Honey-Mediated Wound Healing: H2O2 Entry through AQP3 Determines Extracellular Ca2+ Influx

Since Biblical times, honey has been utilized in “folk medicine”, and in recent decades the positive qualities of honey have been re-discovered and are gaining acceptance. Scientific literature states that honey has been successfully utilized on infections not responding to classic antiseptic and antibiotic therapy, because of its intrinsic H2O2 production. In our study, we demonstrated the involvement of H2O2 as a main mediator of honey regenerative effects on an immortalized human keratinocyte cell line. We observed that this extracellularly released H2O2 could pass across the plasma membrane through a specific aquaporin (i.e., AQP3). Once in the cytoplasm H2O2, in turn, induces the entry of extracellular Ca2+ through Melastatin Transient Receptor Potential 2 (TRPM2) and Orai1 channels. Honey-induced extracellular Ca2+ entry results in wound healing, which is consistent with the role played by Ca2+ signaling in tissue regeneration. This is the first report showing that honey exposure increases intracellular Ca2+ concentration ([Ca2+]i), due to H2O2 production and redox regulation of Ca2+-permeable ion channels, opening up a new horizon for the utilization of the honey as a beneficial tool.


Introduction
Wound healing is of countless significance for skin medicine and a precise focus is set on natural compounds.
Honey has been utilized as medicine in many cultures for a long time [1]. Honey is primarily a sugar-rich food source for bees. There are at least about 200 compounds described in honey, and it consists of water, mainly sugars, and other substances such as organic acids, proteins (enzymes), vitamins (i.e., niacin, vitamin B6, thiamine, riboflavin and pantothenic acid), minerals (including copper, calcium, etc.) [2].
In spite of a wide amount of literature about honey clinical practices, the subjacent mechanisms of action are still largely unclear. Accumulating records propose the contribution of specific physiological mechanisms, such as angiogenesis, granulation, and epithelialization [3].
We have recently elucidated the modulatory effects of different honey types in an in vitro model of keratinocytes and fibroblasts [4]. Our data have shown that honey treatments are poorly cytotoxic on skin cells, confirming that honey can be used safely not only for external applications on healthy skin, but also as a dressing on wounds. In addition, data coming from scratch wound observations as well as from cell migration assays have shown beyond any reasonable doubt that honey produces a striking increase of the wound repair abilities of keratinocytes and fibroblasts [4]. Taken together, our previous observations consistently suggest that honey is generally active in helping wound closure, thus confirming a bulk of anecdotal and scientific evidence [5,6].
The high concentrations of sugars (about 80%) combined with less than 1% of water causes osmotic stress preventing microorganism growth. Moreover, already during the 1960s, hydrogen peroxide (H 2 O 2 ) was recognized as a major antibacterial compound present in honey. In fact, honey continuously provides H 2 O 2 , which is generated by glucose oxidase-mediated conversion of glucose [7].
Specific aquaporins (AQPs) facilitate the passive diffusion of H 2 O 2 across the biological membranes [8]. AQP-mediated transmembrane transport of H 2 O 2 is of physiological importance for further downstream signaling events, such as the onset of intracellular Ca 2+ signals [8].
In wound healing, cellular reactions to injury are associated with an increase in intracellular Ca 2+ concentration ([Ca 2+ ] i ) which activates multiple Ca 2+ -dependent processes associated to tissue regeneration, including proliferation, motility and differentiation [9][10][11]. Accordingly, Ca 2+ has a main role in cellular homeostasis and is utilized by a plethora of extracellular cues to finely regulate cell fate [12][13][14]. Recent studies have demonstrated that also H 2 O 2 may signal through an increase in [Ca 2+ ] i by activating Ca 2+ -permeable channels located either on the endoplasmic reticulum (ER) [15], the largest endogenous Ca 2+ reservoir, or on the plasma membrane. Accordingly, H 2 O 2 has been found to induce extracellular Ca 2+ entry though Orai1 and Melastatin Transient Receptor Potential 2 (TRPM2) [16][17][18][19]. Notably, AQPs have long been known to regulate the [Ca 2+ ] i by modulating the osmotic gradient across the plasma membrane [20]. However, it is still unclear whether AQPs influence intracellular Ca 2+ dynamics by mediating extracellular H 2 O 2 entry.
Herein, we sought to investigate the functional interaction among extracellular H 2 O 2 , AQPs and plasmalemmal Ca 2+ -permeable channels in honey-induced skin regeneration. Our findings shed novel light on the intracellular mechanisms by which honey uses reactive oxygen species (ROS) to induce wound closure and hint at H 2 O 2 as a key signaling messenger to induce tissue remodeling due to its ability to trigger intracellular Ca 2+ signals.

