Skip to main content
Download PDF

Zinc-metal–organic frameworks with tunable UV diffuse-reflectance as sunscreens

Journal of Nanobiotechnology volume 20, Article number: 87 (2022) Cite this article

Abstract

Background

UV exposure continues to induce many health issues, though commercial sunscreens are available. Novel UV filters with high safety and efficacy are urgently needed. Metal–organic frameworks (MOFs) could be a suitable platform for UV filter development, due to their tunable optical, electrical, and photoelectric properties by precise controlled synthesis.

Results

Herein, four zinc-based MOFs with various bandgap energies were chose to investigate their optical behaviors and evaluate their possibility as sunscreens. Zeolitic imidazolate framework-8 (ZIF-8) was found to possess the highest and widest UV reflectance, thereby protecting against sunburn and DNA damage on mouse skin and even achieving a comparable or higher anti-UV efficacy relative to the commercially available UV filters, TiO2 or ZnO, on pig skin, a model that correlates well with human skin. Also, ZIF-8 exerted appealing characteristics for topical skin use with low radical production, low skin penetration, low toxicity, high transparency, and high stability.

Conclusion

These results confirmed ZIF-8 could potentially be a safe and effective sunscreen surrogate for human, and MOFs could be a novel source to develop more effective and safe UV filters.

Graphical Abstract

Introduction

Ultraviolet (UV) radiation produced by sun falls into three categories, including UVA at 400–320 nm, UVB at 320–280 nm, and UVC at 280–200 nm. UV plays a key role for the living things and regulates many biology processes. For example, UV radiation at a suitable dose increases the production of natural endorphins and Vitamin D in the skin. However, excessive exposure of UV would lead to health risks, such as atrophy, melanin deposition, aging and skin carcinogenesis [1]. UVA is reported to induce aging, immunosuppression, and skin carcinogenesis through generation of reactive oxygen species (ROS) or upregulation of immunosuppressive cytokines. UVB potentially causes sunburn, induces DNA damage through increasing the expression of cyclobutane pyrimidine dimers (CPDs), and even leads to non-melanoma skin cancers [2, 3].

Commercially available organic UV filters, including benzophenones, salicylates, octinoxates, etc., have been developed for anti-UV protection with the mechanisms to absorb UV radiation and release the energy through photoreactions, fluorescence, or energy redistribution within UV filter [4]. However, these organic molecules can pass through follicles or stratum corneum, penetrate into skin cells and induce skin complications [5]. Also, these organic compounds would be further absorbed and distributed into the whole body, which potentially results in systemic toxicities (e.g. endocrine disruption) [5]. Alternatively, inorganic sunblock agents, such as titanium dioxide (TiO2) and zinc oxide (ZnO) nanoparticles with UV absorption and scattering characteristics have been used to protect against UV exposure and address the transdermal penetration concern of organic UV filters. However, TiO2 can induce the generation of ROS with UV irradiation, which potentially lead to damage to cells and tissues, and even the development of skin cancer, though carcinogenesis of TiO2 is still controversial [6,7,8]. To address these concerns, current strategies, including decorating the nanoparticles with aluminum oxide or silica oxide or addition antioxidant molecules, have been developed to relieve the photocatalytic activity [6,7,8]. However, the results are not satisfied due to the degradation of antioxidants after exposure to the sun for a long time [9]. ZnO also is used as a surrogate for UV filter with no obvious local toxicity and no penetration of ZnO nanoparticles into viable epidermis [10, 11]. However, Zn content dissociated from ZnO was detected in blood [10], which would potentially cause unknown biological effects, though the reported Zn dose in circulation is relatively low. Moreover, TiO2 and ZnO suspensions are prone to be opaque, which is not favored for cosmetic reasons. Therefore, novel candidates with high anti-UV efficiency, photostability, physiological stability, and transparent characteristics in suspensions are urgently needed for the development of UV filters.

Metal–organic frameworks (MOFs), the hybrid polymers formed with metal nodes and organic ligands through coordination bonds, are predominantly used in gas absorption and separation, catalysis, sensing, and biomedicine due to the properties of high stability and porosity, large surface area, and tunable functionalities. Unfortunately, MOFs have not been well exploited for sunscreen development. Semiconductor MOFs (e.g. Zn, Mg, Cu, Ni, Mn, or Sr-based MOFs) [12] can transport photogenerated charges between metal centers or between the metal ions and linkages for long-distance by hydrogen-bonded vertical π stacks or full π-d conjugation [2, 12,13,14,15,16,17]. Specially, the bandgap of MOFs, which influences the maximal excitation wavelength (λmax), can be tuned by selecting various metal nodes and organic linkers, modifying the conjugation of the linkers, and functionalizing the linkers with other groups, etc. According to photoelectric effect Eq. (1), the light will be absorbed if the wavelength is lower than λmax and scattered or reflected if the wavelength is higher than λmax.[12, 18] So, MOFs with tunable UV absorption and reflection characteristics due to structure variations provide us a novel channel to develop UV filters with on-demand anti-UV effects.

$$Eg = hv = h*\frac{c}{\lambda }\qquad \lambda = \frac{hc}{{Eg}}$$
(1)

where h is Planck constant (6.626 × 10–34 J·S), c is light velocity (3.0 × 108 m s−1).

Herein, four zinc-based MOFs with different bandgap energies (Egs), including zeolitic imidazolate framework-8 (ZIF-8), Zn3L3DMF2, MOF-5 and IRMOF-1 (isoreticular MOF-5), were selected to study their optical characteristics and find the potential sunscreens. Among them, ZIF-8 showed the widest and highest scattering of UVA and UVB, followed by IRMOF-1, MOF-5, Zn3L3DMF2, TiO2 and ZnO (Fig. 1C, D, Additional file 1: Fig. S2), thus ZIF-8 was selected as the final model MOFs to assess its potential as UV filters. ZIF-8 was more transparent compared to ZnO and TiO2 with high physiological stability, low skin toxicity, and weak ROS production (Figs. 1B, E, F, 6A, B). On the mouse skin, ZIF-8 successfully inhibited epidermal hyperplasia and collagen degradation caused by UV exposure. On the Ba-Ma miniature pig skin, a model that correlates well with human skin, ZIF-8 achieved better effects to inhibit epidermal hyperplasia and DNA damage than ZnO, and comparable effect relative to TiO2. Furthermore, negligible ZIF-8 was found to penetrate mouse and pig skin, combined with low ROS generation, making it safer for clinical use. All these results confirmed ZIF-8 could potentially be used as a novel sunscreen surrogate for human. As we are investigating a new use of metal organic frameworks and how they perform on UV absorbance/reflectance with structure variations, we believe that our results are relevant to wide research areas including aesthetic medicine, pharmacy, and materials science.

Fig. 1
figure 1

Physical and chemical characterizations of zinc-based MOFs. A TEM images of TiO2, ZnO and ZIF-8. B Digital photographs of TiO2, ZnO and ZIF-8. ZIF-8 suspension was more transparent relative to TiO2 and ZnO. C Diffuse reflection spectra and D calculated UV–Vis absorbance spectra for TiO2, ZnO and zinc-based MOFs. ZIF-8 showed the widest and highest UVA and UVB scattering with lowest λmax. E ZIF-8 degradation in artificial sweat (pH 6.5, 32 °C). F EPR spectra of POBN-OH spin abduct signal produced by TiO2, ZnO and ZIF-8 suspensions (800 µg mL−1 in ethanol). ZIF-8 induced less EPR signal compared to TiO2 after UV exposure. G Fluorescence spectra of TiO2, ZnO and ZIF-8 after excitation with UV light (308 nm). ZIF-8 showed an emission at 621 nm. H Thermographs and I quantitative analysis of glycerol, TiO2, ZnO and ZIF-8 after UV exposure for 2 h. (RT: room temperature) 2–3 °C temperature increases were observed for all four groups compared to room temperature

Full size image

Results and discussion

Physical and chemical characterizations of Zn-based MOFs

MOFs were successfully synthesized and the structures of MOFs, TiO2 and ZnO were confirmed in the experiment section (Additional file 1: Figs. S1–S16). TEM revealed sizes of 102.1 ± 18.9 nm and 114.8 ± 53.2 nm for TiO2 and ZnO, respectively, whereas ZIF-8 showed a similar size with 82.3 ± 24.5 nm in diameter (Fig. 1A, Additional file 1: Fig. S1), confirming TiO2, ZnO and ZIF-8 are in nanoparticulate form. The sizes for MOF-5, IRMOF-1, and Zn3L3DMF2 were 310.6 ± 142.4 nm, 47.1 ± 13.6 nm, and 256.0 ± 91.3 nm, respectively (Additional file 1: Fig. S13A–F). The zeta potentials also were measured, showing 31.7 ± 0.6 mV, 16.4 ± 0.7 mV, 29.5 ± 0.8 mV, − 9.9 ± 1.5 mV, − 7.0 ± 0.6 mV, and − 5.6 ± 0.6 mV for TiO2, ZnO, ZIF-8, MOF-5, IRMOF-1, and Zn3L3DMF2, respectively (Additional file 1: Fig. S13G). The previously reported bandgaps for MOFs, TiO2, and ZnO are ZIF-8 (5.5 eV) > IRMOF-1 (3.6 eV) > MOF-5 (3.4 eV) > TiO2 (3.2 eV) = ZnO (3.2 eV) > Zn3L3DMF2 (3.1 eV) [12, 19, 20], and the calculated λmax order should be ZIF-8 < IRMOF-1 < MOF-5 < TiO2 = ZnO < Zn3L3DMF2 according to photoelectric effect Eq. (1). However, the diffuse-reflectance results showed ZIF-8 exerted lowest λmax (236 nm), followed by IRMOF-1 (λmax: 290 nm), MOF-5 (λmax: 294 nm), TiO2 (λmax: 340 nm), Zn3L3DMF2 (λmax: 366 nm) and ZnO (λmax: 370 nm) (Fig. 1C, D, Additional file 1: Fig. S2), indicating the acquired λmax order from experiment did not correlate well with the calculated one, which could be due to red and blue shifts caused by the quantum size effect and dielectric confinement effect [21]. As shown in the Eq. (2) for the lowest exciton energy, π2/Ȓ2 could cause a blue shift with the decline of particle size, whereas A1/Ȓ and A0 would result in a red shift with the decreasing of particle size and rising of dielectric-constant ratio ϵ1/ϵ2 [21].

$$Eg\,=\,\frac{{\pi }^{2}}{{\hat R}^{2}}+ \frac{-8\pi {I}_{2}+4{I}_{3}-2{\pi }^{2}{\alpha }_{0}}{{\pi }^{2}}\frac{1}{\hat R}-{\bar \alpha}^{2}+ \varphi \left(\hat R\right)=\frac{{\pi }^{2}}{{\hat R}^{2}}+ \frac{{A}_{1}}{\hat R}+{A}_{0}+ \varphi \left(\hat R\right)$$
(2)

where Ȓ = R / aB* (R is radius of nanoparticle, aB* is the exciton Bohr radius in bulk material; The information of the other quantities in this equation could be found in Additional file 1: Eqs. S2.1–2.7 [21].

