Electrical Stimulation Enhances Epidermal Proliferation in Human

ORIGINAL ARTICLE

Electrical Stimulation Enhances Epidermal Proliferation in Human Cutaneous Wounds by Modulating p53–SIVA1 Interaction Anil Sebastian1,SyedA.Iqbal1, James Colthurst2, Susan W. Volk3 and Ardeshir Bayat1,4,5

Cutaneous wounds establish endogenous ‘‘wound current’’ upon injury until re-epithelialization is complete. Keratinocyte proliferation, regulated partly by p53, is required for epidermal closure. SIVA1 promotes human double minute 2 homolog (HDM2)–mediated p53 regulation. However, the role of SIVA1 in wound healing is obscure. Here, we report that electrical stimulation (ES) accelerates wound healing by upregulating SIVA1 and its subsequent ability to modulate p53 activities. Cultured donut-shaped human skin explants, subjected to ES, exhibited better epidermal stratification, increased proliferation, and upregulation of gene and protein expression of HDM2/SIVA1, compared with non-ES-treated explants. ES significantly increased in vitro keratinocyte proliferation and phospho-p53–SIVA1 interaction; however, this showed stable expression of phospho-p53, which increased significantly in the absence of SIVA1. Here, HDM2 alone was unable to downregulate nuclear- accumulated phospho-p53, which was evident from decreased proliferation and increased sub-G1 population seen by flow cytometry. Further examination of the epidermis of human cutaneous wounds showed higher p53–SIVA1 coexpression and proliferation 7 days after injury in ES-treated wounds compared with control wounds. In summary, ES-inducible SIVA1 modulates p53 activities in proliferating keratinocytes, and exogenous ES affects p53/HDM2/SIVA1 axis leading to increased proliferation during re-epithelialization. This highlights ES as a potential strategy for enhancing cutaneous repair. Journal of Investigative Dermatology (2015) 135, 1166–1174; doi:10.1038/jid.2014.502; published online 8 January 2015

INTRODUCTION However, precise mechanisms by which ES promotes tissue Upon injury to the epithelium, an endogenous transepithelial repair remain poorly defined. potential drives current away from low-resistance regions The role of exogenous ES in accelerating re-epithelialization (Nuccitelli, 2003) and results in the generation of a lateral has been associated with cell migration (Nishimura et al., electric field (wound current; Nuccitelli et al., 2008) within 1996; Ojingwa and Isseroff, 2003; Jennings et al., 2010). the epidermis. Therefore, the rationale for exogenous However, the exact role of cell proliferation has not been application of electrical stimulation (ES) is to mimic the investigated in relation to ES and re-epithelialization, despite in vivo bioelectric system or to augment the endogenous the fact that normal healing requires keratinocyte proliferation wound current that promotes the wound healing process. during re-epithelialization (Amar and Wu, 2014). In mammalian cells, p53 tumor suppressor activates antiproliferative processes in response to stress such as DNA 1Plastic and Reconstructive Surgery Research, Institute of Inflammation and damage (Vogelstein et al., 2000; Vousden and Lane, 2007). Repair, Manchester Institute of Biotechnology, School of Medical and Human Temporal expression of p53 in wound healing has been Sciences, University of Manchester, Manchester, UK; 2Fenzian Ltd, Hungerford, Berkshire, UK; 3Section of Surgery, Department of Clinical investigated primarily in experimental animal models with a Studies, University of Pennsylvania School of Veterinary Medicine, paucity of data available for human epithelial healing Philadelphia, Pennsylvania, USA; 4Centre for Dermatology, Institute of (Antoniades et al., 1994). In vitro datasuggestthatthe Inflammation and Repair, University of Manchester, Manchester, UK and maximal expression of p53 occurs in proliferating 5Manchester Academic Health Science Centre, South Manchester University Hospital Foundation Trust, Wythenshawe Hospital, Manchester, UK keratinocytes (Dazard et al., 2000). Recently, it has been Correspondence: Ardeshir Bayat, Plastic and Reconstructive Surgery Research, shown that mild ES induces p53 phosphorylation via p38 Manchester Institute of Biotechnology, School of Medical and Human mitogen-activated protein kinase signaling and G2 cell cycle Sciences, University of Manchester, 131 Princess Street, Manchester M1 7DN, arrest without apoptosis (Fukuda et al., 2013). UK. E-mail: [email protected] The activation of p53 induces the expression of downstream Abbreviations: DW, degenerate wave; ES, electrical stimulation; HDM2, genes such as SIVA1 and human double minute 2 (HDM2), human double minute 2; PCNA, proliferating cell nuclear antigen which further regulate apoptosis and cell cycle progression Received 29 August 2014; revised 22 October 2014; accepted 12 November 2014; accepted article preview online 28 November 2014; published online (Chen et al., 1994; Beckerman and Prives, 2010). SIVA1 has 8 January 2015 been identified to have both antiapoptotic (Du et al., 2010)

