Royal jelly-derived proteins enhance proliferation and migration of human epidermal keratinocytes in an in vitro scratch wound model

Background Skin injury is inevitable in daily life. In recent years, with the increasing morbidity of diseases such as diabetes and metabolic disorders, chronic wounds have become a considerable challenge in clinical practice. Royal jelly, reported to have multifarious biological and physiological properties, has been used as a remedy for a variety of wounds since ancient times. However, the active components and mechanisms underlying the wound-healing properties of royal jelly are still largely unknown. Methods Water-soluble proteins of royal jelly were fractionated and investigated for the proliferative and migratory effects on human epidermal keratinocytes (HaCaT) in an in vitro wound healing model. The proteins present in bioactive fractions were characterised and quantified using Label-free protein quantification method. The potential functions of these proteins in biological systems were further analysed using bioinformatic tools. Results A protein fraction, mainly containing major royal jelly proteins 2 (MRJP2), MRJP3 and MRJP7, stimulated proliferative and migratory activities in HaCaT cells without visible cytotoxicity. It exerted the greatest effects on the growth of HaCaT cells in the first 48 h. Furthermore, when treated with this protein fraction, the closure rates of the in vitro scratch wound were significantly increased. Functional analysis indicated that MRJP2, MRJP3 and MRJP7 were associated with carbohydrate transport and metabolism. Conclusions We fractionated the water-soluble proteins of royal jelly and identified one fraction (Fraction 2) that induced both proliferative and migratory effects on a human epidermal keratinocyte cell line. Major royal jelly proteins (MRJP2, MRJP3 and/or MRJP7) were speculated to possess potential wound-healing bioactivity. This is the first report that royal jelly may improve wound closure via MRJP-induced cellular proliferation and migration. These proteins may be valuable lead compounds for the development of novel wound healing medications. Our findings would facilitate better understanding of the wound repair mechanisms of royal jelly.


Background
Royal jelly, food given exclusively to larvae and queen bees, is synthesised and secreted by the glands located in the hypopharynx and mandible of nurse honeybees. It is composed of 60-70% water, 11-23% carbohydrate, 9-18% protein, 4-8% lipid and 0.8-3% other compounds such as vitamins, salts, amino acids and minerals [1][2][3][4]. To date, major royal jelly proteins (MRJPs) consisting of nine members (MRJP1-9) with molecular weight from 49 to 87 kDa have been identified in royal jelly and taken up 83-90% of its total protein content [5].
Many studies have reported that royal jelly possesses a plethora of biological activities, including antimicrobial [6][7][8][9], anti-inflammatory [10], anti-tumour [11], antioxidant [12], immunomodulatory [13] and growthpromoting [14] effects. What is more, it has been documented to be used in folk medicine for the facilitation of wound healing since ancientry [15]. Although some studies have indicated that topical application of royal jelly can shorten the period of infected or uninfected wound healing in animal models and even diabetic foot ulcers [16][17][18], the substances and precise mechanisms associated with wound healing are still poorly investigated. Previous studies merely reported that royal jelly components, especially 10-hydroxy-2-decenoic acid (10-HDA) and defensin-1, might accelerate wound healing through anti-inflammation, promoting synthesis of growth factors, or migration of skin fibroblasts or keratinocytes [15,[19][20][21][22]; MRJPs could induce proliferation of several human cell lines [23].
Wound healing generally occurs naturally without any external interference. Nevertheless, if handled improperly or not dealt with promptly, wounds easily become infected. Clinical treatment of wounds may involve cleaning, disinfection, suturing, antibiotic treatment and dermatoplasty. To date, there is no effective medicine or therapeutic method for the treatment of intractable wounds such as gravely infected wounds or diabetic foot ulcers, resulting in significant suffering to patients, seriously affecting their quality of life, and imposing a severe financial burden on both families and society. It is therefore of significant clinical importance to discover pharmacologically-active substances that create conditions conducive to the process of wound healing and to clarify the precise molecular mechanisms related to their actions.
Here, we explored the wound-repairing activity of royal jelly proteins using keratinocytes which are a type of cells prominently spread in epidermis and are critical in wound healing process, particularly in epithelialization. A water-soluble protein fraction mainly consisting of MRJP2, MRJP3 and MRJP7 was found to induce proliferative and migratory effects in human epidermal keratinocytes (HaCaT) without obvious cytotoxicity, implying the potential of MRJPs in the healing of cutaneous wounds. This is the first report that royal jelly may improve wound closure via MRJP-induced cellular proliferation and migration. Our findings facilitate greater understanding of the wound-healing actions of royal jelly and suggest that MRJPs may have potential applications in the treatment of wound healing disorders and diabetic foot ulcers.