ZIF-8 showed the lowest λmax (236 nm) and highest bandgap, meaning ZIF-8 needs more energy to generate electronic transition from full band to conducting band, which resulted in a UV absorption at low wavelength range and exerted a UV reflection at high wavelength range. The wide reflection of UVA, UVB and even some UVC could potentially endow ZIF-8 with a high anti-UV efficacy. So ZIF-8 was selected as the model MOFs for further study.

A widely used in vitro sun protection factor (SPF) assay with timesaving and no human subject involvement properties was performed to compare the photoprotective efficacy of ZIF-8, TiO2 and ZnO [22, 23]. ZIF-8 revealed an SPF value of 15.6, while which were 15.3 and 9.6 for TiO2 and ZnO, respectively, suggesting ZIF-8 could potentially provide a comparable or even higher UV protection effect relative to TiO2 and ZnO, respectively (Additional file 1: Fig. S16). Additionally, the SPF values of TiO2 and ZnO correlated well with that reported in the literatures, with 11.68 for TiO2 (15%) and 8.74 for ZnO (15%) [24, 25].

The digital photographs were taken under normal room light for the suspensions of TiO2 (15 wt%), ZnO (15 wt.%) and ZIF-8 (15 wt%) in glycerol. ZIF-8 yielded a much more transparent suspension in glycerol relative to that of TiO2 and ZnO, which is a competitive advantage from aesthetic considerations (Fig. 1B). ZIF-8 showed a low degradation in artificial sweat (36.2 ± 0.5% of degradation within 24 h) (Fig. 1E), which could potentially lower skin penetration and result in less time for ZIF-8 clearance. Nanoparticles would induce conduction band electron (e−) and valence band hole (h+) after UV radiation, which could further react with surrounding medium and produce free radicals of H+, H2O2, ·OH or HO2. Therefore, we studied the free radical formation of ZIF-8, TiO2, ZnO, MOF-5, IRMOF-1 and Zn3L3DMF2 using electron paramagnetic resonance (EPR) method. Interestingly, ZnO produced most free radical of ·OH (1.3 × 1012 spins/mm3), followed by TiO2 (5.9 × 1011 spins/mm3), MOF-5 (4.1 × 1011 spins/mm3), Zn3L3DMF2 (2.1 × 1011 spins/mm3), IRMOF-1 (6.2 × 1010 spins/mm3) and ZIF-8 (2.3 × 1010 spins/mm3) (Fig. 1F, Additional file 1: Fig. S15), suggesting ZIF-8 induced the least EPR signal compared to TiO2, ZnO, and the other Zn-based MOFs and exhibited weakest photocatalytic activity to generate hydroxyl radical, a hazard to human skin if topically applied. We also measured the temperature change and fluorescence production after the samples were exposed to UV, as the absorbed UV energy would be transferred into heat, fluorescence, or phosphorescence. Only 2–3 °C temperature increases were observed for all four groups compared to room temperature (RT) (Fig. 1H, I), suggesting that ZIF-8, TiO2 and ZnO would not do a second burn harm to the skin. Interestingly, a fluorescence at 621 nm was detected for ZIF-8 after UV exposure, indicating that the absorbed UV energy could partially be released as fluorescence for ZIF-8 (Fig. 1G). All these results above confirmed the UV reflection and absorption of MOFs could be tuned by changing the ligands. The selected MOFs, ZIF-8, showed appealing characteristics compared to commercial UV filters with wider and higher UV reflection, lower ROS production, and more transparent appearance.

In vitro cytotoxicity

The toxicity of ZIF-8 to human immortalized epidermal keratinocytes (HaCaTs) and human epithelial keratinocytes (HEKas) cells was assessed using MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay, a widely used tool to estimate the metabolic activity of living cells, where formazan with intense purple-blue color was formed from the lightly colored tetrazolium salt due to enzymatic reduction [26]. ZIF-8 showed much lower cytotoxicity compared to ZnO on both cell lines. The HaCaT survival rate was as high as 91.3 ± 1.6% for ZIF-8 at 50 µg mL−1, whereas which was only 11.0 ± 3.9% for ZnO at the same concentration. The survival rate of HEKas was decreased from 94.5 ± 2.8% to 79.0 ± 2.8% and 77.5 ± 0.3% for ZIF-8 at 10, 25, and 50 µg mL−1, respectively. On the contrary, the viability rate of cells treated with ZnO was 81.9 ± 2.2%, 36.8 ± 3.0%, and 10.8 ± 0.8%, respectively. No obvious cytotoxicity was observed for TiO2 at these concentrations (Fig. 2A). The cell apoptosis induced by these three nanoparticles also was evaluated by observing the cell and nucleus morphologies or by calculating cell apoptosis rates using flow cytometry. Obvious cell apoptosis with cell shrinkages and chromatin condensation was observed for both HaCaTs and HEKas after ZnO treatment at 60 μg mL−1 for 12 h or 8 h, respectively (Fig. 2B, C). Furthermore, the flow cytometry analysis showed that HaCaT apoptosis rates were 2.36%, 21.78%, and 5.73% after exposure with TiO2, ZnO and ZIF-8, respectively, whereas which were 4.66%, 42.20%, and 15.45% for HEKas (Fig. 2D), indicating that ZIF-8 caused much less cell apoptosis and could be more biocompatible relative to ZnO. However, it is still controversial for the cell toxicity of ZnO, because these assays do not take skin delivery into consideration [27]. Also, toxicity depends on exposure and toxic nature, ZnO toxicity could potentially not mean that much with absence of cellular exposure [10].

Fig. 2
figure 2

Toxicity on HaCaTs and HEKas. A Cytotoxicity of TiO2, ZnO and ZIF-8 at various concentrations toward HaCaTs and HEKas using MTT assay. B Digital images of HaCaTs and HEKas after treatment with TiO2, ZnO or ZIF-8. C Nucleus images with DAPI staining after cells were treated with TiO2, ZnO or ZIF-8. Obvious chromatin condensation was observed for ZnO-treated cells. D Flow cytometry analysis of cell apoptosis induced by TiO2, ZnO and ZIF-8

Full size image

Protection against DNA damage after UV exposure

UVB is reported to induce oxidative DNA damage through the generation of free radicals and ROS [2, 28]. In order to test the protective effect of ZIF-8 against UVB, comet assay was performed, and DNA damage was quantified. Without protection, 71.5 ± 9.3% of HaCaTs possessed DNA tails (single-strand or double-strand DNA breaks), whereas which was significantly reduced to 51.7 ± 3.5% with ZIF-8 protection and 38.5 ± 9.5% with TiO2 protection. HEKas showed a similar result, with 74.9 ± 6.0% for no protection group but 52.7 ± 3.6% and 29.4 ± 2.1% for ZIF-8 and TiO2 groups, respectively (Fig. 3A–C). These results confirmed ZIF-8 could efficiently prevent DNA breaks in skin cells after UV exposure, possibly due to its high UV reflection capability. TiO2 also exerted high DNA protection effects, potentially caused by its high UV absorption [29]. ZnO did not show protective effect against DNA fragmentation, which may be due to UV exposure accelerated the entrance of ZnO to skin cells, resulted in more ROS production due to the increased ZnO level and finally DNA damage [28].

Fig. 3
figure 3

Photoprotective effects on HaCaTs and HEKas. A Representative fluorescence images of cells with DNA tail after pre-protection and subsequent UV exposure. B, C Quantitative analysis of (B) HaCaTs and (C) HEKas with DNA tail. D, E) Cell viability of (D) HaCaTs and E HEKas with pre-protection and subsequent UV exposure

Full size image

The UV shielding capability of ZIF-8, TiO2, and ZnO also was assessed using MTT assay. TiO2 significantly increased HaCaT viabilities to 60.9 ± 8.2% at 50 μg mL−1 and 65.6 ± 3.4% at 60 μg mL−1, compared to cells without protection (51.6 ± 4.2%). ZIF-8 also exerted a protective effect for HaCats, with 63.0 ± 2.7% cell viability at 60 μg mL−1. Unfortunately, ZnO did not protect HaCaTs against UV irradiation, as shown by negligible changes on cell viabilities. For HEKas, only ZIF-8 at 50 μg mL−1 revealed an obviously higher cell viability with 47.6 ± 0.3%, compared to non-protection group (41.6 ± 3.6%) (Fig. 3D, E). ZIF-8, TiO2 and ZnO were not toxic in this test, because the nanoparticles were washed away using PBS and the cells were cultured in fresh medium without nanoparticles after UV exposure. These results confirmed that both TiO2 and ZIF-8, but not ZnO, could efficiently protect cells against UV exposure.

Protection against ROS generations in skin cells

ROS levels in HaCaTs and HEKas were measured using confocal fluorescent microscopy and flow cytometry after UVB or UVA exposures. After UVB irradiation, HaCaTs and HEKas with ZIF-8 pretreatment showed no obvious or less increase of ROS relative to that of the cells without protection or with the protections of TiO2, ZnO, MOF-5, IRMOF-1 or Zn3L3DMF2 (Additional file 1: Figs. S19, 20). After UVA exposure, both cells revealed similar results, with few or less ROS generations in ZIF-8 group, whereas which were elevated in the groups of No protection, TiO2, ZnO, MOF-5, IRMOF-1 and Zn3L3DMF2 (Additional file 1: Figs. S21, 22). All these results confirmed ZIF-8 could effectively reduce ROS productions in skin cells with UVA or UVB exposures.

Protective effects on mouse skin

The anti-UV effects of ZIF-8 was assessed on the dorsal skin of BALB/c mice, which was divided into five parts (1 cm × 1 cm for each part), preincubated with either glycerol, ZnO, TiO2, or ZIF-8 at an optimized dose of 15% for 15 min, and then exposed to UVB (280—320 nm) at an optimized UV dose of 206 mJ m−2. The skin with no protection was used as control (Additional file 1: Figs. S23, 24). After three days, the skin with treatments of ZIF-8, ZnO and TiO2 showed much less ulceration, edema or erythema compared to that of glycerol or no protection groups (Fig. 4A, Additional file 1: Fig. S25), suggesting ZIF-8, ZnO and TiO2 are all protective against macroscopic skin damages from UVB. Also, the skin histological assay by H&E staining revealed that the epidermal thickness was significantly increased after UV exposure from 20.1 ± 2.6 μm for normal skin to 56.6 ± 4.1 μm and 45.5 ± 3.2 μm for no protection and glycerol groups, respectively. However, the epidermal thickness increase was significantly inhibited by TiO2, ZnO and ZIF-8 treatments, with an epidermal thickness of 26.5 ± 4.6 μm, 27.9 ± 2.1 μm, and 30.8 ± 4.1 μm, respectively. No obvious difference for epidermal thickness was observed for these three groups (Fig. 4B, D), confirming ZIF-8 achieved comparable anti-epidermal hyperplasia effects to TiO2 and ZnO on mouse skin, which would contribute to a smoother skin appearance.

