Long-term ultraviolet B irradiation at 297 nm with light-emitting diode improves bone health via vitamin D regulation

Ultraviolet radiation is the primary determinant for vitamin D synthesis. Sunlight is inefficient and poses a risk, particularly for long-term exposure. In this study, we screened the most favorable wavelength for vitamin D synthesis among four types of narrowband light-emitting diodes (LEDs) and then irradiated osteoporosis rats with the optimal wavelength for 3–12 months. The 297 nm narrowband LED was the most efficient. Long-term radiation increased vitamin D levels in all osteoporotic rats and improved bone health. No skin damage was observed during irradiation. Our findings provide an efficient and safe method of vitamin D supplementation.


Introduction
Vitamin D is a type of multifunctional steroid hormone in the process of human physiological activities [1][2][3].However, vitamin D deficiency is a widespread challenge worldwide, existing in all ages, genders, countries, and regions [4][5][6][7].Approximately 91.2% of postmenopausal women in China suffer from vitamin D deficiency or insufficiency [8].Vitamin D deficiency is associated with various diseases, such as osteoporosis, obesity, diabetes, tumors, and immune system diseases [9][10][11][12].Maintaining an adequate vitamin D status is a prerequisite for body health, particularly for the bones.However, humans cannot synthesize vitamin D independently and require long-term supplementation.Therefore, providing adequate vitamin D supplementation to patients with vitamin D deficiency in a convenient, safe, and portable manner is crucial.
Vitamin D levels in the human body are increased using two primary methods: ultraviolet B (UVB) irradiation and oral supplementation.Currently, oral vitamin D is the most common supplementation [13]; however, oral supplementation is less effective in patients with absorption dysfunction.Moreover, owing to the decline in kidney function in elderly populations or patients with renal failure, the activity and expression of the enzyme used to synthesize 1,25(OH) 2 D 3 decrease, resulting in insufficient production of 1,25(OH) 2 D 3 [14][15][16].
When the skin is exposed to UV irradiation, epidermal 7-dehydrocholesterol (7-DHC) is transformed into previtamin D 3 (preVD 3 ), and preVD 3 then leads to the formation of vitamin D 3 through a thermochemical reaction in the epidermis.MacLaughlin et al. found that 297 nm was the most effective wavelength for vitamin D synthesis in the skin [17]; however, UV radiation mainly originates from sunlight or an artificial light source (arc lamp or fluorescent lamp), which has a wide light spectrum (280-400 nm).The optimal wavelength for promoting vitamin D synthesis accounts for only a small proportion of this range; thus, the synthesis efficiency is low.Owing to the limitations of irradiation, the synthesis of vitamin D in the skin is significantly affected.Moreover, adverse reactions such as skin photoaging [18,19], skin cancer [20,21], and skin pigmentation [22] occur upon exposure to UV light.
With the development of semiconductor materials and lighting technologies, light-emitting diodes (LEDs) have emerged as new light sources.The wavelength of the LED light source is customizable, and can be prepared narrowband (NB) (± 5 nm), as well as the LED light source can be made into a wearable device for easy irradiation.Currently, many researchers use LED light sources in the medical field [23][24][25], and UV irradiation has a positive effect on the concentration of vitamin D [26][27][28][29].
However, in the current research, LED light source did not emit light at 297 nm.Our previous study used 293 nm UV light to irradiate rats for two months and found that the device not only effectively increased the serum vitamin D level in rats but also did not cause any damage to the skin [30].However, vitamin D should be administered over a lifetime.It is not clear whether long-term radiation with a wavelength of 297 nm causes damage to the skin or if vitamin D synthesis is efficient.Its effect on bone metabolism is unknown.Furthermore, for patients to accept long-term radiation, the radiation method should be more convenient.
Therefore, we prepared a device with simulated wearable 297 nm NB-UVB irradiation and compared its effectiveness in vitamin D synthesis and safety for the skin with other light sources.Long-term irradiation was performed to observe the effects of the device on vitamin D metabolism, bone metabolism, muscle function, and skin safety.In summary, our study offers a new method for vitamin D supplementation and verifies its effectiveness and safety in animal experiments (Fig. 1).

UV light source
Narrowband UV LED devices (Fig. 2(a)-2(c)) developed by the Institute of Semiconductors, Chinese Academy of Sciences, in collaboration with the Laser Medicine Department of the PLA General Hospital, were used as the UV sources.The LED device spectrum was measured using a UV radiometer (Ocean Optics QE65 Pro).The power density of the devices was measured using a laser power meter (OPHIR, StarLite) to detect the power density of the device.2.2.PreVD 3 relative production preVD 3 relative production from a source of ultraviolet radiation is obtained by weighting the relative spectral irradiance of the radiation at wavelength λ in nm by the effectiveness of radiation of this wavelength, based on the action spectrum recommended by CIE (174-2006) [31] for the conversion of 7-DHC to preVD 3 .Comparison of preVD 3 relative production with different light sources was based on 298 nm (normalized to 100%) as the standard.
where Pλ is the relative spectral irradiance, that is, the proportion of different wavelengths in the total power.Rλ is relative response as CIE (174-2006) described, normalized to relative response 100% at 298 nm.