Royal jelly sample
Fresh royal jelly produced by Fengqiang No. 1 (Apis mellifera ligustica) offspring colonies was harvested from a local apiary in Jiangsu Province, P.R. China, and was deposited in a freezer (− 20°C) prior to use. No permission was necessary for the collection of royal jelly.

Fractionation of royal jelly
Royal jelly (10 g) was homogenized in 40 ml of phosphate buffered saline (PBS). Extraction of water-soluble proteins: the royal jelly homogenate was vortexed vigorously at intervals then left to stand on ice over a period of 30 min. The supernatant obtained following centrifugation at 10,000×g for 40 min at 4°C was dialysed against 2 L of PBS for 3 days at 4°C. The dialysate was further centrifuged as previously, which was followed by lyophilising the final supernatant in a freeze dryer (Thermo Fisher Scientific, USA) and then storing at − 80°C before use. The concentration of water-soluble proteins was determined using BCA protein assays [24]. Approximately 50 mg of lyophilised protein was reconstituted in 10 ml of deionised water, 500 μl of which was subjected to an ÄKTA™pure system (GE Healthcare, USA) fitted with a Tricorn™ Superdex 75 Increase 10/ 300 GL high performance column (10 × 300 mm, 9 μm, GE Healthcare, USA). Deionised water (0.7 ml/min) and λ280 nm wavelength were applied for elution and effluent absorbance detection, respectively. Fractions derived from discrete absorbance peaks were collected and lyophilised.

Electrophoretic analysis of fractions
Fractionated proteins were analysed electrophoretically using 12% SDS-PAGE gels in which Blue Plus Protein Marker (Transgen, China) was included as a molecular weight protein standard. Electrophoresis was followed by staining the gels with Coomassie Brilliant Blue R-250 to preliminarily analyse the protein composition in each fraction.

MTT cell viability assay
HaCaT cells (100 μl) suspended in DMEM at a density of 5 × 10 4 cells/ml were seeded onto each well of 96-well plates and incubated at 37°C for 24 h in a CO 2 incubator. The medium was replaced with serum-free DMEM for cell starvation for 12 h, and the cells were subsequently treated with royal jelly fractions, bovine serum albumin (BSA, 3.2 μg/ml) or serum-free DMEM (control) for 24, 48 or 72 h. Then, they were incubated for a further 4 h after addition of 10 μl of 5 mg/ml 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) solution (Beyotime, China). The generated formazan crystals were dissolved in 100 μl of DMSO following removal of the supernatant. Cell viability was determined by the absorbance at 492 nm in an Infinite F50 plate reader (Tecan, Austria), and expressed as %

Scratch-wound assays
Cell migration effects were evaluated using scratch-wound assays. Culture-Inserts (ibidi, Germany) were placed in each well of 24-well plates. A volume of 70 μl of HaCaT cells (5 × 10 5 cells/ml) was seeded in each cell of the Culture-Insert and cells were cultured as monolayers to confluency for about 18 h in a CO 2 incubator. Culture-Inserts were removed carefully to form an approximately 500-μm-wide cell free gap, and cells were washed twice to remove nonadherent cells. To observe the effects of royal jelly protein fractions on keratinocyte migration, cells were treated with a series of the protein fractions at different concentrations prepared in serum-free DMEM or with serum-free medium (control). The wounded cell monolayer was observed under a phase-contrast microscope (Olympus, Tokyo, Japan) and images (three per well, captured randomly) were taken at 0 and 24 h following the infliction of wounds which was a cell free nick created by Culture-Inserts [25]. Closure of wounds was measured using Image J software (National Institutes of Health, Bethesda, MD, USA). Wound closure rates were expressed as percentages of the wound area closed at 24 h relative to the initial area of the cell-free region at 0 h. Three replications of each treatment were included in each experiment and all experiments were replicated independently three times.

Functional analysis of the proteins identified in bioactive fractions
The structural features and potential functions of the proteins present in the bioactive fractions were analysed and annotated by InterPro [26], Cluster of Orthologous Groups (COG) [27], Gene Ontology (GO) [28], and Kyoto Encyclopedia of Genes and Genomes (KEGG) [29] databases in conjunction with BLAST search [30].

Statistical analysis
GraphPad Prism 5.0 software (GraphPad Software Inc., CA, USA) was employed for statistical analysis, in which one-way ANOVA was used for comparisons of variance. Results with p-values less than 0.05 were regarded to be statistically significant. Values were expressed as means ± SEM.

