Photobiomodulation Inhibits Long-term Structural and Functional Lesions of Diabetic Retinopathy

The pathogenesis of diabetic retinopathy (DR) remains unclear, but hyperglycemia is regarded as the initiating event, and subsequent local oxidative stress and inflammation have been implicated in the development of early stages of the retinopathy (1–3). The focus of interest has been primarily on retinal vascular changes. Nevertheless, abnormalities of the neural retina also are known to develop (4,5), and at least some neuronal cells (photoreceptor cells) likely contribute to the pathogenesis of the observed vascular changes (6–8).

Therapeutic use of light (photobiomodulation [PBM]) in the visible far-red to near-infrared region of the spectrum (590–850 nm) (9,10) is being found to have beneficial effects in a variety of diseases, including wound healing, hypoxic injury, cerebral degeneration, Alzheimer disease, postoperative cognitive dysfunction, and retinal degeneration (10–20). Recent evidence from our group (21,22) also indicates that PBM inhibits the development of molecular abnormalities that contribute to the development of early lesions of DR in animals and resolve early (noncenter-involved) macular edema in two case series of patients with diabetes (23,24).

The mechanism by which PBM mediates observed beneficial effects remains under investigation. A number of investigations have focused on stimulation of cytochrome c oxidase, an enzyme that enhances the energy production by mitochondria in the respiratory electron transport chain (25–28), whereas other studies have shown the beneficial effects of PBM correlate with release of stem cells or progenitor cells, including enhancing proliferation of bone marrow mesenchymal stem cells and adipose-derived stem cells in vitro (29–31). Tuby et al. (32) reported low-level laser therapy recruited mesenchymal stem cells from the circulation to infarcted heart area and attenuated the scarring process. Other evidence (21,33–36) raises the possibility of an indirect effect, because beneficial effects could be observed even if light was blocked from shining directly on a particular area of the body by a shield. Thus, favorable effects of PBM may be mediated by both local and systemic effects.

In this study, our goal is to assess the long-term effects of PBM therapy on 1) the morphologic vascular lesions that characterize early DR, 2) the function of the neural retina via testing of visual psychophysics, and 3) two possible systemic mechanisms by which PBM might mediate its beneficial effects on the retina (release of stem cells and activation of signaling from the vitamin D receptor [VDR]). Vitamin D and signaling via the VDR have strong anti-inflammatory and antioxidant actions (37,38).

Male C57BL/6J mice were purchased from The Jackson Laboratory (Bar Harbor, ME) and housed in ventilated microisolator cages under 12:12-h light/dark cycle. All animal procedures in this study adhered to the National Institute of Health Guide for the Care and Use of Laboratory Animals, the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research, and authorization of the Institutional Animal Care and Use Committee at Case Western Reserve University. Intraperitoneal injection of streptozotocin (STZ) in citrate buffer was used to induce diabetes in mice at the age of 2 to 3 months (60 mg/kg of body weight for 5 consecutive days) (39). Subcutaneous injection of insulin (0–0.2 units) was applied as needed to prevent weight loss while still allowing hyperglycemia. The onset of diabetes was defined as three consecutive measures of nonfasting blood glucose >250 mg/dL (∼14 mmol/L). Both blood glucose and HbA_1c were monitored as described previously (21,40,41).

Cohorts of diabetic and nondiabetic mice were treated with far-red light generated by an array of light-emitting diodes (SpectraLife; Quantum Devices, Inc., Barneveld, WI). This device was determined to deliver 670-nm light at a power of 25 mW/cm² (irradiance) at the 50% intensity setting at a distance of 2 to 3 cm between the device and the animal (measured with spectroradiometer, specbos 1211UV; Dataoptics, Inc., Ypsilanti, MI). The mice were confined to a small plastic box (90 × 135 mm) that blocked ambient light, but did not have a cover between the animals and the light source. The light provided illumination evenly across the entire back of the animals. Under ambient room light in the animal room (250 lux), animals were irradiated for 240 s each day (∼5 J/cm² dose daily). PBM was started immediately after confirmation of diabetes and administered daily between 10:00 and 11:30 a.m. Conditions of light delivery were similar to those in our prior studies (21,22).