UV index
The UV Index is a quantitative index that represents the effective erythema irradiance level of ultraviolet radiation [32].
where K er is a constant equal to 40 m 2 /W.E λ is the spectral irradiance in W•m −2 •nm −1 at wavelength λ in nm and dλ is the wavelength interval used in the summation.The ability of UV radiation to produce erythema in the human skin is highly dependent on the radiation wavelength and is expressed by the erythema action spectrum (S er ).It is represented by relatively simple functions over three clearly defined spectral regions as follows: Ser (λ) = 10 0.094(298−λ) for 298<λ ≤ 328 nm (4) Ser (λ) = 10 0.015(140−λ) for 328<λ ≤ 400 nm (5)

Animals
The animal care and experimental procedures were approved by the Animal Center of the PLA General Hospital (2021-X17-134).To determine the most effective irradiation wavelength for vitamin D synthesis, 30 healthy 10-week-old female Sprague-Dawley (SD) rats were randomly divided into five groups.For long-term irradiation, 99 healthy 10-week-old female SD rats weighing 260-280 g were provided by the Beijing Keyu Animal Breeding Center.The animals were raised at a constant room temperature (25 ± 2 °C), with a relative humidity of 40-60% and good ventilation.They were fed a vitamin-deficient diet that did not contain vitamin D and contained 1.5% calcium, 1.0% phosphorus, and 3.0% total fat.Ninety-nine rats were randomly divided into three groups of 33 rats each: Sham, Ovariectomies (OVX), and LED groups.Bilateral ovariectomies were performed in the OVX and LED groups when rats were 12 weeks old.The hair on the dorsal skin was clipped to form a hair removal area of approximately 5 × 12 cm.After sterilization with an iodine tincture, 2% (mass fraction) pentobarbital sodium was injected into the abdominal cavity for anesthesia, and a longitudinal incision was made down the dorsal midline, with a length of approximately 2-3 cm.After the skin was incised, pink ovaries and closely connected uterine corners were observed below the kidneys on both sides.The uterine corners were ligated and cut, ovaries were removed, and muscles and skin were sutured.

LED irradiation
Rats were anesthetized by isoflurane inhalation, and a 5 × 6 cm dorsal patch of skin was cleanshaven as an area of irradiation (Fig. 2(c)).Before irradiation, the external power supply was used to adjust the voltage and current (Fig. 2(d)), control the power density (Fig. 2(e)).As previously described [29], the power density was 0.4 mW/cm 2 , radiation duration 5 min and radiation frequency 2 times/week.

Real-time qPCR
Total RNA was extracted from the muscle, skin, liver, and kidney using the TRIzol reagent (Cat#15596-026, Invitrogen) based on the manufacturer's instructions.PrimeScript RT Master Mix (RR036A, Takara) was used to transform RNA into cDNA.Real-time PCR was performed using the TB Green (RR820A, Takara) chimeric fluorescence method and CFX96 real-time PCR detection system.The primers used in this study are listed in Table 1.GAPDH was used as the housekeeping gene.Femurs were obtained 3, 6, 9, and 12 months after irradiation, fixed in 4% paraformaldehyde, and scanned using a micro-CT scanner (Inveon, Siemens, Germany).Each scan was performed with a full rotation of 360°, with a source voltage of 80 kV, and current of 500 µA.The exposure time was 1500 ms, and the pixel size was 20 µm.Using Inveon Research Workplace software to calculate the following parameters, bone volume/total volume (BV/TV), trabecular thickness (Tb.T, µm), number (Tb.N, 1/mm), spacing (Tb.Sp, mm), bone mineral density (BMD, mg/cm 3 ).
To determine the morphometric parameters of the trabecular bone, the volume of interest (VOI) started at 1 mm from the growth plate of the femur and extended 2 mm toward the diaphysis (2 mm in height) comprising the trabecular bone and the marrow cavity.
ImageJ software was used to analyze the images.

Muscle histology
Quadriceps were isolated from the rats and weighed.After isolation, the muscles were rapidly fixed in a 4% paraformaldehyde solution.Frozen 8 µm muscle sections were prepared for Masson staining.Cross-sections (mean fiber area) were analyzed using the ImageJ software based on Masson staining.

Grip strength experiment
After the electronic gravity scale was placed horizontally, the rats were gently placed on the contact plate of the grip grid of the tester.When the two forepaws grasped the grid, the rat's tail was grabbed and pulled back gently.After the rat grasped the grip grid, it was pulled until it released its paw with an even force.The instrument automatically recorded the maximum grip strength of each rat.

Rotating rod experiment
The rats were placed on a 9 cm-long rotating rod and crawled along the rod.The parameters of the rotating rod fatigue tester were set as follows: the initial speed was 4 rpm, the rotating speed was directional, the speed was accelerated to 50 rpm within 5 min, the tester stopped rotating, and the time for each rat (the longest time was 5 min) was recorded.

Open field experiment
The open-field experimental site was a square wooden box (100 × 100 × 50 cm) without a cover, and a camera was installed at the top two meters above the center of the site, which was connected to a computer.At the beginning of the experiment, the ANY maze software recorded the movement behaviors of the rats in real time.The test was conducted in a room with weak quiet lights between 18:00 and 20:00.Each rat was tested once every five minutes.A single test was conducted.The rats were placed at the center of the experimental site.The software tracked the rats and recorded the distance between them to evaluate the level of spontaneous activity.

Skin elasticity test
After the rats were anesthetized, the thumb and index finger were used to lift the mouse skin from the midline as much as possible (the hind limb touched the table) for 1s, released, and the time required for the skin to return to its original state was recorded.

Sample size and statistical analysis
The sample size was estimated using F test.In the experiment for screening the optimal wavelength, based on our preliminary experimental measurements of rat 25(OH)D 3 concentration, a sample size of n = 6 in each group was estimated to have sufficient power to detect significant differences among five groups (95% power, α=0.05).In the long-term radiation experiment, a sample size of n = 3∼6 in each group at each time point was estimated to have sufficient power to detect significant differences among three groups (95% power, α=0.05).
The data are expressed as mean ± standard deviation (SD).Two-tailed unpaired Student's t-tests were used for two-group comparisons.One-way analysis of variance (ANOVA) was used to assess the data of different groups, and the least significant difference (LSD) test was used for multiple comparisons.GraphPad Prism software was used for statistical analysis, and differences were considered statistically significant at p < 0.05.