Fig. 4
figure 4

Photoprotective effects on mouse skin. A Representative digital graphs of mouse dorsal skin three days after pre-protection and subsequent UV exposure. ZIF-8 group showed less erythema compared to no protection or glycerol groups. B Microscope photographs of skin with H&E staining. C Microscope photographs of skin with CPD immunohistochemistry. D, E Quantitative analysis of (D) epidermal thickness and E CPD positive cells in the skin after pre-protection and subsequent UV exposure. ZIF-8 inhibited epidermal hyperplasia and CPD formation

Full size image

UVB irradiation induces the formation of cyclobutane pyrimidine dimers (CPDs), causes DNA lesions, blocks normal DNA replication if DNA damage is not repaired and leads to mutagenesis and even carcinogenesis [30]. So CPDs formed in the skin after UV irradiation was stained using immunohistochemistry (IHC). Obvious CPDs were observed for the skin without protection (30.3 ± 0.5 CPD+ cells per mm2) and glycerol pretreatment (30.4 ± 4.1 CPD+ cells per mm2), but which was obviously reduced for the skins with ZIF-8, ZnO or TiO2 protections (10.6 ± 2.5, 15.7 ± 2.1, or 8.6 ± 0.5 CPD+ cells per mm2, respectively). ZIF-8 even achieved a significant reduction of CPD compared to ZnO, confirming ZIF-8 could protect skin against UV-caused DNA damage on mouse skin (Fig. 4C, E). The lowered CPD formation could be due to the reduced ROS formation for the skin with ZIF-8 shielding. After UVB irradiation, the ROS level was (2.4 ± 0.3) × 105 for the skin with ZIF-8 protection, which was significantly lower relative to that of TiO2 ((3.4 ± 0.6) × 105) and ZnO ((4.4 ± 1.1) × 105) groups, respectively, and even comparable to that of normal skin ((2.5 ± 0.2) × 105) without UVB exposure, suggesting ZIF-8 could effectively protect the skin against ROS generation caused by UVB irradiation (Additional file 1: Fig. S26).

Collagen is an essential component to maintain skin structure, and the degradation of which after UV radiation would cause wrinkling and laxity [31]. Thus, we performed Masson’s trichrome staining for paraffin sections of the skin. Compared to normal skin, a reduced collagen density and irregularly distributed collagen fiber were observed after UV exposure for no protection and glycerol groups, but not for ZIF-8, TiO2 and ZnO groups, confirming ZIF-8 exerted a similar protective effect to TiO2 and ZnO against collagen degradation induced by UV irradiation (Additional file 1: Fig. S27A). The collagen protective effect of ZIF-8 would be beneficial to the maintenance of a smooth skin after UV exposure.

UVB exposure is also reported to cause pro-inflammatory responses with elevated levels of pro-inflammatory cytokines (such as IL-1β, TNF-α and IL-6) and lead to inflammatory disequilibrium in the skin [32,33,34,35]. We found that IL-1β expression in the skin was increased for no protection and glycerol groups after UV exposure, while which was not observed for ZIF-8, TiO2 and ZnO groups, suggesting ZIF-8 could inhibit IL-1β expression, thereby mitigating pro-inflammatory responses (Additional file 1: Fig. S27B). Overexpression of IL-1β would result in overproduction of MMP-2, MMP-9 and MMP-12, which would further lead to the degradation of elastin and collagens I, II and III [32, 36,37,38]. The inhibition of ZIF-8 on IL-1β overexpression would contribute to the maintenance of collagen morphology and well collagen distribution in the skin, endowing a smooth, full and firm skin appearance.

Protective effects on pig skin

In order to evaluate the clinical translational potential of ZIF-8 as sun-screening agent, we further investigated the protective effect of ZIF-8 on Ba-Ma miniature pig, which shows similar skin histology compared to that of human in terms of horny layer thickness, epidermal and dermal thicknesses, epidermis/dermis ratio, and collagen physicochemical properties [39]. Also, Ba-Ma miniature pig exhibits some special characteristics, including heterogeneity of basal cells, granules of mast cells, serrated pattern for epidermal-dermal interface, and developed vascular system, which are found only on the skin of humans, nonhuman primates, and pigs [39]. The dorsal skin of Ba-Ma miniature pig was demarcated into squares (1 cm × 1 cm), randomly pretreated with glycerol, TiO2, ZnO, or ZIF-8 for 15 min, and exposed to UVB radiation at the optimized dose (544 mJ cm−2) with an optimized ZIF-8 dose of 15% (Additional file 1: Figs. S28, 29). After UV exposure, the skin with protections of ZIF-8, TiO2, or ZnO exhibited much less erythema compared to that of the skin without protection or with protection of glycerol (Fig. 5A, Additional file 1: Fig. S30). Also, the epidermal layer thickness was increased from 33.5 ± 0.9 μm for normal skin to 79.9 ± 6.7 μm or 74.1 ± 3.6 μm for no protection or glycerol groups. However, the epidermal layer thickness was significantly lowered by TiO2 (43.1 ± 0.9 μm), ZnO (53.7 ± 6.6 μm), or ZIF-8 (44.4 ± 2.5 μm) groups, suggesting all of them could inhibit the epidermal hyperplasia after UV exposure. Interestingly, the epidermal thickness of ZIF-8 group was similar to that of TiO2, which was much thinner than that of ZnO group, confirming ZIF-8 achieved comparable anti-epidermal hyperplasia effects to TiO2 and even better effects than ZnO (Fig. 5B, D).

Fig. 5
figure 5

Photoprotective effects on pig skin. A Representative digital images of pig skin after pre-protection and subsequent UV exposure. Fewer erythema was observed for ZIF-8 group, compared to no protection group. B Photographs of skin with H&E staining after pre-protection and subsequent UV irradiation. (Red arrows point to parakeratosis.) C γ-H2AX immunofluorescence images of skin sections after pre-protection and subsequent UV exposure. (Nucleus: blue; γ-H2AX: red; White arrows point to γ-H2AX+ cells.) D, E Quantitative analysis of (D) epidermal thickness and (E) γ-H2AX positive cells on pig skin with pre-protection and subsequent UV exposure. ZIF-8 inhibited epidermal hyperplasia and γ-H2AX formation

Full size image

UV exposure would induce skin DNA damage with the formation of DNA double-strand breaks (DSB), which further induces the phosphorylation of histone H2AX on serine 139 [40]. Thus, γ-H2AX formation is considered as a DSB-response marker and was assessed using immunofluorescence staining. UV irradiation induced remarkable γ-H2AX on no protection skin and glycerol pretreated skin, with 48.4 ± 7.3 and 37.8 ± 7.7 γ-H2AX positive (γ-H2AX+) cells per mm2, respectively. However, after pretreatment with TiO2, ZnO or ZIF-8, γ-H2AX+ cells were significantly reduced to 20.6 ± 1.2, 28.5 ± 3.2, and 22.6 ± 3.2 per mm2 (Fig. 5C, E), suggesting ZIF-8 achieved comparable effects to TiO2 and ZnO on inhibiting UV-induced DNA damage.

Porcine skin is confirmed to be a suitable model for sunscreen protection efficacy assessment because the universal sun protection factor obtained using porcine skin correlates well with that on human skin (correlation factor R2 = 0.98) [41]. The comparable or even higher anti-UV efficacy of ZIF-8 relative to TiO2 or ZnO on pig skin suggests that ZIF-8 could be a potential effective sunscreen surrogate for human.

Long-term in vivo toxicity

We studied the long-term toxicity by applying either glycerol, TiO2, ZnO or ZIF-8 to mouse dorsal skin once every other day for a total of 6 applications. The skin showed no obvious signs of acute histology toxicity or long-term inflammation for all four groups. The epidermis structures, including corneum structure, epidermis thickness, skin follicles and sebaceous structure, also were not affected for all groups (Fig. 6A, B). Also, the liver function and kidney function were normal and the tissues, including heart, liver, spleen, lung, and kidney, were not affected after treatments with TiO2, ZnO, or ZIF-8 (Additional file 1: Figs. S31, 32).

Fig. 6
figure 6

Long term toxicity and penetration into mouse or pig skins. After repeated dosing for total six times in half a month, the skin was collected and subjected for H&E staining. A Representative H&E staining images of mouse skin. B Quantitative analysis of epidermal thickness. No obvious epidermal thickness was increased after long-term application of ZIF-8. C–F Penetrations of (C, E) ZIF-8, ZnO, and (D, F) TiO2 into (C, D) mouse and E, F pig skin. Skin was treated with glycerol, ZIF-8, ZnO, and TiO2 for 6 h, then the Zn2+ and Ti4+ in skin were detected by ICP-MS. ZnO easily penetrated mouse and pig skin, but not for ZIF-8 and TiO2

Full size image

We further investigated the penetration of TiO2, ZnO, or ZIF-8 into mouse or pig skin and their accumulations in mouse blood, heart, liver, spleen, lung, and kidney by measuring Zn or Ti contents remaining in the tissues using inductively coupled plasma-mass spectrometry (ICP-MS). The normal Zn content in mouse skin was 13.5 ± 1.5 μg g−1 tissue, which was significantly increased to 18.2 ± 3.0 μg g−1 tissue for ZnO group. This correlates well with previous reports that Zn ions dissociated from ZnO could get access into human skin after multiple uses [10, 42]. However, the Zn content in mouse skin was kept at normal level after ZIF-8 treatment (15.0 ± 0.2 μg g−1) (Fig. 6C), possibly due to its hydrophobic surface [43]. No obvious increases of Zn level were observed in mouse blood and main organs (Additional file 1: Fig. S33). The Ti content in normal mouse skin with no treatments was 0.10 ± 0.03 μg g−1 tissue, however, which was slightly but not significantly increased to 0.30 ± 0.14 μg g−1 tissue after TiO2 exposure (Fig. 6D). Ti levels in blood and main organs were not affected with TiO2 treatment. (Additional file 1: Fig. S33). We also confirmed skin penetrations on pig, a model for human skin due to the similar skin morphology and permeability characteristics [44], which yielded similar results under the same treatment. The Zn level for ZnO group was significantly increased to 45.2 ± 3.5 μg g−1 tissue from 17.6 ± 6.6 μg g−1 tissue in normal pig skin. There was no obvious Zn level increase in the pig skin with glycerol (23.5 ± 4.3 μg g−1 tissue) or ZIF-8 (11.5 ± 1.8 μg g−1 tissue) treatments (Fig. 6E). TiO2 also did not increase Ti level on pig skin, compared to no treatment or glycerol groups (Fig. 6F). The lower skin penetration of ZIF-8 was potentially caused by its hydrophobicity [43], which would potentially contribute to a lower topical toxicity due to the low ZIF-8 level in skin, and further result in a lower systemic risk with less Zn ions in circulation before ZIF-8 was cleared completely.

Conclusions

Exposure to solar UV is still a significant health risk for human, and there is still urgent need for safe and effective sunscreens due to the safety and efficacy concerns of current commercially available sunscreens. Herein, we synthesize a serial of Zinc-based MOFs and confirm the UV absorption and reflection can be tuned by choosing different organic ligands. Due to the wide and high UV reflection, ZIF-8 is selected as model MOFs and the potential as UV filter is evaluated. ZIF-8 exhibits good biocompatibility, low radical production, weak skin penetration, and achieves a high anti-UV efficacy on both mouse and pig skin, respectively, suggesting ZIF-8 could be a potential sunscreen surrogate for human with high efficacy and safety. All these results demonstrate that Zinc-based MOFs could potentially be a suitable platform to develop sunscreens through tuning UV reflectance and other characteristics, such as hydrophobicity, stability, and photocatalytic activity.