Honey Induces Ca 2+ Signals in a Dose-Dependent Manner
First at all, we decided to investigate if different honey types are able to induce variations in [Ca 2+ ] i . Therefore, we assessed intracellular Ca 2+ variations induced after honey exposure, by using time-lapse confocal microscopy imaging of HaCaT cells loaded with the fluorescent Ca 2+ probe Fluo-3/AM. For honey concentration of 4% v/v, the calcein-AM cytotoxicity assay carried out on HaCaT cells previously demonstrated that this dose did not affect cell viability [4].
Preliminary observations showed that the [Ca 2+ ] i , sampled at 10-s intervals, did not undergo spontaneous oscillations in control conditions ( Figure 1A). The analysis of confocal imaging of Fluo-3/AM loaded cells exposed to 4% v/v honey samples revealed that manuka honey triggered a single, large [Ca 2+ ] i spike. The spike started a few seconds after honey exposure, reached a peak within 60-90 s, and decayed to a plateau level of intermediate amplitude in approximately 100-200 s ( Figure 1A,B).
Then, we tested the effects on [Ca 2+ ] i after treatment with other honey samples, i.e., buckwheat and acacia honey type ( Figure 1A,B). Buckwheat honey triggered a biphasic increase in [Ca 2+ ] i which was similar to that induced by manuka honey, but displayed a lower amplitude ( Figure 1A,B). Conversely, acacia honey evoked a slow increase in [Ca 2+ ] i which was significantly (p < 0.05) reduced as compared to manuka-and acacia-induced intracellular Ca 2+ variations ( Figure 1A,B). Based on these evidences, we reasoned that manuka honey was the most suitable type of honey to investigate the role of intracellular Ca 2+ signaling in honey-induced wound repair. We also evaluated the dose-response relationship of the Ca 2+ response to manuka honey by examining a range of concentrations (1, 2 and 4% v/v). As shown in Figure 1C,D, 4% v/v manuka honey induced the largest increase in [Ca 2+ ] i in HaCaT cells and was, therefore, employed throughout the remainder of this investigation ( Figure 1C,D).

The Ca 2+ Response to Manuka Honey Requires Extracellular Ca 2+ Entry and the Intracellular Production of Hydrogen Peroxide
Intracellular Ca 2+ signals can be generated by the opening of Ca 2+ -permeable channels which are located either on the plasma membrane or are embedded within the membrane of intracellular organelles, such as the ER [12,14]. We found that the Ca 2+ response to 4% v/v manuka honey disappeared in Ca 2+ -free medium (Figure 2A,B). Therefore, extracellular Ca 2+ entry is the main pathway underlying-induced elevation in [Ca 2+ ]i in HaCaT cells. It is already known that honey samples induce H2O2 production in cell cultures, thereby causing an increase in intracellular H2O2 levels [16]. By using the xylenol orange assay, we evaluated the dose-dependent production of H2O2