1166 Journal of Investigative Dermatology (2015), Volume 135 & 2015 The Society for Investigative Dermatology A Sebastian et al. SIVA1 Modulates p53 in Wound Healing

and proapoptotic roles through multiple signaling pathways release assay. This assay did not show any significant increase (Chu et al., 2004; Resch et al., 2009), whereas HDM2 has in cell death until beyond day 16 in the cultured tissues been shown to act as the principal p53 antagonist (E3 ligase compared with baseline values (Supplementary Figure S2c for p53) and it limits the duration and intensity of p53 online). Progression of healing between groups was compared activation (Michael and Oren, 2003). In normal/unstressed on days 6 and 16 (Figure 1a and Supplementary Figure S2a cells, SIVA1 promotes p53-HDM2 interaction, thereby online). ES treatment accelerated tissue ingrowth into the enhancing p53 ubiquitination and degradation without wound center compared with the control explants in our affecting the stability of HDM2 (Du et al., 2010). ex vivo model (Supplementary Figure S2d online). To further The ability of human keratinocytes to upregulate prolifera- explore the ability of ES to maintain epidermal stratification tion-related subunits while simultaneously increasing p53 and and integrity over 16 days, cytokeratin 10 immunohistochem- HDM2 expressions has been shown previously (Flatt et al., istry was performed (Figure 1b). Cytokeratin 10 stains supra- 1998). In one study, instead of directly affecting p53 protein basal epidermal layers, and it is one of the preliminary proteins levels, the downstream target genes of p53 including those for to be detected when keratinocytes commit to differentiate apoptosis were significantly reduced (transcriptional (Patel et al., 2006; Lamb and Ambler, 2013). mRNA transcript repression) by overexpressing HDM2 (Ohkubo et al., 2006). quantification of cytokeratin 10 and the epidermal thickness However, the role of SIVA1 as a mediator for modulating p53 measurements showed a significant increase (P ¼ 0.001 and activities remains ill-defined. Collectively, as (i) ES increases P ¼ 0.006, respectively) on day 16 ES explants compared with p53 phosphorylation (Fukuda et al., 2013), (ii) p53 differential day 16 control explants (Figure 1c and d). The results indicated expression is observed in cellular proliferation (Dazard et al., the potential advantage of electric fields in maintaining the 2000; Ohkubo et al., 2006), and (iii) SIVA1 enhances p53 epidermal integrity of the wounded explants. degradation, we hypothesized that ES can accelerate the rate of re-epithelialization through its ability to modulate p53 in a ES upregulates PCNA, HDM2, and SIVA1 in the epidermis of SIVA1-dependent manner and subsequently increase ex vivo–cultured human skin explants proliferation. Therefore, the aim of this study was to To test the hypothesis that ES promotes keratinocyte prolifera- determine whether SIVA1 modulation of p53 expression tion, which is known to contribute to both epidermal main- drives keratinocyte proliferation during normal cutaneous tenance and re-epithelialization through its ability to modulate healing and whether ES drives this proregenerative response p53 activity, we examined the expression of proliferating cell in healing human skin. Our results showed that increased nuclear antigen (PCNA), phospho-p53, HDM2, and SIVA1 in expression of SIVA1 significantly modulates p53 activities and the wounded human explants exposed to exogenous electric enhances keratinocyte proliferation during ES. field. A significant upregulation (Po0.05) of PCNA on day 16 in ES explants was seen, predominantly within the stratum RESULTS basale of the epidermis, compared with day 16 of control ES preserves epidermal integrity in ex vivo human organ culture explants (Figure 2a and b, Supplementary Figure S3a and b tissue online). Here, a stable expression of phospho-p53 was observed To determine the role of ES in human cutaneous repair, an in stratum basale despite elevated proliferation levels (Figure 2a ex vivo donut-shaped wound healing model was developed. and c, Supplementary Figure S3c and d online). HDM2 showed Here, human skin tissues (n ¼ 10) were collected and significant overexpression (Po0.05) in ES explants compared ‘‘wounded’’ using a standardized punch biopsy (full-thickness with controls on day 16 (Figure 2a and d, Supplementary Figure wounds; 4 mm wound inside an 8 mm tissue biopsy; S3e and f online). HDM2 was present in the stratum spinosum Supplementary Materials and Methods online). Tissues were and the stratum basale in normal human skin epidermis on day divided into three categories: (i) non-ES treated (control), (ii) 0 (Supplementary Figure S3e online). However, HDM2 was direct current treated (monophasic ES treatment (Sebastian confined only to the stratum basale in the control explants on et al., 2011a); Supplementary Figure S1a online), and (iii) day 16, while it was also expressed in the suprabasal layers of degenerate wave (DW) treated (biphasic ES treatment the ES explants similar to normal skin on day 0. This differential (Sebastian et al., 2011b); Supplementary Figure S1b; localization of HDM2 expression on day 16 suggests a Supplementary Materials and Methods online). Both categories supportive role for ES in keratinocyte differentiation and subjected to ES were exposed to 100 mV mm 1 of electric maintenance of normal epidermal architecture. Significant field for 30 min every day for 16 days (Supplementary Figure upregulation (Po0.05) of SIVA1 was also observed on day 16 S2a online). The applied electric field was similar to the in ES explants compared with control explants (Figure 2a and e, endogenous ‘‘current of injury’’ prevailing during the initiation Supplementary Figure S3g and h online). SIVA1 was also of granulation tissue formation and wound contraction, further expressed predominantly in the stratum basale of the epidermis, leading to innate wound closure (Nuccitelli et al., 2008, 2011). similar to PCNA, phospho-p53, and HDM2. Matrigel was added to the center of the explants, as mammalian cell growth has been previously reported in this ES upregulates PCNA, HDM2, and SIVA1 in human three-dimensional basement membrane gel (Muthuswamy keratinocytes in vitro, and promotes physical interaction of et al., 2001; Supplementary Figure S2b online). Cell viability phospho-p53 with SIVA1 within cultured explants over the period examined in this study We further investigated the expression of PCNA, phospho- was confirmed using mitochondrial lactate dehydrogenase p53, HDM2, and SIVA1 at the cellular level following ES. For

www.jidonline.org 1167 ASebastianet al. SIVA1 Modulates p53 in Wound Healing

Day 0 control Control DC DW

Day 16

Control day 0 Suprabasal S. basale layers

Control day 16 DC day 16 DW day 16

Cytokeratin10 100 0.0006 80 0.0004 60 40 0.0002 Epidermal 20 0 thickness ( μ m) 0 Control Control DC day DW day

