Green light exposure elicits anti-inflammation, endogenous opioid release and dampens synaptic potentiation to relieve post-surgical pain

Abstract

Light therapy improves multiple conditions such as seasonal affective disorders, circadian rhythm dysregulations, and neurodegenerative diseases. However, little is known about its potential benefits in pain management. While current pharmacologic methods are effective in many cases, the associated side effects can limit their use. Non-pharmacological methods would minimize drug dependence, facilitating a reduction of the opioid burden. Green light therapy has been shown to be effective in reducing chronic pain in humans and rodents. However, its underlying mechanisms remain incompletely defined. In this study, we demonstrate that green light exposure reduced post-surgical hypersensitivity in rats. Moreover, this therapy potentiated the antinociceptive effects of morphine and ibuprofen on mechanical allodynia in male rats. Importantly, in female rats, GLED potentiated the antinociceptive effects of morphine but did not affect that of ibuprofen. We showed that green light increases endogenous opioid levels while lessening synaptic plasticity and neuroinflammation. Importantly, this study reveals new insights into how light exposure can affect neuroinflammation and plasticity in both genders. Clinical translation of these results could provide patients with improved pain control and decrease opioid consumption. Given the noninvasive nature of green light, this innovative therapy would be readily implementable in hospitals.

Graphical abstract:

green light exposure (GLED) modulates three different systems that are interconnected, providing a reduced inflammatory environment that can act on synaptic transmission. Capitalizing on several mechanisms to control pain may have a better outcome for postoperative pain, therefore reducing drug reliance.

Introduction

Approximately 310 million major surgeries are performed annually worldwide ³³. Chronic pain patients present greater challenges and require more opioids for postoperative pain control ²⁴. Furthermore, patients with uncontrolled postoperative pain may experience increased hospital stays with physiological and psychological comorbidities ³⁹. Therefore, adequate and effective postoperative pain control methods are essential for surgical planning.

There are several contributors to postoperative pain severity ³⁴. For example, preexisting pain and chronic opioid use increase postoperative pain ⁵. The nature of surgical procedures and altered afferent input also result in hypersensitivity to tactile stimuli ³⁴. Finally, injuries promote the release of inflammatory mediators that decrease sensory thresholds, promote spontaneous activity, and enlarge the nociceptors’ receptive fields ^{7, 8, 13, 15, 101}.

Opioids are prescribed for postoperative pain ³⁹. However, many side effects are associated even with short-term opioid use, such as addiction, sedation, nausea and respiratory depression ¹¹. These side effects are even more pronounced in the elderly, leading to decreased mobility, reduced independence, and potential morbidity ⁶². Collectively, this contributes to the frailty of the elderly ⁶⁹. Therefore, a new strategy is needed to decrease postoperative pain.

Previous studies have shown that different light colors alter nociception and provide significant pain relief in rodents and humans ^{22, 44, 58, 61, 72, 74, 78, 79}. A pre-clinical study in rodents suggests that exposing rats to 8 hours a day for 5 days with green light-emitting diode (GLED) induces analgesia in acute and chronic pain models ⁵⁸. Results demonstrated an increase in endogenous opioid mRNAs, though the mechanisms remain unknown. On the other hand, exposing rats to red light-emitting diodes induces thermal and mechanical hypersensitivity in a rodent injury-free model ⁶¹. Subsequent studies confirmed the stimulation of the endogenous opioid system following green light exposure; the GLED effect requires spinal expression of both μ- and δ-opioid receptors ⁷³.

Neuroinflammation has become the subject of numerous studies, from pain to neurodegeneration ^{20, 49}, and there is an intricate relationship between the inflammatory and endogenous opioid systems ^{70, 97, 104}. Indeed, both regulate each other; on the one hand, opioids can activate glial cells by direct action on μ-opioid receptor and TLR4 ¹⁰⁴. On the other hand, cytokines are potent regulators of μ-opioid receptors and modulate their expression: Interleukin-6 (IL-6) and tumor necrosis factor-alpha (TNFα) up-regulate the receptor, while interferon-gamma (IFNγ) induces its down-regulation ⁶³. In order to delve deeper into the mechanisms of pain and its potential treatments, we now need to consider pain as a multisystemic process. Inflammatory neuromodulators secreted during acute stages of neuroinflammation sensitize both nociceptors and peripheral glial cells ^{9, 17, 35}. Microglia became of crucial interest as they regulate both central inflammation and synaptic transmission. Indeed, microglial mediators such as TNFα or IL-10 are powerful modulators of pain signaling through neuron-glia interactions ²⁰. A main feature of the microglial response concerns their rapid morphological changes; upon activation, they will gradually acquire an amoeboid shape characterized by the retraction of their processes ⁴⁸ Although the role of microglia has been well characterized in chronic pain, their involvement in acute pain conditions still needs to be characterized ²⁰. Importantly, microglia modulate synaptic activity to change pain behavior within tens of minutes ¹², and the microglial release of IL-1β enhances glutamatergic activity in the spinal cord ¹⁰². Furthermore, NMDA receptor-dependent synaptic plasticity in the spinal cord contributes to enhanced sensory responses following injury ¹⁰⁶. In rodents, TNFα increases the activity of N-methyl-D-aspartate (NMDA) and α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptors, which contribute to increased surgical pain ^{42, 81}. These aspects demonstrate how neuroinflammation modulates pain processing through the glutamatergic system and subsequent neuronal transmission.

Here, we examine GLED’s mechanisms and assess its benefits in managing postoperative pain. Our hypothesis is that GLED attenuates postoperative thermal and mechanical hypersensitivity in rats by acting on the endogenous opioid system and modulating inflammatory and plasticity processes. Our previous studies demonstrate GLED’s potential in managing chronic pain. However, as little is known about dose-effect properties, we first conducted behavioral experiments to evaluate exposure time-dependent outcomes. As GLED stimulates the endogenous opioid system, we analyzed its potential drug-sparing properties. Following the characterization of the antinociceptive effects of GLED, we investigated the possible mechanisms by analyzing multiple systems involved in pain processing. We first analyzed endogenous opioid levels and receptor expression in the spinal cord using immunosorbent assays and western blots. Then, we evaluated the impact of GLED on neuroinflammation, using 3D-modeling approaches to analyze the morphology of microglia, an indicator of their activation status. Finally, as microglia modulate synaptic plasticity, we studied the activation and synaptic expression of glutamatergic receptors in the dorsal horn of the spinal cord.

Methods

Animals

Specific pathogen-free, Sprague–Dawley rats (male weight at testing 300-600g, female weight at testing 200-350g; Envigo, Indianapolis, IN) or C57B16/J mice (male weight at testing 25-29g, The Jackson Laboratory, Bar Harbor, ME) were housed in climate-controlled rooms on a 12-h light/dark cycle and were allowed to have food and water ad libitum. All procedures were approved by the University of Arizona Animal Care and Use Committee and conform to the guidelines for the use of laboratory animals of the National Institutes of Health (Public Health Service Policy on Humane Care and Use of Laboratory Animals, 2015 revision; and Guide for the Care and Use of Laboratory Animals, eight edition, 2011). Males and females were tested in different rooms, with specific equipment for each sex. A total of 303 males and 101 females (388 rats and 16 mice) were used for this study.

Blinding protocols

For behavioral studies, animals were exposed to light in an isolated room. Only MD was allowed to enter the room. Behavioral testing was conducted in a separate room by LFM or KC. MD/MK brought the cages 20 to 30 minutes before behavioral testing. LFM never approached animals while they were exposed. After data quantification, results were sent to SMW, who unblinded the study.