The Ca 2+ Response to Manuka Honey Requires Extracellular Ca 2+ Entry and the Intracellular Production of Hydrogen Peroxide
Intracellular Ca 2+ signals can be generated by the opening of Ca 2+ -permeable channels which are located either on the plasma membrane or are embedded within the membrane of intracellular organelles, such as the ER [12,14]. We found that the Ca 2+ response to 4% v/v manuka honey disappeared in Ca 2+ -free medium (Figure 2A,B). Therefore, extracellular Ca 2+ entry is the main pathway underlying-induced elevation in [Ca 2+ ] i in HaCaT cells. It is already known that honey samples induce H 2 O 2 production in cell cultures, thereby causing an increase in intracellular H 2 O 2 levels [16]. By using the xylenol orange assay, we evaluated the dose-dependent production of H 2 O 2 induced by the different honey types (i.e., acacia, buckwheat and manuka, Figure 3A). 5% manuka honey releases in the culture medium around 50 µM H 2 O 2 . We also explored intracellular ROS levels by recording the fluorescence of DHR-123 loaded HaCaT cells in a microplate reader. As shown in Figure 3B, 4% v/v manuka honey induced a sustained rise in intracellular ROS levels. induced by the different honey types (i.e., acacia, buckwheat and manuka, Figure 3A). 5% manuka honey releases in the culture medium around 50 µM H2O2. We also explored intracellular ROS levels by recording the fluorescence of DHR-123 loaded HaCaT cells in a microplate reader. As shown in Figure 3B, 4% v/v manuka honey induced a sustained rise in intracellular ROS levels. We then evaluated whether the addition of exogenous catalase (CAT), which scavenges H2O2, reduces manuka honey-induced Ca 2+ signals. Confocal imaging showed that, pre-treatment of cells with different doses of CAT, reduces (500 U) or abrogates (1000 U) the Ca 2+ response to manuka honey ( Figure 3C,D). Therefore, H2O2 is the most likely candidate to trigger manuka honey-induced Ca 2+ signaling. To further corroborate this notion, we assessed the Ca 2+ variations induced by 50 and 100 µM H2O2. Confocal imaging showed that 50 µM H2O2 induces a biphasic increase in [Ca 2+ ]i similar to that induced by manuka honey, although the time to peak was slightly slower and the early phase of the plateau phase slightly higher. 100 µM H2O2 triggered a progressive and sustained rise of [Ca 2+ ]i, reaching values that were remarkably higher than those achieved by manuka honey and not compatible with cell survival ( Figure 3E,F). We then evaluated whether the addition of exogenous catalase (CAT), which scavenges H 2 O 2 , reduces manuka honey-induced Ca 2+ signals. Confocal imaging showed that, pre-treatment of cells with different doses of CAT, reduces (500 U) or abrogates (1000 U) the Ca 2+ response to manuka honey ( Figure 3C,D). Therefore, H 2 O 2 is the most likely candidate to trigger manuka honey-induced Ca 2+ signaling. To further corroborate this notion, we assessed the Ca 2+ variations induced by 50 and 100 µM H 2 O 2 . Confocal imaging showed that 50 µM H 2 O 2 induces a biphasic increase in [Ca 2+ ] i similar to that induced by manuka honey, although the time to peak was slightly slower and the early phase of the plateau phase slightly higher. 100 µM H 2 O 2 triggered a progressive and sustained rise of [Ca 2+ ] i , reaching values that were remarkably higher than those achieved by manuka honey and not compatible with cell survival ( Figure 3E,F). induced by the different honey types (i.e., acacia, buckwheat and manuka, Figure 3A). 5% manuka honey releases in the culture medium around 50 µM H2O2. We also explored intracellular ROS levels by recording the fluorescence of DHR-123 loaded HaCaT cells in a microplate reader. As shown in Figure 3B, 4% v/v manuka honey induced a sustained rise in intracellular ROS levels. We then evaluated whether the addition of exogenous catalase (CAT), which scavenges H2O2, reduces manuka honey-induced Ca 2+ signals. Confocal imaging showed that, pre-treatment of cells with different doses of CAT, reduces (500 U) or abrogates (1000 U) the Ca 2+ response to manuka honey ( Figure 3C,D). Therefore, H2O2 is the most likely candidate to trigger manuka honey-induced Ca 2+ signaling. To further corroborate this notion, we assessed the Ca 2+ variations induced by 50 and 100 µM H2O2. Confocal imaging showed that 50 µM H2O2 induces a biphasic increase in [Ca 2+ ]i similar to that induced by manuka honey, although the time to peak was slightly slower and the early phase of the plateau phase slightly higher. 100 µM H2O2 triggered a progressive and sustained rise of [Ca 2+ ]i, reaching values that were remarkably higher than those achieved by manuka honey and not compatible with cell survival ( Figure 3E,F).

Role of Specific Inhibitors of Ca 2+ Entry
To disclose the mechanism(s) responsible for of the Ca 2+ response to manuka honey, we focused on the role of TRPM2 and Orai1. Accordingly, H2O2 has long been known to activate TRPM2 [17,18]; in addition, recent studies revealed that H2O2 may also activate Orai1-dependent Ca 2+ entry [21,22]. Confocal imaging showed that HaCaT cell preincubation with econazole, a TRPM2 inhibitor [23,24], was able to reduce the [Ca 2+ ]i increase caused by manuka exposure. Econazole (10 µM, 30 min) significantly (p < 0.05) reduced the amplitude of both the peak and the plateau phase of the Ca 2+ response to manuka honey ( Figure 4A,B). As econazole is not selective to TRPM2 channels, we downregulated its expression with a specific RNAi selectively targeting TRPM2 ( Figure 4D), and observed an almost complete abrogation of the Ca 2+ response to manuka honey exposure ( Figure 4E,F).

Role of Specific Inhibitors of Ca 2+ Entry
To disclose the mechanism(s) responsible for of the Ca 2+ response to manuka honey, we focused on the role of TRPM2 and Orai1. Accordingly, H 2 O 2 has long been known to activate TRPM2 [17,18]; in addition, recent studies revealed that H 2 O 2 may also activate Orai1-dependent Ca 2+ entry [21,22]. Confocal imaging showed that HaCaT cell preincubation with econazole, a TRPM2 inhibitor [23,24], was able to reduce the [Ca 2+ ] i increase caused by manuka exposure. Econazole (10 µM, 30 min) significantly (p < 0.05) reduced the amplitude of both the peak and the plateau phase of the Ca 2+ response to manuka honey ( Figure 4A,B). As econazole is not selective to TRPM2 channels, we down-regulated its expression with a specific RNAi selectively targeting TRPM2 ( Figure 4D), and observed an almost complete abrogation of the Ca 2+ response to manuka honey exposure ( Figure 4E,F).  Likewise, Pyr6 (1 µM, 30 min), a selective inhibitor of Orai1 channels, caused a significant decrease of both phases of manuka honey-induced elevation in [Ca 2+ ] i ( Figure 4A,B). As Pyr6 selectively targets Orai1, but not TRP channels [11,22,25], we did not silence Orai1. These data strongly suggest that TRPM2 and Orai1 mediate manuka honey-induced extracellular Ca 2+ entry. In agreement with this hypothesis, when HaCaT cells were pretreated with both econazole and Pyr6, the Ca 2+ response to manuka honey was fully abrogated. Cell incubation with vehicle DMSO, used alone, was ineffective on intracellular Ca 2+ dynamics (not shown).
In order to extend at molecular levels the findings obtained during Ca 2+ imaging recordings, we used qPCR to confirm that Stim1-2, Orai1-3 and TRPM2 transcripts are expressed in HaCaT cells ( Figure 4C).