Selected the optimal wavelength for promoting vitamin D synthesis
To determine the most effective irradiation wavelength for vitamin D synthesis, 280, 293, 297, and 310 nm narrowband (± 5 nm) LED were selected (Fig. 3(a)).Based on the action spectrum recommended by CIE for the conversion of 7-DHC to preVD 3 (CIE 174, 2006), normalized to relative response 100% at 298 nm, 293 nm, and 297 nm have a similar effect on the formation of preVD 3 , being 1.76 times and 3.53 times more effective compared to 280 nm and 310 nm, respectively (Fig. 3(b)).The UV index of the 297 nm LED was approximately 5, which belongs to medium-intensity exposure and causes little damage to the skin; however, 293 and 280 are greater than 6, which indicates high-intensity exposure (Fig. 3(c)).Further, 280, 293, 297, and 310 nm LED radiation was applied to the dorsal skin of the rats.After 4 weeks of irradiation, rats receiving 297 nm radiation had the highest concentration of 25(OH)D 3 compared to the other groups (Fig. 4(a)) (p < 0.05), and no significant difference was observed in 1,25(OH) 2 D 3 concentration between the irradiated and non-irradiated rats (Fig. 4(b)) (p > 0.05).Considering both its effectiveness and safety, we used a 297 nm device for subsequent experiments.

Effects of 297 nm NB-UVB LED long-term irradiation on vitamin D metabolites
Before ovariectomy, the concentrations of 25(OH)D 3 and 1,25(OH) 2 D 3 were not significantly different between the three groups (Sham, OVX and LED) (p > 0.05).After 3 months (0 mon), the 25(OH)D 3 concentration of all rats exhibited a downward trend, with the OVX and LED groups being significantly lower than the Sham group (p < 0.05).The 1,25(OH) 2 D 3 concentration in all rats also displayed a downward trend; however, no significant difference was observed among the three groups (p > 0.05).Subsequently, at 3 months (3 mon), 6 months (6 mon), and 9 months (9 mon), the concentration of 25(OH)D 3 in the LED group continued to increase, while the 25(OH)D 3 concentration in the Sham and OVX groups decreased continuously.The 25(OH)D 3 concentration in the LED group was significantly higher than those in the Sham and OVX groups at 3, 6, 9, and 12 months (p < 0.05).(Fig. 5(a)) The concentration of 1,25(OH) 2 D 3 decreased continuously at 3, 6, 9, and 12 months in the Sham and OVX groups; however, the rats in the LED group maintained normal levels.At 3 months, the 1,25(OH) 2 D 3 concentration in the LED group was significantly higher than that in  the OVX group (p < 0.05) and did not differ significantly from that in the sham group (p > 0.05).At 6, 9, and 12 months, the 1,25(OH) 2 D 3 concentration in the LED group was significantly higher than that in the OVX and sham groups (p < 0.05) (Fig. 5(b)).These results demonstrate that 297 nm UVB radiation can increase vitamin D levels in rats (Fig. 5(a), 5(b)).
In this study, we observed that the concentration of 25(OH)D 3 in the LED group rats did not show a linear increase, whereas the concentration of 1,25(OH) 2 D 3 remained stable.This indicates that the metabolism of vitamin D in the body is complex and that the supplementation dose may not align with the expected degree of increase.Vitamin D, which has a biological activity ( 5(c)), and estrogen levels did not appear to affect their expression (Fig. 5(d)).Moreover, the concentrations of 25(OH)D 3 and 1,25(OH) 2 D 3 affected the expression of related enzymes (Fig. 5(d)).In the skin and liver, the mRNA levels of CYP27A1 in the LED group had higher expression than in the OVX group at 3 months (p < 0.001).However, CYP27B1 mRNA levels between LED and OVX groups showed different trends in skin and kidney (Fig. 5(d)).CYP24A1 is a 24 hydroxylase, responsible for the hydroxylation of 1,25(OH) 2 D 3 to 1,24,25(OH) 3 D 3 (no biological activity).the mRNA level of CYP24A1 in LED group was higher than OVX groups (p < 0.01) (Fig. 5(d)).

Effects of 297 nm NB-UVB LED long-term irradiation on bone formation
Figure 6(a) and 7(a) show representative images of the bone morphology and histopathology, respectively.The number of bone trabeculae in all rats decreased over time, and the continuity of the bone trabeculae gradually deteriorated, particularly at 6-12 months.We further analyzed the bone morphological parameters using micro-CT (Fig. 6(a)).The bone mineral density (BMD) of rats in the LED and OVX groups decreased with age.At 3 months, the BMD of the LED group was significantly higher than that of the OVX group (p < 0.001).However, no significant differences were observed between the LED and OVX groups at 6-12 months (p > 0.05) (Fig. 6(b)).We also analyzed the other bone morphological parameters, such as bone volume/total tissue volume (BV/TV) (Fig. 6(c)), trabecular number (Tb.N) (Fig. 6(d)), trabecular thickness (Tb.Th) (Fig. 6(e)) and trabecular separation (Tb.Sp) (Fig. 6(f)) between OVX and LED groups.Except for Tb.Th, the other bone morphological parameters in the LED group were significantly better than those in the OVX group in the early stage of osteoporosis, and no significant difference was observed between the two groups in the late stage.
Bone morphological parameters were analyzed using hematoxylin and eosin staining (Fig. 7(a)).The BV/TV and number of osteocytes in the two groups decreased with age.The BV/TV ratio of the LED group was significantly higher than that of the OVX group at 3, 6, and 12 months (p < 0.05) (Fig. 7(b)).Although no statistically significant difference was noted between the LED and OVX groups at nine months (p > 0.05), we observed a higher trend in the LED group.The number of osteocytes in the LED group was higher than that in the OVX group at 3-9 months (Fig. 7(c)).Although no difference was observed between the LED and OVX groups at 12 months, we observed a higher trend in the LED group (Fig. 7(c)).These results showed that 297 nm UVB radiation can slow bone loss.