For immunohistochemistry and ELISA, neither SMW nor LFM were in contact with the animals while they were exposed. MD/MK brought numbered rats from the exposure room to the dissecting room. LFM and SMW collected samples, unaware of the conditions. After ELISA analysis and modeling of microglia, results were sent to KC, who unblinded the study.

Drug administration

For subcutaneous injections, saline solution or two different morphine sulfate doses were injected (1 mg/kg or 5 mg/kg, diluted in 0.9% saline). The doses of medications chosen were based on previously published data in animals ^{50, 56, 77}.

For oral administration, rats were physically restrained, and the feeding tubes (FTP-13-90) were inserted to the left part of the mouth. The doses are described in Table 1.

Table 1.

Drugs used in the study

Drug	Source	Dose	Route
Morphine	Sigma M8777	10 mg.kg−1	Oral
		5 mg.kg−1	Subcutaneous
		1 mg.kg−1	Subcutaneous
Ibuprofen	Sigma I4883	100 mg.kg−1	Oral
Gabapentin	Sigma G154	100 mg.kg−1	Oral
Acetaminophen	Sigma A3035	200 mg.kg−1	Oral

Surgery

To induce acute post-surgical pain, animals were subjected to an incisional surgery based on the procedure previously described by Brennan et al.¹⁴. Briefly, animals were anesthetized with 2% isoflurane in O₂ anesthesia (total time under anesthesia was < 20 minutes). After scrubbing the left hind paw with chlorhexidine and 70% ethanol, an incision was performed through the skin and fascia from the heel toward the toes on the plantar aspect of the left hind paw, exposing the underlying muscle. The flexor muscle was then elevated and longitudinally incised, leaving the muscle origin and insertion intact. After hemostasis with gentle pressure, the skin was closed with two mattress sutures using 5-0 nylon on a curved needle.

Light-emitting diodes (LED)

All visible spectrum LED flex strips were purchased from ledsupply.com (VT, USA). The specifications of the LEDs were: (i) #LS-AC50-GR-006, 525-nanometer wavelength (i.e., green), 8 Watts, 120 Volts, 120-degree beam angle; and (ii) #LS-AC50-WW-006, white, 9.6 Watts, 120 Volts, 120-degree beam angle. Our own evaluation of green light specifications with a spectrometer (Supp. Fig. 1) revealed a center wavelength of 515 nm, with a Full-Width Half Max span of 35 nm. This indicates a reduction in the intensity of 50% at approximately 497 nm and 533 nm. The measurements for the white LED strip showed a peak intensity wavelength of approximately 605 nm and a Full-Width Half Max span of approximately 90 nm. This indicates a reduction in the intensity of 50% at approximately 560 nm and 650 nm.

For exposure, LED strips were affixed on the top of racks where animals were housed, allowing global diffusion of light. Animals were exposed to the LED in these cages with full access to food and water in a dark room devoid of any other source of light. Following behavioral assessment, the animals were returned to their cages for additional LED exposure. At the end of daily testing, the animals were returned to their regular animal room where they were exposed to room light illuminated with Sylvania Octron 3500K F032/835 model which is 48” in length and power output of 32 Watt florescent bulbs producing intensity of 750 Lux. A Lux meter (Tondaj LX1010B, Amazon.com) was used to determine the illuminance and luminous emittance of the LED strips. We chose different times and intensities of exposure to characterize the antinociceptive effect of GLED (Fig. 1 and Supp. Fig. 2). For our main experiments, we opted for an exposure of 4 days, based on our own observations. As peak pain behavior is observed in the 2 days following surgery ¹⁴, we opted to cover this period with green light exposure. Hence, our study design for all experiments in our main manuscript corresponds to an exposure of 2 days before and 2 days after surgery for a total of 4 days of green light exposure. Animals were exposed to green or white light 8 hours per day from 7 am to 3 pm, after which they were returned to the vivarium facility which to maintain a 12h/12h light:dark cycle.

Figure 1. Exposure to green LED attenuates thermal and mechanical hypersensitivity in a model of acute post-surgical pain.

(A) Experimental design for thermal and mechanical hypersensitivity assessment, depending on the time of exposure. Male rats were exposed to green LED (GLED, λ=525 nm; 0, 2, or 4 days, 8h/day, 100 Lux) or white LED (WLED; 4 days, 8h/day, 100 Lux) prior to paw incision surgery (Sx), and then GLED or WLED for 2 days post-surgery. (B) Following measurement of baseline (BL), thermal hypersensitivity was assessed before (Presurg) and two days after surgery (n=6-10, two-way ANOVA followed by Tukey’s posthoc test). GLED attenuated thermal hypersensitivity (F_{(3, 96)}=9.744, p<0.0001). (C) Area under the curve analyses (from day 1 to 2) for figure 1B, corresponding to paw withdrawal latencies among different durations of GLED and WLED exposure (n=6-10, Shapiro-Wilk test followed by Kruskal Wallis and Dunn’s posthoc tests). Attenuation of thermal hypersensitivity following exposure to GLED depends on exposure duration. (D) Mechanical hypersensitivity was assessed before light exposure, before surgery, and for 7 days after surgery (n=6-10, two-way ANOVA followed by Tukey’s posthoc test). GLED attenuated mechanical hypersensitivity (F_{(3, 234)}=41.68, p<0.0001). (E) Area under the curve analyses (from day 1 to 7) for figure 1D, corresponding to paw withdrawal threshold among different durations of GLED exposure and WLED exposure (n=6-10, Shapiro-Wilk test followed by Kruskal Wallis and Dunn’s posthoc tests). Attenuation of mechanical hypersensitivity following exposure to GLED depends on exposure duration. (F) Experimental design for mechanical hypersensitivity assessment, with different gaps prior to surgery. Rats were exposed to either 4 days of WLED or GLED before surgery, with a 2- or 4-day gap between exposure termination and surgery. (G) Mechanical hypersensitivity was assessed before surgery and for 4 days after surgery (n=8, two-way ANOVA, F_{(2, 105)}=23.55, p<0.0001). (H) Area under the curve analyses (from day 1 to 4) for figure 1G, representing paw withdrawal threshold among different post-exposure gaps before surgery (n=8, Shapiro-Wilk test followed by Kruskal Wallis and Dunn’s posthoc tests). GLED exposure resulted in a long-lasting antinociceptive effect. Each panel corresponds to different sets of animals, except for panel B&C, D&E and G&H. All results represent mean±SEM with the p-values obtained from statistical analysis above the graphs.

Thermal sensoiy thresholds

Paw withdrawal latencies were determined as described by Hargreaves et al. ⁴⁷. Rats were acclimated within Plexiglas enclosures on a clear Plexiglass plate maintained at room temperature. A radiant heat source (high-intensity projector lamp) was focused onto the plantar surface of the hind paw. When the paw was withdrawn, a motion detector halted the stimulus and a timer. A maximal cutoff of 33.5 sec was used to prevent tissue damage.

Tactile thresholds

The assessment of tactile sensory thresholds was determined by measuring the withdrawal response to probing the hind paw with a series of calibrated fine (von Frey) filaments. Each filament was applied perpendicularly to the plantar surface of the paw of animals held in suspended wire mesh cages. Primary hyperalgesia was evaluated according to previously described methods ¹⁰³. The withdrawal threshold was determined by sequentially increasing and decreasing the stimulus strength (the “up and down” method). Data were analyzed with the nonparametric method of Dixon, as described by Chaplan and colleagues ¹⁹, and expressed as the mean withdrawal threshold.