Role of AQP3 in H 2 O 2 Influx
As discussed in Section 2.2, the Ca 2+ response to manuka honey requires an increase in intracellular ROS levels which follows H 2 O 2 liberation in the culture medium. Therefore, we hypothesized that, once released in the extracellular milieu, H 2 O 2 diffuses across the plasma membrane to increase intracellular ROS concentration and activate external Ca 2+ entry. AQPs are integral membrane proteins acting as channels in the transfer of water across the membrane, playing an important role in skin hydration [26][27][28]. In addition, AQP3 is able to mediate the transport of H 2 O 2 in different tissues and conditions [29][30][31]. Therefore, we first decided to quantify the basal expression of some AQPs and to assess the variation in their expression levels after honey exposure, by means of qPCR analysis and immunoblotting. In Figure 5A,B, we analyzed the basal expression of AQP-1, -3, -4, -5, -6, -8 and -9 transcripts in keratinocytes, but only the expression of AQP3 was remarkable enhanced after honey exposure, in particular after treatment with manuka honey. Immunoblotting results confirmed the AQP3 upregulation after manuka treatment. Therefore, we decided to investigate the role of AQP3 as a mediator of H 2 O 2 signaling in the Ca 2+ response to manuka honey by using a specific RNAi. As expected, silencing of AQP3 prevented the increase in intracellular ROS despite manuka honey exposure ( Figure 5G), thereby completely abrogating the rapid increase in [Ca 2+ ] i ( Figure 5C,D).  Figure 4A,B). As Pyr6 selectively targets Orai1, but not TRP channels [11,22,25], we did not silence Orai1. These data strongly suggest that TRPM2 and Orai1 mediate manuka honey-induced extracellular Ca 2+ entry. In agreement with this hypothesis, when HaCaT cells were pretreated with both econazole and Pyr6, the Ca 2+ response to manuka honey was fully abrogated. Cell incubation with vehicle DMSO, used alone, was ineffective on intracellular Ca 2+ dynamics (not shown).
In order to extend at molecular levels the findings obtained during Ca 2+ imaging recordings, we used qPCR to confirm that Stim1-2, Orai1-3 and TRPM2 transcripts are expressed in HaCaT cells ( Figure 4C).