Experimental section

Materials and animals

Zinc oxide (ZnO, ≥ 99.9%), titanium oxide (TiO2, ≥ 99.9%), zinc acetate dihydrate (Zn(Ac)2·2H2O, ≥ 99.9%), terephthalic acid (H2BDC, 99%), N,N-dimethylformamide (DMF, 99.8%) and 2-methylimidazole (2-MeIM) were obtained from J&K Chemicals (Beijing, China). Dimethyl sulfoxide-d6 (DMSO-d6, ≥ 99.9%) were purchased from Tian in Fuyu Fine Chemical Co., Ltd (Tianjin, China). 4,4’-stilbenedicarboxylic acid (LH2) was bought from Flurochem (UK). Dulbecco’s modified eagle medium (DMEM), minimum essential medium (MEM) and fetal bovine serum (FBS) were purchased from Thermo Fisher Scientific (Waltham, MA, USA). Thiazolyl blue (MTT) was bought from Biosharp (Hefei, China). IL-1β polyclonal antibody was obtained from Signalway Antibody LLC (College Park, USA). Phospho-histone H2AX (Ser139) antibody (γ-H2AX) was bought from Affinity Biosciences (Cincinnati, USA). Anti-thymine dimer antibody (cyclobutane pyrimidine dimers, CPDs) was obtained from Sigma-Aldrich (Louis, USA). Masson’s trichrome staining kit was purchased from Beijing Solarbio Science & Technology (Beijing, China). Annexin V-FITC/PI apoptosis detection kit was bought from Procell Life Science & Technology (Wuhan, China). 2,7-Dichlorofluorescein Diacetate (DCFH-DA) probe kit and DNA damage assay kit were obtained from Nanjing Jiancheng Bioengineering Institute (Nanjing, China). UVB light with a peak emission at 308 nm was bought from Sankyo Denki Co. (Tokyo, Japan).

Human immortalized epidermal keratinocytes (HaCaTs) were purchased from China Center for Type Culture Collection (CCTCC, Wuhan, China) and cultured in DMEM with 10% FBS and 1% antibiotics. Human epithelial keratinocytes (HEKas) were obtained from Jennio Biotech (Guangzhou, China) and reserved in MEM medium with 10% FBS and 1% antibiotics. Both cells were placed in a humidified incubator with 5% CO2 at 37 °C.

Male BALB/c mice (6 weeks) were obtained from Laboratory Animal Centre of Guangzhou University of Chinese Medicine. Male Ba-Ma miniature pig (2–3 months) was bought from Dongguan Songshan Lake Laboratory Animal Technology Co., Ltd (Guangdong, China). All the protocols for animal experiments were approved by the Animal Ethics Committee of Southern Medical University, China. (Approval number of the laboratory: L2018244).

Instruments and methodologies

Powder X-ray diffraction (PXRD) was performed on Bruker D8 advance (Bruker Corporation, Billerica, USA) at the speed of 10° min−1 with an angle range of 5°–60°. 1H NMR spectra were recorded on 400 MHz Bruker (Bruker Corporation, Billerica, USA). Fourier transform infrared (FT-IR) spectra from KBr pellets were carried out using a Nicolet iS10 spectrometer (Thermo, Waltham, MA, USA). Elemental analysis was performed on an Elementar Vario-EL Cube CHNS elemental analyzer (Vario EL cube, Hanau, Germany). Transmission electron microscope (TEM) imaging was acquired on Hitachi H-7650 microscope (80 kV, Hitachi, Tokyo, Japan). N2 isotherm measurements were carried out on ASAP 2460 (Micromeritics instrument Ltd., GA, Norcross, USA) at 77 K with 20–100 mg of sample per measurement. Inductively coupled plasma mass spectrometry (ICP-MS) was performed on Agilent 7700 (Agilent Technologies, Inc., Santa Clara, USA). X-ray photoelectron spectroscopy (XPS) were measured using K-Alpha (Thermo Scientific, Waltham, MA, USA). Thermogravimetric analysis (TGA) was acquired on TGA 5500 from 30 to 780 °C at a speed of 5 °C min−1 (TA Instruments, New Castle, USA). UV–visible diffuse-reflectance spectrum were performed from BaSO4 pellets on Lambda 950 (PerkinElmer Inc., Waltham, USA) equipped with photometric integrating sphere (150 mm Int. sphere). The zeta potentials of Zn-based MOFs were measured using Zetasizer Nano ZS (Malvern Panalytical, UK). To measure electron paramagnetic resonance (EPR), TiO2, ZnO, or ZIF-8 (2 mL, 800 μg mL−1) were mixed with free radical catcher α-(4-Pyridyl N-oxide)-N-tert-butylnitrone (POBN, 38.8 mg), irradiated with UV (200–400 nm) for 10 min at 25 ± 0.1 °C under standard atmospheric pressure, and analyzed using a A300 spectrometer (Bruker Corporation, Billerica, USA). The samples at 800 μg mL−1 but not 50 μg mL−1, were selected for EPR measurement, because ROS signal was too low to be detected at 50 μg mL−1, though ZIF-8 and TiO2 could protect skin cells against UV damage at 50 μg mL−1. (Additional file 1: Fig. S14).

The synthesis and characterization of MOFs

ZIF-8: ZIF-8 with various sizes were synthesized according to previously reported methods with minor modifications [45, 46]. Briefly, Zn(Ac)2·2H2O (2.28 × 10–1 mol L−1 in DMF, 2 mL) were added with various ratios of Zn2+ to 2-MeIM (4 mL of DMF, 2.28 × 10–1 mol L−1 for ZIF-8 1:2, 5.71 × 10–1 mol L−1 for ZIF-8 1:5, 9.14 × 10–1 mol L−1 for ZIF-8 1:8, 18.28 × 10–1 mol L−1 for ZIF-8 1:16) under stirring at 450 rpm to synthesize ZIF-8 with different sizes. After 7.5 h, the reaction solution was centrifuged at 8000 rpm for 2 min and washed with DMF (5 mL) and ethanol (5 mL) for 2 times, respectively. The samples were resuspended with ethanol and stored at – 80 °C for further use.

TEM, 1H NMR, PXRD, XPS, and N2 isotherm were conducted to confirm the structure of ZIF-8.

ZIF-8 1:2: 1H NMR (DMSO-d6/D2SO4 (9:1, v/v)): 7.49 (s, 2H, Imidazole H), 2.60 (s, 3H, –CH3). Yield: 29.7%. BET surface area was 1237.8 m2 g−1. Pore size: 9.3–15.0 Å. XPS data also revealed the successful synthesis of ZIF-8 1:2 (Additional file 1: Fig. S3).

ZIF-8 1:5: 1H NMR (DMSO-d6/D2SO4 (9:1, v/v)): 7.49 (s, 2H, Imidazole H), 2.56 (s, 3H, –CH3). Yield: 19.1%. BET surface area was 1306.9 m2 g−1. Pore size: 9.3–15.0 Å. XPS data also revealed the successful synthesis of ZIF-8 1:5 (Additional file 1: Fig. S4).

ZIF-8 1:8: 1H NMR (DMSO-d6/D2SO4 (9:1, v/v)): 7.38 (s, 2H, Imidazole H), 2.48 (s, 3H, –CH3). Yield: 19.0%. BET surface area was 1566.8 m2 g−1, which was similar to the reported values, potentially due to the high crystallinity and the excellent activation before N2 isotherm measurement. [47, 48] Pore size: 9.3–16.0 Å. XPS data also revealed the successful synthesis of ZIF-8 1:8 (Additional file 1: Fig. S5).

ZIF-8 1:16: 1H NMR (DMSO-d6/D2SO4 (9:1, v/v)): 7.49 (s, 2H, Imidazole H), 2.50 (s, 3H, –CH3). Yield: 15.7%. BET surface area was 1267.4 m2 g−1. Pore size: 9.3–15.9 Å. XPS data also revealed the successful synthesis of ZIF-8 1:16 (Additional file 1: Fig. S6).

ZIF-8 was observed using TEM and the particle sizes were plotted thereafter. The particle size were 164.8 ± 32.6 nm, 102.5 ± 26.8 nm, 82.3 ± 24.5 nm, and 80.0 ± 37.7 nm for ZIF-8 1:2, ZIF-8 1:5, ZIF-8 1:8, ZIF-8 1:16, respectively, confirming the size was decreased with the ratios of Zn2+ to 2-MeIM decreasing (Fig. 1A, Additional file 1: Figs. S1, S7A, B). These were similar to the previous report that ZIF-8 sizes were decreased from 110 to 40 nm with the ratios of Zn2+ to 2-MeIM decreasing from 1:2 to 1:16 [46]. However, no obvious ZIF-8 morphology variations were observed, (Fig. 1A, S7A) which were not consistent with the reports that morphologies of ZIF-8 turned from cube to sphere with the ratios of Zn2+ to 2-MeIM decreasing [46]. Possibly due to the decline of particle sizes, UV reflectance, especially for UVB and UVC, was enhanced with the decrease of Zn2+ to 2-MeIM ratios, which reached to the highest value for ZIF-8 1:8. No further enhancement was observed for ZIF-8 1:16, (Additional file 1: Fig. S7C) possibly because size of ZIF-8 1:16 was similar to that of ZIF-8 1:8. So ZIF-8 1:8 was selected as model ZIF-8 in the following experiments.

To study the energy release ways of ZIF-8 after UV exposure, TiO2, ZnO, or ZIF-8 (150 mg mL−1 in glycerol, 1 mL) were irradiated with UVB at the dose of 6.408 × 104 J m−2, thermal images were then taken by FLIR C2 Compact Thermal Camera (FLIR Systems, Wilsonville, USA). After irradiation, the fluorescence intensity of TiO2, ZnO, and ZIF-8 (50 μg mL−1 in ethanol) were obtained by F97 pro fluorescence spectrophotometer (Lengguang Tech., Shanghai, China) with excitation wavelength at 308 nm.

TEM, PXRD, XPS also were used to confirm the structures of TiO2 and ZnO (Fig. 1A, Additional file 1: Figs. S8, S9).

MOF-5: MOF-5 was synthesized as previously reported [49]. Zn(NO3)2·6H2O (290 mg, 1 mmol) was added into H2BDC (5.0 × 10–2 mol L−1 in DMF, 10 mL) and heated at 120 °C for 21 h. The reaction solution was centrifuged at 8000 rpm for 2 min and washed with DMF (5 mL) for 3 times. The sample was stored in ethanol at − 80 °C for further use. 1H NMR (DMSO-d6/D2SO4 (9:1, v/v)): 8.07 (s, 4H, Ar H). Yield: 14.2%. BET surface area: 741.5 m2 g−1. Pore size: 6.5—15.0 Å. XPS data confirmed the successful synthesis of MOF-5 (Additional file 1: Figs. S10, S13).

IRMOF-1: IRMOF-1 was synthesized according to previously reported method [50, 51]. Briefly, Zn(NO3)2·6H2O (10 mg, 0.034 mmol) was added into H2BDC solution (2.7 × 10–2 mol L−1 in DMF, 10 mL), and heated at 100 °C for 18 h. The crystals were centrifuged at 8000 rpm for 2 min and washed with DMF (5 mL) and ethanol (5 mL) for 3 times, respectively. The sample was stored in ethanol at − 80 °C for further use. 1H NMR (DMSO-d6/D2SO4 (9:1, v/v)): 8.07 (s, 4H, Ar H). Yield: 55.2%. BET surface area: 799.2 m2 g−1. Pore size: 6.0–25.0 Å. XPS data also revealed the successful synthesis of IRMOF-1 (Additional file 1: Figs. S11, S13).