Generation of spinal cord lysates

Spinal cords were harvested by hydraulic extrusion from Sprague-Dawley male rats, and tissue lysates were generated by homogenization and trituration in lysis buffer (50 mM Tris-HCl, 150 mM NaCl, 1% Triton-X-100, 0.5% sodium deoxycholate,1 mM EDTA, 0.1% SDS, pH 7.4). To isolate ipsilateral dorsal horns, the lumbar segment (L4-L6) of the spinal cord was hemisected prior to homogenization. Lysis buffer was supplemented with protease inhibitor cocktail (Bimake, B14002), phosphatase inhibitors (Bimake, B15002), and Pierce universal nuclease (Fisher Scientific, PI88701). Protein concentrations were determined using BCA protein assay (ThermoFisher Scientific, PI23225).

Synaptic fractionation

Male rats were killed by isoflurane overdose, followed by decapitation. Lumbar spinal cords were collected by hydraulic extrusion, and the ipsilateral part of the lumbar dorsal horn was then dissected. The synaptic fractionation protocol was adapted from a previously described method ²⁸. In brief, samples were homogenized in ice-cold buffer (sucrose 0.32 M, HEPES 10 mM, pH 7.4) and then centrifuged at 1000 g for 10 minutes to remove nuclei and large debris. The remaining supernatants were centrifuged at 12,000 g for 20 minutes to obtain a crude membrane fraction. Pellets were resuspended in hypotonic EDTA buffer (4mM HEPES, 1mM EDTA, pH 7.4) to chelate calcium and again centrifuged at 12,000 g for 20 minutes to pellet the synaptosomal fractions. Synaptosomes were then incubated in a low-triton buffer (20 mM HEPES, 100 mM NaCl, 0.5% Triton X, pH 7.2) for 15 minutes on ice and centrifuged at 12,000×g for 20 minutes at 4°C. The supernatant contained the presynaptic membrane fraction, referred to as the triton-soluble fraction. The resulting pellets were further extracted with a high-detergent buffer (20 mM HEPES, 0.15 mM NaCl, 1% triton X100, 1% deoxycholic acid, 1% SDS, pH 7.5) for 1 hour and then centrifuged for 15 minutes at 10,000 g to obtain the postsynaptic density fraction remaining in the supernatant. The enrichment quality of each fraction was assessed by immunoblotting for PSD95, a postsynaptic-specific structure, and synaptophysin, which is not found in the postsynaptic element. All buffers were supplemented with nuclease, protease, and phosphatase inhibitor cocktails. BCA protein assay was used to analyze protein concentrations.

ELISA assays

Rat ELISA assays were purchased from MyBioSource (San Diego, CA) to measure serum, cerebrospinal fluid (CSF), and spinal cord lysate levels of endogenous opioids (β-endorphin, #MBS452166; Proenkephalin, #MBS726498 and Dynorphin, #MBS720677). ELISA assays for IL-10 were purchased from RayBiotech (#ELR-IL10-1) and for TNFα from Invitrogen (#KRC3011). Procedures were conducted according to the manufacturers’ instructions. Colorimetric detection was based on H2O2/TMB reaction. To determine the optical density of each well, a Biotek Epoch microplate reader was set to 450 nm. Determination of endogenous opioid and cytokine levels involved triplicate determinations for each sample. For lysate samples, each concentration was normalized to the weight of the different spinal cords after collection.

Western blotting

For detection of AMPA and NMDA subunits, CamKII, Src family kinases, protein kinase C gamma (PKCγ), and their respective phosphorylated forms, 10 μg (PSD/Non-PSD fractions) to 40 μg (total lysate) of proteins were loaded into wells of 3-8% gradient SDS-PAGE gels (Criterion XT, Biorad). Gels were transferred to 00.25 PVDF membranes after activation in 100% methanol. Membranes were immunoblotted with antibodies described in supplementary Table 1 after a blocking step for 45 minutes in Tris-buffered saline, 0.1% Tween, 5% milk or BSA, pH 7.6. Membranes were then incubated overnight with antibodies diluted in Tris-buffered saline, 0.1% Tween, pH 7.6, with 2% milk for total protein or 2% BSA for phosphorylated protein detection. Goat anti-rabbit HRP or Goat anti-mouse HRP (1/10000 or 1/20000, Jackson Immunoresearch; room temperature, 2h) were used as secondary antibodies. Protein bands were detected with Azure Sapphire Biomolecular Imager (Azure Biosystems) after applying a chemiluminescent reagent for 2 minutes (ThermoFisher Scientific). Bands were quantified densitometrically with Image J software (National Institutes of Health). Detection of PSD95 and Synaptophysin proteins was used as a control for fractionation protocol, and actin protein levels were quantified in total lysates for loading control.

Immunohistochemistry

Animal perfusion and spinal cord collection: to collect spinal cords from Sprague-Dawley male rats, animals were anesthetized with 1-5% isoflurane delivered in O₂ and then perfused with phosphate-buffered saline (PBS) and 4% paraformaldehyde (PFA). A single 4 cm lateral incision was made through the integument and the abdominal wall just beneath the rib cage. A cut was then made through the rib cage up to the clavicle. A 20-gauge needle was inserted through the left ventricle into ascending aorta. Approximately 60 mL of PBS and 180 mL of 4% PFA were injected through the heart at a steady flow rate. After fixation, rats were decapitated, and lumbar spinal cords were collected by hydraulic extrusion. Spinal cords were postfixed in 4% PFA for 1-2 hours and then immersed in 30% sucrose in PBS for 24 hours. Spinal cords were then flash-frozen into Tissue-Tek and stored at −80°C. 15 μm frozen sections were sliced using Cryostar NX50.

Microglia immunostainings: briefly, slices were permeabilized in 0.3% Triton-X-100 with 1% BSA diluted in PBS, and proteins of interest were stained overnight at 4°C in the same buffer with antibodies directed against Iba1 (Supplementary Table 1). After primary probing, slices were rinsed and incubated for 2 hours at room temperature with corresponding fluorescent conjugated secondary antibodies (Jackson Immunoresearch). A DAPI counterstaining was performed after secondary antibody incubation, and slices were mounted with FluorSave reagent (Millipore #345789).

Microglia image acquisition and 3D modeling

Images were obtained utilizing a Zeiss LSM 880 confocal microscope. Identical acquisition parameters were applied for all conditions. Each slide was randomly encoded to undergo a full analytical process in a blinded manner. Images were obtained with a Plan-Apochromat 63x/1.40 Oil objective. All planes of the same image were acquired with a 400 nm Z step in the laminae I and II from the dorsal horns of male rats that underwent paw incision surgery (ipsilateral region). The experimenter was blinded to the treatment conditions and selected one to two microglia per slice, as long as their processes and morphology were not damaged by freezing and cutting steps. Laser power and gain were adapted to obtain the best signal without saturation. Microglia morphology was analyzed with Imaris software (Bitplane). Briefly, microglial processes were selected and modeled with a filament tracer. Cell somas were selected and modeled with the Volume tool. All thresholds were automatically selected by the software. A total of 55 microglia from 6 independent experiments were analyzed (26 males, 29 females). All parameters for each microglia were automatically measured by the software and then exported to an Excel file for further analysis. Branch orders correspond to the level of process ramification: branches of the first order correspond to the processes between somas and first ramification points, whereas branches of the tenth order correspond to the processes between the ninth and tenth ramification points. To reflect microglial complexity, we analyzed the percentages of each branch order and compared them between our two conditions.

