ZnO NPs delay the recovery of psoriasis-like skin lesions through promoting inflammation and keratinocyte apoptosis via nuclear translocation of phosphorylated NF-κB p65 and cysteine deficiency CURRENT STATUS: POSTED

This study aimed to confirm the safety and risk of applying zinc oxide nanoparticles (ZnO NPs) to pathological skin, such as psoriasis-like skin. The majority of previous studies confirmed the safety of applying ZnO NPs to normal skin. However, we know very little about the risks of using sunscreen, cosmetics and topical drugs containing ZnO NPs for individuals with skin diseases. In addition, some studies claimed that ZnO NPs can penetrate normal or pathological skin, and ZnO NPs have frequently been reported to have proinflammatory and lethal effects in vitro. Therefore, it is necessary to evaluate the safety of applying ZnO NPs to pathological skin. the cysteine level dramatically decreased, and the proportion of apoptotic cells revealed by Annexin V/PI analysis was significantly increased. These findings show that 30 ~ 40 µg/mL is the initial dose at which ZnO NPs can simultaneously promote inflammation and apoptosis. This result is consistent with the phenomena observed in vivo; thus, this dose was adopted for the subsequent experiments. T-GSH: glutathione; GPx: glutathione peroxidase; MDA: malonaldehyde; ROS: reactive oxygen species; NF-κBnuclear factor-κB; MAPK: mitogen-activated protein kinase; CBS: cystathionine β-synthase; CTH: γ-cystathionase; NAC: N-acetylcysteine; Fer-1: ferrostatin-1; DEP: deferiprone; ZVF: Z-VAD-FMK; HRP: horseradish peroxidase stress in mice. (A~B) H&E staining showed that the number of inflammatory cells in the dermis was increased in the ZnO NP-treated group (n=6). evaluations showed that the expression levels of TNF-α, IL-1β, IL-6 and COX2 in epidermis were increased in the ZnO NP-treated group (n=3). were increased was 8-hydroxy-2 deoxyguanosine; oxidative pathway: CBS and CTH); and the effects of antioxidants and VZF on ZnO NP-induced alterations in those pathways. (L~N) Fer-1 and NAC inhibited ZnO NP-induced secretion of TNF-α, IL-1β and IL-6; ZVF inhibited ZnO NP-induced secretion of IL-1β and IL-6. (a compared to TZ0, b compared to TZ40. *P<0.05, ** P<0.01, ***P<0.001, ****P<0.0001, n=3) Abbreviations: MAPK: mitogen-activated protein kinase (including p-44/42 MAPK (ERK), p38 MAPK and JNK), NAC: N-acetylcysteine; Fer-1: ferrostatin-1, DEP: deferiprone, ZVF: Z-VAD-FMK, CBS: cystathionine β-synthase, CTH: γ-cystathionase.


Abstract Background
This study aimed to confirm the safety and risk of applying zinc oxide nanoparticles (ZnO NPs) to pathological skin, such as psoriasis-like skin. The majority of previous studies confirmed the safety of applying ZnO NPs to normal skin. However, we know very little about the risks of using sunscreen, cosmetics and topical drugs containing ZnO NPs for individuals with skin diseases. In addition, some studies claimed that ZnO NPs can penetrate normal or pathological skin, and ZnO NPs have frequently been reported to have proinflammatory and lethal effects in vitro. Therefore, it is necessary to evaluate the safety of applying ZnO NPs to pathological skin.

Results
ZnO NPs passed through gaps between keratinocytes and entered stratum basale of epidermis and dermis in imiquimod (IMQ)-induced psoriasis-like skin lesions. Application of a ZnO NP-containing suspension for 3 connective days delayed the healing of the epidermal barrier; increased the expression levels of inflammatory cytokines; promoted keratinocyte apoptosis and disturbed redox homeostasis. In vitro, ZnO NPs promoted TNF-α, IL-1β and IL-6 secretion and apoptosis of recombinant-human-TNF-α-stimulated HaCaT cells. NF-κB, ERK, p38 and JNK inhibitors blocked ZnO NP-induced inflammation. JSH-23, an inhibitor of the nuclear translocation of p-NF-κB p65, and NAC, an acetylated precursor of L-cysteine, not only inhibited the ZnO NP-induced inflammation but also inhibited apoptosis and cysteine deficiency. Neither erastin nor RSL3 induced p-NF-κB p65 nuclear translocation, but they did reduce cysteine biosynthesis. Additionally, ferropstatin-1, an inhibitor of lipid peroxidation, partially rescued ZnO NP-induced decreases in cell viability and cysteine content.