Zn3L3(DMF)2: Zn3L3(DMF)2 was synthesized following previously reported methods [12]. Zn(NO3)2·6H2O (209.2 mg, 0.7 mmol) was added to LH2 (9.14 × 10–1 mol L−1 in DMF, 20 mL), heated at 75 °C for 16 h and followed by heating at 85 °C for 4 h. Thereafter, the samples were centrifuged at 8000 rpm for 2 min and the precipitate was washed with DMF (5 mL) and ethanol (5 mL) for 3 times, respectively. The sample was stored in ethanol at − 80 °C for further use. 1H NMR (DMSO-d6/D2SO4 (9:1, v/v)): 7.94 (s, 2H, Ar H), 7.82 (s, 2H, Ar H), 7.49 (s, 2H, C=CH), 2.93 (S, –CH3, 3H), 2.77 (S, 3H, –CH3). Yield: 12.8%. BET surface area: 910.1 m2 g−1. Pore size: 6.4–11.8 Å. XPS data also revealed the successful synthesis of Zn3L3(DMF)2 (Additional file 1: Fig. S12, 13).

The degradations of MOFs: The degradation of Zn-based MOFs in artificial sweat was assessed as previously reported. Briefly, ZIF-8 1:8 (10 mg, 0.5 mg), MOF-5 (0.5 mg), IRMOF-1 (0.5 mg), and Zn3L3(DMF)2 (0.5 mg) dispersed in glycerol were sealed into dialysis tubes, followed by an incubation in artificial sweat (0.5% NaCl, 0.1% lactic acid, and 0.1% urea, pH = 6.5, 32 °C ) at 100 rpm. At the predetermined time intervals, artificial sweat samples were collected and renewed with fresh media. The MOF ligands in artificial samples were measured with NanoPhotometer (NP80 Touch, IMPLEN, Germany). ZIF-8 (10 mg) showed the lowest degradation rate (36.2 ± 0.5% of degradation within 24 h), followed by ZIF-8 (0.5 mg) (46.6 ± 6.2% in 4 h, 72.0 ± 8.3% in 24 h), Zn3L3DMF2 (0.5 mg) (46.8 ± 6.8% in 4 h, 80.7 ± 9.2% in 24 h), MOF-5 (0.5 mg) (78.3 ± 5.0% in 4 h), and IRMOF-1 (0.5 mg) (86.8 ± 5.9% in 4 h) (Fig. 1E, Additional file 1: Fig. S13H).

In vitro SPF measurement: TiO2, ZnO, and ZIF-8 were dispersed in ethanol (15% by weight), respectively, and treated with ultrasonication for 10 min. The absorption at 290–320 nm was assessed using NanoDrop 1000 UV–VIS spectrophotometer (Additional file 1: Fig. S16). SPF values were calculated according to Mansur Eq. (3) below [22]:

$${SPF}_{\mathrm{spectrophotometric}}=\mathrm{CF} \times \sum_{290}^{320}\mathrm{EE }\left(\uplambda \right)\times \mathrm{I }\left(\uplambda \right)\times \mathrm{Abs }\left(\uplambda \right)$$
(3)

where CF = correction factor (10), EE (λ) = erythemal effect caused by the radiation with λ wavelength, I (λ) = solar intensity with λ wavelength, Abs (λ) = absorbance of samples. EE × I are constants as previously reported [23].

In vitro cytotoxicity assay HaCaTs or HEKas were seeded (2 × 104 cells per well) into 96-well plates and incubated for 24 h. Thereafter, the cells were treated with saline or various concentrations (1—100 μg mL−1) of TiO2, ZnO, or ZIF-8 for 24 h and then incubated with MTT solution (0.5 mg mL−1) at 37 °C for 4 h. Finally, DMSO (200 μL) was added to dissolve the resulted crystal and the absorbance at 570 nm were measured using a microplate reader (Multiskan FC, Thermo Scientific, Waltham, MA USA). Cell viability was expressed as a percentage of the absorbance to that of the control experiment without treatment.

Cell apoptosis assay Cells were seeded into 12-well plates (3 × 105 cells per well) and incubated for 24 h. After treated with TiO2, ZnO, or ZIF-8 (60 μg mL−1) (HaCaTs for 12 h and HEKas for 8 h), cells were photographed using a microscope with white light. Thereafter, cells were harvested, fixed with 4% paraformaldehyde at 4 °C for 30 min, permeabilized in 1% Triton-X 100 for another 30 min, and then stained with DAPI for 20 min. The stained cells were examined using fluorescence microscope (DMI8, Leica, Germany).

Cell apoptosis also was assessed using flow cytometry. Cells were grown in 6-well plates at a density of 1 × 106 cells per well and incubated to complete adhesion. Then, the cells were treated with TiO2, ZnO or ZIF-8 (60 μg mL−1, HaCaTs for 12 h and HEKas for 8 h). Thereafter, the cells were detached with trypsin, centrifuged at 300 g for 5 min, stained with Annexin V-FITC and propidium iodide (PI) for 20 min, and analyzed by flow cytometer (CytoFLEX LX, Beckman Coulter Biotechnology Co., California, USA).

Protection against UV-induced cell death The cell viability of HaCaTs or HEKas after exposure with various UV doses for 24 h was first assessed by MTT assay. Around 50% cell growth inhibition was achieved at UV doses of 35 mJ cm−2 for HaCaTs and 75.6 mJ cm−2 for HEKas (Additional file 1: Fig. S17). Also, the optimized UV doses are similar to the previously reported UVB irradiation doses (30 or 50 mJ cm−2) for cells [52, 53]. Thus, the two UV doses were used in the following protection experiments. HaCaTs or HEKas (30 μL of medium per well) in 96-well plates were treated with TiO2, ZnO or ZIF-8 at concentrations of 50 μg mL−1 or 60 μg mL−1 for 15 min, irradiated with UV lamp (emission peak 308 nm, Sankyo Denki Co., Taiwan, China) at the optimized doses, washed with PBS to remove the nanoparticles and incubated in fresh complete medium for another 24 h. The cell viability was then assessed using MTT assay.

Protection against DNA damage caused by UV irradiation Comet assay was used to determine the photoprotective effect of ZIF-8 to HaCaTs or HEKas. To optimize UV doses, cells were seeded into 12-well plates (5 × 105 cells per well) and incubated for 24 h. Thereafter, cells were exposed to four UV doses (84, 98, 114 or 126 mJ cm−2 for HaCaTs and 38.5, 52.5, 66.5 or 84 mJ cm−2 for HEKas) and incubated with 1 mL of complete medium for another 2 h. Cells were collected, mixed with 0.7% low melting point agarose. The cell suspensions (40 μL) were added onto slides with 1% normal melting point agarose. The slides were soaked in lysis buffer (4 °C ) for 1 h, gently washed with PBS, immersed in electrophoresis buffer for 18 min to allow DNA denaturation, and subjected to electrophoresis for 30 min (25 V, 200 mA, Horizontal electrophoresis system, DYY-6C, Beijing Six-one Instrument plant, Beijing, China). Subsequently, the samples were neutralized with Tris–HCl (pH 7.5, 5 min × 3 times), stained with PI for 10 min, photographed by Leica fluorescence microscope, and analyzed by Comet Assay Software Project (CASP). Obvious DNA tails could be observed for HaCaTs and HEKas after UV exposure at 114 mJ cm−2 and 66.5 mJ cm−2, respectively (Additional file 1: Fig. S18). So, the corresponding UV doses were selected for the following cell protection test.

To assess the protective effect against DNA damage, HaCaTs or HEKas were pretreated with TiO2, ZnO, or ZIF-8 at the concentration of 60 µg mL−1 for 15 min and irradiated with UVB (114 mJ cm−2 for HaCaTs, 66.5 mJ cm−2 for HEKas). Comet assay was then performed as above mentioned.

Intracellular production of ROS HaCaT cells and HeKa cells (2.5 × 105) were seeded in 24-well plates and incubated for 24 h, pretreated with TiO2, ZnO, or ZIF-8 at the concentration of 60 µg mL−1 for 15 min and irradiated with UVB or UVA (UVB: 350 mJ cm−2 for HaCaTs, 180 mJ cm−2 for HEKas; UVA: 100 mJ cm−2 for HaCaTs, 60 mJ cm−2 for HEKas). Thereafter, the cells were treated with 2′,7′—dichlorofluorescein diacetate (DCFH-DA) reactive oxygen species assay kits following the manufacture instructions. The intracellular ROS levels were assessed using confocal fluorescent microscopy and flow cytometry. (Negative control: Cells without UV exposure and without UV protection; Positive control: Cells with UV exposure but without UV protection.)

Protective effects on mice and pig skin The dorsal skin of each male BALB/c mouse (6-weeks old) was demarcated into six squares (1 cm × 1 cm), exposed to UV at various doses (172, 206, or 240 J m−2, respectively) to optimize UV dose. After three days, obvious erythema was observed for the skin with UV exposure at dose of 206 J m−2 and this UV dose was selected for further use (Additional file 1: Fig. S23). To optimize ZIF-8 dose, the skin squares were treated with ZIF-8 (1.5 μL) for 15 min with concentrations of 10%, 15%, or 20%, respectively, and exposed to UV at 206 J m−2. After three days, no obvious erythema was observed for the skin with ZIF-8 protection at the dose of 15%, so this ZIF-8 dose was selected for further use (Additional file 1: Fig. S24). To assess anti-UV effect, the skin squares were randomly treated with glycerol, TiO2, ZnO, or ZIF-8 (15%, 1.5 μL) for 15 min, exposed to UVB radiation (206 mJ cm−2) and photographed after 3 days. The mice were fed separately during the experiment. At the end time point, the skin was fixed with 4% paraformaldehyde, embedded in paraffin, sectioned with a thickness of 4 µm and subjected to Hematoxylin and Eosin (H&E) staining, Masson’s trichrome staining, or immunohistochemistry for CPD and IL-1β, respectively. The skin without protection was used as control. In these tests, the mouse and pig skin were pretreated with filters for 15 min following the World Health Organization (WHO) recommendation with minor changes, with the purpose to not affect the SPF value of these sunscreens [54, 55].

In vivo ROS in mouse skin after UV exposure The dorsal skin (1 cm × 1 cm) of each mouse was treated with glycerol (1.5 μL), TiO2, ZnO, and ZIF-8 (15%, 1.5 μL) for 15 min, respectively. Thereafter, the skin was exposed with UVB (18 mJ m−2) and collected after 3 h. After that, the cells were detached from the skin and stained using DCFH-DA. The ROS level was assessed using flow cytometry.

In vivo penetration into mouse skin The dorsal skin of the mice was demarcated into five squares (1 cm × 1 cm) after the hair was removed and randomly treated for 6 h with glycerol (1.5 μL), TiO2, ZnO, and ZIF-8 (15%, 1.5 μL), respectively. During the experiment, the mice were fed with 0.1 mL of water per hour by intragastric administration and heated on a pad at 37 °C. After that, the skins were topically washed with PBS (37 °C, 5 min × 3 times) and dried. The skin samples were collected and stripped for 30 times with tape, wiped with ethanol swabs for 3 times, weighed, and lysed with 70% HNO3 for 12 h. The levels of Zn2+ and Ti4+ in the skin were measured using ICP-MS. The skin without treatment was used as control.