Isolation and electrophoretic analysis of proteins from royal jelly
Isolation of royal jelly proteins by size exclusion chromatography using an ÄKTA™pure system (GE Healthcare, USA) revealed that two major protein peaks (Fraction 1 and Fraction 2) were eluted at retention times of about 9.5 and 18 min, respectively (Fig. 1a). SDS-PAGE analysis indicated that Fraction 1 was composed of a single protein band with molecular weight of around 55 kDa, and Fraction 2 of two protein bands with molecular weights of 50-60 kDa (Fig. 1b).

Proliferative effect of protein fractions on human epidermal keratinocytes
The proliferative effect of the royal jelly proteins in these fractions on HaCaT cells was examined. Regardless of the duration of treatment, Fraction 1 was found to be devoid of growth-promoting activity on the cell line at concentrations up to 40 ng/ml and demonstrated severe cytotoxicity at high concentrations ( Fig. 2a), which was similar to the effects of MRJP1 (data not shown). By contrast, Fraction 2 exhibited the efficacy of facilitating the proliferation of HaCaT cells at 0.8 and 1.6 μg/ml without causing obvious toxicity at any concentration.   When treated with 0.8 μg/ml of Fraction 2, cell growth continued for 48 h (Fig. 2b).
Cell migratory effects of protein fractions in an in vitro scratch wound model Fraction 2, which showed proliferative effects and was nontoxic in keratinocytes, was investigated further for its migratory effects on HaCaT cells in an in vitro scratch wound model. Activity on cell migration was measured as the rate of wound coverage over a period of 24 h. As the micrograph in Fig. 3a, at the end of the 24-h treatment, the scratch wound treated with Fraction 2 at the tested concentrations appeared to be much narrower than the vehicle control. The healing rate of keratinocytes treated with 1.6 μg/ml of Fraction 2 (53%) was dramatically higher than that of the medium-only control at 24 h (Fig. 3b), implying that Fraction 2 could facilitate keratinocyte migration and may possess potential in wound healing.

Identification and quantification of bioactive protein fractions
The proteins present in Fraction 2 were successfully identified and quantified using Label-free protein quantification in which absorption peak areas were used for protein quantification and MS/MS fragment ions were used for protein identification. It resulted in identification of a total of 63 proteins including MRJPs, protein kinase and ATP synthase, among which MRJP2 (60%), MRJP3 (30%), MRJP7 (2%), MRJP1 (1%) were the most abundant and characteristic ( Table 1). The total ions chromatograph (TIC) of Fraction 2 and the canonical MS/MS fragmentation sequencing spectrums of the identified MRJPs are shown in Fig. 4.

Bioinformatic analysis of the proteins identified in fraction 2
InterPro protein domain classification revealed that the majority of proteins in Fraction 2 shared the typical domain with major royal jelly proteins (Fig. 5), which was consistent with the mass spectrometric results. According to the COG functional annotation, most proteins were related to translation, ribosomal structure and biogenesis (function class J), posttranslational modification, protein turnover, chaperones (function class O), and carbohydrate transport and metabolism (function class G) ( Fig. 6 and Table 2), suggesting that these proteins might promote wound healing through the enhancement of protein synthesis. The GO analysis interpreted the roles of proteins in cells, indicating that the most enriched proteins contributed to the translationassociated biological process (Fig. 7). The KEGG pathway analysis illustrated the potential functions and metabolic  procedures of proteins in biological system. In accordance with COG and GO analysis, many proteins were concerned with transport and catabolism (cellular processes), and translation (genetic information processing) (Fig. 8). It is noteworthy that masses of proteins (21 proteins) were in connection with human diseases such as Non-alcoholic fatty liver disease, Alzheimer's disease, Parkinson's disease and HTLV-I infection.