Relative mRNA levels Control Control DC day DW day day 0 day 16 16 16 day 0 day 16 16 16

Figure 1. Maintainence of human skin explants in tissue culture. (a) Within the limited exposure to electrical stimulation (ES; 100 mV mm 1 every day for 30 min), ES explants had partial dermal wound closure compared with control explants, on day 16. (b) Immunohistochemical analysis of cytokeratin 10 on day 0 and day 16 explants, with and without ES. (c) Messenger RNA (mRNA) transcript quantiﬁcation of cytokeratin 10. There was nearly a threefold increase in mRNA expression of cytokeratin 10 in day 16 ES explants compared with day 16 control explants. (d) Epidermal thickness was nearly twofold in day 16 ES explants compared with day 16 control explants, showing similarity of ES explants to normal skin. DC, direct current; DW, degenerate wave. Bar ¼ 1mmfor(a)and50mmfor(b).

this, we subjected a monolayer of human primary keratino- keratinocyte proliferation was increased with a coincident cytes to similar electric field for 1 h and assessed the mRNA significant increase in SIVA1 induction. transcripts by quantitative real-time reverse-transcriptase–PCR. Next, we looked for any physical interaction between We observed a significant increase (Po0.05) in mRNA phospho-p53 and SIVA1 post ES. Endogenous phospho-p53 transcripts of SIVA1 and PCNA (nearly twofold and threefold was immunoprecipitated by an anti-SIVA1 antibody but not by increase, respectively) in keratinocytes subjected to ES com- a control antibody and reciprocally, which showed an pared with non-ES-treated keratinocytes (Supplementary enhanced interaction between phospho-p53 and SIVA1 post Figure S4a online). In addition, protein analysis also confirmed ES (Figure 3c). significant upregulation of PCNA, HDM2, and SIVA1 (Po0.05 compared with control explants; Figure 3a and b), but with a Loss of SIVA1 augments the antiproliferative activity of p53 post stable expression of phospho-p53 post ES (P40.05). Taking ES together the mRNA and the protein results, the in vitro ES Next, we investigated the role of SIVA1 in augmenting conditions could maintain p53 levels, but with a significant keratinocyte proliferation. For this, we transfected primary increase of SIVA1. Here, similar to the explant study, human keratinocytes with single-interference RNA for SIVA1

1168 Journal of Investigative Dermatology (2015), Volume 135 A Sebastian et al. SIVA1 Modulates p53 in Wound Healing

PCNA Phospho-p53 35 NS 30 25 40 20 30 15 20 10 10 ControlDC day day 0DW 16 dayControl 16 day 16 5 band intensity Percentage of 0 band intensity 90 kDa 0 Percentage of HDM2 Control DC DW Control Control DC DW Control phospho- day 0 day 16 day 16 day 16 day 0 day 16 day 16 day 16 55 kDa p53

35 kDa PCNA 40 HDM2 SIVA1 35 15 kDa SIVA1 30 40 25 20 30 40 kDa β-actin 15 20 10 band intensity Percentage of 5 10 band intensity 0 Percentage of 0 Control DC DW Control Control DC DW Control day 0 day 16 day 16 day 16 day 0 day 16 day 16 day 16

Figure 2. Electrical stimulation (ES) upregulates proliferating cell nuclear antigen (PCNA), human double minute 2 (HDM2), and SIVA1 in human skin explants. Protein analysis of (a, b) PCNA, (a, c) Phospho-p53, (a, d) HDM2, and (a, e) SIVA1 on day 0 and day 16 skin explants, with and without ES. On day 16, PCNA, HDM2, and SIVA1 were signiﬁcantly upregulated (Po0.05) in explants subjected to ES compared with control explants. However, there was a stable expression of phospho-53 (P40.05). Immunohistochemical analysis (Supplementary Figure S3) also supported this observation. DC, direct current; DW, degenerate wave; NS, nonsigniﬁcant.

50 NS Control DC DW 45 40 55 kDa HDM2 Input 35 DW ESWild-typeKeratinocytesIgG control 30 phospho- 55 kDa Phospho- IP: anti-SIVA1 55 kDa 25 p53 p53 20 35 kDa SIVA1 PCNA 15 IP: anti-p53 15 kDa 10 β-actin SIVA1 5 40 kDa 15 kDa 0 Percentage of band intensity DC DC DC DC DW DW DW β-actin DW 40 kDa Control Control Control Control 4 HDM2 Phospho- PCNA SIVA1 3.5 p53 3 2.5 2 45 Control DC DW 1.5 SIVA1 40 siRNA + + + 35 1 30 55 kDa HDM2 25 0.5 20 Phospho- 0 55 kDa 15 Fold change of gene expression DC DC DC DC DC p53 DC DW DW DW DW DW 10 DW 35 kDa 5 Control Control Control Control Control PCNA Control 0