Effects of 297 nm NB-UVB LED irradiation on the balance of osteogenesis and osteoclasis, bone senescence, and bone mineralization
Osterix plays a key role in bone formation [33].Therefore, we measured the protein expression of osterix in the femur using immunofluorescence to observe the bone-forming ability of the  two groups of rats. Figure 8(a) shows that the expression of osterix protein was low in the OVX and LED groups at 12 months; however, the fluorescence intensity of the LED group was 1.82 times than that of the OVX group (p < 0.001) (Fig. 8(b)).The number of osteoclasts in the LED group was significantly lower than that in the OVX group at 12 months (p < 0.05) (Fig. 8(c),(d)).
The mineral apposition rate (MAR) in the OVX group was significantly lower than that in the LED group (p < 0.001) (Fig. 8(e),(f)).Aging is a major cause of osteoporosis [34]; therefore, we measured aging-related genes and the senescence-associated secretory phenotype (SASP) in the bone tissue.The gene expression of p21 and p16 in the OVX group were significantly higher than LED group (Fig. 8(g), (h)), TNF-α, IL-6, IL-1α, MMP3, and MMP13 were also significantly increased in the OVX group compared with LED group (p < 0.05) (Fig. 8(i)).

Effects of 297 nm NB-UVB LED long-term irradiation on muscle tissue
The muscle is an important organ that maintains body activity and balance.Decreased estrogen levels and aging can affect the muscle tissue.In this study, the quadriceps femoris weight in the LED group was significantly higher than that in the OVX group after LED irradiation (p < 0.05) (Fig. 9(a)).Masson staining showed that the quadriceps muscle fibers in the LED group were thicker and more closely arranged than those in the OVX group at 12 months (Fig. 9(b)), and the mean fiber area was significantly higher than that in the OVX group (p < 0.05) (Fig. 9(c)).Further analysis of muscle synthesis and age-related genes in the two rat groups The expression of myogenin in the LED group was higher than that in the OVX group (p < 0.05), and myostatin gene expression was lower in the LED group (p < 0.05) at 3, 9, and 12 months (Fig. 9(d),(e)).The expression of p21 and p16 was higher in the OVX group (p < 0.05) at 9 and 12 months (rats were 15 and 18 months old) (Fig. 9(f),(g)).In this study, LED radiation did not affect upper limb grip strength at 3 months (p > 0.05); however, significantly improved grip strength at 9 months (p < 0.05) (Fig. 9(h)).The activity and balance abilities of the rats were observed using a rotating rod experiment.Rats in the OVX group had shorter residence times on the rod at 3 and 9 months (p < 0.05) (Fig. 9(i)).Open-field experiment showed the same trend in the two groups at different time points as in the rotarod experiment (Fig. 9(j)).

Safety of 297 nm NB-UVB LED long-term irradiation on the rats
The long-term oral administration of vitamin D can cause poisoning, leading to damage to the body.Disorders of calcium and phosphorus metabolism are the most common complications of vitamin D poisoning.To verify whether long-term exposure to LED noninvasive phototherapy equipment could also lead to metabolic disorders of calcium and phosphorus in the body, serum calcium and phosphorus concentrations in rats were measured after 12 months of exposure.No significant differences were observed in calcium (p > 0.05) and phosphorus (p > 0.05) concentrations between the Con and LED group (Fig. 10(a), 9(b)).
UV rays can easily damage the skin, which is an obstacle to the use of UV light for medical treatment.This study used a 297 nm narrowband LED non-invasive phototherapy device to irradiate the dorsal skin of rats for 12 months and observed the effects of the device on rat skin after long-term irradiation.Similar to the OVX group, the skin of the LED group was smooth and no pigmentation, damage, or wrinkles were observed on its surface (Fig. 10(c)).Compared with the skin of the same site as the OVX groups, the skin of the LED group rats at the irradiation site showed a complete structure of the epidermal layer after 3, 6, 9, and 12 months of exposure, with an orderly arrangement of epidermal cells and collagen fibers in the dermis; no inflammatory cell infiltration was observed throughout the entire layer (Fig. 10(d)).The skin epidermal thickness was not significantly different among the three groups at 3 and 6 months; however, it was thicker in the LED group than in the OVX group at 9 (p < 0.05) and 12 months (p < 0.01) (Fig. 10(e)).A skin elasticity test was performed to detect photoaging.The recovery times of the OVX and LED groups were not significantly different (p > 0.05) (Fig. 10(f)).Immunohistochemistry revealed no melanocytes or keratinocytes with melanin pigmentation in the epidermal areas of the two groups.p53-, H2aX-, and Ki67-positive cells were not found in the LED group after 12 months of continuous irradiation (Fig. 10(g)).No significant differences were observed in IL-6, TNF-α and MMP3gene expression between the two groups at 3, 6, 9, and 12 mon (p > 0.05) (Fig. 10(h)-10(j)).

