PBMT did not alter the metabolic parameters linked to STZ-induced type-1 diabetes since the clinical signals of type-1 diabetes (hyperglycemia, weight loss, polyuria, polydipsia, and polyphagia) in the STZ-PBMT group remained equal to those of untreated STZ group. Type-1 diabetes induction protocol through low doses of STZ (five low doses, a single dose of 25 mg/kg per day) was suitable for installing irreversible hyperglycemia. All rats submitted to STZ injections (STZ and STZ + PBMT groups) became hyperglycemic (≥ 250 mg/dL of blood glucose concentration; 349.07 ± 48.23 mg/dL) after five STZ-low doses, reaching the threshold for diabetes between the fourth and fifth days (Fig. 1, panels A and B). Diabetic rats also presented other characteristic metabolic alterations, such as polyuria, polyphagia, and polydipsia (data not shown), and stopped gaining weight with a slight loss (in grams) throughout the experimental period (Fig. 1, panel C).
STZ is a diabetogenic antibiotic known for its selective capacity to kill the pancreatic beta cells (β-cells), commonly used for type-1 diabetes induction in animal models67–70. STZ is taken up by the β-cells glucose transporter GLUT2 and triggers immune mechanisms68,69. According to Wang and Gleichmann (1995; 1998)68,69, STZ restricts GLUT2 expression in vivo and in vitro when administered through multiple low-doses protocol, which is a method for producing fewer STZ side effects, such as neurotoxicity. In general, rats submitted to STZ injections show deficits in insulin production, leading to hyperglycemia and, consequently, polydipsia and polyuria70. All those diabetes signals were observed in rats submitted to the low doses STZ protocol (STZ and STZ + PBMT groups), which characterizes a reproducible model of diabetes induction, as shown in previous studies from our group66,71−74.
Diabetic rats, submitted or not to PBMT (STZ + PBMT and STZ groups, respectively), showed similar levels of hyperglycemia on days 21, 24, and 28 of the experimental protocol (454.94 ± 37.10 mg/dL). The same was observed regarding the rats' weight, once this parameter was only dependent on the diabetic condition and not on the laser treatment. In like manner, PBMT did not cause any influence over healthy rats (control, non-diabetic and non-hyperalgesic; SCB + PBMT group), as shown in Fig. 1 (panels B and C). Our results using NIR light (904 nm; 70 mW; 2.03 J; 29 s) corroborate with the study performed by Peplow and colleagues (2012)75, in which PBMT (660 nm; 100 mW; 4.7–6.3 J/cm²; 20 s) applied for wound healing in diabetic patients, showed no interference neither on hyperglycemia nor on patients' metabolic status.
PBMT leads to an anti-hyperalgesic effect against the DN mechanical hyperalgesia induced by low doses STZ, which was unrelated to a putative influence of PBMT in the metabolic and clinical signals type-1 diabetes, as shown above. PBMT reduced the STZ-induced diabetic hyperalgesia. The data reproduced previous results obtained from our research group66.
On the 21st day (after the first PBMT session), there was no reduction in the mechanical hyperalgesia intensity (Δ withdrawal threshold, g). However, on the 24th and 28th days, a time-significant reduction (p < 0.01 and p < 0.001, respectively) was observed in the intensity of hyperalgesia of STZ + PBMT group when compared to STZ group, characterizing an anti-hyperalgesic effect promoted by PBMT. Nonetheless, when PBMT was applied to control rats (SCB + PBMT), there was no change in the mechanical withdrawal thresholds in a similar condition observed in SCB (vehicle) and naïve groups, as shown in Fig. 1 (panels D and E; the latest in detail for the PBMT period).
PBMT has long been used for the clinical treatment of neuropathic pain showing satisfactory results76,77, but its analgesic mechanisms are not entirely understood. The inhibition of the neuronal hyperactivity by the infrared light seems to be one of the ways that PBMT acts directly over the neurons78,79, and consequently the pain. Holanda et al. (2016 and 2017)52,80 suggested a neuromodulation effect once they showed that patients with back pain submitted to PBMT (808 nm; 100 mW; 8.4 J; 84 s; a single session) on L4-L5 levels presented a significant pain relief. This modulatory effect could be the key to regulating the neuronal activity severely affected by hyperglycemia to avoid painful sensibility.