Percentage of band intensity SIVA1 SIVA1 SIVA1 DC DC β DC DW DW -actin DW 40 kDa WT KD WT KD WT KD Control Control Control p53 HDM2 PCNA HDM2 Phospho- PCNA p53

Figure 3. Proliferation was decreased in keratinocytes subjected to electrical stimulation (ES), post SIVA1 single-interferenceRNA (siRNA) transfection. (a) Western blotting showed upregulation of proliferating cell nuclear antigen (PCNA), human double minute 2 (HDM2), and SIVA1 expression post ES. (b) Densitometric analysis of band intensities in a showed signiﬁcant increase in PCNA, HDM2, and SIVA1 (Po0.05). (c) Coimmunoprecipitation (IP) followed by western blotting of phospho-p53 and SIVA1. (d) Western blotting of keratinocytes subjected to SIVA1 siRNA transfection. (e) Densitometric analysis of band intensities in d showed the upregulation of HDM2 and downregulation of proliferation in keratinocytes subjected to ‘SIVA1 siRNA transfection þ ES’ compared with siRNA-transfected keratinocytes. (f) Fold change in messengerRNA expression of ‘keratinocytes þ ES’, ‘keratinocytes þ SIVA1 siRNA’, and ‘keratinocytes þ SIVA1 siRNA þ ES’, compared with wild-type keratinocytes. NS represents nonsigniﬁcant. DC, direct current; DW, degenerate wave; IP, immunoprecipitation. Bar ¼ 50 mm.

(Supplementary Figure S4b and c online) and subsequently (P40.05; Supplementary Figure S4d online), protein analysis applied ES. Here, loss of SIVA1 expression in keratinocytes showed a significant decrease (Po0.05) in PCNA expression post ES induced higher mRNA expression of p53 and HDM2 (Figure 3d and e). HDM2 protein expression was also compared with non-ES-treated keratinocytes (Po0.05; significantly upregulated (Po0.05) in siRNA-transfected kera- Supplementary Figure S4d online). Although the decline in tinocytes post ES compared with control; however, the mRNA expression of PCNA in SIVA siRNA-treated keratino- antiproliferative effect of enhanced phospho-p53 was evident cytes post ES was nonsignificant compared with controls after ES treatment (Figure 3d and e). This shows decreased

www.jidonline.org 1169 ASebastianet al. SIVA1 Modulates p53 in Wound Healing

DAPI Phospho-p53 SIVA1 Merge 400 G1 Wild-type keratinocytes 300 Wild-type Sub- G1 keratinocytes 200 S Count G2/M 100

0 10 200.000 400.000 654.678

DW 400 DW G1 300

200 Sub- Count G1 S G /M DW 100 2 SIVA1 siRNA 0 66.813 200.000 400.000 600.000 774.923

400 SIVA1 siRNA G1 DW 60 300 Sub- 40 G 200 1 S Wild-type Count SIVA1 siRNA G /M keratinocytesDW ES + DW ES 20 1 phospho- 100 55 kDa 0

p53 Band intensities

55 kDa HDAC1 0 DW ES 10 200.000 400.000 735.974 Wild-type Propidium iodide + DW ES SIVA1 siRNA keratinocytes

DAPI HDM2 SIVA1 Merge

Wild-type 60 keratinocytes 50

30 DW 20

Percentage of cells 10

0 2n < G S G /M DW 1 2 SIVA1 siRNA Cell cycle

Figure 4. Upregulation of p53 in keratinocytes subjected to ‘SIVA1 single-interferenceRNA (siRNA) transfection þ ES’. (a) Intracellular localization of SIVA1 and p53 by double immunolabeling. SIVA1 knocked-down keratinocytes post direct current (DW) electrical stimulation (ES) showed stronger signals of p53 accumulated in nuclear and perinuclear regions. (b) Phospho-p53 from nuclear lysate analyzed by western blotting using histone deacetylase (HDAC1) as the internal control. (c) Higher expression of human double minute 2 (HDM2) was observed in the cytoplasm and the nucleus of SIVA1 knocked-down keratinocytes post DW ES, compared with ‘keratinocytes þ DW ES’. (d) Cell cycle analysis showed an increase of S phase in ‘keratinocytes þ DW ES’ compared with wild-type keratinocytes. In ‘keratinocytes þ SIVA1 siRNA þ DW ES’, there was an increase in sub-G1 and decrease in G1 and S phases compared with ‘keratinocytes þ DW ES’. DAPI, 40,6-diamidino-2-phenylindole.

degradation efficiency of HDM2 or increased transcriptional ubiquitination processes (O’Brate and Giannakakou, 2003). activity of phospho-p53 in keratinocytes subjected to ES in the In wild-type keratinocytes, SIVA1 was rarely detected as a absence of SIVA1. This suggests that HDM2-modulated nuclear protein; however, weak cytoplasmic localization was activities of p53 in wound healing, such as proliferation of observed (Figure 4a). ES (DW) significantly increased phos- human keratinocytes, is largely a SIVA1-dependent process. pho-p53 and SIVA1 expression levels in the cytoplasm (Figure 4a). Interestingly, here, SIVA1 was also expressed in Phospho-p53 accumulates more in the cell nucleus post ES in the nucleus where a near absence of phospho-p53 was SIVA1 knocked-down keratinocytes observed. However, loss of SIVA1 induced an increase in Having established that downregulation of SIVA1 expression nuclear expression and a slight decrease in the cytoplasmic could both significantly increase p53 and decrease PCNA expression of phospho-p53 in keratinocytes post ES. In addi- expression, we investigated whether ES could also induce tion, an increase in phospho-p53 levels was also evident from differential intracellular localization of phospho-p53, HDM2, the protein analysis of nuclear lysates post ES (Figure 4b). and SIVA1. Generally, p53 shuttles between the cytoplasm In wild-type keratinocytes, HDM2 was mainly localized in and the nucleus, which is controlled by several mechanisms the cytoplasm. This suggests degradation of cytoplasmic including potential transactivation, protein stability, and phospho-p53 in wild-type keratinocytes under normal