Assessment of diabetes-induced retinal capillary disease has been previously described (39). Briefly, at 8 months of diabetes, mouse eyes were fixed in 10% buffered formalin (Sigma-Aldrich, St. Louis, MO) for at least 1 week, and retinas were isolated, washed overnight in water, digested in Tris buffer containing 40 units/mL elastase overnight, and mechanically brushed to dislodge retinal neurons from the vasculature. When totally cleaned of neural cells, the isolated vasculature was laid out on a glass microscope slide and stained with hematoxylin and periodic acid Schiff reagent. Degenerate (acellular) capillaries were quantitated in six to seven field areas corresponding to the midretina (original magnification ×200) in a masked manner. Acellular capillaries (reported per square millimeter of retinal area) were identified as capillary-sized vessel tubes having no nuclei along their length, as previously reported (42–44).

Accumulation of the blood protein, albumin, in the neural retina has been viewed as a marker of increased vascular permeability (41,45). At 8 months of diabetes, sterile fluorescein isothiocyanate (FITC)–BSA (50 µg/µL; Sigma-Aldrich) in PBS (NaCl, 0.138 mol/L; KCl, 0.0027 mol/L; pH 7.4) was injected into the tail vein of mice at 100 µg/g body weight. After 20 min, mice were euthanized, and their eyes were enucleated, fixed in 4% paraformaldehyde for 5 h, followed by incubation in increasing concentrations of sucrose (up to 20%; at least 2 h each), and then frozen in Optimal Cutting Temperature compound (Sakura Finetek USA, Torrance, CA). At least three noncontiguous cryosections of retina were cut from each eye, and average image pixel density in the neural retina was measured by computer-assisted fluorescence microscopy. Visible blood vessels were excluded from areas where fluorescence (pixel density) was measured. Leakage of albumin was estimated from measurements of fluorescence in the inner and outer plexiform layers and the inner nuclear layer of the neural retina. Average values of relative fluorescence units were normalized by expressing tissue fluorescence for each animal as a ratio to the relative fluorescence units of blood plasma from each animal. Vascular permeability in diabetes was expressed as the ratio of FITC-dextran concentration in neural retina of diabetic animals relative to that in nondiabetic animals.

The spatial frequency threshold (an estimate of visual acuity) and contrast sensitivity threshold were measured with the Virtual Optokinetic system (OptoMotry; CerebralMechanics, Inc., Medicine Hat, Alberta, Canada), as reported previously (40,46,47). The maximum spatial frequency capable of driving head tracking was determined as the spatial frequency threshold. The contrast sensitivity at 8 months of diabetes was measured at six spatial frequencies (0.031, 0.064, 0.092, 0.103, 0.192, and 0.272 cycles/degree) to detect functional defects in spatially sensitive retinal cells or in higher visual pathways. This was determined as the inverse of Michelson contrast without correction for luminance of the monitors. Each measurement was repeated several times to evaluate the reproducibility of responses. The experimenter was masked as to the identity of the experimental animals and groups.

Blood was collected between 9:00 and 10:00 a.m. from anesthetized animals via heart puncture with syringes containing EDTA, immediately transferred to 7.2-mg EDTA vacutainers (BD Biosciences, San Jose, CA), and stored on wet ice for up to 10 min. Erythrocytes were lysed by the addition of red blood cell lysis buffer (Invitrogen, Carlsbad, CA), and remaining cells were washed and analyzed with flow cytometry (BD Accuri C6 cytometer; BD Biosciences). Cells were stained with FITC–anti-mouse c-Kit and phycoerythrin anti-mouse CD45 antibodies (BD Pharmingen, San Jose, CA). Identically labeled isotype control antibodies (BD Pharmingen) served as negative controls. Circulating stem cells were gated as CD45^−/low/c-Kit⁺ cells (48) and expressed as a percentage of all CD45⁺ cells. c-Kit⁺ cells also were identified in retinal whole mounts and counted manually. The antibody used was the same as that described above, and the immunohistochemical method used was described previously (44).