Statistical analysis

Unless stated otherwise, all data were expressed as mean ± standard error of the mean (SEM). Statistical analysis was run using GraphPad Prism software 8.0 (San Diego, CA). All data were first tested for Gaussian distribution and heteroscedasticity, respectively, using Shapiro-Wilk and Bartlett’s tests. The statistical significance of differences between the means or the medians was determined by parametric and non-parametric analysis followed by post-hoc comparisons. Non-parametric tests (Mann-Whitney or Kruskal-Wallis) were used when data distribution was not normal or homoscedastic. Differences were considered significant if the probability value was p≤0.05. No outlier data were removed. All area under the curve (AUC) analyses correspond to the integral of the function representing the data between the x values, from post-incision to end of testing. All data were plotted using Prism software. Supplemental Table 2 describes all the statistical tests that were used in the figures, with the corresponding p-values.

Data Statement

The data that support the findings of this study are available from the corresponding author, Mohab Ibrahim, upon request.

Results

The long-lasting analgesic effect of GLED is dependent on exposure duration and proximity to surgery

Given that analgesia was more pronounced when we exposed rats to 100 Lux compared to 4 Lux (Supp. Fig. 2), we decided to use 100 Lux intensity exposure for testing the behavioral and biological effects of white and green LED on rats. To identify the exposure protocol that provides the best results, we first exposed rats that had undergone incisional surgery of the hind paw to different durations of pre- and post-surgical GLED exposure (Fig. 1A). Data demonstrated that the antinociceptive effects of GLED on both thermal and mechanical hypersensitivity are dependent on the duration of GLED exposure (Fig. 1B, C, D and E). Furthermore, 4 days of GLED exposure that ended 2 days prior to surgery was more efficacious than 4 days of GLED exposure that ended 4 days prior to surgery (Fig. 1F, G and H). As peak pain behavior is observed in the 2 days following surgery ¹⁴, we opted to cover this period with green light exposure. Based on these results (Fig. 1 & Supp. Fig. 2), our main study design corresponds to an exposure of 2 days before and 2 days after surgery for a total of 4 days of green light exposure. Importantly, our results were reproduced in male mice, where GLED reduced post-surgical hypersensitivity after 3 days of exposure, indicating that these findings are not species-specific (Supp. Fig. 3).

GLED potentiates morphine reversal of postoperative thermal and mechanical hypersensitivity in rats

Opioids are one of the most commonly prescribed medications for postoperative pain ³⁹, and our previous study revealed an increase in the endogenous opioid release after exposure to green light ^{58, 73}. Hence, we sought to investigate their potential synergistic effect when used in combination. As nociceptive signal processing is sexually dimorphic ^{4, 21, 90}, we conducted our main experiment in both sexes. Male rats received a short exposure to GLED or white light-emitting diodes (WLED) for 2 days after surgery (Fig. 2A). We opted for this sub-therapeutic exposure as 2 days of GLED exposure did not result in any significant changes in pain thresholds compared to WLED-exposed animals in our first experiment. After the 2-day light exposure (two days after surgery), rats were orally administered with saline, morphine (10mg/kg), Ibuprofen (100mg/kg), Acetaminophen (200mg/kg) or Gabapentin (100mg/kg). Mechanical hypersensitivity was assessed over 2h and then over 5 days after drug administration.

Figure 2. Synergistic effect of green light exposure (GLED) with different types of common drugs used for pain management in male rats.

(A) Experimental design for hypersensitivity assessment (Von Frey filaments) to examine potential synergy between GLED and different drugs used for pain management. Following paw incision surgery, male rats were exposed to white LED (WLED) or green LED (GLED), 100 Lux 8h/day over two days. Saline, Morphine (10 mg/kg), Ibuprofen (100mg/kg), Acetaminophen (200mg/kg), or Gabapentin (100mg/kg) were administered orally on day 2 after surgery. (B) Mechanical hypersensitivity was assessed before surgery (pre), before injections (post), over two hours post-injections, and daily until 5 days post-surgery (n=6-8, two-way ANOVA, followed by Sidak’s posthoc test, *p<0.05). Two-way ANOVAs revealed a significant effect of the treatment for morphine (F_(1,42)= 10.60, p=0.0022, n=8) and ibuprofen injections (F_(1,42)=11.40, p=0.0016, n=8). (C) Area under the curve (from baseline post-surgery to 2h) analyses for paw withdrawal thresholds presented in figure 2B, among saline- and drug-injected animals exposed to either WLED or GLED (n=6-8, Shapiro-Wilk test followed by Mann Whitney test). GLED increased mechanical thresholds in rats injected with morphine and ibuprofen. Panels B and C corresponds to the same set of animals. All results represent mean±SEM with the p-values obtained from statistical analysis above the graphs.

In saline-injected groups, GLED exposure did not result in any significant effect, demonstrating that 2 days of exposure is not enough to induce antinociception in this acute postoperative pain model. When GLED was administered in combination with morphine, this “subthreshold dose” of GLED potentiated morphine-induced antinociceptive effect in males (Fig. 2B&C). Interestingly, GLED only potentiated the antinociceptive effect of ibuprofen in males but not in females (Fig. 3B&C), suggesting that different underlying mechanisms between sexes are engaged. Finally, exposure to GLED did not affect acetaminophen and gabapentin treatments.

Figure 3. Synergistic effect of green light exposure (GLED) with different types of common drugs used for pain management in female rats.

(A) Experimental design for hypersensitivity assessment (Von Frey filaments), to examine potential synergy between GLED and different drugs used for pain management. Following paw incision surgery, female rats were exposed to white LED (WLED) or green LED (GLED), 100 Lux 8h/day over two days. Saline, Morphine (10 mg/kg), Ibuprofen (100mg/kg), Acetaminophen (200mg/kg), or Gabapentin (100mg/kg) were administered orally on day 2 after surgery. (B) Mechanical hypersensitivity was assessed before surgery (pre), before injections (post), over two hours post-injections, and daily until 5 days post-surgery (n=5-8, two-way ANOVA, followed by Sidak’s posthoc test, **p<0.01). Two-way ANOVAs revealed a significant effect of the morphine treatment (F_(1,42)=7.733, p=0.0081, n=8). (C) Area under the curve (from baseline post-surgery to 2h) analyses for paw withdrawal thresholds presented in figure 3B, among saline- and drug-injected animals exposed to either WLED or GLED. GLED increased mechanical thresholds in rats injected with morphine (n=8, Shapiro-Wilk and Bartlett tests followed by unpaired T-test). Panels B and C correspond to the same set of animals. All results represent mean±SEM with the p-values obtained from statistical analysis above the graphs.

To confirm the potential of GLED to reduce opioid requirements, we designed another experiment comparing the effect of light exposure on two different doses of morphine (low: 1mg/kg, and high: 5mg/kg, subcutaneous injections). In this experiment, rats were exposed to GLED over 4 days, and we compared the efficacy of each treatment or combination of treatments on thermal and mechanical hypersensitivity (Fig. 4A). In saline-injected groups, GLED-exposed animals exhibited significantly higher post-surgery paw withdrawal thresholds from Days 1 to 5 (Fig. 4B&C). Thermal sensitivity in the GLED group was significantly lower on days 1 to 2 when compared to that of the WLED group (Fig. 4D&E). These results confirm that GLED reduces hypersensitivity in rats that had received paw incision surgery.

Figure 4. Green light-emitting diode (GLED) potentiates morphine reversal of postoperative thermal and mechanical hypersensitivity in male rats.