1170 Journal of Investigative Dermatology (2015), Volume 135 A Sebastian et al. SIVA1 Modulates p53 in Wound Healing

20 PCNA 18 80 16 60 14 40 12 20 10 0 8 Percentage of positive wound epidermis cells in wound epidermis ES ES ES 6

Normalskin Non-ES Non-ES Non-ES 4

Day 0 Day 3 Day 7 Day 10 Percentage of positive cells in 2 0 HDM2 HDM2 HDM2 HDM2 HDM2 HDM2 HDM2 SIVA1 SIVA1 SIVA1 SIVA1 SIVA1 SIVA1 SIVA1 Pho-p53 Pho-p53 Pho-p53 Pho-p53 Pho-p53 Pho-p53 Pho-p53

Normal Day 3 Day 7 Day 10 90 skin 80 Increased dissociation Decreased downstream Normal of SIVA1 from the transcriptional keratinocyte 70 ternary complex repression proliferation 60 50 p53 Normal healing SIVA1 p53 40 HDM2 30 20

Percentage of co-localization 10 p53 0 p53 Human cutaneous SIVA1 HDM2 wound healing on day 7

Exogenous Decreased dissociation or Increased downstream Increased application of better maintenance of transcriptional keratinocyte

HDM2-SIVA1 HDM2-SIVA1 HDM2-SIVA1 HDM2-SIVA1 HDM2-SIVA1 HDM2-SIVA1 HDM2-SIVA1 electrical stimulation SIVA1 in the repression proliferation Pho-p53-SIVA1 Pho-p53-SIVA1 Pho-p53-SIVA1 Pho-p53-SIVA1 Pho-p53-SIVA1 Pho-p53-SIVA1 Pho-p53-SIVA1 ternary complex

Normal Day 3 Day 7 Day 10 skin

Figure 5. Proliferation, phospho-p53–SIVA1 coexpression, and ternary complex interactions were upregulated on day 7 in healing epidermis treated with electrical stimulation (ES). (a) ES-treated wounds showed higher proliferation than non-ES-treated wounds until day 7. (b) Quantitative analysis of phospho-p53 and SIVA showed that the percentage of positive cells increased until day 7. (c) Analysis of coexpression of phospho-p53–SIVA1 and human double minute 2 (HDM2)-SIVA1 showed higher percentage of coexpression until day 3. On further days, there was higher downregulation of p53–SIVA1 coexpression in non-ES- treated wounds compared with ES-treated wounds. (d) Ternary complex on healing day 7. We suggest higher transcriptional repression activities of HDM2 in association with SIVA1 on p53 targets in ES-treated wounds than in non-ES-treated wounds. ( ) Normal healing wounds; (....) ES-treated wounds. PCNA, proliferating cell nuclear antigen; Pho, phospho. — conditions by coordinated HDM2 and SIVA1 interaction. After SIVA1 modulates cell cycle progression of primary keratinocytes ES, HDM2 was strongly induced in both the nucleus and the post ES cytoplasm of keratinocytes (Figure 4c). When correlated to Given the known role of p53 activation on cell cycle phospho-p53 cellular localization post ES, this suggests pro- progression and apoptosis (Chen et al.,1994),we teosomal degradation of nuclear phospho-p53 by combined investigated the role of SIVA1 in p53-dependent modulation HDM2 and SIVA1 interaction. In SIVA1-deficient keratino- of keratinocyte cell cycle in wound healing electric fields. cytes post ES, an upregulation of both nuclear and cytoplasmic Here, propidium iodide staining for DNA content was HDM2 was observed. Here, when correlated to phospho-p53 performed. Compared with wild-type keratinocytes (non-ES cellular localization post ‘SIVA1 knock down þ ES’, the treated), DW ES significantly increased S phase population decrease in cytoplasmic phospho-p53 degradation could (P ¼ 0.0037; Figure 4d); however, attenuation in S phase was largely be because of HDM2 alone (if any), but this was not observed. In addition, there was only a mild increase totally unseen in the nucleus. The increased nuclear localiza- (P40.05) in G2/M phase, which indicated the possibility for tion of phospho-p53 could also be because of nuclear import further initiation of cellular differentiation. Collectively, of phospho-p53 (Liang and Clarke, 1999). However, wound healing electric field of 100 mV/mm may increase HDM2 alone was unable to decrease nuclear phospho-p53 DNA synthesis with a likely representation of DNA replication levels. Collectively, the in vitro staining suggests (i) rather than DNA repair in human primary keratinocytes. In cytoplasmic degradation of phospho-p53 under normal SIVA1 knocked-down keratinocytes post ES, sub-G1 popula- conditions by HDM2 and SIVA1, (ii) enhanced degradation tion was significantly increased when compared with wild- of nuclear phospho-p53 by HDM2-SIVA1 coexpression post type keratinocytes and ‘wild-type keratinocytes þ DW ES’ ES, and (iii) significant decrease in the ability to degrade (P ¼ 0.0004). Here, G1 and S phase cells were significantly nuclear phospho-p53 by HDM2 alone in the absence of decreased (P ¼ 0.0044 and P ¼ 0.008, respectively). This SIVA1, post ES. statistically significant increase in sub-G1 post ‘SIVA1