Discussion
Cutaneous wound healing impairment has always been a serious health problem as it may result in ulcers, relapse, amputation and even death. Over the past decades, with the high incidence of diabetes, an increasing number of patients are suffering from one of the most typical complications-foot ulcers. Only about half of all such ulcers can be cured and there is a high risk of relapse. It has been reported that one amputation occurs every 30 s and as many as 2,500 patients undergo some form of    MRJPs and the predicted functions are in bold typeface amputation every single day [31]. According to the International Diabetes Federation (IDF), the diabetic population worldwide is predicted to increase to 592 million in 2035. At present, suturing and skin-grafting are the primary therapies used to treat deep or large defective wounds, and antibiotic therapies are usually applied to infected wounds. There is still no effective agent that promotes the healing of either intractable wounds or diabetic foot ulcers. The discovery and development of novel wound repairing drugs and treatments is thus of significant importance.
Wound healing is a complicated dynamic process which generally involves three consecutive and interrelated phasesinflammatory reactions, proliferative stage and tissue remodelling [32]. An inflammatory reaction occurs in the early stages of wound formation, in which inflammatory cells phagocytose bacteria and cell debris from necrotic tissue, serving as a form of wound cleaning. Subsequently, fibroblasts proliferate to promote the synthesis of extracellular matrix, while granulation tissue grows at the bottom and edges of the wound, filling up the incisions that will be covered by proliferative and migrating basal cells to form an epithelial monolayer. Keratinocytes then proliferate and move towards the centre of the wound, thereby covering the wound surface and completing wound healing. The proliferation, migration and differentiation of fibroblasts and keratinocytes thus plays a vital role in wound reepithelialization and reparation [33][34][35][36].
In this study, we attempted to evaluate the efficacy of royal jelly proteins on human epidermal keratinocytes to identify bioactive components of royal jelly with woundhealing activity and their mechanisms. Following fractionation of royal jelly water-soluble proteins, SDS-PAGE analysis showed that Fraction 1 was a relatively pure protein with a mass of 55 kDa, most likely MRJP1, and Fraction 2 was a mixture of proteins with molecular weights ranging from 50 to 60 kDa (Fig. 1). Fraction 1 was ineffective in wound healing and was even toxic to the growth of HaCaT cells, while Fraction 2 induced some cell growth (Fig. 2). Compared with being incubated with BSA, cells incubated with Fraction 2 showed a significant increase in cell growth (Fig. 2b), suggesting that the proliferative effects of Fraction 2 were different from the nutritive effects of BSA. In addition, as pure MRJP1 and Fraction 1 had very similar profiles and biological effects on the cell viability of HaCaT (data not shown), Fraction 1 was further confirmed as MRJP1. It was surprising that MRJP1, the most abundant MRJP [37], reported to possess various bioactivities [14,38,39] including weak up-regulation of mRNA expression of cytokines (TNF-α, IL-1β and TGF-β) in cultured human primary keratinocytes [40], was found to be devoid of having an effect on HaCaT cell growth. Owing to the toxicity and inefficacy of MRJP1, we only investigated Fraction 2 further for its cell migratory effects on an in vitro scratch wounding model. The proportion of wound healing induced by Fraction 2 (1.6 μg/ml) was significantly higher after 24 h incubation than that of the vehicle control (Fig. 3b). As shown in Fig. 2b, however, 1.6 μg/ml of Fraction 2 had little effect on the growth of HaCaT cells at 24 h; moreover, fraction samples used in the wound scratching assay were reconstituted in serum-free medium which had no effect on cell growth. Therefore, effects of Fraction 2 on cell growth and medium-induced cell proliferation in the wound closure observed can be ignored. Fraction 2 may thus mediate wound healing via keratinocyte migration.
Interestingly, Fraction 2 did not function in a strict dose-dependent manner concerning its proliferative and migration efficacy towards keratinocytes, efficacy declining at higher concentrations (Figs. 2b and 3). This phenomenon might be caused by the antagonistic effect of the complicated compositions within Fraction 2, and thus further purification of the fraction or expression of each component as a pure recombinant protein would be desirable to facilitate further functional investigation. In addition, as the procedure of wound healing is complex, involving various categories of cells, cytokines, growth factors, and many other intracellular/extracellular components [41], it will be important to explore the regulatory effects of these proteins on other types of cells such as macrophages.
Qualitative and quantitative analysis of Fraction 2 revealed that the main biologically-active components were MRJP2, MRJP3 and MRJP7. COG analysis demonstrated that MRJP2, MRJP3 and MRJP7 were associated with carbohydrate transport and metabolism ( Fig. 6 and Table 2), which might be beneficial to the conversion into nutrition such as proteins to promote wound healing. However, the precise mechanisms underlying the wound healing activity need to be further investigated to facilitate better understanding of the wound repair functions of royal jelly in the future. This is the first experimental evidence that MRJP2, MRJP3 and/or MRJP7 may possess potential wound healing function, providing valuable lead compounds to be developed into novel wound repairing medications.

Conclusions
In this study, the potential wound healing activity of royal jelly proteins was first time to be concerned. As a result, MRJP2, MRJP3 and MRJP7 present in a watersoluble protein fraction were speculated to possess potential wound repairing activity. Our finding that MRJPs may modulate wound healing by stimulating keratinocyte growth and migration suggests that these naturally-occurring proteins would be biomaterials with potential for development as novel wound-healing agents.