(A) Experimental design for mechanical and thermal hypersensitivity assessment to examine potential synergy between GLED and morphine reversal of postoperative hypersensitivity. Following measurement of baseline (BL), male rats were exposed to white led (WLED), or green LED (GLED), 100 Lux 8h/day, for two days before paw incision surgery (Sx) and two days after surgery. Saline or morphine solution (1 or 5 mg/kg) was injected subcutaneously 24h after surgery. (B) Mechanical hypersensitivity was assessed before light exposure, before morphine/saline injection (inj), every 0.5h after injection until 2h, and daily until 6 days post-surgery (n=6, two-way ANOVA, followed by Tukey’s posthoc). GLED exposure resulted in potentiation of morphine reversal of postoperative mechanical hypersensitivity. (C) Area under the curve (from BL to Day 6) analyses for paw withdrawal thresholds presented in figure 4B, among saline- and morphine-injected animals exposed to either WLED or GLED (n=6, Mann Whitney - saline conditions - or Kruskal Wallis test followed by Dunn’s posthoc test - morphine conditions). (D) Thermal hypersensitivity was assessed before light exposure, before morphine/saline injection, every 0.5h after injection until 2h, and daily until 3 days post-surgery (n=7-9, two-way ANOVA, followed by Tukey’s posthoc). GLED exposure resulted in potentiation of morphine reversal of postoperative thermal hypersensitivity. (E) Area under the curve (from BL to Day 6) analyses for paw withdrawal latencies presented in figure 4D, among saline- and morphine-injected animals exposed to either WLED or GLED (n=7-9, Mann Whitney - saline conditions - or Shapiro-Wilk and Bartlett tests followed by one-way ANOVA and Tukey’s posthoc tests - morphine conditions). Panels B-C and D-E correspond to different groups of animals. All results represent mean±SEM with the p-values obtained from statistical analysis above the graphs.

In morphine-injected animals, thermal and mechanical sensitivities were reduced compared to their saline-injected counterparts (Fig. 4). Notably, combined GLED exposure and injection of 1mg/kg of morphine had a similar or greater reversal of mechanical hypersensitivity when compared to that of WLED + 5mg/kg morphine. These data collectively show that GLED potentiates the antinociceptive effect of morphine and that a low dose of morphine used in conjunction with GLED can achieve similar or greater levels of antinociception than using a higher dose of morphine by itself.

GLED exposure increases endogenous opioid levels but does not modulate the expression of their receptors

Naloxone was shown to reverse the antinociceptive effects of GLED. Given that GLED exposure increases endogenous opioid levels in chronic pain models ^{58, 73}, we wanted to investigate the effect of GLED exposure following surgery on endogenous opioid levels. Thus, we analyzed endogenous opioid levels and the expression of their receptors in male rats. After surgery, 4 days of GLED exposure did not affect the levels of endogenous opioids in serum fractions (Fig. 5A, B and C), but resulted in increased central nervous system (CNS) levels of both β-endorphin and Proenkephalin, without affecting dynorphin levels (Fig. 5D, E and F). This suggests that GLED-induced antinociception is dependent on central rather than peripheral mechanisms. Rats exposed to GLED also did not display any differences in expression of μ- (MOR), δ- (DOR), and κ-opioid receptors (KOR) (Fig. 5G, H, I and J). These data suggest that attenuation of hypersensitivity by GLED exposure is due, at least in part, to endogenous opioid release.

Figure 5. Endogenous opioid levels and receptor expression in male rats after exposure to white (WLED) or green LED (GLED).

Male rats were exposed to green LED (GLED; 2 days; 8h/day, 100 Lux) or white LED (WLED; 2 days; 8h/day, 100 Lux) prior to paw incision surgery, and then GLED or WLED for 2 days post-surgery. Blood and spinal cords were collected at the end of the second day post-surgery. Levels of β-endorphin were analyzed through ELISA assays in both serum (A) and spinal cord lysates (B), as well as Proenkephalin (C and D) and Dynorphin (E and F) (n=5, Mann-Whitney test). GLED exposure resulted in increased levels of β-endorphin and Proenkephalin in spinal cord lysates, but not in the serum. (G) Representative western blots of μ-, δ-, and κ-opioid receptor (OR) expression in the dorsal horn of the lumbar spinal cord after exposure to WLED or GLED. The histogram represents the quantification of μ-OR (H), δ-OR (I), and κ-OR (J) expression (n=8). No differences in their expression were observed between WLED and GLED conditions. Panels A-F and G-J correspond to different sets of animals. All results represent mean±SEM with the p-values obtained from statistical analysis above the graphs.

GLED reduces neuroinflammation in rat spinal cord

Surgeries typically increase the levels of TNFα ⁵⁹, which shift the balance of cytokines to promote a pro-inflammatory status ¹⁵. Additionally, IL-10 has analgesic effects in chronic pain conditions, including inflammatory and neuropathic pain ¹⁰. To further elucidate the mechanisms of GLED in reversing postoperative hypersensitivity, we investigated levels of inflammatory cytokines in the serum and CSF of male rats. Animals received paw incision surgery and were exposed to either GLED or WLED. Animals exposed to GLED either had 0 or 4 days of pre-surgery exposure, as well as 2 days post-surgery exposure (Fig. 6A). At the end of exposure, blood and CSF were collected to quantify anti-inflammatory IL-10 and pro-inflammatory TNFα by ELISA. No significant differences in IL-10 levels were found in the serum between groups (Fig. 6B). Importantly, we were not able to detect TNFα in the serum with our ELISA assays. Conversely, CSF IL-10 levels in the 6-day GLED group were significantly higher than that of the 6-day WLED group (Fig. 6C). Furthermore, CSF TNFα levels in the 6-day GLED group were significantly lower than that of the 6-day WLED group (Fig. 6D). Overall, these results show that a 6-day GLED exposure decreases levels of TNFα and increases levels of IL-10 in the CSF of male rats but not in the serum. In conjunction with increased levels of endogenous opioids, specifically in the CSF, these results further demonstrate the effects of GLED exposure on the central nervous system.

Figure 6. Inflammation-involved cytokine levels after exposing male rats to GLED.

(A) Experimental design to analyze cytokine levels in serum and cerebrospinal fluid (CSF). Male rats were exposed to green LED (GLED; 0 or 4 days) or white LED (WLED; 4 days), 8h/day 100 Lux, prior to paw incision surgery (Sx), and then GLED or WLED for 2 days post-surgery. Blood and cerebrospinal fluid (CSF) were collected at the end of the second day post-surgery. (B) ELISA assay results representing the average IL-10 concentration in serum samples (n=6-11, Kruskal Wallis test, followed by Dunn’s posthoc). (C) ELISA assay results representing the average IL-10 concentration in CSF samples (n=5-6, Kruskal Wallis test, followed by Dunn’s posthoc). (D) ELISA assay results representing the average TNFα concentration in CSF samples (n=5-6, mean±SEM, Kruskal Wallis test, followed by Dunn’s posthoc). GLED exposure resulted in increased levels of IL-10 and decreased levels of TNFα only in CSF. Panels B and C-D corresponds to different groups of animals. All results represent mean±SEM with the p-values obtained from statistical analysis above the graphs.

Microglia are one of the main sources of TNFα in the central nervous system ⁴⁶ and regulate central inflammation ⁹². Furthermore, microglia play an essential role in regulating inflammation related to pain ²⁰. Given their crucial role in pain, we sought to investigate if GLED exposure had any effects on microglial activation in the spinal cord that could be contributing to the induced antinociception. To initially evaluate any changes in microglial activation, total lysates of lumbar dorsal horns were analyzed by western blot for expression of Iba1, a marker of microglial activation ^{53, 54}. Iba1 expression in the post-surgery WLED group was significantly higher than that of the pre-surgery control group, indicating that paw incision surgery resulted in microglial activation (Fig. 7A). However, when comparing Iba1 expression between post-surgery GLED and pre-surgery control groups, no significant differences were found, indicating that GLED exposure may attenuate microglial activation by paw incision surgery (Fig. 7A).