www.jidonline.org 1171 ASebastianet al. SIVA1 Modulates p53 in Wound Healing

knockdown þ ES’ indicated a gradual transition of the prolif- wounds on day 7 compared with non-ES-treated wounds (ES erating cells in G2/M population to potential cell death. These treatment day 7:normal skin healing day 7 ¼ 34%, P ¼ 0.001). data support a role for SIVA1 in modulating the antiprolifera- Consistent with the role of ES in upregulating SIVA1 during tive and proapoptotic effect of p53 in human keratinocytes cutaneous healing, the observed increment in the colocaliza- post ES. tion of phospho-p53–SIVA1 in ES-treated wounds has shown to enhance proliferation on day 7. HDM2-SIVA1 temporal ES increases epidermal proliferation in vivo in human cutaneous coexpression was also similar to phospho-p53–SIVA1 coex- healing wounds pression on the respective healing days, with no significant To further characterize the in vivo role of SIVA1 in modulating difference between the ES-treated and the non-ES-treated keratinocyte proliferation in response to exogenous ES during wounds. This validates the known role of HDM2 in modulat- human skin repair, we examined phospho-p53 /HDM2/SIVA1 ing p53 along with SIVA1. Taking together proliferation and expression in biopsies from volunteers (n ¼ 20) undergoing phospho-p53-HDM2-SIVA1 coexpression during ES, we sug- normal healing (non-ES treated) and ES-assisted healing (DW gest that SIVA1 may also have a critical role to enhance ES treatment; Supplementary Figure S5; Supplementary in vivo proliferative mechanisms in association with HDM2, Materials and Methods online). After establishing that ES but without p53 degradation, as previously reported (Ohkubo upregulated proliferation within the epidermis of human et al., 2006). explants ex vivo, we investigated temporal induction of proliferation within the epidermis during both normal and DISCUSSION ES-assisted wound healing processes (Figure 5a and In this study, we showed that exogenous ES can effectively Supplementary Figure S6a online). As anticipated, keratino- maintain epidermal integrity ex vivo and enhance prolifera- cyte proliferation in ES-treated wounds was higher compared tion/SIVA1 expression in keratinocytes while maintaining the with non-ES-treated wounds throughout the healing process level of p53. Our in vitro results were consistent with ex vivo (until day 10). Although proliferation peaked on day 3 in both results; however, loss of SIVA1 in keratinocytes resulted in an ES and non-ES-treated wounds (as evidenced by the percen- increase of p53 and attenuation of proliferation post ES. tage of cells stained positive for PCNA in Figure 5a), ES Finally, a significant increase in the coexpression of phos- treatment increased cellular proliferation by 19 and 38% on pho-p53–SIVA1, and proliferation during ES-assisted wound healing days 3 and 7, respectively (P ¼ 0.002 for ES treatment healing in human epidermis, was shown in our in vivo healing day 7 versus normal skin healing day 7). In addition, the data. Collectively, our findings identified that SIVA1 was number of epidermal cells were significantly higher in ES- upregulated in vitro, ex vivo,andin vivo post ES application treated wounds on both healing days 3 (increase ofB62%, and that the enhanced SIVA1 modulated p53 activities with- P ¼ 0.001) and 7 (increase ofB27%, P ¼ 0.02) compared with out p53 degradation. normal healing wounds (Supplementary Figure S6b online). In normal conditions, SIVA1 promotes p53-HDM2 complex Consistent with this finding, the thickness of healing epithe- interaction and enhances HDM2-mediated p53 ubiquitination lium was also slightly higher (P40.05) in ES-treated wounds and degradation (Du et al., 2010). However, when cells are compared with non-ES-treated wounds (increment of nearly subjected to DNA damage, this ternary complex is broken B18% on day 3 andB11% on day 7, respectively; down and p53 is stabilized (Mei and Wu, 2012). However, Supplementary Figure S6c online). few studies indicate an inhibition of upregulated p53 activities by enhanced HDM2 protein levels (Chen et al.,1994).Even ES treatment increases p53–SIVA1 colocalization in vivo though both HDM2 and phospho-p53 were upregulated Next, in order to define SIVA1-mediated p53 modulation by following SIVA1 knock down in keratinocytes, the HDM2 in vivo, colocalization of phospho-p53, HDM2, and degradation-promoting effect of HDM2 alone on p53 in SIVA1 was investigated. Specifically, we examined the colo- physiological wound healing is questionable. calization of (i) phospho-p53–SIVA1 and (ii) HDM2-SIVA1 by Chen et al. (1994) has shown the inability of cells to arrest double immunolabeling the neo-epidermis (Supplementary in G1 phase following overexposure of HDM2 as it Figure S7a and b online, respectively). In both normal and downregulates p53. Accordingly, in SIVA1 knocked down ES-treated healing processes, the percentage of cells stained cells, HDM2 was unable to decrease p53 post-ES and positive for phospho-p53 and SIVA1 were increased until day consequently G1 population was decreased, which is also in 7 and decreased on further days (Figure 5b). Notably, there accordance with Flatt et al. (1998). Our observations on was a significant increase in SIVA1 expression on days 3 and 7 inefficient p53 regulation by HDM2 alone is further in ES-treated wounds compared with non-ES-treated wounds strengthened by increased nuclear localization of phospho- (increment ofB34%, P ¼ 0.009 and increment ofB41%, p53 post ES in the absence of SIVA1. After ES, SIVA1 and P ¼ 0.004, respectively, on days 3 and 7; Figure 5b). The HDM2 were expressed in the nucleus, where an absence of colocalization of phospho-p53–SIVA1 and HDM2-SIVA1 was phospho-p53 was observed, indicating successful p53 at its highest on day 3 for both the healing mechanisms (ES degradation. Here, cytoplasmic localization of phospho-p53 treated and non-ES treated), which correlated to the highest was also detected in the presence of HDM2 and SIVA1, proliferation rate on day 3 when compared with subsequent supporting different mechanisms of p53 accumulation and wound healing days (Figure 5c). However, the colocalization degradation in nuclear and cytoplasmic segments (O’Brate of phospho-p53–SIVA1 was significantly higher in ES-treated and Giannakakou, 2003).