Discussion
Although the major source of vitamin D needed by humans comes from UVB irradiation [35,36], the insufficiency or deficiency of vitamin D remains a global challenge, especially in the elderly.This is primarily because sunlight exposure is affected by climate, altitude, season, and other factors.Artificial UVB light sources can only be used in hospitals, which severely limits patient access to adequate UVB radiation.Meanwhile, in special populations, such as the elderly, the expression of vitamin D synthesis-related enzymes decreases [15,16], the epidermal layer of the skin becomes thinner, and stores of 7-DHC decrease, leading to a reduction in the ability of the skin to synthesize vitamin D [37,38].Based on these challenges, portable UV irradiation equipment with high vitamin D synthesis efficiency is required.
MacLaughlin et al. used different UV wavelengths to irradiate human skin and found that 297 nm was the most beneficial for converting 7-DHC to preVD 3 in skin tissue [17], as the maximum absorption peak of 7-DHC was approximately 297 nm.LED can produce narrowband UV-B light sources that satisfy these requirements.Based on our previous research [30] and the action spectrum recommended by the CIE [31] for the conversion of 7-DHC to preVD 3 , we compared the effects of different light sources on the synthesis of preVD 3 based on their spectra.The results showed that 297 and 293 nm had similar synthesis effects on preVD 3 , which were significantly better than those at 280 and 310 nm.We further compared the effects of 280, 293, 297, and 310 nm radiation on the concentration of 25(OH)D 3 in rats.Interestingly, the improvement effect at 297 nm on the concentration of 25(OH)D 3 was significantly greater than that at 293 nm, which was inconsistent with the preVD 3 relative production which was calculated based on the action spectrum.The process from preVD 3 to 25(OH)D 3 is complex.PreVD 3 should be converted into vitamin D 3 in the skin, and then vitamin D 3 enters the liver through blood vessels to be converted into 25(OH)D 3 .The synthesis site of preVD 3 and vitamin D 3 is mainly the epidermal layer [39]; however, the epidermal layer does not contain blood vessels, which mainly exist at the junction of the dermis and epidermis [40].Therefore, we speculated that the synthesized vitamin D 3 in the epidermal layer closer to the dermis was more likely to enter blood vessels, thereby making it easier to convert into 25(OH)D 3 .The longer the wavelength, the deeper the penetration [41], which is more conducive to the entry of vitamin D 3 into blood vessels at the junction of the dermis and epidermis.This also explains why the synthetic effect of preVD 3 at 310 nm was not as strong as that at 280 nm; however, the concentration of 25(OH)D 3 was similar to that at 280 nm.Meanwhile, the 297 nm artificial light source had a much higher synthesis effect on preVD 3 than sunlight.Therefore, considering the synthesis of preVD 3 and the subsequent transport of vitamin D 3 , 297 nm is the optimal wavelength for vitamin D supplementation.
Vitamin D supplementation is essential because the body cannot synthesize it internally.Consequently, continuous external supplementation was required.However, the long-term efficacy of vitamin D supplementation with UVB irradiation remains unclear.In this study, we utilized a 297 nm NB LED to administer prolonged, continuous radiation to rats and measured their 25(OH)D 3 concentrations at 3, 6, 9, and 12 months.Our findings revealed a significantly higher concentration of 25(OH)D 3 in the rats that received LED irradiation than in those that did not.While other researchers have also used narrowband UV LED to irradiate rats or mice and observed a significant increase in serum 25(OH)D 3 concentration, their irradiation periods were comparatively short regarding the subjects' lifespan [26][27][28]42,43].
We also found that continued irradiation with a 297 nm narrowband LED did not increase the concentration of 25(OH)D 3 after reaching a certain level; however, rather maintained its concentration.Simultaneously, the concentration of 1,25(OH) 2 D 3 in the LED group remained at normal levels without significant improvement.The conversion of vitamin D 3 to 25(OH)D 3 and then to 1,25(OH) 2 D 3 is regulated by a variety of hormones and enzymes, which ultimately maintain the concentrations of 25(OH)D 3 and 1,25(OH) 2 D 3 within safe ranges.CYP27A1 and CYP27B1 are enzymes involved in the synthesis of 25(OH)D 3 and 1,25(OH) 2 D 3 respectively.In the present study, UV irradiation increased the expression of CYP27A1 in the liver and decreased the expression of CYP27B1 in the kidneys.We speculated that when vitamin D 3 enters the liver, high CYP27A1 expression is beneficial for converting vitamin D 3 into its stored form (25(OH)D 3 ), reducing vitamin D 3 waste.At the same time, the downregulation of CYP27B1 was still beneficial for maintaining the stored form without excessively increasing the concentration of 1,25(OH) 2 D 3 .This can reduce the risk of vitamin D poisoning, while effectively maintaining vitamin D levels.This study observed the additional advantages of vitamin D supplementation through UVB radiation.Keratinocytes in the skin contain CYP27A1 and CYP27B1, which work together to convert vitamin D into 1,25(OH) 2 D 3 [44][45][46].After the skin of rats was exposed to UVB, the expression of CYP27A1 and CYP27B1 in the skin tissue was upregulated, which was conducive to the synthesis of 1,25(OH) 2 D 3 .Ala-Houhala [47] and Krause [48] also found that UVB radiation can stimulate the expression of hydroxylases related to 1,25(OH) 2 D 3 synthesis and increase the ability of the skin to locally synthesize 1,25(OH) 2 D 3 .Although the amount of 1,25(OH) 2 D 3 synthesized by the skin is relatively low, these findings are particularly meaningful in elderly patients and those with renal failure.Krause et al. suggested that UVB irradiation is superior to oral administration in patients with chronic kidney disease [48].
Although improvements in vitamin D concentrations through UV radiation have been confirmed, many studies have attempted to improve the bone morphology of osteoporosis model animals or human subjects using UV radiation [26][27][28]42,43].Only a few studies have reported improved BMD in young rats [43].Furthermore, it is unclear whether long-term exposure to UV radiation can improve BMD in hypoestrogenic and aging rats.In this study, we found that 297 nm narrowband LED radiation lasting for 3 months in estrogen-reduced young rats improved their BMD; however, with increasing age, continuous UV radiation (6-12 months) did not improve BMD.Although we did not observe significant improvements in bone morphology after 6-12 months of continuous irradiation, we found that LED group showed significantly higher expression of Osterix protein and MAR in bone tissue compared to the OVX group, while the number of osteoclasts and aging-related markers (p16 and p21 gene expression) were significantly lower at 12 months.This indicates that 297 nm narrowband LED radiation promotes bone formation and mineralization, as well as inhibits bone resorption and aging at the cellular and molecular levels.Morita et al. [42] also found that UV radiation reduced the number of osteoclasts and increased bone mineralization ability but did not improve BMD and bone morphology.Based on guidelines for the treatment of osteoporosis [49][50][51][52], vitamin D is the basic drug used for treatment, and patients still need to be supplemented with other anti-osteoporosis drugs.