Role of AQP3 in H2O2 Influx
As discussed in Section 2.2, the Ca 2+ response to manuka honey requires an increase in intracellular ROS levels which follows H2O2 liberation in the culture medium. Therefore, we hypothesized that, once released in the extracellular milieu, H2O2 diffuses across the plasma membrane to increase intracellular ROS concentration and activate external Ca 2+ entry. AQPs are integral membrane proteins acting as channels in the transfer of water across the membrane, playing an important role in skin hydration [26][27][28]. In addition, AQP3 is able to mediate the transport of H2O2 in different tissues and conditions [29][30][31]. Therefore, we first decided to quantify the basal expression of some AQPs and to assess the variation in their expression levels after honey exposure, by means of qPCR analysis and immunoblotting. In Figure 5A,B, we analyzed the basal expression of AQP-1, -3, -4, -5, -6, -8 and -9 transcripts in keratinocytes, but only the expression of AQP3 was remarkable enhanced after honey exposure, in particular after treatment with manuka honey. Immunoblotting results confirmed the AQP3 upregulation after manuka treatment. Therefore, we decided to investigate the role of AQP3 as a mediator of H2O2 signaling in the Ca 2+ response to manuka honey by using a specific RNAi. As expected, silencing of AQP3 prevented the increase in intracellular ROS despite manuka honey exposure ( Figure 5G), thereby completely abrogating the rapid increase in [Ca 2+ ]i ( Figure 5C,D).  the Ca 2+ response to 4% v/v manuka honey recorded at the peak (light bars) and at the plateau (dark bars) under the designate treatments. Number of cells as in C. Different letters above bars indicate statistical difference determined by One-way ANOVA followed by Bonferroni post-test (p < 0.01). (E) Effect of AQP3 RNAi on honey-induced scratch wound repair of HaCaT monolayers. Wound closure rate is expressed as the difference between wound width at 0 and 24 h. Data were recorded 24 h after scratch wound healing of cells exposed to 0.1% manuka honey. Bars represent mean ± SEM of wound closure inhibition deriving from two independent experiments, each with n = 20. Different letters on bars indicate significant differences according to Tukey's test (p < 0.01). (F) Aquaporin-3 (AQP3) protein expression in HaCaT cells after AQP3 RNAi. Blots representative of three were shown. Lanes were loaded with 15 µg of proteins, probed with anti-AQP3 rabbit polyclonal antibodies and processed as described in Materials and Methods. The same blots were stripped and re-probed with anti-β-actin polyclonal antibody, as housekeeping (* p < 0.001, t-test). (G) Fluorescence values recorded at 10 min. in scrambled or silenced (RNAi AQP3) cells, in control conditions (dark bars) or incubated with 4% manuka honey (light bars). Data are presented as means ± SD of rhodamine 123 fluorescence expressed in arbitrary units; n = 16 microplate wells from two different experiments. Different letters on bars indicate significant differences according to Tukey's test (p < 0.01).
To assess the role played by AQP3 in manuka honey-induced wound repair, we performed the scratch wound assay after silencing of AQP3 by RNAi (as shown in Figure 5F). In silenced response to 4% v/v manuka honey recorded at the peak (light bars) and at the plateau (dark bars) under the designate treatments. Number of cells as in C. Different letters above bars indicate statistical difference determined by One-way ANOVA followed by Bonferroni post-test (p < 0.01). (E) Effect of AQP3 RNAi on honey-induced scratch wound repair of HaCaT monolayers. Wound closure rate is expressed as the difference between wound width at 0 and 24 h. Data were recorded 24 h after scratch wound healing of cells exposed to 0.1% manuka honey. Bars represent mean ± SEM of wound closure inhibition deriving from two independent experiments, each with n = 20. Different letters on bars indicate significant differences according to Tukey's test (p < 0.01). (F) Aquaporin-3 (AQP3) protein expression in HaCaT cells after AQP3 RNAi. Blots representative of three were shown. Lanes were loaded with 15 µg of proteins, probed with anti-AQP3 rabbit polyclonal antibodies and processed as described in Materials and Methods. The same blots were stripped and re-probed with anti-β-actin polyclonal antibody, as housekeeping (* p < 0.001, t-test). (G) Fluorescence values recorded at 10 min. in scrambled or silenced (RNAi AQP3) cells, in control conditions (dark bars) or incubated with 4% manuka honey (light bars). Data are presented as means ± SD of rhodamine 123 fluorescence expressed in arbitrary units; n = 16 microplate wells from two different experiments. Different letters on bars indicate significant differences according to Tukey's test (p < 0.01).
To assess the role played by AQP3 in manuka honey-induced wound repair, we performed the scratch wound assay after silencing of AQP3 by RNAi (as shown in Figure 5F). In silenced keratinocytes (see Figure 5E), despite honey treatment, we could not observe any significant variation in the rate of wound closure, confirming the pivotal role of AQP3 in mediating wound closure.
To disclose the functional role of manuka honey-induced Ca 2+ signals in tissue regeneration, we performed the scratch wound assay in presence or not of econazole and Pyr6 ( Figure 6A,C). We observed that these inhibitors significantly reduced the wound healing rate, thereby confirming the role of TRPM2 and Orai1 in manuka honey-dependent tissue regeneration. Moreover, also CAT significantly (p < 0.05) reduced manuka honey-induced wound repair, in agreement with the role played by H 2 O 2 in TRPM2 and ORA1 activation ( Figure 6B,C).
keratinocytes (see Figure 5E), despite honey treatment, we could not observe any significant variation in the rate of wound closure, confirming the pivotal role of AQP3 in mediating wound closure.
To disclose the functional role of manuka honey-induced Ca 2+ signals in tissue regeneration, we performed the scratch wound assay in presence or not of econazole and Pyr6 ( Figure 6A,C). We observed that these inhibitors significantly reduced the wound healing rate, thereby confirming the role of TRPM2 and Orai1 in manuka honey-dependent tissue regeneration. Moreover, also CAT significantly (p < 0.05) reduced manuka honey-induced wound repair, in agreement with the role played by H2O2 in TRPM2 and ORA1 activation ( Figure 6B,C). Each inhibitor was used according to the procedures, which were previously shown to inhibit manuka honey-induced extracellular Ca 2+ entry: 10 µM Econazole, 1 µM Pyr6. We also used (B) 1000 U CAT. Wound closure rate is expressed as the difference between wound width at 0 and 24 h. Data were recorded 24 h after scratch wound healing of cells exposed to 0.5% manuka honey, in the presence or absence of various inhibitors. Bars represent mean ± SEM of wound closure inhibition deriving from two independent experiments, each with n = 20. For each inhibitor, different letters above bars indicate significant differences according to Tukey's test (p < 0.01). (C) Micrographs of scratch wounded HaCaT monolayers incubated under control conditions or in the presence of manuka honey and inhibitors and then stained with blue toluidine and observed 24 h after wounding. Scale bar, 200 µm.