In vivo long-term toxicity The mice were randomly divided into five groups as above mentioned. The dorsal skin was gently outlined into 2 cm × 2 cm squares using purple surgical marker and treated with glycerol (6 μL), TiO2, ZnO or ZIF-8 (15%, 6 μL) for 15 min every other day and 6 times in total. Three days after the last treatment, the skin, heart, liver, spleen, lung, kidney was collected, the paraffin section and the subsequent H&E staining were performed. Also, the Ti or Zn levels in heart, liver, spleen, lung, kidney, or blood were measured using ICP-MS. Additionally, blood biochemical parameters were measured to assess the system toxicity, including alkaline phosphatase (AKP), aspartate aminotransferase (AST), and alanine transaminase (ALT) for liver function, and creatinine (CRE) and serum urea nitrogen (BUN) for kidney function.

In vivo anti-UV effect on pig Ba-Ma miniature pig (60–90 days old, male) was anesthetized with 3.5% sodium pentobarbital (0.3 mL kg−1) and 10% xylazine hydrochloride injection (0.3 mL kg−1). During the experiment, the pig was given with supplemental anesthetics when necessary. The skin was demarcated into squares (1 cm × 1 cm) after the dorsal hair was removed. Thereafter, the skin squares were exposed to UV at various doses (442, 544, or 646 J m−2, respectively) to optimize UV dose. After one day, obvious erythema was induced by UV exposure at dose of 544 J m−2 and this UV dose was selected for further use (Additional file 1: Fig. S28). To optimize ZIF-8 dose, the skin squares were treated to ZIF-8 for 15 min with 1.5 μL of ZIF-8 at concentrations of 10%, 15%, or 20%, respectively, and exposed to UV at 544 J m−2. After one day, erythema was successfully inhibited by ZIF-8 protection with concentration of 15%, however, obvious erythema could still be seen for the skin with ZIF-8 treatment at 10%, so ZIF-8 dose at 15% was selected for further use (Additional file 1: Fig. S29). To assess anti-UV effect of ZIF-8 on pig, the skin squares were randomly pretreated with glycerol (1.5 μL), TiO2, ZnO, or ZIF-8 (15%, 1.5 μL) for 15 min, exposed to UVB radiation (544 mJ cm−2) and photographed again after 24 h. The skin was collected, and the paraffin section was performed, followed by H&E staining, Masson’s trichrome staining, and γH2AX immunofluorescence staining. The skin without protection was used as control.

Ex vivo penetration into pig skin Fresh pig skin was cut into pieces (1 cm × 1 cm), topically treated with glycerol, TiO2, ZnO or ZIF-8 (1.5 μL, 15%), respectively, and incubated at 32 °C in a humidity chamber for 6 h. The levels of Zn2+ and Ti4+ in the skins were assessed using ICP-MS after treatment following the methods in the section of “In vivo ZIF-8 penetration into mouse skin”. The skin with no treatment was used as control.

Abbreviations

MOFs:

Metal–organic frameworks

ROS:

Reactive oxygen species

Egs:

Bandgap energies

SPF:

Sun protection factor

RT:

Room temperature

ZnO:

Zinc oxide

TiO2 :

Titanium oxide

Zn(Ac)2·2H2O:

Zinc acetate dihydrate

H2BDC:

Terephthalic acid

DMF:

N,N-dimethylformamide

2-MeIM:

2-Methylimidazole

DMSO-d 6 :

Dimethyl sulfoxide-d6

LH2 :

4,4’-Stilbenedicarboxylic acid

DMEM:

Dulbecco’s modified eagle medium

MEM:

Minimum essential medium

FBS:

Fetal bovine serum

MTT:

Thiazolyl blue

γ-H2AX:

Phospho-histone H2AX (Ser139) antibody

DCFH-DA:

2,7-Dichlorofluorescein diacetate

HaCaTs:

Human immortalized epidermal keratinocytes

HEKas:

Human epithelial keratinocytes

PXRD:

Powder X-ray diffraction

TEM:

Transmission electron microscope

ICP-MS:

Inductively coupled plasma mass spectrometry

XPS:

X-ray photoelectron spectroscopy

TGA:

Thermogravimetric analysis

EPR:

Electron paramagnetic resonance

POBN:

α-(4-Pyridyl N-oxide)-N-tert-butylnitrone

PI:

Propidium iodide

H&E:

Hematoxylin and Eosin

References

  1. D’Orazio J, Jarrett S, Amaro-Ortiz A, Scott T. UV radiation and the skin. Int J Mol Sci. 2013;14(6):12222–48.

    PubMed  PubMed Central  Google Scholar 

  2. Tian Z, Yao T, Qu C, Zhang S, Li X, Qu Y. Photolyase-like catalytic behavior of CeO(2). Nano Lett. 2019;19(11):8270–7.

    CAS  PubMed  Google Scholar 

  3. Lo JA, Fisher DE. The melanoma revolution: from UV carcinogenesis to a new era in therapeutics. Science (New York, NY). 2014;346(6212):945–9.

    CAS  Google Scholar 

  4. McSweeney PC. The safety of nanoparticles in sunscreens: an update for general practice. Aust Fam Phys. 2016;45(6):397–9.

    Google Scholar 

  5. Deng Y, Ediriwickrema A, Yang F, Lewis J, Girardi M, Saltzman WM. A sunblock based on bioadhesive nanoparticles. Nat Mater. 2015;14(12):1278–85.

    CAS  PubMed  PubMed Central  Google Scholar 

  6. Livraghi S, Corazzari I, Paganini MC, Ceccone G, Giamello E, Fubini B, Fenoglio I. Decreasing the oxidative potential of TiO(2) nanoparticles through modification of the surface with carbon: a new strategy for the production of safe UV filters. Chem Commun (Camb). 2010;46(44):8478–80.

    CAS  Google Scholar 

  7. Pan Z, Lee W, Slutsky L, Clark RA, Pernodet N, Rafailovich MH. Adverse effects of titanium dioxide nanoparticles on human dermal fibroblasts and how to protect cells. Small. 2009;5(4):511–20.

    CAS  PubMed  Google Scholar 

  8. Aditya A, Chattopadhyay S, Gupta N, Alam S, Veedu AP, Pal M, Singh A, Santhiya D, Ansari KM, Ganguli M. ZnO nanoparticles modified with an amphipathic peptide show improved photoprotection in skin. ACS Appl Mater Interfaces. 2019;11(1):56–72.

    CAS  PubMed  Google Scholar 

  9. Caputo F, De Nicola M, Sienkiewicz A, Giovanetti A, Bejarano I, Licoccia S, Traversa E, Ghibelli L. Cerium oxide nanoparticles, combining antioxidant and UV shielding properties, prevent UV-induced cell damage and mutagenesis. Nanoscale. 2015;7(38):15643–56.

    CAS  PubMed  Google Scholar 

  10. Mohammed YH, Holmes A, Haridass IN, Sanchez WY, Studier H, Grice JE, Benson HAE, Roberts MS. Support for the safe use of zinc oxide nanoparticle sunscreens: lack of skin penetration or cellular toxicity after repeated application in volunteers. J Invest Dermatol. 2019;139(2):308–15.

    CAS  PubMed  Google Scholar 

  11. Mohammed YH, Haridass IN, Grice JE, Benson HAE, Roberts MS. Bathing does not facilitate human skin penetration or adverse cellular effects of nanoparticulate zinc oxide sunscreens after topical application. J Investig Dermatol. 2020;140(8):1656–9.

    CAS  PubMed  Google Scholar 

  12. Usman M, Mendiratta S, Lu KL. Semiconductor metal-organic frameworks: future low-bandgap materials. Adv Mater (Deerfield Beach, Fla.). 2017; 29(6):1605071.

    Google Scholar 

  13. Dong R, Han P, Arora H, Ballabio M, Karakus M, Zhang Z, Shekhar C, Adler P, Petkov PS, Erbe A, Mannsfeld SCB, Felser C, Heine T, Bonn M, Feng X, Cánovas E. High-mobility band-like charge transport in a semiconducting two-dimensional metal-organic framework. Nat Mater. 2018;17(11):1027–32.

    CAS  PubMed  Google Scholar 

  14. Yang C, Dong R, Wang M, Petkov PS, Zhang Z, Wang M, Han P, Ballabio M, Bräuninger SA, Liao Z, Zhang J, Schwotzer F, Zschech E, Klauss HH, Cánovas E, Kaskel S, Bonn M, Zhou S, Heine T, Feng X. A semiconducting layered metal-organic framework magnet. Nat Commun. 2019;10(1):3260.

    PubMed  PubMed Central  Google Scholar 

  15. Arora H, Dong R, Venanzi T, Zscharschuch J, Schneider H, Helm M, Feng X, Cánovas E, Erbe A. Demonstration of a broadband photodetector based on a two-dimensional metal-organic framework. Adv Mater (Deerfield Beach Fla). 2020;32(9): e1907063.

    Google Scholar 

  16. Papurello RL, Lozano LA, Ramos-Fernández EV, Fernández JL, Zamaro JM. Post-synthetic modification of ZIF-8 crystals and films through UV light photoirradiation: impact on the physicochemical behavior of the MOF. ChemPhysChem. 2019;20(23):3201–9.

    CAS  PubMed  Google Scholar 

  17. Haider G, Usman M, Chen T-P, Perumal P, Lu K-L, Chen Y-F. Electrically driven white light emission from intrinsic metal-organic framework. ACS Nano. 2016;10(9):8366–75.

    CAS  PubMed  Google Scholar 

  18. Alvaro M, Carbonell E, Ferrer B, Llabrési Xamena FX, Garcia H. Semiconductor behavior of a metal-organic framework (MOF). Chemistry (Weinheim an der Bergstrasse Germany). 2007;13(18):5106–12.

    CAS  Google Scholar 

  19. Rehman S, Ullah R, Butt AM, Gohar ND. Strategies of making TiO2 and ZnO visible light active. J Hazard Mater. 2009;170(2–3):560–9.

    CAS  PubMed  Google Scholar 

  20. Kuc A, Enyashin A, Seifert G. Metal-organic frameworks: structural, energetic, electronic, and mechanical properties. J Phys Chem B. 2007;111(28):8179–86.

    CAS  PubMed  Google Scholar 

  21. Takagahara T. Effects of dielectric confinement and electron-hole exchange interaction on excitonic states in semiconductor quantum dots. Phys Rev B Condensed Matter. 1993;47(8):4569–84.

    CAS  Google Scholar 

  22. Marto J, Gouveia LF, Gonçalves L, Chiari-Andréo BG, Isaac V, Pinto P, Oliveira E, Almeida AJ, Ribeiro HM. Design of novel starch-based pickering emulsions as platforms for skin photoprotection. J Photochem Photobiol B Biol. 2016;162:56–64.

    CAS  Google Scholar 

  23. Sayre RM, Agin PP, LeVee GJ, Marlowe E. A comparison of in vivo and in vitro testing of sunscreening formulas. Photochem Photobiol. 1979;29(3):559–66.

    CAS  PubMed  Google Scholar 

  24. Tyagi N, Srivastava SK, Arora S, Omar Y, Ijaz ZM, Al-Ghadhban A, Deshmukh SK, Carter JE, Singh AP, Singh S. Comparative analysis of the relative potential of silver, zinc-oxide and titanium-dioxide nanoparticles against UVB-induced DNA damage for the prevention of skin carcinogenesis. Cancer Lett. 2016;383(1):53–61.