Osteoporosis-related fractures are the most important factors affecting the quality of life of patients and can even lead to death.Fractures are associated with bone loss and falls can be caused by muscle dysfunction.Vitamin D is essential for maintaining muscle function.Vidal et al. found that 1,25(OH) 2 D 3 concentration positively correlated with muscle mass [53].Girgis et al. found that vitamin D deficiency resulted in reduced grip strength in mice [54].In this study, we found that maintaining the concentration of 1,25(OH) 2 D 3 at a normal level could increase myogenin gene expression and reduce myostatin gene expression, thereby enhancing the muscle's synthetic capacity.Consequently, the muscle can exhibit increased muscle fiber area and muscle mass, which positively affect muscle function.This study also validated through grip strength experiments, rotating rod experiments, and open field experiments that the LED group rats showed significantly higher muscle strength, balance, and activity abilities than the OVX group.As this study used rats as the research subjects, it was unable to assess the impact of increased muscle function on the risk of falls in rats.However, the literature has indicated that muscle mass [55], muscle strength [56] are closely related to the occurrence of fractures.Based on the improvements in muscle strength, balance, and activity abilities observed in the LED group of rats in this study, it suggests that 297 nm narrowband LED not only improves vitamin D levels but also enhances muscle function, which may help reduce the incidence of fractures.
Frequent and long-term exposure to certain UV wavelengths can lead to photoaging.In this study, long-term irradiation with 297 nm narrowband LED on rats showed that the skin at the irradiated site was smooth, with no roughness or wrinkles observed.HE staining revealed neatly arranged collagen fibers, and the skin elasticity test showed normal resilience.Therefore, the long-term irradiation in this study did not induce skin aging in rats.Ultraviolet radiation induces photoaging of the skin primarily by increasing the expression of inflammatory factors in the skin after UV exposure, leading to an increase in MMP expression.The high expression of MMPs degrades the collagen fibers in the skin tissue, thereby causing skin aging [107].The energy density used to induce photoaging in SD rats with UVB irradiation in current studies is significantly higher than the dose used in this study.Liang et al.
[108] used a UVB light source to irradiate rats for 8 weeks, with 5 irradiations per week.The energy densities used for each irradiation in the first three weeks were 60 mJ/cm 2 , 120 mJ/cm 2 , and 180 mJ/cm 2 , respectively, and 240 mJ/cm 2 for each irradiation in the subsequent 5 weeks.Seven days after the end of the irradiation, the MMP3 gene expression in the rats treated by Liang et al. was more than 5 times higher than that in normal rats.In this study, we used an energy density of 120 mJ/cm 2 , twice a week, and during the long-term irradiation process, there was no observed increase in skin inflammatory factors.Therefore, we believed that the use of the radiation parameters in this study will not induce skin aging in rats.
For pigmentation and skin cancer, our research and Li et al. [30] did not find pigmentation and skin cancer through immunohistochemistry (Melan-A, Ki67, H2aX, and p53) after 297 nm or 293 nm radiation.Morita et al. [42] irradiated mouse skin with an LED device using a wavelength of 305 nm and found that the skin tissue of mice showed pigmentation.Melanocytes mainly exist in the lowest layer of the epidermis (the basal layer).The longer the ultraviolet wavelength, the easier it reaches the basal layer, thus stimulating melanocytes to produce excess melanin.Meanwhile, wavelengths of 260-280 nm are the most harmful to skin nucleic acids and proteins because the maximum absorption wavelengths of nucleic acids are 260 nm and 280 nm, respectively.Excessive protein and nucleic acid damage can cause skin cancer.Based on these risk factors, it is important to select appropriate wavelength to ensure the safety of UVB radiation.A wavelength of 297 nm avoids 260-280 nm, and because the wavelength is short, inducing pigmentation is unlikely.In fact, at the same power density, the UV index of the 297 nm narrowband LED was lower than that at 280 nm or 293 nm.Although the 310 nm UV index was the lowest, the device may cause problems such as melanosis, and the synthesis efficiency of preVD 3 is very low.Power density, radiation duration, and radiation frequency are factors that affect skin safety.Hearn et al. [57] observed the incidence of skin cancer in 3867 patients treated with narrowband UVB and found no significant correlation between NB-UVB treatment and basal cell carcinoma (BCC), squamous cell carcinoma (SCC), or melanoma.This finding implies that the optical parameters currently used in UVB for clinical treatment are safe and do not cause skin cancer.Our study and others confirmed that UVB radiation does not cause skin cancer under reasonable radiation parameters.
Excessive vitamin D supplementation has several disadvantages.Burt et al. [58] administered three doses of vitamin D orally, followed up for a long duration time found that the incidence of serious adverse events, such as hypercalcemia, was 8%-15%.As the digestive tract quickly absorbs vitamin D, the body can quickly convert vitamin D into 1,25(OH) 2 D 3 in its biological form.High concentrations of 1,25(OH) 2 D 3 can have serious adverse effects.In this study, long-term ultraviolet radiation kept 1,25(OH) 2 D 3 at normal levels and did not cause calcium or phosphorus metabolism disorders.This was related to the synthesis and transport modes.The skin can convert 7-DHC into preVD 3 immediately after receiving ultraviolet radiation, whereas the conversion of preVD 3 into vitamin D 3 takes approximately 1-3 days [39].Therefore, we believe that UV radiation is safer than oral radiation for maintaining vitamin D concentrations.
Long-term 297 nm narrowband LED radiation in rats requires a comprehensive consideration of its effectiveness in increasing vitamin D levels and its safety for the skin.Selecting appropriate power density, radiation duration, and radiation frequency is crucial, and this will also provide theoretical support for the next step in clinical applications.In this study, the selection of 0.4 mW/cm 2 was based on the our preliminary research.In the preliminary study [30], irradiated rats with different power densities ranging from 0.1 to 0.8 mW/cm 2 and found that within the range of 0.1 to 0.4 mW/cm 2 , the higher the power density, the more significant the increase in serum 25(OH)D 3 in rats; within the range of 0.4 to 0.8 mW/cm 2 , there was no significant difference in the effect of increasing serum 25(OH)D 3 in rats.Therefore, we chose a power density of 0.4 mW/cm 2 .On this basis, to ensure skin safety, we adjusted the radiation duration and finally determined to control the radiation duration at 5 minutes, which corresponds to an energy density of 120 mJ/cm 2 .At this energy density, our preliminary experiment did not observe any skin damage in rats.Therefore, considering the effectiveness and safety of UV radiation, we have chosen a power density of 0.4 mW/cm 2 .and radiation duration of 5 minutes.This frequency of 2 times per week can provide the skin with sufficient time to recover after each light exposure, reducing the risk of cumulative UV radiation damage and improving the safety of the UV exposure.Additionally, according to the research of Holick [39], it takes about 1-3 days for preVD 3 in the skin to be fully converted to vitamin D 3 .During this time, additional light exposure may decompose the already synthesized vitamin D 3 [59], this will cause waste of vitamin D 3 in the skin.Therefore, for the same considerations of effectiveness and safety, we irradiated the rats twice a week.
In addition to safety and effectiveness, poor convenience is the main factor that hinders people from receiving UV radiation.The skin of any part of the human body produces vitamin D after UV irradiation.The most exposed parts in daily life are the face, head, neck, and limbs.However, excessive exposure to UV radiation can lead to photoaging and melanin deposition in the face, head, and neck [60][61][62], affecting the beauty of patients and can also cause severe damage to the eyes.Although the limbs were easily exposed, they were not sufficiently flat to cause uneven radiation.At present, the UV irradiation light source is mainly sunlight or fluorescent lamps, which makes irradiation inconvenient.It should be irradiated outdoors or in professional institutions, which makes chest or dorsal skin irradiation problematic.Significant differences were noted in the synthesis of preVD 3 , owing to the influence of region, season, and time of sunlight exposure.LED can adjust their power density and shape freely and simultaneously, advantages that sunlight and fluorescent lamps do not possess.Wearable irradiation devices can be prepared based on the patient's needs.