Conclusions
ZnO NPs delay the recovery of psoriasis-like skin lesions through promoting inflammation and keratinocyte apoptosis via the nuclear translocation of phosphorylated NF-κB p65 and cysteine deficiency. This work reminds the public that ZnO NPs are not safe for pathological skin, especially in inflammatory skin diseases such as psoriasis, and has revealed a partial mechanism by which ZnO NPs delay the recovery of pathological skin, promoting the appropriate use of ZnO NPs. Background For many years, zinc oxide (ZnO) has been widely used in food, pharmaceuticals, cosmetics and other chemicals used on a daily basis, and formulations of these products containing ZnO nanoparticles (ZnO NPs) are more popular than conventional formulations [1][2][3][4]. The content of ZnO NPs in certain topical drugs and cosmetics that frequently contact the skin is high. Previous studies claimed that ZnO NPs can penetrate normal skin, but this finding was not confirmed by most other studies; additionally, pathological damage has not been found to result from topical ZnO NP application [5][6][7].
Therefore, it is generally believed that the topical application of ZnO NPs to normal human skin is safe. In fact, individuals with skin diseases, such as patients with psoriasis, atopic dermatitis (AD), acne and rosacea, use ZnO NP preparations at almost the same frequency as or at an even higher frequency than those with normal skin. These individuals often need to use medicinal or cosmetic ZnO NP preparations to treat diseases or cover unsightly skin. However, discussion on whether ZnO NPs are safe for pathological skin is limited, and the relevant guidelines do not provide definitive recommendations.
In skin diseases, the defects of skin barrier and the abnormal expression of barrier-related proteins in the epidermis are common [8,9]. For example, both atopic dermatitis (AD) and psoriasis lead to the abnormal expression of filaggrin and loricrin, which are vital for the integrity of the physical barrier of the epidermis [10][11][12]. Theoretically, a defective barrier resulting from abnormal expression of barrierrelated proteins is more permeable; moreover, in an AD mouse model, ZnO NPs have been reported to enter the dermis through the epidermis, inducing vigorous IgE production and suppressing inflammation [13]. This study indicated that ZnO NPs possess promising anti-inflammatory effects and can be used to treat inflammatory skin diseases. However, such results have been rarely observed in vitro. Most studies have reported that ZnO NPs have proinflammatory and lethal effects on keratinocytes and that oxidative stress is the main mechanism by which ZnO NPs injure cells [14][15][16][17].
These inconsistencies may be attributed to of the use of distinct ZnO NP dispersal systems or the features of different disease models.
In this study, psoriasis-like skin models were adopted to further elucidate the potential effects of ZnO NPs on pathological skin. Psoriasis is an inflammatory skin disease characterized by epidermal hyperplasia and excessive proliferation of keratinocytes [18]. The tumour necrosis factor-α (TNF-α)-nuclear factor-κB p65 (NF-κB p65) axis plays an important role in the inflammatory response and the excessive proliferation of keratinocytes in psoriasis [19]. The imiquimod (IMQ)-induced mouse model and TNF-α-stimulated HaCaT cell line are two classical models used to study psoriasis [20,21].
Research on the IMQ-induced mouse model has demonstrated that ZnO NPs affect the progression or recovery of psoriasis-like skin lesions through their involvement in the inflammatory response, oxidative stress and apoptosis. In TNF-α-stimulated HaCaT cells, the effects of ZnO NPs on inflammatory keratinocytes have been evaluated by determining the levels of proteins and signalling molecules highly related to psoriasis and oxidative stress damage. The potential mechanism underlying the effects of ZnO NPs has also been discussed and mainly involves the nuclear translocation of phosphorylated NF-κB p65 (p-NF-κB p65) and cysteine deficiency. We believe that this work is enlightening, improves our understanding of the pathogenicity of ZnO NPs in abnormal skin and will be helpful for individuals with pathological skin conditions.