    CAS  PubMed  PubMed Central  Google Scholar 

  25. Gutiérrez-Hernández JM, Escalante A, Murillo-Vázquez RN, Delgado E, González FJ, Toríz G. Use of agave tequilana-lignin and zinc oxide nanoparticles for skin photoprotection. J Photochem Photobiol B Biol. 2016;163:156–61.

    Google Scholar 

  26. Kumar P, Nagarajan A, Uchil PD. Analysis of cell viability by the MTT assay. Cold Spring Harb Protoc. 2018. https://doi.org/10.1101/pdb.prot095505.

    Article  PubMed  Google Scholar 

  27. Yamada M, Mohammed Y, Prow TW. Advances and controversies in studying sunscreen delivery and toxicity. Adv Drug Deliv Rev. 2020;153:72–86.

    CAS  PubMed  Google Scholar 

  28. Pal A, Alam S, Mittal S, Arjaria N, Shankar J, Kumar M, Singh D, Pandey AK, Ansari KM. UVB irradiation-enhanced zinc oxide nanoparticles-induced DNA damage and cell death in mouse skin. Mutation Res Genet Toxicol Environ Mutagenesis. 2016;807:15–24.

    CAS  Google Scholar 

  29. Egerton TA, Tooley IR. UV absorption and scattering properties of inorganic-based sunscreens. Int J Cosmet Sci. 2012;34(2):117–22.

    PubMed  Google Scholar 

  30. Mao P, Smerdon MJ, Roberts SA, Wyrick JJ. Chromosomal landscape of UV damage formation and repair at single-nucleotide resolution. Proc Natl Acad Sci USA. 2016;113(32):9057–62.

    CAS  PubMed  PubMed Central  Google Scholar 

  31. Ablon G. Safety and effectiveness of an automated microneedling device in improving the signs of aging skin. J Clin Aesthet Dermatol. 2018;11(8):29–34.

    PubMed  PubMed Central  Google Scholar 

  32. Kong SZ, Li DD, Luo H, Li WJ, Huang YM, Li JC, Hu Z, Huang N, Guo MH, Chen Y, Li SD. Anti-photoaging effects of chitosan oligosaccharide in ultraviolet-irradiated hairless mouse skin. Exp Gerontol. 2018;103:27–34.

    CAS  PubMed  Google Scholar 

  33. Fang IM, Yang CM, Yang CH. Chitosan oligosaccharides prevented retinal ischemia and reperfusion injury via reduced oxidative stress and inflammation in rats. Exp Eye Res. 2015;130:38–50.

    CAS  PubMed  Google Scholar 

  34. Han S, Kang SM, Oh JH, Lee DH, Chung JH. Src kinase mediates UV-induced TRPV1 trafficking into cell membrane in HaCaT keratinocytes. Photodermatol Photoimmunol Photomed. 2018;34(3):214–6.

    PubMed  Google Scholar 

  35. Li M, Lin XF, Lu J, Zhou BR, Luo D. Hesperidin ameliorates UV radiation-induced skin damage by abrogation of oxidative stress and inflammatory in HaCaT cells. J Photochem Photobiol B Biol. 2016;165:240–5.

    CAS  Google Scholar 

  36. Misawa E, Tanaka M, Saito M, Nabeshima K, Yao R, Yamauchi K, Abe F, Yamamoto Y, Furukawa F. Protective effects of Aloe sterols against UVB-induced photoaging in hairless mice. Photodermatol Photoimmunol Photomed. 2017;33(2):101–11.

    CAS  PubMed  Google Scholar 

  37. Fang L, Teuchert M, Huber-Abel F, Schattauer D, Hendrich C, Dorst J, Zettlmeissel H, Wlaschek M, Scharffetter-Kochanek K, Kapfer T, Tumani H, Ludolph AC, Brettschneider J. MMP-2 and MMP-9 are elevated in spinal cord and skin in a mouse model of ALS. J Neurol Sci. 2010;294(1–2):51–6.

    CAS  PubMed  Google Scholar 

  38. Taddese S, Jung MC, Ihling C, Heinz A, Neubert RH, Schmelzer CE. MMP-12 catalytic domain recognizes and cleaves at multiple sites in human skin collagen type I and type III. Biochem Biophys Acta. 2010;1804(4):731–9.

    CAS  PubMed  Google Scholar 

  39. Liu Y, Chen JY, Shang HT, Liu CE, Wang Y, Niu R, Wu J, Wei H. Light microscopic, electron microscopic, and immunohistochemical comparison of Bama minipig (Sus scrofa domestica) and human skin. Comp Med. 2010;60(2):142–8.

    CAS  PubMed  PubMed Central  Google Scholar 

  40. Oh K-S, Bustin M, Mazur SJ, Appella E, Kraemer KH. UV-induced histone H2AX phosphorylation and DNA damage related proteins accumulate and persist in nucleotide excision repair-deficient XP-B cells. DNA Repair. 2011;10(1):5–15.

    CAS  PubMed  Google Scholar 

  41. Weigmann HJ, Schanzer S, Patzelt A, Bahaban V, Durat F, Sterry W, Lademann J. Comparison of human and porcine skin for characterization of sunscreens. J Biomed Opt. 2009;14(2): 024027.

    PubMed  Google Scholar 

  42. Gulson B, Wong H, Korsch M, Gomez L, Casey P, McCall M, McCulloch M, Trotter J, Stauber J, Greenoak G. Comparison of dermal absorption of zinc from different sunscreen formulations and differing UV exposure based on stable isotope tracing. Sci Total Environ. 2012;420:313–8.

    CAS  PubMed  Google Scholar 

  43. Zhan WW, Kuang Q, Zhou JZ, Kong XJ, Xie ZX, Zheng LS. Semiconductor@metal-organic framework core-shell heterostructures: a case of ZnO@ZIF-8 nanorods with selective photoelectrochemical response. J Am Chem Soc. 2013;135(5):1926–33.

    CAS  PubMed  Google Scholar 

  44. Heylings JR, Davies DJ, Burton R. Dermal absorption of testosterone in human and pig skin in vitro. Toxicol In Vitro. 2018;48:71–7.

    CAS  PubMed  Google Scholar 

  45. Hunter-Sellars E, Saenz-Cavazos PA, Houghton AR, McIntyre SR, Parkin IP, Williams DR. Sol-gel synthesis of high-density zeolitic imidazolate framework monoliths via ligand assisted methods: exceptional porosity, hydrophobicity, and applications in vapor adsorption. Adv Funct Mater. 2021; 31(5):2008357.

    CAS  Google Scholar 

  46. Zheng WJ, Ding R, Yang K, Dai Y, Yan XM, He GH. ZIF-8 nanoparticles with tunable size for enhanced CO2 capture of Pebax based MMMs. Sep Purif Technol. 2019;214:111–9.

    CAS  Google Scholar 

  47. Park KS, Ni Z, Côté AP, Choi JY, Huang R, Uribe-Romo FJ, Chae HK, O’Keeffe M, Yaghi OM. Exceptional chemical and thermal stability of zeolitic imidazolate frameworks. Proc Natl Acad Sci USA. 2006;103(27):10186–91.

    CAS  PubMed  PubMed Central  Google Scholar 

  48. Shen K, Zhang L, Chen X, Liu L, Zhang D, Han Y, Chen J, Long J, Luque R, Li Y, Chen B. Ordered macro-microporous metal-organic framework single crystals. Science (New York, NY). 2018;359(6372):206–10.

    CAS  Google Scholar 

  49. Hafizovic J, Bjorgen M, Olsbye U, Dietzel PDC, Bordiga S, Prestipino C, Lamberti C, Lillerud KP. The inconsistency in adsorption properties and powder XRD data of MOF-5 is rationalized by framework interpenetration and the presence of organic and inorganic species in the nanocavities. J Am Chem Soc. 2007;129(12):3612–20.

    CAS  PubMed  Google Scholar 

  50. Winston EB, Lowell PJ, Vacek J, Chocholousova J, Michl J, Price JC. Dipolar molecular rotors in the metal-organic framework crystal IRMOF-2. Phys Chem Chem Phys. 2008;10(34):5188–91.

    CAS  PubMed  Google Scholar 

  51. Eddaoudi M, Kim J, Rosi N, Vodak D, Wachter J, O’Keeffe M, Yaghi OM. Systematic design of pore size and functionality in isoreticular MOFs and their application in methane storage. Science (New York, NY). 2002;295(5554):469–72.

    CAS  Google Scholar 

  52. Balupillai A, Nagarajan RP, Ramasamy K, Govindasamy K, Muthusamy G. Caffeic acid prevents UVB radiation induced photocarcinogenesis through regulation of PTEN signaling in human dermal fibroblasts and mouse skin. Toxicol Appl Pharmacol. 2018;352:87–96.

    CAS  PubMed  Google Scholar 

  53. Zhong J, Li L. Skin-derived precursors against UVB-induced apoptosis via Bcl-2 and Nrf2 upregulation. Biomed Res Int. 2016;2016:6894743.

    PubMed  PubMed Central  Google Scholar 

  54. Petersen B, Wulf HC. Application of sunscreen—theory and reality. Photodermatol Photoimmunol Photomed. 2014;30(2–3):96–101.

    PubMed  Google Scholar 

  55. Kim SY, Lamichhane S, Jung-Hun J, Yun J. Protective effect of octylmethoxycinnamate against UV-induced photoaging in hairless mouse via the regulation of matrix metalloproteinases. Int J Mol Sci. 2018;19(7):1836. https://doi.org/10.3390/ijms19071836.

    Article  CAS  PubMed Central  Google Scholar 

Download references

Acknowledgements

This study was supported by National Natural Science foundation of China (No. 31971321, No. 81801815, No. 32001007, No. 31901002), Natural Science Foundation of Guangdong Province (No. 2020A1515010826), Guangzhou Science, Technology and Innovation Commission (No. 202002030498).

Author information

Author notes

  1. Jisheng Xiao and Haishan Li contribute equally to this work

Authors and Affiliations

  1. Translational Medicine Research Center, Zhujiang Hospital, Southern Medical University/The Second School of Clinical Medicine, Southern Medical University, Guangzhou, 510515, Guangdong, China

    Jisheng Xiao

  2. Cancer Research Institute, School of Basic Medical Sciences, Southern Medical University, Guangzhou, 510515, China

    Jisheng Xiao, Chengyuan Cai, Tingting You, Zhenyu Wang, Jinmei Cheng, Jiaxin Li & Xiaopin Duan

  3. Science and Technology Innovation Center, Guangzhou University of Chinese Medicine, Guangzhou, 510405, China

    Jisheng Xiao, Haishan Li & Wanling Zhao

  4. Department of Pharmacology, School of Pharmacy, Fudan University, Minhang Hospital, Shanghai, 201203, China

    Mengling Wang

  5. Artemisinin Research Center, Institute of Science and Technology, The First Affiliated Hospital, The First Clinical Medical School, Lingnan Medical Research Center, Guangzhou University of Chinese Medicine, Guangzhou, 510405, China

    Feng Zeng

Authors
  1. Jisheng Xiao
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Haishan Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Wanling Zhao
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Chengyuan Cai
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Tingting You
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Zhenyu Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Mengling Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Feng Zeng
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Jinmei Cheng
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Jiaxin Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Xiaopin Duan
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

XD and JX jointly developed the idea of this project. XD conceived, directed, and oversaw the project. JX and XD designed the experiments and wrote the manuscript. JX did the data analysis of UV reflectance and SPF measurement. HL, CC, TY, ZW performed the experiments with the help of WZ, MW, FZ and JC. HL, CC, TY did the data analysis. JL provided TOC figures. All authors read and approved the final manuscript.

