of M2 macrophages. 12 To confirm the direct effect of LLLT on the regulation of M1 macrophages, we have innovatively established an in vitro LLLT-M1 macrophage irradiation model. LPS and IFN-γ were used to obtain M1 macrophages, which predominate in the SCI pathology environment.
The M1 macrophages were irradiated, and different irradiation parameters (0-10 J) were compared. The results showed that 4 J F I G U R E 6 Forty-eight hours after irradiation, ELISA was used to investigate the secretion quantities of BDNF, NGF, CNTF and LIF. The secretion quantities of the four investigated neurotrophic factors were increased in the M1+LLLT group compared with the M1 group. A, Secretion quantity of BDNF. B, Secretion quantity of NGF. C, Secretion quantity of CNTF. D, Secretion quantity of LIF. *P < .05, **P < .01, ***P < .001 compared with the M1 group (ANOVA)  There is abundant literature on the pathways involved in the secretion of neurotrophic factors, and CREB phosphorylation has been shown to have a key function in many of these processes.

TA B L E 3 Secretion of neurotrophic factors
F I G U R E 9 A, Immunofluorescence staining was used to analyse the length of axons, with βIII-tubulin as an axonal marker and NeuN as a neuronal body marker. B, The axon length of each group was compared according to inhibitor presence. *P < .05, **P < .01, ***P < .001

A B
Consequently, promotion of the phosphorylation of CREB can boost the levels of many neurotrophic factors within the nervous system, including BDNF, NGF, CNTF and LIF. CREB phosphorylation is regulated by proteins from many upstream pathways, and changes in PKA, ERK and similar proteins can promote CREB phosphorylation and consequently influence the production of neurotrophic factors. Jana et al 26

| CON CLUS ION
Through this study, we verified the hypothesis that LLLT can promote spinal cord repair and functional rehabilitation in rats with SCI.
LLLT can inhibit the macrophages polarized to M1-type, promote the expression of neurotrophic factors by M1 macrophages via PKA/ CREB signalling and then promote the axonal growth of DRG neurons. This study adds further details on the mechanism of action of LLLT in SCI from a completely new point of view, laying a foundation for its eventual clinical application.

ACK N OWLED G EM ENTS
This work was supported by the National Natural Scientific

CO N FLI C T O F I NTE R E S T
The authors declared that they have no conflicts of interest to this work.