Results
Characterization of ZnO NPs X-ray photoelectron spectroscopy (XPS) data were used to determine the composition of the ZnO NPs adopted in this study (Fig. 1A). Transmission electron microscopy (TEM) and scanning electron microscopy (SEM) images showed that the ZnO NPs were prism-shaped with a primary size of 41.34 ± 9.41 nm (Fig. 1B&C). The hydrodynamic sizes of the ZnO NPs in the suspension used for the animal experiments, complete medium and distilled water (DW) were 810.3 ± 116.5 nm, 189.0 ± 63.3 nm, and 266.7 ± 69.2 nm, respectively ( Fig. 1D ~ F). The zeta potentials of the ZnO NPs in the suspension used for the animal experiments, complete medium and DW were 13.1 ± 0.5 mV, 11.9 ± 0.9 mV and 10.9 ± 0.2 mV, respectively (Table 3).
ZnO NPs delay the recovery of psoriasis-like skin lesions and penetrate the dermis IMQ application for 6 consecutive days caused scales (Fig. 1G&H) and epidermal hyperplasia ( Fig. 1I&J) on the dorsal skin. TEM images showed the intercellular gaps of approximately 2 µm between keratinocytes in IMQ-induced psoriasis-like skin lesions (Fig. 1K). These gaps disappeared after treatment with the suspension without ZnO NPs for 3 consecutive days ( Fig. 1L) but remained when the suspension containing ZnO NPs was used. ZnO NPs entered epidermis through stratum corneum and were discovered in the nuclear of keratinocytes, stratum basale of epidermis and dermis ( Fig. 1M ~ O). ICP-MS analysis confirmed that the elemental Zn content in the dermis of the ZnO NPtreated group was significantly higher than that in the negative control group (Fig. 1P).
ZnO NPs promote inflammation, apoptosis and oxidative stress in psoriasis-like skin lesions The haemoxylin and eosin (H&E) staining results were similar for the negative control group and ZnO NP-treated group, but the ZnO NP-treated group exhibited a significantly greater number of inflammatory cells in the dermis than the negative control group (Fig. 2A&B). Immunohistochemical Cysteine deficiency is involved in ZnO NP-induced apoptosis and is mediated by ROS and lipid peroxidation.
The CCK-8 assay revealed that cell viability was decreased by RSL3 but not by erastin (Fig. 5I). Both erastin and RSL3 decreased the cysteine content, but cysteine levels were lower in the RSL3-treated group than in the erastin-treated group (Fig. 5J). Erastin and RSL3 did not increase the nuclear p-NF-κB p65 level, but RSL3 increased the ratio of BAX/Bcl-2. Erastin, RSL3 and ZnO NPs did not affect xCT expression and decreased CD98 expression. The decrease in CD98 expression induced by erastin was less pronounced than that induced by RSL3 or ZnO NPs. The expression of CBS was significantly increased by erastin and was only slightly increased by RSL3 and ZnO NPs. The expression of CTH was significantly increased by erastin and ZnO NPs and was strongly increased by RSL3 (Fig. 5K).
Ferrostatin-1 (Fer-1) and NAC inhibited the ZnO NP-induced decline in cell viability and cysteine content, but Fer-1 was less effective than NAC (Fig. 5I&J). Fer-1 and NAC also inhibited ZnO NPinduced alterations, including the increases in the nuclear p-NF-κB p65 level; caspase 3 activation; the BAX/Bcl-2 ratio; and TNF-α, IL-1β and IL-6 secretion and the decrease in the CD98 level (