Total RNA was isolated from mouse skin and retina with TRIzol reagent (Ambion, Carlsbad, CA) and RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturers’ instructions. A total of 350 ng total RNA was reverse transcribed into cDNA using Iscript Reverse Transcription Supermix (Bio-Rad, Hercules, CA). Quantitative real-time PCR was performed using FastStart Universal SYBR Green Master Mix (Roche, Mannheim, Germany) on an ABI StepOnePlus Real-Time PCR System (Applied Biosystems, Foster City, CA) in triplicate. PCR conditions were 400 nmol primer for 10 min at 95°C, 45 cycles of 15 s at 95°C, and at 57°C for 30 s. The fold-change expression of target genes was calculated using the 2^−ΔΔCt fold-change formula normalized to β-actin and nondiabetic untreated controls. The following primer sets were used in this study: VDR (sense: 5′-CTTCCTAAGAGACTTCCCGAGAGA-3′ and antisense: 5′-GGCATTTATTTACAGCGGTACTTGT-3′); Cyp24a1 (sense: 5′-CATCGACCACCGCCTAGAGA-3′ and antisense: 5′-ACAGCAGCGTACAGTTCTTTC-3′); and β-actin (sense: 5′-GATTCCATACCCAAGAAGGAAGGCTG-3′ and antisense: 5′-AGCTGAGAGGGAAATCGTGCGT-3′).

Statistical analysis was performed using ANOVA followed by Fisher test. A P value ≤0.05 was considered statistically significant. Comparisons of contrast sensitivity data between groups were analyzed by ANOVA with the repeated-measures test. Mann–Whitney test (P ≤ 0.05) was used to compare the flow cytometry results between groups.

Diabetes was successfully induced in mice by injection with STZ based on HbA_1c and blood glucose levels. There were significant differences between STZ-treated mice and nondiabetic control mice in nonfasting blood glucose levels, HbA_1c, and body weight (Table 1). PBM application to diabetic animals had no significant effect on body weight, nonfasting blood glucose, or HbA_1c compared with the untreated diabetic control group.

Although long-term (8 months) diabetes did not result in any change in retinal thickness or loss of photoreceptors, it did result in a significant (P < 0.0001) increase in the number of degenerate (acellular) capillaries in the retina of untreated animals (Fig. 1). Diabetes of 8 months also resulted in a significant (P < 0.01) increase in the amount of albumin extravasated into the nonvascular retina (i.e., in the neural retina between vessels) in each of the three layers of the retinal vascular plexus (in the inner and outer plexiform layers and at the border of the inner plexiform and inner nuclear layers) (Fig. 2). Daily illumination of the animals with 670-nm light for 4 min/day significantly inhibited the diabetes-induced degeneration of retinal capillaries and albumin accumulation in the inner nuclear and outer plexiform layers of the retina.

To determine whether the diabetes-induced reduction in visual function is influenced by PBM, an optokinetic system was used to measure spatial frequency threshold and contrast sensitivity at six different spatial frequencies (Fig. 3) in mice diabetic for 8 months. As reported previously (47), diabetic mice displayed significantly lower spatial frequency threshold and contrast sensitivity than normal diabetic mice (P < 0.001). Both measures of visual function were largely preserved in diabetic mice treated daily with PBM for the entire duration of diabetes, with the effect being statistically different from that in the diabetic control group (P < 0.001). There was no marked difference in visual function between the untreated nondiabetic group and nondiabetic mice receiving PBM. These results indicate that PBM can play an important role in preserving normal neural function in diabetes.

Because our previous study of PBM in DR (22) provided no support for the prevailing hypothesis that beneficial effects of PBM were mediated via stimulation of mitochondrial cytochrome oxidase activity, we conducted pilot studies to determine if other possible effects of the light therapy (increasing numbers of circulating stem cells and stimulation of vitamin D signaling) were associated with the observed beneficial effects on functional and structural end points of the retinopathy.