Discussion
The essential role of intracellular Ca 2+ in wound repair and regeneration processes has been suggested by in vitro studies on scratch wounded layers of fibroblasts and endothelial cells [10,32].
Even more striking evidence has come from an in vitro study on HaCaT keratinocytes, showing that the cell-permeant Ca 2+ chelator BAPTA inhibits both the wound closure and chemoattraction effects [33].
Wound-induced intracellular Ca 2+ waves have been also observed in endothelial cells [34], whereas [Ca 2+ ]i rises have been found to promote cell growth and movement in endothelial repair [35]. Finally, in vivo evidence of the regulatory role of [Ca 2+ ]i in wound healing comes from the finding of high calmodulin levels in mitotic basal keratinocytes during the proliferative phase of rat wounds [36]. Each inhibitor was used according to the procedures, which were previously shown to inhibit manuka honey-induced extracellular Ca 2+ entry: 10 µM Econazole, 1 µM Pyr6. We also used (B) 1000 U CAT. Wound closure rate is expressed as the difference between wound width at 0 and 24 h. Data were recorded 24 h after scratch wound healing of cells exposed to 0.5% manuka honey, in the presence or absence of various inhibitors. Bars represent mean ± SEM of wound closure inhibition deriving from two independent experiments, each with n = 20. For each inhibitor, different letters above bars indicate significant differences according to Tukey's test (p < 0.01). (C) Micrographs of scratch wounded HaCaT monolayers incubated under control conditions or in the presence of manuka honey and inhibitors and then stained with blue toluidine and observed 24 h after wounding. Scale bar, 200 µm.