Motor function dysfunction related to DN and detected by CatWalk system was amended by PBMT, likely due to PBMT anti-hyperalgesic property. Our analysis was based on a previous study identifying dynamic motor function alterations related to STZ-induced DN74. The results showed that PBMT was able to improve the Maximum Contact Area (cm²) and the Print Area (cm²) (Fig. 2, panels A and B, respectively) of the rats' hind paws after 4 (24th day) and 8 (28th day) laser irradiation sessions, likely suggesting analgesia amelioration ought to improvement of nerve conduction. Statistical differences were observed between STZ + PBMT and STZ groups in such periods. No difference was observed for the Stride Length (cm) parameter (Fig. 2, panel C), except between both neuropathic groups (STZ and STZ + PBMT) vs. control groups (Naïve, SCB, and SCB + PBMT). Areas of the digits and plantar pad (glabrous) appeared well delimited in the footprints of the right hind paw (RH) from rats of the STZ + PBMT group on days 24th and 28th (Fig. 2, panel D: e), which were similar to the observed in Naïve and SCB groups (Fig. 2, panel D: a, b, d), thus differing from the STZ group (Fig. 2, panel D: c). The latest showed a reduction in the contact area with a low resolution of the footprints on the 28th day.
According to Zochodne et al. (2008)81, non-controlled diabetes leads to damage of sensory neurons before the involvement of the motor ones. DN is associated with postural changes in this process, involving the alteration of foot positioning during gait82 and alterations in pressure applied throughout the stand phase83,84. Vieira et al. (2020)74 demonstrated that Maximum Intensity (a. u.) was the main parameter to demonstrate alteration on the 14th day after starting STZ injections, which corroborates the data obtained by Benitez et al. (2015)84. Diabetic animals (STZ group), which presented significant changes in the mechanical threshold on the 14th day, as observed in the electronic von Frey test results for mechanical hyperalgesia, applied less pressure during the paws' contact with the glass floor of the CatWalk XT system. These findings follow the pattern of sensory changes in DN, also known as "stocking-and-glove," which occurs as a sensory disorder of the extremities of limbs, compromising functionality85,86.
Few studies evaluated the influence of PBMT on motor parameters in animal models, primarily when the CatWalk XT system was used. Vieira and colleagues (2016)60 analyzed the influence of low-level laser therapy (LLLT) using the GaAs laser (904 nm; 4 J/cm²) on motor changes resulting from snake venom-induced myonecrosis. At 3 h after the venom injection, mice presented a semi-inflected posture of the right hind limb. When PBMT was performed, for the same period (3 h), animals presented values of Maximum Intensity (u. a.), Stand (s), and Balance (s) similar to control. The semi-flexion posture could represent a protective reflex to a painful stimulus. In the DN model, rats did not present a hind limb semi-flexion; however, they applied less pressure (represented by less intensity and smaller paw contact area) during gait. In this sense, PBMT could normalize part of the altered gait we observed in DN, which may be related to the anti-hyperalgesic effect promoted by light therapy. Altogether, we suggest that type-1 diabetes-derived neuropathic pain likely results from impairment in sensory and motor neurons.
PBMT was able to reduce the levels of cytokines. To further investigate whether the anti-hyperalgesic effect promoted by PBMT could be linked to reducing the DN-associated cytokines, we studied the concentration of TNF-α, IL-1β, IL-6, and CINC-1 by ELISA immunoassay. DRG from diabetic neuropathic rats, PBMT-untreated (STZ group), showed increased concentrations of IL-1β and IL-10 (Fig. 3, panels B and E); levels of TNF-α, IL-6, and CINC-1 showed no alterations regarding the DN setting (Fig. 3, panels A, C, and D), but presented alterations when the groups were exposed to PBMT (Fig. 3, panels A, C, and D). Specifically, there was a significant increase in IL-1β in the STZ group compared to all the other groups (p < 0.05), especially in the STZ + PBMT group (p < 0.05), as shown in Fig. 3 (panel B). As for IL-1β, the levels of IL-10 (an anti-inflammatory cytokine) showed increased levels in the STZ group but not in the controls or the STZ + PBMT group (p < 0.05) (Fig. 3, panel E). Differently, levels of TNF-α showed no alterations caused by the diabetic neuropathy (STZ group), which were the same for control groups, except for STZ + PBMT group, which showed a significant reduction (p < 0.01) in the concentration of TNF-α, compared to STZ (without PBMT) (Fig. 3, panel A). CINC-1 concentrations reduced significantly (p < 0.05) only for the SCB + PBMT group (One-way ANOVA followed by Bonferroni post-hoc test) (Fig. 3, panel D).