Corresponding authors

Correspondence to Jisheng Xiao or Xiaopin Duan.

Ethics declarations

Ethics approval and consent to participate

All the protocols for animal experiments were approved by the Animal Ethics Committee of Southern Medical University, China. (Approval number of the laboratory: L2018244).

Consent for publication

All authors have seen the manuscript and approved the submission.

Competing interests

The authors have declared that no competing interest exists.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Additional file 1: Figure S1.

Particle size of TiO2, ZnO and ZIF-8. Particle size was measured using TEM and plotted thereafter. The peak sizes for TiO2, ZnO, and ZIF-8 were 102.1 nm, 114.8 nm and 82.3 nm, respectively. Figure S2. Enlarged figures for (A) Fig. 1C and (B) Fig. 1D. Figure S3. Characterization of ZIF-8 1:2. (A) PXRD pattern. (B) 1H NMR spectrum. (C) XPS spectrum. (D) N2 adsorption and desorption isotherms. (E) Pore size distribution. Figure S4. Characterization of ZIF-8 1:5. (A) PXRD pattern. (B) 1H NMR spectrum. (C) XPS spectrum. (D) N2 adsorption and desorption isotherms. (E) Pore size distribution. Figure S5. Characterization of ZIF-8 1:8. (A) PXRD pattern. (B) 1H NMR spectrum. (C) XPS spectrum. (D) N2 adsorption and desorption isotherms. (E) Pore size distribution. Figure S6. Characterization of ZIF-8 1:16. (A) PXRD pattern. (B) 1H NMR spectrum. (C) XPS spectrum. (D) N2 adsorption and desorption isotherms. (E) Pore size distribution. Figure S7. Physical and chemical characterizations of ZIF-8 (1:2, 1:5, 1:16). (A) TEM images of ZIF-8 (1:2, 1:5, 1:16). (B) Particle sizes of ZIF-8 (1:2, 1:5, 1:16). Particle size was measured using TEM and plotted thereafter. The sizes for ZIF-8 (1:2, 1:5, 1:16) were 164.8 ± 32.6 nm, 102.5 ± 26.8 nm, and 80.0 ± 37.7 nm, respectively. (C) Diffuse reflection spectra for ZIF-8 (1:2, 1:5, 1:16). UV reflectance, especially for UVB and UVC, was enhanced with decreasing the ratios of Zn2+ to 2-MeIM and reached to the highest value for ZIF-8 1:8. UV reflectance was not further increased ZIF-8 1:16. Figure S8. Characterization of TiO2. (A) PXRD pattern. (B) XPS spectrum. Figure S9. Characterization of ZnO. (A) PXRD pattern. (B) XPS spectrum. Figure S10. Characterization of MOF-5. (A) PXRD pattern. (B) 1H NMR spectrum. (C) XPS spectrum. (D) N2 adsorption and desorption isotherms. (E) Pore size distribution. Figure S11. Characterization of IRMOF-1. (A) PXRD pattern. (B) 1H NMR spectrum. (C) XPS spectrum. (D) N2 adsorption and desorption isotherms. (E) Pore size distribution. Figure S12. Characterization of Zn3L3DMF2. (A) PXRD pattern. (B) 1H NMR spectrum. (C) XPS spectrum. (D) N2 adsorption and desorption isotherms. (E) Pore size distribution. Figure S13. Physical and chemical characterizations of zinc-based MOFs. (A-C) TEM images. (A) MOF-5. (B) IRMOF-1. (C) Zn3L3DMF2. (D-F) Particle size distributions of MOF-5, IRMOF-1 and Zn3L3DMF2. Particle size was measured using TEM and plotted thereafter. The sizes were 310.6 ± 142.4 nm, 47.1 ± 13.6 nm, and 256.0 ± 91.3 nm for MOF-5, IRMOF-1, and Zn3L3DMF2, respectively. (G) The zeta potentials of TiO2, ZnO, ZIF-8, MOF-5, IRMOF-1, and Zn3L3DMF2 were 31.7 ± 0.6 mV, 16.4 ± 0.7 mV, 29.5 ± 0.8 mV, − 9.9 ± 1.5 mV, − 7.0 ± 0.6 mV, and − 5.6 ± 0.6 mV, respectively. (H) The degradation of zinc-based MOFs in artificial sweat (pH 6.5, 32 °C). ZIF-8 (0.5 mg) exhibited the lowest degradation rate relative to that of MOF-5 (0.5 mg), IRMOF-1 (0.5 mg), and Zn3L3DMF2 (0.5 mg). Figure S14. EPR spectra of POBN-OH spin abduct signal produced by TiO2 suspensions in ethanol. (A) TiO2 at 800 µg mL−1, (B) TiO2 at 50 µg mL−1. No obvious EPR signal was detected for TiO2 at 50 µg mL−1. Figure S15. EPR spectra of POBN-OH spin abduct signal produced by suspensions of TiO2, ZnO, and Zn-based MOFs at 800 µg mL−1 in ethanol. (A) TiO2, (B) ZnO, (C) ZIF-8, (D) MOF-5, (E) IRMOF-1, (F) Zn3L3DMF2. ZnO produced most free radical of ·OH (1.3 × 1012 spins/mm3), followed by TiO2 (5.9 × 1011 spins/mm3), MOF-5 (4.1 × 1011 spins/mm3), Zn3L3DMF2 (2.1 × 1011 spins/mm3), IRMOF-1 (6.2 × 1010 spins/mm3) and ZIF-8 (2.3 × 1010 spins/mm3), suggesting Zn-based MOFs induced much less EPR signal compared to TiO2 and ZnO after UV exposure. Figure S16. UV absorbance of ZIF-8, TiO2, ZnO and 2-MeIM. ZIF-8 showed a higher UV absorbance compared to ZnO. Also, ZIF-8 revealed a higher UVB absorption relative to TiO2, though UVA absorption of is lower. Figure S17. Cell viability of (A) HaCaTs or (B) HEKas after exposed with UV in various doses. Figure S18. Fluorescence images of DNA tail after (A) HaCaTs or (B) HEKas were exposed to UV in various doses. Figure S19. ROS levels in HaCaTs after UVB exposure. (A) Confocal fluorescent images (Blue, nucleus; Green, ROS positive.). (B) Flow cytometry analyses of free radical levels in HaCaTs with/without protections. HaCaTs with ZIF-8 pretreatment showed no obvious increase of ROS. However, ROS were elevated for the cells without protection or with the protections of TiO2, ZnO, MOF-5, IRMOF-1 or Zn3L3DMF2. Figure S20. ROS levels in HEKas after UVB exposure. (A) Confocal fluorescence images (Blue, nucleus; Green, ROS positive.) and (B) flow cytometry analyses of free radicals in HEKas with/without protections. HaCaTs with ZIF-8 pretreatment showed less ROS production relative to that for the cells without protection or with the protections of TiO2, ZnO, MOF-5, IRMOF-1 or Zn3L3DMF2. Figure S21. ROS levels in HaCats after UVA exposure. (A) Confocal fluorescent images (Blue, nucleus; Green, ROS). (B) Flow cytometry analyses of free radical level in HaCats with/without protections. HaCaTs with ZIF-8 pretreatment showed no obvious ROS production. More ROS production were observed in the cells without protection or with the protections of TiO2, ZnO, MOF-5, IRMOF-1 or Zn3L3DMF2. Figure S22. ROS measurement in HEKas after UVA exposure. (A) Confocal fluorescent images (Blue, nucleus; Green, ROS.). (B) Flow cytometry analyses of free radicals in HEKas with/without protections. HEKas with ZIF-8 protection showed less ROS production relative to that for the groups of No protection, IRMOF-1, TiO2, ZnO, MOF-5, or Zn3L3DMF2. Figure S23. UV dose optimization on mouse skin. Images of mouse dorsal skin three days after UV exposure at different doses. UV dose of 206 J m−2 was selected for further use because erythema was observed at this dose. Figure S24. ZIF-8 dose optimization. Images of mouse skin three days after UV exposure with protections of TiO2, ZnO, and ZIF-8 at different doses. The dose of 15% was selected for further in vivo mouse study, because some sunburn was observed for TiO2 and ZnO mice at this dose, while no obvious erythema was observed for ZIF-8 mice at dose of 15%. Figure S25. Digital graphs of mouse dorsal skin three days after UV exposure with the protections of TiO2, ZnO, or ZIF-8. ZIF-8 group showed less ulceration, edema or erythema compared to no protection or glycerol group. Figure S26. ROS in mouse skin after UV exposure. (A) Quantitative and (B) qualitative analyses of free radical level in the skin with/without protections. Figure S27. Microscope photographs of the mouse skin with (A) Masson’s Trichrome staining or (B) IL-1β immunohistochemistry. UV disturbed collagen distribution and decreased density for no protection and glycerol groups, while ZIF-8 group showed a normal collagen appearance. ZIF-8 also inhibited the skin expression of IL-1β after UV exposure. Figure S28. UV dose optimization on pig skin. Digital images of pig dorsal skin 24 h after UV exposure. UV at 544 J m−2 could induce obvious erythema. Figure S29. ZIF-8 dose optimization against UV exposure on pig skin. Digital graphs of pig dorsal skin 24 h after UV exposure with protections of ZIF-8 at different doses. ZIF-8 at 15% obviously inhibited erythema formation. Figure S30. Digital graphs of pig dorsal skin 24 h after UV exposure with the protections of TiO2, ZnO, or ZIF-8. ZIF-8 obviously inhibited erythema formation. Figure S31. Blood biochemical analyses after mice were treated with TiO2, ZnO, or ZIF-8 for six times in 15 days. (A-C) Serum levels of (A) AKP, (B) ALT, and (C) AST for liver function analyses. (D-E) Serum levels of (D) BUN, and (E) CRE for kidney function analyses. Both liver and kidney functions were not affected by TiO2, ZnO, or ZIF-8. Figure S32. H&E images of main tissues after mice were treated with TiO2, ZnO, or ZIF-8 for six times in 15 days. No obvious tissue damage was observed for heart, liver, spleen, lung, and kidney in all these groups. Figure S33. The accumulations of Ti or Zn in blood and tissues after TiO2, ZnO, or ZIF-8 were applied for six times in 15 days. (A, B) Ti levels in (A) heart, liver, spleen, lung, and kidney and (B) blood for the mice with TiO2 treatment. (C, D) Zn levels in (C) heart, liver, spleen, lung, and kidney and (D) blood for the mice with ZnO or ZIF-8 treatments.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Xiao, J., Li, H., Zhao, W. et al. Zinc-metal–organic frameworks with tunable UV diffuse-reflectance as sunscreens. J Nanobiotechnol 20, 87 (2022). https://doi.org/10.1186/s12951-022-01292-1

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12951-022-01292-1

Keywords

Journal of Nanobiotechnology

ISSN: 1477-3155

Contact us