Discussion
To confirm the effects of ZnO NPs on pathological skin, an IMQ-induced psoriasis-like skin model was adopted in this study. After IMQ was topically applied for connective 6 days, the skin was covered with scales, and H&E staining showed epidermal hyperplasia; these findings demonstrated that a psoriasis-like skin model had been established. TEM images revealed abundant gaps of approximately epidermis and dermis in this IMQ-induced psoriasis-like skin model. Similarly, a previous study reported that ZnO NPs enter the dermis through the epidermis in an AD model [13]. Although characterization of the ZnO NPs indicated that they tended to agglomerate and that ZnO NP suspensions should be ultrasonically dispersed before use, the agglomerated ZnO NPs were still able to penetrate through these gaps in the epidermis and reached to stratum basale of epidermis and dermis. Additionally, TEM images of cells displayed the ability of keratinocytes to phagocytose ZnO NPs. All of the above results indicate the risks of applying topical drugs and cosmetics containing ZnO NPs to pathological skin.
In a further experiment, we applied a ZnO NP suspension to the IMQ-induced psoriasis-like skin model at a single exposure dose of 1.67 mg/cm 2 for 3 consecutive days. This dose is lower than the effective concentration of ZnO NPs in sunscreen (2 mg/cm 2 ) [22]. The IMQ-induced psoriasis-like skin model is reversible [23,24]. This means that the defective epidermal barrier can spontaneously recover. We found that application of ZnO NPs delayed the recovery of psoriasis-like skin lesions and induced inflammation, apoptosis and oxidative stress in the epidermis and dermis. In the negative control group, the gaps between keratinocytes disappeared after 3 days of treatment, whereas these gaps remained in the ZnO NP-treated group. In psoriasis, IL-1 is likely to initiate the inflammatory response and cause dysfunction of keratinocytes; furthermore, TNF-α, IL-1 and IL-6 derived from dysregulated keratinocytes amplify the inflammatory response through promoting the differentiation of myeloid dendritic cells [19,25,26]. IL-17A and IL-22 released by activated lymphocytes mediate chronic inflammation in psoriasis[21, [27][28][29]. In psoriasis, high expression of COX2 is associated with a persistent inflammatory response [19,30]. The enhancement of TNF-α, IL-1β, IL-6 and COX2 expression levels and the increase in the transcription of inflammatory cytokines, including TNF-α, IL-1β, IL-6, IL-17A and IL-22, in the ZnO NP-treated group indicated that ZnO NPs may hinder the antiinflammatory effect of drugs used to treat these diseases. More importantly, the higher level of TUNEL staining in the ZnO NP-treated group implied that keratinocyte apoptosis was increased and that the histological foundation for epidermal renewal capacity was impaired. Oxidative stress is generally thought to play a pivotal role in the mechanism by which nanomaterials damage cells [31]. The higher level of 8-OHdG staining in the ZnO NP-treated group revealed that ZnO NPs promoted DNA damage related to oxidative stress in pathological keratinocytes. GSH prevents proteins from being damaged by ROS by oxidizing their own thiol groups [32]. The MDA level indicates the extent of lipid peroxidation [31]. The disturbance of GSH antioxidant system and the increased MDA level in the ZnO NP-treated group revealed the potential mechanism by which ZnO NPs promote inflammation and keratinocyte apoptosis in pathological skin. TNF-α inhibition or antagonism is an important therapeutic strategy for the treatment of psoriasis and the production and secretion of TNF-α is associated with NF-κB and MAPK pathways [33,34]. The TNFα-NF-κB axis is involved in the epidermal hyperplasia and inflammatory response in psoriasis [35,36].
MAPK not only participates in the inflammatory response in psoriasis but also plays a part in the nanomaterial-induced stress response [31]. Treatment with 40 µg/mL ZnO NPs for 6 h induced strong nuclear translocation of p-NF-κB p65 and MAPK activation; enhanced phosphorylation of ERK, p38 and JNK was observed and was accompanied by enhancement of the inflammatory response. These results suggest that ZnO NPs aggravate psoriasis by promoting the inflammatory response via NF-κB and MAPK activation. Indeed, our results indicated that inhibitors of NF-κB and MAPK block the proinflammatory effect of ZnO NPs on rh-TNF-α-stimulated HaCaT cells.
The inflammatory response is responsible for keratinocyte proliferation and epidermal hyperplasia in psoriasis. The roles of NF-κB in promoting inflammation and inhibiting apoptosis have been acknowledged, and inhibitors of NF-κB are frequently used to protect against the inflammatory response and excessive proliferation [19,36]. However, we unexpectedly found that two inhibitors of NF-κB, QNZ and JSH-23, had opposing effects on the ZnO NP-induced decline in cell viability.
Generally, NF-κB activation begins with IκBα degradation, which is followed by a regulated series of steps in a cascade in the cytoplasm and the nuclear translocation and binding of NF-κB to DNA.
According to the literature, QNZ is a potent inhibitor of NF-κB and inhibits the production of TNFα [37]. In contrast, JSH-23 inhibits the nuclear translocation of NF-κB p65 without affecting IκBα degradation [38]. In this study, QNZ had a more thorough inhibitory effect on the NF-κB activation and p-NF-κB p65 nuclear translocation than JSH-23. As a result, the decline in cell viability and apoptosis mediated by the mitochondrial pathway (increased BAX/Bcl-2 ratio) and the caspase family (increased cleaved-caspase 3 level) were rescued by JSH-23 but not QNZ. This suggests that the strong nuclear translocation of p-NF-κB p65 is involved in ZnO NP-induced apoptosis and that excessive inhibition of the NF-κB pathway is not conducive to preventing the pro-apoptotic effect of ZnO NPs on keratinocytes. Moreover, NF-κB inhibitors are commonly used for the treatment of inflammatory skin diseases. Thus, the addition of ZnO NPs to sunscreens or medicines may affect the efficacy of these drugs.
It has been acknowledged that nanomaterials tend to mediate oxidative stress, and this effect may explain the potential risk of nanomaterials or be used to promote their application in some fields [39][40][41][42]. NAC is a general inhibitor of ROS. More precisely, NAC is an acetylated precursor of cysteine, which is deacetylated in cells and become cysteine [43][44][45]. In this study, NAC efficiently inhibited ZnO NP-induced inflammation, apoptosis and activation of related pathways, including the increase in the level of nuclear p-NF-κB p65. In addition, we found that ZnO NP-induced cysteine deficiency was partially reversed by JSH-23, an inhibitor NF-κB p65 nuclear translocation. Therefore, it can be inferred that ZnO NP-induced cysteine deficiency is related to the nuclear translocation of p-NF-κB p65. However, erastin and RSL3 did not induce the nuclear translocation of p-NF-κB p65 in rh-TNF-α-stimulated HaCaT cells though they inhibited cysteine biosynthesis. This suggests that cysteine deficiency is involved in ZnO NP-induced apoptosis and is mediated by the strong nuclear translocation of p-NF-κB p65 induced by ZnO NPs. Additionally, erastin and RSL3 did not promote inflammatory cytokine secretion.
Cysteine is an essential donor of sulfhydryl in the biosynthesis of GSH and plays a decisive role in cell survival [32]. ZnO NP-induced cysteine deficiency means the imbalance of antioxidant system. The cysteine/glutamate amino acid transporter system x c − , which consists of CD98 and xCT, is the main system by which cysteine is synthesized [46]. Certain types of cells can biosynthesize cysteine through the transsulfuration pathway, of which CBS and CTH are the rate-limiting enzymes of the first and last steps, respectively [47][48][49]. In this study, erastin, RSL3 and ZnO NPs decreased CD98 expression. However, erastin significantly increased CBS and CTH levels and did not cause cell death though reducing cysteine content. RSL3 and ZnO NPs increased the level of CTH and caused apoptosis and a relatively severe cysteine deficiency. These results suggest that when system x c − is impaired, keratinocytes can utilize the transsulfuration pathway to generate cysteine; however, cysteine deficiency and related cell death cannot be prevented by promoting only the last step of the transsulfuration pathway. Besides, previous studies claimed that cysteine is an endogenous metal ion chelator [50,51]. Therefore, NAC not only maintains the antioxidant system by providing sufficient sulfhydryl, but also can chelate ZnO NPs or Zn 2+ to alleviate ZnO NP-induced cell damage. In addition, Fer-1, an inhibitor of lipid peroxidation, is effective for preventing ZnO NP-induced cell damage. All of the above results imply that oxidative stress and lipid peroxidation are involved in ZnO NP-induced damage to keratinocytes and that an available antioxidant system is essential for the survival and health of keratinocytes exposed to ZnO NPs.