As reported by Hsieh and collaborators (2012)87, PBMT (660 nm; 30 mW; 9 J/cm²; 60 s; 7 consecutive days; 63 J/cm² cumulative energy density) applied for the treatment of chronic pain induced by sciatic nerve constriction, caused a reduction in the concentration of proinflammatory cytokines, such as TNF-α, IL-1β, and HIF-1α (hypoxia-inducible factor-1α), corroborating our results. The same reduction in TNF-α levels promoted by PBMT (950 nm; 2.5 J/cm2; 32 s; 15 consecutive days; 37.5 J/cm² cumulative energy density) was observed in a mice model of sciatic nerve crush injury, reported by Cidral-Filho et al. (2013)88. These shreds of evidence bring the discussion on the high anti-inflammatory capacity exhibited by PBMT. In this regard, the anti-inflammatory effects promoted by PBMT were previously observed in carrageenan-induced inflammation in rats' paws, in which both 660 nm and 684 nm low-level lasers (30 mW; 7.5 J/cm²; 196 s; a single point) reduced the edema formation and inflammatory cell migration at 4 h after carrageenan injections89.
Inflammatory cytokines can activate diverse cell membrane receptors, thus transmitting environmental signals, and in several cell lines, activation of G-protein-coupled-receptors (GPCRs) and receptor tyrosine kinases (RTKs) activate MAPKs (for review, see Kholodenko, 2012)90. MAPK cascades signaling are evolutionarily conserved pathways related to the intracellular signal transduction in response to various extracellular stimuli (for review, see Plotnikov et al., 2011)91. MAPK controls many cellular processes, such as growth, proliferation, differentiation, motility, stress response, survival, and apoptosis92. All three groups of MAPK (p38; ERK1/2; JNK) can be activated by osmotic perturbations derived from glucose, polyol pathway, oxidative stress, and advanced glycation end products (AGE)93. These events are highly involved in the etiology of the DN1.
PBMT reverses the STZ-increased of p38 MAPK gene- and protein expression. To examine a possible interaction between the inflammation and MAP kinases, we further assessed MAPKs gene- and protein expression in L4-L5 DRG, once these molecular targets are supposed to be altered in DN, due to a condition imposed by the uncontrolled diabetes93. Before the Real-Time quantitative PCR experiments, primers for target- (p38, ERK1/2, and JNK) and housekeeping genes (endogenous control, Arfgef1, and Serpinb6) were tested for their efficacy by the standard curve construction. It was done through serial dilutions of naïve cDNA (1:4; 1:8; 1:16; 1:32; 1:64) and a fixed primer concentration (100 nM). High efficiency was observed for all tested primers, including the endogenous controls (99.9; r² 0.98) (data not shown). Then, the expression of target genes was normalized by the mean values of Ct (cycle threshold) from a pool of Arfgef1 and Serpinb6 gene expression.
It was observed a significant increase in the expression of p38 mRNA in the STZ group compared to Naïve and SCB (vehicle) (p < 0.01) and SCB + PBMT groups (p < 0.001; unpaired t-test). In addition, p38 mRNA expression in the hyperalgesic group (STZ) was significantly higher in comparison to the laser-exposed one (STZ + PBMT) (p < 0.05; unpaired Student t-test), as shown in Fig. 4, panel A. Slight differences between the groups regarding ERK1/2 and JNK mRNA expression were not significant (Fig. 4, panels B and C).