Discussion
The essential role of intracellular Ca 2+ in wound repair and regeneration processes has been suggested by in vitro studies on scratch wounded layers of fibroblasts and endothelial cells [10,32].
Even more striking evidence has come from an in vitro study on HaCaT keratinocytes, showing that the cell-permeant Ca 2+ chelator BAPTA inhibits both the wound closure and chemoattraction effects [33].
Wound-induced intracellular Ca 2+ waves have been also observed in endothelial cells [34], whereas [Ca 2+ ] i rises have been found to promote cell growth and movement in endothelial repair [35].
Finally, in vivo evidence of the regulatory role of [Ca 2+ ] i in wound healing comes from the finding of high calmodulin levels in mitotic basal keratinocytes during the proliferative phase of rat wounds [36].
However, until now, no data are available about component(s) (other than antibacterial) of the honey able to induce/accelerate wound healing. Honey samples are able to release, also in in vitro conditions, H 2 O 2 [37], and this H 2 O 2 can contribute to their antibacterial properties [38].
In our study, we demonstrated the involvement of H 2 O 2 as a main mediator of honey regenerative effects on HaCaT cells, a spontaneously immortalized human keratinocyte line We confirmed that this extracellularly released H 2 O 2 could pass across the plasma membrane through a specific aquaporin (i.e., AQP3). AQP3 has been already demonstrated to play a pivotal role in hydration of mammalian skin as well as in the regulation of proliferation and differentiation of keratinocytes [27]. Once in the cytoplasm, H 2 O 2 induces the entry of extracellular Ca 2+ through TRPM2 and Orai1 channels by a ROS-dependent mechanism [39]. The following pieces of evidence support this model. First, manuka honey failed to elicit a detectable increase in [Ca 2+ ] i in the absence of extracellular Ca 2+ . Second, the Ca 2+ response to manuka honey was mimicked by 50 µM H 2 O 2 and inhibited in a dose-dependent manner by CAT, an H 2 O 2 scavenger. Accordingly, honey samples may release H 2 O 2 that is produced by the enzyme glucose oxidase and is responsible for the antiseptic effect of honey dressing [40]. Despite the production of H 2 O 2 by honey, a harmful oxidizing effect is not observed on skin cells due to the honey polyphenolic component, which can antagonize this pro-oxidant action. In agreement with this hypothesis, the use of generic fluorescent H 2 O 2 indicators revealed local sites of H 2 O 2 production within the cytoplasm, including ER, mitochondria and plasma membrane [41][42][43]. This mechanism could also explain the different kinetics and amplitude of the Ca 2+ response induced by the three different types of honey we probed. Accordingly, the H 2 O 2 -releasing ability is manuka > buckwheat >> acacia. Third, genetic silencing of AQP3 with a selective RNAi prevented the increase in intracellular ROS levels and abrogated the Ca 2+ response to manuka honey. Accordingly, AQP3 has recently been shown to mediate H 2 O 2 uptake and accumulation within the cytosolic leaflet of the plasma membrane, thereby generating a signaling nanodomain that could gate adjacent H 2 O 2 -sensitive channels [41]. Conversely, AQP3 is not permeable to extracellular Ca 2+ , while there is not report about the recruitment of an intracellular signaling pathway leading to H 2 O 2 production upon honey exposure. These data, therefore, strongly suggest that AQP3 mediates H 2 O 2 entry into the cytosplasm. Fourth, H 2 O 2 may activate both TRPM2-and Orai1-mediated extracellular Ca 2+ entry by acting from the cytosolic side. As to TRPM2, it is still unclear whether H 2 O 2 directly gates TRPM2 [44] or does so by inducing ADP ribose (ADPr) mobilization from mitochondria [45,46]. H 2 O 2 could then facilitate ADPr-dependent TRPM2 activation [45]. Orai1, in turn, mediates the so-called store-operated Ca 2+ entry (SOCE), which is activated upon depletion of the ER Ca 2+ store [25]. The drop in ER Ca 2+ levels is sensed by STIM1, an ER-resident single-pass protein which oligomerizes and redistributes towards ER-plasma membrane junctions, known as puncta, thereby binding to and gating ORAI1 [25]. In agreement with our observations, earlier studies revealed that H 2 O 2 may activate Orai1 by promoting Stim1 oligomerization and relocalization from perinuclear ER to plasma membrane puncta through S-glutathionylation of cysteine 56 in COS7 cells [47]. Likewise, H 2 O 2 was found to stimulate Orai1 in a Stim1-dependent manner in HEK293 and DT40 cells [21] and in coronary vascular smooth muscle cells [22].
Honey-induced Ca 2+ entry through TRPM2 and Orai1 channels, in turn, accelerated wound closure in human keratinocytes. We have recently demonstrated that honey boosted the rate of wound repair in HaCaT cells by promoting migration rather than proliferation [48]. Of note, we found that honey-induced wound closure was inhibited by BAPTA, a membrane permeable buffer of intracellular Ca 2+ levels [4,10,49], and PD98059, which selectively blocks the extracellular signal-regulated kinase (ERK) pathway [5,50]. Herein, we demonstrated that honey-induced wound healing was severely inhibited by blocking extracellular Ca 2+ entry by the genetic suppression of AQP3 (which prevents H 2 O 2 entry), H 2 O 2 scavenging (with CAT) and the pharmacological blockade of TRPM2 and Orai1 channels (with econazole and Pyr6, respectively). Therefore, TRPM2 and Orai1 provide the major sources of Ca 2+ to recruit the Ca 2+ -dependent machinery driving HaCaT cell migration. Other works have demonstrated that intracellular Ca 2+ signaling may induce cell migration by engaging the ERK pathway [10,[50][51][52][53]. For instance, ORAI1-mediated Ca 2+ entry stimulated human melanoma cell migration by recruiting Ca 2+ /Calmodulin-dependent kinase II (CaMKII) to induce Raf-1 phosphorylation, thereby activating the ERK pathway [53]. Likewise, TRPM2 has been shown to drive migration in several cell types [54,55], although the downstream signaling machinery remains to be elucidated.
In conclusion, we demonstrated that the presence of H 2 O 2 in the extracellular space (due to honey's glucose oxidase enzymatic activity [7]) causes a change in the intracellular levels of Ca 2+ .
The present results can be explained by the following model (see Figure 7): • (CaMKII) to induce Raf-1 phosphorylation, thereby activating the ERK pathway [53]. Likewise, TRPM2 has been shown to drive migration in several cell types [54,55], although the downstream signaling machinery remains to be elucidated.
In conclusion, we demonstrated that the presence of H2O2 in the extracellular space (due to honey's glucose oxidase enzymatic activity [7]) causes a change in the intracellular levels of Ca 2+ .
The present results can be explained by the following model (see Figure 7): • Honey produces H2O2 in the extracellular space • AQP3 mediates the H2O2 entry in to the cytosol [41] • H2O2 activates TRPM2 and Orai1 determining Ca 2+ entry [56] from the outside that induces wound closure. Figure 7. Diagram depicting the mechanism of action of manuka honey on HaCaT keratinocytes, as characterized in the study. Data showed that honey could trigger intracellular Ca 2+ changes. The levels and profile were dependent on the source of honey and the source of the Ca 2+ is extracellular. Honey induced the production of micromolar levels of H2O2. Catalase blocks the Ca 2+ response indicating that it is a function of H2O2 production. Pharmacological and genetic (siRNA) inhibition of TRPM2 inhibited the Ca 2+ response to honey. Pharmaceutical inhibition of Orai1 also diminished the Ca 2+ response to honey. Inhibition of both types of channels completely abrogated the Ca 2+ response. Examination of aquaporins expression found that AQP3 was upregulated upon honey exposure. siRNA for AQP3 blocked the Ca 2+ increase elicited by honey treatment.
This is the first report showing that honey exposure affects [Ca 2+ ]i regulation, due to H2O2 production and redox regulation of ion channels.
These data lay the foundations for scientific and rational use of honey as a wound repair agent, with the understanding of cellular and molecular mechanisms underlying the effects on skin cells.
Moreover, our observations open up a new horizon for the use of the honey as a beneficial tool in the management of skin disorder by the modulation of aquaporin expression.