Figure 7. GLED exposure decreases microglial activation in male rats after surgery.

Analysis of neuroinflammation status before and after exposure to white (WLED) or green (GLED) LED. Male rats were exposed 2 days before and after paw incision surgery. (A) Spinal cords were collected before exposure (Preop. Control) and after surgery in WLED and GLED groups (Postop. White and Postop. Green, respectively). Total lysates from the dorsal horn of the lumbar spinal cord were then analyzed in western blot to evaluate Iba1 expression (n=4-5, Kruskal Wallis, followed by Dunn’s posthoc). Surgeries resulted in increased Iba1 expression when rats were exposed to WLED but not GLED. (B) Representative images of microglia after exposure to WLED or GLED and their respective 3D modeling. Scale bar=10μm. (C) After modeling, microglia complexity was assessed by quantifying the number of branches of each order. Branches from the first order correspond to segments starting from the soma. Branches from the second-order correspond to segments after the first ramification. Results illustrate the distribution of branches in WLED and GLED conditions at the end of exposure (n=13 microglia, two-way ANOVA, F_{(10, 264)}=91.87, p<0.0001, followed by Sidak’s posthoc). (D) Microglia soma volume after exposure to WLED or GLED (n=13 microglia per condition, mean±SEM, Shapiro-Wilk and Bartlett tests followed by Mann Whitney test). (E) Microglia process length after exposure to WLED or GLED (n=13 microglia per condition, Shapiro-Wilk and Bartlett tests followed by unpaired T-test). GLED exposure resulted in increased microglial complexity and smaller soma volume. Panels C to E correspond to the same set of animals. All results represent mean±SEM with the p-values obtained from statistical analysis above the graphs.

Microglia morphology reflects their activation states ²⁷. As Iba1 expression is decreased in the dorsal horn of the spinal cord in males exposed to GLED, we sought to further investigate the effects of GLED exposure on microglial morphology by 3D morphometric analyses. After image acquisition of microglia from male rats, 3D modeling was performed to analyze the complexity of microglial process branching (Fig. 7B). Microglial complexity was assessed by analyzing the distribution of their processes in relation to branching points. First-order branches referred to process segments between somas and first ramifications. Seventh-order branches referred to process segments between the 6^th and 7^th branching points. This distribution analysis allowed us to have a clear visualization of microglia complexity and arborization. Exposure to GLED resulted in a less ramified arborization of microglia (Fig. 7C). Furthermore, soma volumes were significantly lower than that of the WLED group (Fig. 7D), and the total length of processes was significantly higher in the GLED group (Fig. 7E). These changes in microglia morphology show that GLED exposure results in reduced microglial activation in male rats.

Microglia involvement in pain processing seems to be sex-dependent ⁹⁰. Furthermore, GLED potentiated the effect of ibuprofen only in males, suggesting that GLED elicits sexually dimorphic mechanisms. In order to identify the mechanisms underlying such differences, we also evaluated microglia activation in females. Analysis of microglial complexity reveals that GLED also decreased microglial activation in females (Fig. 8). Importantly, female microglia were characterized by a more complex arborization and longer processes compared to males. Altogether, these results confirmed that exposure to GLED decreases neuroinflammation in both male and female rats subjected to incisional surgery.

Figure 8. GLED exposure decreases microglial activation in female rats after surgery.

Analysis of neuroinflammation status before and after exposure to white (WLED) or green (GLED) LED. Female rats were exposed 2 days before and after paw incision surgery. (A) Representative images of microglia after exposure to WLED or GLED and their respective 3D modeling. Scale bar=15μm. (B) After modeling, microglia complexity was assessed by quantifying the number of branches of each order. Branches from the first order correspond to segments starting from the soma. Branches from the second-order correspond to segments after the first ramification. Results illustrate the distribution of branches in WLED and GLED conditions, at the end of exposure (n=13 microglia, two-way ANOVA, F_{(17, 486)}=45.86, p<0.0001, followed by Sidak’s posthoc). (C) Microglia soma volume after exposure to WLED or GLED (n=14-15 microglia per condition, Mann Whitney test). (D) Microglia process length after exposure to WLED or GLED (n=14-15 microglia per condition, Shapiro-Wilk and Bartlett tests followed by unpaired T-test). GLED exposure resulted in increased microglial complexity and smaller soma volume. Panels B to D correspond to the same set of animals. All results represent mean±SEM with the p-values obtained from statistical analysis above the graphs.

GLED modulates postsynaptic expression of glutamatergic receptors in male rats

GLED exposure resulted in increased levels of endogenous opioids, specifically in the CNS, suggesting that GLED-induced antinociception may rely more on central than on peripheral mechanisms. Furthermore, GLED induced long-lasting antinociception, suggesting a potential effect on synaptic plasticity ⁵⁸. Thus, we evaluated the effect of GLED exposure on plasticity by analyzing the expression and activation of glutamatergic receptors in the dorsal horn of spinal cords collected from rats that underwent paw incision surgery. Importantly, AMPA receptors containing GluR1, GluR2, and GluR3 subunits are largely expressed in the dorsal horn, and NR2B-containing receptors dominate synaptic responses in the adult spinal cord ^{51, 85}. Dorsal horns of lumbar spinal cords were collected from male rats after a 4-day exposure to WLED or GLED. These samples were then lysed and fractionated to obtain pre- and post-synaptic fractions. In total lysates, no significant differences were found in the expression of GluR1 (Fig. 9A&B). However, phosphorylation of GluR1 serine 831 was significantly decreased, whereas no effects were observed on serine 845. CamKII phosphorylation was not affected (Fig. 9C), but analysis of PKC phosphorylation on threonine 514 revealed a decrease activation of the kinase (Fig. 9D), which is involved in the phosphorylation of GluR1 subunit on serine 831. Regarding NMDA receptors, expression of NR1 subunit remained unchanged following GLED exposure (Fig. 9E). But western blot analysis of NR2B phosphorylation demonstrated a decreased phosphorylation of the NR2B subunit on tyrosine 1472 (Fig. 9F). These results reveal that, after performing surgery on male rats, GLED exposure was able to decrease the activation of both NMDA and AMPA receptors, key receptors involved in plasticity modulation in the brain. Many kinases regulate synaptic expression of glutamatergic receptors during synaptic potentiation. Amongst them, CaMKII, PKC, and PKA play a prevalent role in phosphorylating the GluR1 subunit, leading to increased synaptic expression and transmission potentiation ^{26, 28}. Src family kinases (SFKs) are also crucial for NMDA receptors and modulate their trafficking through phosphorylation ⁸⁷. Notably, NR2B receptors have increased phosphorylation by PKC and SFKs in the spinal cord from the rodent model of chronic pain ⁵². Our results confirmed that GLED-induced plasticity modulation involves these key regulators. Indeed, GLED decreased SFK (Fig. 9G) and PKC activation, thus leading to reduced phosphorylation of NMDA and AMPA receptors. As glutamatergic receptor phosphorylation controls their trafficking and synaptic expression, we analyzed the synaptic expression of both NMDA and AMPA receptors. By analyzing post-synaptic density (PSD) content in western blots, we found that both NR2B and GluR1 subunits were less expressed at the synapse following GLED exposure (Fig. 9H, I, J and K). No effect was observed on GluR3. Interestingly, when female rats were exposed to GLED, only the synaptic expression of NMDA receptors was affected, and GLED did not decrease GluR1 synaptic expression (Fig. 10). Overall, these data show neuronal plasticity changes in glutamatergic receptors that suggest decreased glutamatergic transmission contributing to the long-lasting antinociceptive effects of GLED in rats.