The immunofluorescence experiments also found a higher increase in the expression of activated MAPK, especially p-p38, in the STZ group (Fig. 5, panels H and I). The fluorescence regarding p-38 phosphorylation in the hyperalgesic rats was reduced by the PBMT treatment, as shown in the STZ + PBMT group, in which there was some p-38 activation but in a lower number of neurons in comparison to STZ group (Fig. 5, panels K and L). For both STZ and STZ + PBMT groups, the activation of p38 was highly concentrated in the nuclei of the DRG afferent neurons, as shown by DAPI co-staining (Fig. 5, panels I and L in detail). Besides the lower number of p-p38 positive neurons in the STZ + PBMT group, it was also possible to observe as another difference in comparison to the STZ group a higher intensity of fluorescence (represented by green) scattered in neurons cytoplasm with some vesicle-like conformations (Fig. 5, panels K and L in detail).
Also corroborating gene expression results, low phosphorylation of ERK1/2 protein was detected. However, for both laser-treated groups (SCB + PBMT and STZ + PBMT; Fig. 6, panels E and K, respectively), the levels of fluorescence-related p-ERK1/2 activation were even lower in comparison to the non-irradiated groups (SCB and STZ; Fig. 6, panels B and H, respectively). An interesting observation regarding the STZ + PBMT group is the presence of p-ERK1/2 accrued in the vicinity of the plasma membrane, as shown in Fig. 6 (panels K and L in detail).
In line with the low phosphorylation of ERK1/2 for SCB + PBMT and STZ + PBMT groups, a very low fluorescence intensity regarding p-JNK staining was observed in such irradiated groups (Fig. 7, panels E and K, respectively). For SCB and STZ groups, there was a basal expression of p-JNK, as observed in a diffused way in the cytoplasm, as shown in Fig. 7 (panels B and H, respectively).
Studies have shown that MAPK activation is linked to allodynia and hyperalgesia in different disease conditions94–96. p38-MAPK, the most studied member of MAPK family, is typically activated by extracellular stress and proinflammatory cytokines, with a prominent role in the inflammatory process once there is a significant reduction in inflammation after the systemic administration of p38 pharmacological inhibitor97. A study of 8-12-week STZ-administered rats showed activation of all the main MAPKs (p38; JNK; ERK1/2) in L4-L5 DRG, although with delayed activation of JNK, relative to p38 and ERK1/241. It is well known that TNF-α and IL-1β play a key role in developing and maintaining pain after peripheral nerve injury98. Consequently, as p38 regulates TNF-α and IL-1β biosynthesis, both reductions culminate in anti-inflammatory effects with a positive reflex in analgesia99. Therefore, our data showed that PBMT, which can be considered a photophysical and photochemical effector of cell events, promoted a reduction in the concentrations of TNF-α and IL-1β, associated with less activation (phosphorylation) of p38, which can help explain the analgesic effect of the therapy. p38 has several proinflammatory roles, and systemically administered p38 inhibitors produce anti-inflammatory effects by reducing the synthesis of TNF-α and IL-1β, as well as COX-2 induction of inflammatory cells28,99. Hence, the pharmacological and/or non-pharmacological therapies that reduce the concentration of proinflammatory molecules might reduce the painful condition.
The light modulation against MAPK activation caused by PBMT (He-Ne; 632.8 nm; 4.5 mW; 3 s each 1.8 mm x 1.8 mm square of the tissue-culture plate) was studied by Shefer and colleagues (2001)100 in skeletal muscle cells, which showed increased levels of ERK1/2 under laser irradiation, and no effect over p38 and JNK expression. Aleksic and collaborators (2010)101 also verified activation of ERK1/2 in osteoblast cell cultures submitted to laser irradiation (Er: YAG; 2.94 µm; fluence between 0.7–17.2 J/cm²), with no effect on p38 and JNK expression. According to Ji et al. (2002)99, p38 activation in the DRG is initiated by retrograde transport of nerve growth factor (NGF) release from inflamed tissue. It increases both the translation and the transport of TRPV1 (a polymodal receptor channel) to the peripheral nociceptor terminal, contributing to the maintenance of inflammatory heat pain hypersensitivity. In addition, p-p38 was found mainly in small neurons of the DRG, suggesting a higher activation in type-C sensory fibers.