To investigate whether stem cells are associated with the beneficial effects of PBM, blood was collected from mice diabetic for 3 months (5 months of age; the duration of diabetes when blood was collected for this initial study was arbitrary), and numbers of CD45⁻/c-Kit⁺ cells in peripheral blood were determined by flow cytometry. CD45⁻/c-Kit⁺ cells were found to be decreased in circulating blood from untreated diabetic mice compared with nondiabetic controls (Fig. 4 and Supplementary Fig. 1). Diabetic mice that had received PBM soon after the onset of diabetes, in contrast, had a moderate increase of CD45⁻/c-Kit⁺ cells (P < 0.05) compared with untreated diabetic mice. Efforts to detect and quantitate c-Kit⁺ cells in the retina in diabetes showed that the c-Kit⁺ cells were very rare in all experimental groups (zero to one immunopositive cells per retina), and no obvious effects of diabetes or PBM were observed. Thus, although the PBM-induced increase in circulating CD45⁻/c-Kit⁺ in diabetes was associated with the observed beneficial effects of PBM on the retinopathy, the lack of significant accumulation of these cells in the retina in any of the groups offers no support to the postulate that c-Kit⁺ cells played an important role in the inhibition of the retinopathy by PBM.

Production of vitamin D by sunlight and subsequent signaling via the VDR is well recognized to have important biologic effects (49,50). To explore the possibility that PBM might act systemically by activating the VDR signaling, we measured expression of VDR and a gene regulated by VDR signaling (Cyp24a1) in both skin and retina. As shown in Fig. 5A and B, VDR mRNA levels did not exhibit any significant change due to diabetes or PBM in either skin or retina. In contrast, Cyp24a1 mRNA levels in retina were significantly decreased in diabetic controls compared with nondiabetic controls, and exposure of diabetic animals to PBM inhibited the diabetes-induced loss of Cyp24a1 (Fig. 5D). Thus, preservation of retinal levels of Cyp24a1 by PBM at this duration was at least associated with the observed beneficial effects of PBM on structural and functional lesions of DR.

The current clinical approach to inhibit development of retinopathy and other complications of diabetes uses exogenous insulin or oral medications to maintain blood glucose concentrations as close to normal as possible. Such good glycemic control is difficult to achieve and maintain for many patients with diabetes, so there has been continued interest in identification of alternate ways to inhibit the development of diabetic complications.

Increased permeability of the retinal vasculature, progressive degeneration of the retinal capillaries, and altered function of the neural retina, as evidenced by altered visual function, are major abnormalities that characterize the early stages of DR. All three abnormalities develop together in diabetes, but it is unclear whether or not the vascular and neural abnormalities of diabetes share a similar pathogenesis (47). The current study shows that PBM therapy for only 4 min/day significantly inhibited all three of these clinically important lesions of DR. PBM exerts these beneficial effects noninvasively, is inexpensive, and is easy to administer. The 8-month duration of daily PBM therapy in the current study is among the longest durations in which PBM has been evaluated, and it is important that no adverse effects of the therapy were observed.

Mechanisms by which PBM affect cellular biologic processes are still under investigation. A number of studies have demonstrated a beneficial effect of PBM on activity of cytochrome oxidase (51,52), a critical enzyme in the mitochondrial electron transport chain. In retinas of diabetic animals, however, we previously reported that PBM inhibited the diabetes-induced increases in retinal oxidative stress and inflammation by mechanisms that seemed not to involve cytochrome oxidase (22). Thus, we looked for other changes mediated by PBM that might give a clue pertaining how the light therapy exerts the observed beneficial effects.

PBM has been shown by our group (21) to have beneficial effects on the retina in diabetes even if there is not direct exposure of the light with eye, and others have made similar observations in other tissues (33–36,53,54). We reasoned that if observed beneficial effects of PBM were not directly mediated on the eye, then skin would be a logical place where the light energy would be absorbed. Thus, we investigated the possibility that the PBM might exert beneficial effects in vivo via VDR signaling in skin. Our initial studies focused on effects of diabetes and PBM on expression of the VDR and Cyp24a1, which encodes the hydroxylase that converts 1,25-dihydroxyvitamin D3 into inactive 24-hydroxylated products. The wavelength, which most strongly stimulates generation of vitamin D (∼300 nm [55]), is much shorter than the wavelength used by us in this study, suggesting that direct signaling through the VDR as a result of PBM therapy is unlikely. Nevertheless, we did find evidence that expression of a gene regulated by vitamin D and the VDR (Cyp24a1) was subnormal in diabetes in the retina but preserved by PBM. Neither the inhibitory effects of diabetes nor the beneficial effects of PBM on retinal Cyp24a1 were shared by skin.