Conclusions
In summary, this work demonstrated that ZnO NPs can delay the recovery of psoriasis-like skin lesions through promoting the inflammatory response and keratinocyte apoptosis and that ZnO NPs can lead to further inflammation and TNF-α-stimulated keratinocyte apoptosis via strong nuclear translocation Animals and treatments 6-week-old female C57BL/6 mice were purchased from the Animal Center of Southern Medical University (Guangdong, China). The mice were housed in a room free of specific pathogens (23 ± 1 °C room temperature, 60 ± 10% relative humidity) and underwent an adaptation period for one week before treatment.  Table 1.

Measurements of oxidative stress biomarkers
Skin tissues that had been frozen in liquid nitrogen and stored at -80 °C were homogenized in precooled NS using a tissue homogenizer. The homogenates (1:10 w/v) were centrifuged at 3000 rpm

Statistical analysis
All the data are presented as the mean ± standard deviation (SD) and were analysed with SPSS 22.0 software. Comparisons among each group were assessed using one-way ANOVA when the variance in the data was homogenous or using a non-parametric test when the variance in the data was not homogenous. Differences for which P < 0.05 were considered to be statistically significant.

Consent for publication
Not applicable.

Availability of data and materials
The datasets generated and analysed during the current study are available from the corresponding author on reasonable request. Online supporting information are available.

Competing interests
The authors declare that they have no competing interests.

Authors' contributions
All authors contributed to this study. XL, XF, LL and SL contributed to the idea and design of this  Tables   Table 1. Primer sequences specific for mice used in the qRT-PCR analysis   Gene  Forward primer  Reverse primer  TNF-α  AGGCACTCCCCCAAAGATG  GCTCCTCCACTTGGTGGTTT  IL-1β  TGCCACCTTTTGACAGTGATG  ATGTGCTGCTGCGAGATTTG  IL-6  TTGCCTTCTTGGGACTGATG  CAGAATTGCCATTGCACAACTC  IL-17A  CAAAGCTCAGCGTGTCCAAAC  CTATCAGGGTCTTCATTGCGG  IL-22  GCGGTGACGACCAGAACATC  GGAAGGAGCAGTTCTTCGTTTTC  GAPDH CTTCGGGCCACGCTAATCTC ATGAAGGGGTCGTTGATGGC     Effects of inhibitors of signalling pathways and antioxidants on ZnO NP-induced