Honey Samples
Honey samples were obtained from Yamada Apiculture Center, Inc., Tomata-Gun, Okayama (Japan). Diagram depicting the mechanism of action of manuka honey on HaCaT keratinocytes, as characterized in the study. Data showed that honey could trigger intracellular Ca 2+ changes. The levels and profile were dependent on the source of honey and the source of the Ca 2+ is extracellular. Honey induced the production of micromolar levels of H 2 O 2 . Catalase blocks the Ca 2+ response indicating that it is a function of H 2 O 2 production. Pharmacological and genetic (siRNA) inhibition of TRPM2 inhibited the Ca 2+ response to honey. Pharmaceutical inhibition of Orai1 also diminished the Ca 2+ response to honey. Inhibition of both types of channels completely abrogated the Ca 2+ response. Examination of aquaporins expression found that AQP3 was upregulated upon honey exposure. siRNA for AQP3 blocked the Ca 2+ increase elicited by honey treatment. This is the first report showing that honey exposure affects [Ca 2+ ] i regulation, due to H 2 O 2 production and redox regulation of ion channels.
These data lay the foundations for scientific and rational use of honey as a wound repair agent, with the understanding of cellular and molecular mechanisms underlying the effects on skin cells.
Moreover, our observations open up a new horizon for the use of the honey as a beneficial tool in the management of skin disorder by the modulation of aquaporin expression.

Cell Culture and Reagents
All reagents were from Sigma-Aldrich (St. Louis, MO, USA), unless otherwise indicated. HaCaT cells are immortalized human skin keratinocytes that mimic many properties of normal epidermal keratinocytes, are not invasive, and can differentiate under appropriate experimental conditions [33].
After probe loading and washing, cells were analyzed through confocal time-lapse analysis, using a Zeiss LSM 510 confocal system interfaced with a Zeiss Axiovert 100 M microscope (Carl Zeiss Inc., Oberkochen, Germany).
Excitation was obtained by the 488 nm line of an Ar laser, and emission was collected using a 505-550 bandpass filter. The laser power was reduced to 15% in order to lower probe bleaching.
Confocal imaging was performed with a resolution of 512 × 512 pixels at 256 intensity values, with a framing rate of 1 frame/5 s.
Several cells were viewed together through a 20× Plan-Neofluar Zeiss objective (0.5 NA). Fluo-3/AM fluorescence was analyzed in digitized images as the average value over defined contours of individual cells, using the ROI-mean tool of the Zeiss LSM 510 2.01 software (Carl Zeiss Microscopy, Feldbach, Switzerland). Fluo-3 calibration was achieved by the following Equation (1) [59].
where K d = 400 nmol/L. Fmax and Fmin are maximum and minimum fluorescence intensities obtained by Fluo-3/AM calibration after cell exposure to 500 µM A23187 for about 10 min, followed by addition of 20 mM EDTA.

H 2 O 2 Assays
The formation of H 2 O 2 induced by manuka in culture medium was determined by xylenol orange, a colorimetric assay, as previously described [58,60].

Measurement of Intracellular ROS
Intracellular levels of ROS were assessed using the fluorescent dye precursor dihydrorhodamine (DHR) 123, which is converted to fluorescent rhodamine 123 upon reaction with ROS. Cells were seeded in 96-well plates, allowed to settle overnight, and loaded for 30 min at room temperature in the dark with DHR-123 (30 µM) in a loading buffer consisting of (mM) 10 glucose, 10 Hepes, 140 NaCl, 2 CaCl 2 , 1 MgCl 2 , and 5 KCl, pH 7.4. Cells were then washed with loading buffer and fluorescence collected in a fluorescence microplate reader, by using 485-nm excitation and 530-nm emission filters. Data of ROS production were expressed as fluorescence arbitrary units [58].

Quantitative Reverse Transcriptase PCR (qRT-PCR) and RNA Interference (siRNA)
Cells were settled in multi-well plates for 24 h and then subjected to the indicated experimental conditions. NucleoSpin RNAII Kit (Macherey-Nagel, Düren, Germany) was then utilized to extract total RNA. Complementary DNA was synthesized from RNA using the Transcriptor First Strand cDNA Synthesis Kit (Roche Diagnostics GmbH, Penzberg, Germany).
Scramble siRNA was obtained using commercial non-targeting siRNA (MISSION siRNA Universal Negative Control).
Cells were collected at 24 h after transfection and used for the experiments.

Immunoblotting
Cells were lysed in RIPA buffer (supplemented with a protease and phosphatase inhibitor cocktail) and homogenates were solubilized in Laemmli buffer [61]. 15-30 µg solubilized proteins were subjected to 12.5% SDS-polyacrilamide gel electrophoresis and transferred to the Hybond-P PVDF Membrane (GE Healthcare, Chicago, IL, USA) by electroelution. The membranes were incubated overnight with anti-AQP3 antibody produced in rabbit (SAB5200111; Sigma, Italy) diluted 1:1000 in TBS and 0.1%