Conclusion
Vitamin D deficiency is a widespread challenge worldwide owing to the limitations of ultraviolet (UV) irradiation.This paper reports a novel 297 nm narrowband LED device for the efficient and safe synthesis of vitamin D 3 .Wearable devices can also be fabricated based on the small and flexible characteristics of LED, further improving radiation convenience.Furthermore, the use of this device to increase vitamin D levels in rats effectively reduced bone loss.Improving vitamin D levels is crucial for maintaining the health of multiple organs.Therefore, these findings provide a fresh perspective for vitamin D supplementation and new opportunities for innovative device development for osteoporosis but are not limited to osteoporosis.

Fig. 1 .
Fig. 1.Schematic of the advantages and therapeutic effects of 297 nm narrowband LED for vitamin D deficiency and osteoporosis.

Fig. 2 .
Fig. 2. Irradiation equipment for animal experiments.(a) Radiation bulb and animal box for animal experiment.(b) Assembled irradiation bulb and animal placement box, with the black block on the back as the radiator.(c) Inside the animal box.(d) Anesthetic state of animals before radiation.(e) Voltage and current controller.(f) Detection wavelength and power density before radiation.

Fig. 3 .
Fig. 3. Theoretical calculation of the effectiveness and safety of different LEDs.(a) Spectrum of the LEDs in this study.(b) preVD 3 relative production of different spectra of 280, 293, 297, and 310 nm NB-UVB LED devices.(c) UV index of different spectra of 280, 293, 297, and 310 nm NB-UVB LED devices.

Fig. 4 .
Fig. 4. Optimal wavelength screening by rats experiments.(a) Effect of different wavelength NB-UVB radiation on the concentration of 25(OH)D 3 in rats over four weeks.(b) Effect of different wavelength NB-UVB radiation on the concentration of 1,25(OH) 2 D 3 in rats over four weeks.Data are presented as mean ± SD, n = 6, *p < 0.05 vs. the Con group, # p < 0.05 vs. the 280 nm group, & p < 0.05 vs the 293 nm group, and p < 0.05 vs the 297 nm group and ns p > 0.05.