The p-p38 staining is co-localized to TRPV1 in DRG. Complementarily to the immunofluorescence staining of p-p38 in DRG, it was further analyzed whether its expression was involved in a particular type of nerve fibers, i.e., type-C fibers (small DRG neurons; unmyelinated; TRPV1+). It was found that TRPV1+ fibers in all experimental groups and most of the afferent neurons present in DRG sections were classified as type-C, with higher intensity of fluorescence and occurrence in the STZ and STZ + PBMT groups (Fig. 8, panels K and O, respectively). The co-staining of p-p38 and TRPV1+ fibers was widely distributed (Fig. 8, panels L and P, respectively), with a prevalence of TRPV1+ signal (red) over p-p38 (green) in the STZ + PBMT group, as shown in Fig. 8 (panel P, in detail).
The relation between TRPV1 and MAPK is because TRPV1 leads to proliferation through Ca2+ entry, ATP release, membrane P2Y2 purinergic receptor activation, and the transactivation of epidermal growth factor receptor (EGFR). Thus, the increase in [Ca2+]i and the binding of ATP to P2Y2 upregulate the intracellular IP3 via phospholipase C (PLC). The upregulation of IP3 leads to the opening of store-operated channels (SOC), which causes Ca2+ release from the endoplasmic reticulum (ER). The previous TRPV1-mediated EGFR transactivation prompts Ras/Raf/MAPK signaling (for review, see Zhai et al., 2020)102.
TRPV1 is highly expressed in cutaneous nociceptive nerve endings103, being also easily exposed to sunlight and activated by ultraviolet (UV) light. Lee et al. (2009)104 showed that UV light activates Ca2+ influx and non-selective cationic current in immortalized human keratinocytes (HaCaT cells), and this activation was suppressed by capsazepine, a TRPV1 antagonist, thus showing an interaction between the Light and TRPV1 channels. In this sense, Wang et al. (2017)105 showed that PBMT through a 980 nm laser device (3 J/cm²; continuous wave) induced a thermal effect and, consequently, the TRPV1 activation in adipose-derived stem cells. Different from our study, during in vitro studies, local increases in temperature of the cell membrane might be caused by laser irradiation, even using low-level devices. Our histological sections from DRG show activation of TRPV1 in diabetic hyperalgesic rats submitted to PBMT. However, further studies should be done to investigate better the involvement of TRPV1 in the anti-hyperalgesic effect of PBMT observed in our STZ-induced DN model.
PBMT influenced Ca 2+ dynamics in DRG-cultured neurons. Inline, the activation of MAPK pathway and its modulation by PBMT seems to be related to the Ca2+ dynamics in the DRG environment. An increase in the fluorescence intensity of FLUO-4 AM-stained DRG neurons was observed after the stimuli with 15 mM KCl and, especially, 50 mM KCl for all the groups (Fig. 9, panel A). Remarkably, the increased fluorescence regarding the Δ[Ca2+]i after a 15 mM KCl stimulus was mainly observed in neurons previously maintained in the hyperglycemic medium (High-glucose group) (Fig. 9, panels B and C). DRG-neurons kept in hyperglycemic medium (55 mM glucose) and exposed to PBMT (High-glucose + PBMT group) showed significant (p < 0.001) less intensity of fluorescence during the 15 mM KCl stimulus in comparison to the High-glucose group (Fig. 9, panels B and C). During the 50 mM KCl peak, which serves as a control for stimulating the Ca2+ influx in responsive neurons, the main difference (p < 0.05) was observed between the fluorescence intensity of the Low-glucose group in comparison to the Low-glucose + PBMT group (Fig. 9, panels D and E). When the 50 mM KCl stimulus ceased, the fluorescence intensity decreased in both hyperglycemic groups, although it did not return to basal levels (Fig. 9, panels D and E).
Beyond the fluorescence intensity (Δ[Ca2+]i) we also quantify the percentage (%) of 15 mM KCl-responsive cells concerning the total number of neurons responsive to 50 mM KCl (Fig. 9, panel F). The percentage (%) of neurons that responded to the 15 mM KCl molarity concerning all responsive ones was higher in the hyperglycemic group in comparison to the normoglycemic (p < 0.001) and to the hyperglycemic + PBMT group (p < 0.05). This result suggests that a hyperglycemic medium (55 mM of glucose) sensitizes the DRG-neurons and that sensitization is reduced by the PBMT exposure, which represents one of the possible analgesic mechanisms of the therapy.