1172 Journal of Investigative Dermatology (2015), Volume 135 A Sebastian et al. SIVA1 Modulates p53 in Wound Healing

Endogenous phospho-p53 was increased until day 7 in Establishment of primary keratinocytes human cutaneous wound healing. However, phospho-p53– Primary keratinocytes were isolated from normal skin tissues obtained SIVA1 coexpression was reduced on day 7 compared with from abdominoplasties. Detailed methodology is provided in day 3, which indicated the stabilization of p53. The intensity Supplementary Materials and Methods online. of this dissociation was greater in normal healing (non-ES- treated wounds), which reflected the diminished prolifera- Preparation of human skin explants, maintenance of organ tion rate on day 7 relative to ES-treated wounds. When culture as air-exposed in collagen gel matrix, and separation of correlated to in vitro ES results, the observed upregulation of the epidermis from the dermis in human explants SIVA1 in ES-treated wounds on day 7 might have augmented The normal skin tissues obtained from abdominoplasties were the increased interaction between phospho-p53, HDM2, processed to design wound explants. Detailed methodology is and SIVA. However, here, phospho-p53 expression was provided in Supplementary Materials and Methods online. not decreased. Albeit there was an increase in SIVA1 and HDM2 post ES, the decreased interaction of phospho-p53– In vitro ES apparatus and application SIVA1-HDM2 on day 7 compared with day 3 opens the A simple and precise apparatus for ES was designed for monolayer of discussion on whether the primary role of ES is to sustain the cells and tissues (Sebastian et al., 2011a). Detailed methodology is association between phospho-p53–SIVA1-HDM2 molecules provided in Supplementary Materials and Methods online. (ternary complex) rather than increasing the number of Tissue preparation for wax embedding, sectioning, and ternary complexes. In one of the studies, downstream targets hematoxylin and eosin staining of p53 were significantly impaired by an overexpression of Explants for wax embedding were fixed in 10% neutral-buffered HDM2 without affecting p53 protein levels (Ohkubo et al., formalin (F5304; Sigma-Aldrich, Dorset, UK) at 41Cfor3–4daysand 2006). Here, a promoter-selective effect on p53 activity at processed as described previously (Sebastian et al., 2011b). Further the transcriptional level was exhibited by HDM2. From our explanation of the methodology is provided in Supplementary in vivo coexpression studies, it is highly possible that Materials and Methods online. involvement of SIVA1 in the ternary complex enhances this transcriptional repression mediated by HDM2, leading RNA isolation, cDNA synthesis, and quantitative real-time to an increase in proliferation. This also is the closest reverse-transcriptase–PCR possibility to demonstrate physiologically enhanced Keratinocytes were processed for RNA isolation, cDNA synthesis, and proliferation with simultaneously higher HDM2 and SIVA1 quantitative real-time reverse-transcriptase–PCR, as described pre- levels (Figure 5d). viously (Sebastian et al., 2011a). Further explanation of the Exogenous application of ES could be an efficient strategy to methodology is provided in Supplementary Materials and Methods maintain the physiological properties in an in vivo wound online. The primers used are detailed in Supplementary Table S2 environment or to deliver additional current to the wound to online. enhance the healing process, especially when endogenous wound current is disturbed owing to injury or concurrent Immunohistochemistry, immunocytochemistry, and image pathology. By comparing ES strategies (direct current and analysis DW), we suggest that electrical waveforms may be able to Immunohistochemical and immunocytochemical staining were per- drive supra-physiological responses to accelerate wound formed as detailed in our previous reports (Sebastian et al., 2011a; healing. In summary, the results presented here signify the Sebastian et al., 2011b). Further explanation of the methodology is presence of wound electric fields in maintaining the integrity detailed in Supplementary Materials and Methods online. Antibodies of the epidermis and the critical role of SIVA1 in regulating and incubation parameters are detailed in Supplementary Table S3 p53 activities in human epithelial repair. online. The stained slides were scanned and images were analyzed on the Definiens Immunohistochemistry software (Definiens tissue MATERIALS AND METHODS studio version 3.51, Definiens AK, Munchen, Germany) to quantify Patient selection and recruitment the results. This study was conducted in accordance with the ethical principles of Good Clinical Practice and the Declaration of Helsinki Principles. For Coimmunoprecipitation, western blotting, and band analysis both human skin explant (ex vivo; n ¼ 10; Supplementary Table S1a Coimmunoprecipitation of phospho-p53 and SIVA1 and online) and human ES studies (in vivo; n ¼ 20; Supplementary Table protein analysis of whole-cell and nuclear lysates of primary S1b online), we received ethical approval from the local research keratinocytes and human explant epidermis by western blotting are committee (Manchester, UK), and all subjects gave full written, detailed in Supplementary Materials and Methods online. Antibodies informed consent. and incubation parameters are detailed in Supplementary Table S4 online. ES device used on human volunteers TheESdeviceusedforin vivo ES-assisted wound healing studies was siRNA gene knockdown the Fenzian system (Fenzian, Hungerford, UK), which is registered by Ambion siRNA target online tool (Silencer Pre-Designed and Vali- the Food and Drug Administration 510(k) and approved by Con- dated siRNAs) was used to select siRNA sequences that target human formite Europenne. ES methodology adopted on human volunteers is SIVA1 siRNA (s20746; Ambion, Austin, TX). Further methodology is detailed in Supplementary Materials and Methods online. provided in Supplementary Materials and Methods online.