Some studies have suggested that stem cells can be mobilized by PBM or low-level laser therapy and thus might participate in beneficial effects of the light therapy (29,32,34,56). Consistent with that, it has been reported that light therapy increased the proliferation of c-Kit⁺ cells in the infarcted heart (32) and other tissues. In our hands, PBM did not increase the number of circulating stem cells in nondiabetic animals, potentially because the animals in our study were considerably older than those in most studies when the blood was collected for analysis. To determine if availability of hematopoietic stem cells in response to PBM is associated with the observed protective effects of PBM in diabetic mice, c-Kit⁺ cells in the circulation were assessed by flow cytometry, and the number of c-Kit⁺ cells in the retina was quantitated in retinal whole mounts. Diabetes significantly reduced the number of c-Kit⁺ cells in the circulation, whereas diabetic mice treated with PBM had a higher number of these cells in the circulation compared with control diabetic mice. It is not clear whether the observed reduction in circulating CD45⁻/c-Kit⁺ cells in diabetes was a result of a true reduction in circulating numbers of this cell population that was enhanced by PBM or was instead a phase shift in the diurnal release of those cells from bone marrow that was corrected by the PBM. However, essentially no c-Kit⁺ cells were observed within the neural retina or retinal vasculature in any groups. The lack of accumulation of these cells within the neural retina is not altogether surprising, because we previously reported that diabetes did not increase the accumulation of any leukocyte cell types in the neural retina (44), and stem cells likely lose their stem markers within the environment of a tissue. Nevertheless, the lack of accumulation of these cells within the retina does not provide support for a possible contribution of stem cells in the observed beneficial effect of PBM on the retinopathy.

Previously, we reported that PBM therapy from the onset of disease or as an intervention therapy inhibited the oxidative stress and inflammation that develop in the retina in diabetes, as well as upregulating survival pathways (21,22). Others have reported similar findings in other conditions (52), and these effects might be mediated by effects on signaling molecules and activation of transcription factors, including nuclear factor-κB. These effects of PBM to inhibit the diabetes-induced increase in oxidative stress and inflammation-associated proteins likely play an important role in the observed inhibition of at least the dysfunction and degeneration of the retinal vasculature in DR, because genetic or pharmacologic inhibition of oxidative stress or inflammation previously was shown to inhibit development of these vascular lesions (1,43,44,57–64). Thus, our present working hypothesis is that the observed beneficial effects of PBM against the development of retinopathy in diabetic animals likely are secondary to PBM-induced changes in transcription or activity of critical enzymes having antioxidant and anti-inflammatory effects.

In summary, PBM for only 4 min/day significantly inhibited the development of several clinically accepted and important lesions of DR. In particular, daily administration of PBM for 8 months significantly inhibited the diabetes-induced increase in vascular permeability and capillary degeneration that is characteristic of DR. Moreover, PBM protected against the reduction in visual function that is caused by diabetes. These findings have important implications for PBM as a potential adjunct therapy to inhibit DR and possibly other degenerative processes. It seems likely that PBM is having beneficial effects on the retina via multiple mechanisms.

Acknowledgments. The authors thank Chieh Allen Lee (Case Western Reserve University, Cleveland, OH), who maintained the diabetic mouse colony, and Katie Franke and Heather Butler (Case Western Reserve University, Cleveland, OH) for measurements of visual function.

Funding. This work was supported by National Institutes of Health grants EY-022938 and R24-EY-024864, U.S. Department of Veterans Affairs grants BX002117-01 and BX003604 (to T.S.K.), and National Eye Institute core grant P30-EY-011373.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. Y.C., Y.D., H.L., and A.V. performed molecular and physiologic analyses and analyzed data. Y.C., A.V., and T.S.K. wrote the manuscript. J.T. evaluated histopathology. T.S.K. designed the experiments and analyzed data. T.S.K. is the guarantor of this work and, as such, had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.