Figure 9. GLED exposure modulates synaptic expression of glutamatergic receptors in male rats in a model of acute post-surgical pain.

Male rats were exposed two days to WLED or GLED before paw incision surgery, and two days after surgery. Spinal cords (dorsal horns) were then collected after rats were exposed to WLED or GLED. (A) Western blots illustrate the expression of glutamatergic receptors and kinases involved in their phosphorylation, as well as their phosphorylated forms. (B) The histogram represents the quantification of GluR1 total expression and its level of phosphorylation on both S831 and S845. (n=8, Shapiro-Wilk and Bartlett tests followed by unpaired T-test). (C) Quantification of CamKII total expression and its level of phosphorylation on Thr286. (n=9, Shapiro-Wilk test followed by Mann Whitney test). (D) Quantification of PCKγ total expression and its level of phosphorylation on Thr514. (n=8, Shapiro-Wilk followed by unpaired T-test). (E) Quantification of NR1 total expression. (n=8, Shapiro-Wilk tests followed by Mann Whitney test). (F) Quantification of NR2B total expression and its level of phosphorylation on Tyr514. (n=8, Shapiro-Wilk and Bartlett tests followed by unpaired T-test). (G) Quantification of Src family kinase total expression and their level of phosphorylation on Tyr416. (n=8, Shapiro-Wilk and Bartlett tests followed by unpaired T-test). (H) Representative western blots of PSD95, Synaptophysin (SYP), NR2B, GluR1, and Glur3. Western blots illustrate the synaptic expression of glutamatergic receptors. Dorsal horns were dissected, and cellular lysates were fractionated to obtain presynaptic and post-synaptic fractions. Expression of PSD95, specific from PSD, and SYP, absent from the PSD, were analyzed to evaluate the enrichment quality. (I) The histogram represents the quantification of NR2B synaptic expression (n=10, Shapiro-Wilk and Bartlett tests followed by unpaired T-test). (J) The histogram represents the quantification of GluR1 synaptic expression (n=10, Shapiro-Wilk and Bartlett tests followed by unpaired T-test). (K) The histogram represents the quantification of GluR3 synaptic expression (n=10, Shapiro-Wilk and Bartlett tests followed by unpaired T-test). Panels B-G and I-K used different sets of animals. All results represent mean±SEM with the p-values obtained from statistical analysis above the graphs.

Figure 10. GLED exposure modulates synaptic expression of glutamatergic receptors in female rats in a model of acute post-surgical pain.

Female rats were exposed two days to WLED or GLED before paw incision surgery, and two days after surgery. Spinal cords (dorsal horns) were then collected after rats were exposed to WLED or GLED. (A) Western blots illustrate the expression of glutamatergic receptors. (B) The histogram represents the quantification of NR2B synaptic expression (n=9, Shapiro-Wilk and Bartlett tests followed by unpaired T-test). (C) The histogram represents the quantification of GluR1 synaptic expression (n=9, Shapiro-Wilk and Bartlett tests followed by unpaired T-test). GluR1 subunits were not detected in the total lysate fraction. Panels B and C correspond to the same set of animals. All results represent mean±SEM with the p-values obtained from statistical analysis above the graphs.

Discussion:

Our results characterize the antinociceptive effect of GLED phototherapy on postoperative pain¹⁴. GLED attenuated incisional hypersensitivity in rats. The attenuation of hypersensitivity heightened with increasing exposure time. This effect was still observed 3 days after GLED termination. These results demonstrate that GLED possesses high potential as a postoperative pain management tool.

Our previous study revealed an increase in endogenous opioid levels after GLED exposure ⁷³. Thus, we evaluated GLED’s potential to reduce opioid need. In male rats, GLED decreased the amount of opioids and ibuprofen needed for hypersensitivity reduction. Interestingly, we did not observe any potentiation of the ibuprofen effect in female rats. However, pain processing is different between the sexes. For instance, estrogen and prolactin pathways may enhance central nervous system sensitization ^{4, 21}. The discrepancy in these results between males and females requires further analysis. Notably, human studies revealed that GLED was analgesic in women ⁷⁴.

To understand how GLED decreases opioid requirements, we explored different systems involved in pain modulation. Antinociceptive mechanisms rely partly on modulation of the endogenous opioid system ⁸⁶. Hence, we first investigated whether GLED exposure could increase endogenous opioid levels or receptors. Our results showed an increase in β-endorphin and proenkephalin in the CNS in male rats, but the expression of their receptors remained unchanged. We did not observe any differences in endogenous opioid levels in serum, suggesting the possibility that GLED’s effect depends on central mechanisms.

Inflammation plays a key role in pain signaling and modulates glutamatergic receptor activity. TNFα increases glutamatergic receptor activity ^{18, 42, 81}, whereas IL-10 decreases postsynaptic expression of AMPA receptors ⁷⁶. Surgeries increase TNFα levels ⁵⁹, shifting the cytokine balance to promote a pro-inflammatory status ¹⁵. Consequently, we measured serum and CSF levels of these cytokines in male rats after surgery. GLED exposure increased IL-10 and decreased TNFα levels in the CSF. This suggests that decreasing TNFα levels may improve postoperative pain. IL-10 is also a key factor in postoperative pain reduction; overexpression of IL-10 decreases surgical inflammation and promotes tissue regeneration ⁸³. As microglia are a major source of cytokines in the CNS ⁴⁶, we also analyzed their activation following GLED exposure in male rats. Microglia morphology is used to analyze their activation; increases in the soma size, roundness, and simple arborization correspond to a more active microglia ^{6, 27}. Our observations reveal that microglia are less activated following GLED exposure. Importantly, microglial modulation of nociceptive signals is different between male and female mice ⁹⁰. We observed different results between males and females when ibuprofen was administered in combination with GLED. Thus, we hypothesized that GLED would not affect microglia activation in females. Surprisingly, exposure to green light also decreased microglial activation in females following incisional surgery, suggesting that GLED modulates neuroinflammation in both sexes. Nevertheless, we observed important differences in microglia morphometry in females, who presented much more complex arborization than males. Although unexpected, these results confirm current knowledge about microglia differences between sexes. In males, microglia are more prone to inflammatory activation ⁴⁵, a phenomenon that could explain why our results showed a simpler arborization in male rats. Overall, surgeries are associated with microglial activation ⁹⁸, and their activation promotes TNFα release during postoperative neuroinflammation ³. Therefore, preoperative strategies to decrease microglial activation may result in less inflammatory mediator release and decreased pain. This raises the possibility that GLED exposure may be responsible for an improved inflammatory environment, lowering pain sensitivity and promoting healing.