Fig. 5 .
Fig. 5. Effects of 297 nm NB-UVB LED irradiation for vitamin D level and relative enzymes.(a) Effect of 297 nm UVB radiation on the concentration of 25(OH)D 3 in rats at different time points.(b) Effect of 297 nm UVB radiation on the concentration of 25(OH)D 3 in rats at different time points.(c) Expression of different genes of the Sham group at different ages.(d) Expression of different genes at different groups at 3 mon.All data are presented as mean ± SD, @ p < 0.05 vs. Sham and LED group, & p < 0.05 vs. OVX and LED group, # p < 0.05 vs. Sham and OVX group, ns p > 0.05, *p < 0.05, **p < 0.01, and ***p < 0.001.All the experiments were repeated independently for at least three times.
1,25(OH) 2 D 3 ), can only be synthesized after multistep hydroxylation.CYP27A1 is 25-hydroxylase responsible for the hydroxylation of vitamin D 3 to 25(OH)D 3 , and CYP27B1 is 1α-hydroxylase responsible for the hydroxylation of 25(OH)D 3 to 1,25(OH) 2 D 3 .The liver contains CYP27A1, the kidney contains CYP27B1, and keratinocytes in the skin contain all the enzymes required to synthesize 1,25(OH) 2 D 3 .The expression of Vitamin D metabolite-related enzymes in different organs decreased with age (Fig.

Fig. 6 .
Fig. 6.Micro-CT detection of rat bone tissue.(a) Micro-CT representative pictures of OVX and LED groups at different time points.(b)-(f) Quantitative analysis BMD, BV/TV, Tb.N, Tb.Th and Tb.sp of the two groups of rats based on micro-CT at different time points, respectively.All data are presented as mean ± SD, n = 6, ns p > 0.05, *p < 0.05 and ***p < 0.001 vs. the OVX group at the same time point.

Fig. 7 .
Fig. 7. HE staining of rat bone tissue.(a) HE staining representative pictures of trabecular bones between the OVX and LED groups at different time points.The red arrow indicates osteocyte.Red scale bar = 200 µm, black scale bar = 100 µm.(b) Quantitative analysis BV/TV of the two groups of rats based on HE staining at different time points.(c) Quantitative analysis of the number of osteocytes of the two groups of rats based on HE staining at different time points, respectively.All data are presented as mean ± SD, n = 6, ns p > 0.05, *p < 0.05, **p < 0.01, and ***p < 0.001 vs. the OVX group at the same time point.

Fig. 8 .
Fig. 8. Effects of 297 nm NB-UVB LED irradiation on bone formation, senescence, and mineralization in rats.(a) Immunofluorescence representative pictures for Osterix of femur among the three groups of rats at different time points.The red arrow indicates Osterix positive cell.Red scale bar = 100 µm, yellow scale bar = 25 µm.(b) Osterix fluorescence intensity of OVX and LED groups based on immunofluorescence.(c) Immunohistochemistry representative pictures for TRAP of the femur between OVX and LED groups at different time points.Red scale bar = 100 µm, yellow scale bar = 25 µm.(d) Number of osteoclasts of the two groups based on immunohistochemistry at different time points.n = 6.(e) Representative micrographs of calcein/xylenol orange dual-labeling at 9 mon.Scale bar = 10 µm.(f) MAR of the two groups based on calcein/xylenol orange dual-labeling at 9 mon.(g),(h) Expression of p21 (g) and p16 (h) gene in the femur at 9 and 12 mon.(i) Expression of SASP genes in the bones.All data are presented as mean ± SD, ns p > 0.05, *p < 0.05, **p < 0.01, and ***p < 0.001 vs. the OVX group.All the experiments were repeated independently for at least three times.

Fig. 9 .
Fig. 9. Effects of 297 nm NB-UVB LED irradiation on muscle tissue.(a) Muscle weight of quadriceps femoris between the two groups of rats at different time points.(b) Masson staining representative pictures of quadriceps femoris between the two groups of rats at 12 mon.Scale bar = 50 µm.(c) Quantitative analysis of the mean fiber area based on Masson staining.(d),(e) Gene expression of myostain and myogenin in the quadriceps femoris at different time points.(f),(g) Expression of p21 and p16 gene in quadriceps femoris at different time points.n = 3. (h)-(j) Muscle function test based on grip strength experiment (h), rotating rod experiment (i), and open field experiment (j).All data are presented as mean ± SD, ns p > 0.05, *p < 0.05, **p < 0.01, and ***p < 0.001 vs. the OVX group at the same time point.All the experiments were repeated independently for at least three times.

Fig. 10 .
Fig. 10.Safety of 297 nm NB-UVB LED irradiation for rats.(a),(b) Serum calcium and phosphorus levels between Con and LED group rats.(c) Skin appearance of the irradiated area between the two groups of rats at 12 mon.(d) HE staining representative pictures of skin between two groups of rats at different time points.Scale bar = 200 µm.(e) Quantitative analysis of the thickness of epidermis based on HE staining.(f) Results of skin elasticity test between the two groups of rats at 12 mon.(g) Immunohistochemistry representative pictures for Melan-A, Ki67, H2aX, and p53 skin between two groups of rats at 12 mon.Scale bar =200 µm.(h)-(j) Expression of IL-6, TNF-α and MMP3 gene the skin at different time points.All data are presented as mean ± SD, and ns p > 0.05 vs. the OVX or Con group, *p < 0.05, **p < 0.01 vs. the OVX group.All the experiments were repeated independently for at least three times.

Funding.
National Key Research and Development Program of China (2017YFB0403801).