To gain insights into GLED’s central mechanisms, we delved into its effect on plasticity mechanisms. Prolonged postoperative pain is related to central sensitization ^{84, 96}. Glutamatergic receptors are key modulators of central sensitization, and blocking AMPA receptors reduces postoperative hypersensitivity ^{43, 64, 65, 93}. Based on these observations, we analyzed the effect of GLED exposure on glutamatergic receptors in male rats. GLED reduced phosphorylation and synaptic expression. Several studies have shown that activation and expression of AMPA receptors are increased in multiple pain models ^{37, 38, 55, 68, 93, 95}. Consequently, the decreased phosphorylation and synaptic expression of GluR1 explain how GLED could dampen the transmission of acute pain signals by modulating plasticity mechanisms in the spinal cord. Furthermore, studies suggest that spinal cord central sensitization shares common mechanisms with synaptic potentiation in the hippocampus ^{36, 37, 60}. Our observations conclude a decreased phosphorylation of Serine⁸³¹, but not Serine⁸⁴⁵, after GLED exposure. In the hippocampus, S845 phosphorylation is crucial for both LTP and LTD, whereas S831 phosphorylation is required for LTP ^{66, 67}. This raises the hypothesis that GLED modulates plasticity by reducing synapses’ ability to potentiate rather than inducing synaptic depression. Our results also demonstrate that GLED abates NMDA receptor synaptic expression and phosphorylation. NMDA receptor activation and phosphorylation are required for NMDA-dependent synapse potentiation, and blocking NMDA receptors reduces postoperative pain ^{71, 99}. By reducing the synaptic expression of NMDA receptors in both sexes, GLED would reduce the synapse’s ability to potentiate, thus dampening synaptic potentiation and pain signal transmission. Nevertheless, we observed discrepancies between males and females regarding the synaptic expression of the GluR1 subunit of AMPA receptors. Recent studies have shown that glutamatergic receptors are differently involved in pain processing between males and females ^{29, 31}. In a model of inflammatory pain, Hildebrand’s team demonstrated increased phosphorylation of NMDA receptors only in males ³¹. Although our results presented a difference in AMPA receptor synaptic expression, these findings shed light on the necessity to delve further into sex differences when therapies affect glutamatergic signaling. Interestingly, a human study concluded that NK1 receptor expression in different brain regions is sex-dependent ⁸⁰. Thus, GLED-induced antinociception could involve tachykinins and their receptors, as they are strongly related to inflammatory pain and involved in long-term changes in spinal sensitivity through modulation of glutamatergic transmission ^{41, 75, 91}. This aspect suggests a potential involvement of NK1 receptors and will require further studies, as GLED could modulate spinal plasticity mechanisms through modulation of NK1 receptors or substance P. Overall, our results demonstrated how light stimulation modulates CNS plasticity in rats.

Altogether, our results suggest that GLED alleviates post-surgical pain in rats by acting on multiple systems, including increased endogenous opioid release, modulating CNS plasticity by reducing the expression of receptors involved in pain transmission, and providing an anti-inflammatory environment. Importantly, all these systems are interconnected.

Increasing evidence shows that TNFα is a key modulator of pain processing, and a reciprocal connection between TNFα and the endogenous opioid system is evident. However, TNFα possesses a paradoxical effect on endogenous opioid release; TNFα inhibits morphine analgesia after acute administration, whereas administration of morphine inhibits TNFα production ⁵⁷. Cytokines not only modulate the opioid system but also regulate microglia activity ⁴⁶, which is crucial for neuronal transmission. Indeed, microglia activation induces an increase in spontaneous excitatory postsynaptic currents, and recent studies suggest that microglia-derived BDNF activates Src kinase, which increases NMDA receptor transmission in rat pain models ^{30,32,82, 105}. Importantly, increasing microglial activation alters LTP, which requires AMPA receptors, and elevates the levels of TNFα which stimulates AMPA receptor endocytosis ¹⁰⁰. Microglia also affect the serotoninergic projections from the rostroventral medulla (RVM), and their activation in chronic pain models increases the release of serotonin (5-HT) in the spinal cord ²⁵. Following the release of 5-HT, spinal NMDA receptor activity is increased through stimulation of 5-HT_2b receptors, facilitating pain signal transmission ². Thus, the 5-HT pathway seems to link microglia activity and glutamatergic receptor activation; GLED could potentially act on this system, reducing the activity of the serotoninergic descending pathway. Such hypotheses will require further investigation.

Other studies reported how light exposure could modulate inflammation and plasticity: exposing mice to flickering lights at 40 Hz drives neural activity and recruits microglia ⁴⁰. Light flickering stimulation in Alzheimer’s disease models also showed reduced inflammation and preservation of neuronal and synaptic density across multiple brain areas ¹.

Managing inflammation in the elderly has become crucial and gives rise to the concept of “senoinflammation” or “inflammaging”, which describes chronic inflammation related to aging ^{23, 88}. This chronic inflammation exacerbates aging processes and age-related chronic diseases. Given that GLED exposure reduces neuroinflammation, this new therapy may offer a way to control inflammation-related pain in the elderly. Capitalizing on several mechanisms to control pain may result in improved outcomes for postoperative pain in the elderly. Nevertheless, whether these mechanisms act as a discretely unique pathway or as a cascade of simultaneous events remains unknown. A study led by Burstein suggested that migraine photophobia could originate in cone-driven retinal pathways and be tuned in thalamic neurons outside the main visual pathway ⁷⁹. Strong evidence also suggests that the suprachiasmatic nucleus (SCN), the site of the master circadian clock, regulates inflammatory responses, and monochromatic light can affect cognitive processes almost instantaneously ^{89, 94}. These effects are mediated by a melanopsin-based photoreceptor system, which is known to transmit irradiance signals to the SCN ⁹⁴. We can hypothesize that GLED modulates SCN functions, thus affecting circadian rhythms that are highly involved in regulating inflammatory processes related to pain conditions ⁸⁹. Furthermore, Bruguerolle & Labrecque described a circadian rhythm of endogenous opioids, demonstrating that β-endorphin levels are highest in the morning ^{16, 89}. By modulating functions of the SCN and circadian rhythms, GLED could affect inflammation and endogenous opioid release. In conjunction with microglial deactivation, decreased levels of TNFα and increased levels of IL-10 may modulate pain signaling transmission by acting on glutamatergic receptors ^{18, 42, 76, 81}. Therefore, GLED may reduce opioid requirements to control postoperative pain, and such a strategy may be especially crucial for elderly patients.

Supplementary Material

1

Supplementary Figure 1. Plug-in green and white light intensity vs. wavelength measurements. The measurements were made with a Spectrometer to provide the relative intensity vs. wavelength. The measurements for the green LED strip show a center wavelength of 515 nm, with a Full Width and Half Max span of 35nm. This indicates a reduction in the intensity of 50% at approximately 497 nm and 533 nm. The measurements for the white LED strip show a peak intensity wavelength of approximately 605 nm and a Full Width Half Max span of approximately 90 nm. This indicates a reduction in the intensity of 50% at approximately 560 nm and 650 nm.

Supplementary Figure 2. Difference in thermal pain sensitivity after exposure to different intensities of green light. Male rats were exposed to green LED (λ=525 nm) or white LED (8h/day, 100 Lux) for 5 days and tested each day to evaluate thermal sensitivity. Two different intensities of GLED were compared (4 and 100 Lux). N=6-7, mean±SEM, two-way ANOVA followed by Turkey’s posthoc test, gray asterisks illustrate significant differences to saline control, green asterisk illustrates significant difference to 4 Lux condition, *p<0.05, **p<0.01. Increasing GLED intensity produced a faster decrease in thermal pain sensitivity.

Supplementary Figure 3. Green light-emitting diode (GLED) attenuates postoperative mechanical hypersensitivity in male mice. Following incisional surgery, male mice were exposed to white led (WLED), or green LED (GLED), 100 Lux 8h/day, for three days. Mechanical hypersensitivity was assessed over 5 days (n=5-6, mean±SEM, two-way ANOVA, F_{(1, 45)}=7.233, p=0.0100).

Highlights.

• Green light (GLED) potentiates ibuprofen- and morphine-induced antinociception

• GLED exposure reduces the required morphine dose to achieve antinociception

• Spinal neuroinflammation is reduced after GLED exposure

• GLED exposure decreases dorsal horn glutamatergic receptor expression

Perspective:

This study provides a potential additional therapy to decrease post-surgical pain. Given the safety, availability, and the efficacy of green light therapy, there is a significant potential for advancing the green light therapy to clinical trials and eventual translation to clinical settings.