www.jidonline.org 1173 ASebastianet al. SIVA1 Modulates p53 in Wound Healing

Flow cytometry Fukuda R, Suico MA, Koyama K et al. (2013) Mild electrical stimulation Flow cytometry analysis of primary keratinocytes post ES is provided at 0.1-ms pulse width induces p53 protein phosphorylation and in Supplementary Materials and Methods online. G2 arrest in human epithelial cells. JBiolChem288:16117–26 Jennings JA, Chen D, Feldman DS (2010) Upregulation of chemokine (C-C Lactate dehydrogenase cytotoxicity assay motif) ligand 20 in adult epidermal keratinocytes in direct current electric fields. Arch Dermatol Res 302:211–20 Lactate dehydrogenase assay was performed according to the man- Lamb R, Ambler CA (2013) Keratinocytes propagated in serum-free, feeder-free ufacturer’s instructions (11644793001; Cytotoxicity Detection Kit, culture conditions fail to form stratified epidermis in a reconstituted skin Roche Mannheim, Sandhofer Strae, Germany). The end point model. PloS One 8:e52494 absorbance of the samples was measured at 490 and 600 nm using Liang SH, Clarke MF (1999) A bipartite nuclear localization signal is required a microplate reader. for p53 nuclear import regulated by a carboxyl-terminal domain. JBiol Chem 274:32699–703 Statistical analysis Mei Y, Wu M (2012) Multifaceted functions of Siva-1: more than an Indian Data are presented as mean± SD from three independent experi- God of destruction. Protein Cell 3:117–22 ments performed in triplicate (n ¼ 3). Statistical analysis was calcu- Michael D, Oren M (2003) The p53-Mdm2 module and the ubiquitin system. lated using one-way analysis of variance for comparison between Semin Cancer Biol 13:49–58 three groups with Tukey’s post hoc test and Student’s t test for Muthuswamy S, Li D, Lelievre S, Bissell M et al. (2001) ErbB2, but not ErbB1, reinitiates proliferation and induces luminal repopulation in epithelial comparison between two groups. Confidence intervals of 95% with a acini. Nat Cell Biol 3:785–92 corresponding P-value of 0.05 were chosen throughout the analysis. Nishimura KY, Isseroff RR, Nuccitelli R (1996) Human keratinocytes migrate to the negative pole in direct current electric fields compar- CONFLICT OF INTEREST able to those measured in mammalian wounds. J Cell Sci 109: JC is an employee of Fenzian Ltd. The remaining authors state no conflict of 199–207 interest. Nuccitelli R (2003) A role for endogenous electric fields in wound healing. Curr Top Dev Biol 58:1–26 ACKNOWLEDGMENTS Nuccitelli R, Nuccitelli P, Li C et al. (2011) The electric field near human skin We thank Jenifer Mendoza for assisting with tissue processing, Sara Ud-din for wounds declines with age and provides a noninvasive indicator of wound punch biopsy collection and transfer, and Gary Sidgwick for Human Tissue healing. Wound Repair Regen 19:645–55 Act, 2004, documentation. Nuccitelli R, Nuccitelli P, Ramlatchan S et al. (2008) Imaging the electric field associated with mouse and human skin wounds. Wound Repair Regen SUPPLEMENTARY MATERIAL 16:432–41 Supplementary material is linked to the online version of the paper at http:// O’Brate A, Giannakakou P (2003) The importance of p53 location: nuclear or www.nature.com/jid cytoplasmic zip code? Drug Resist Update 6:313–22 Ohkubo S, Tanaka T, Taya Y et al. (2006) Excess HDM2 impacts cell cycle and apoptosis and has a selective effect on p53-dependent transcription. JBiol REFERENCES Chem 281:16943–50 Amar MB, Wu M (2014) Re-epithelialization: advancing epithelium frontier Ojingwa JC, Isseroff RR (2003) Electrical stimulation of wound healing. J Invest during wound healing. J R Soc Interface 11:20131038 Dermatol 121:1–12 Antoniades HN, Galanopoulos T, Neville-Golden J et al. (1994) p53 expression Patel GK, Wilson CH, Harding KG et al. (2006) Numerous keratinocyte during normal tissue regeneration in response to acute cutaneous injury in subtypes involved in wound re-epithelialization. J Invest Dermatol swine. JClinInvest93:2206–14 126:497–502 Beckerman R, Prives C (2010) Transcriptional regulation by p53. Cold Spring Resch U, Schichl YM, Winsauer G et al. (2009) Siva1 is a XIAP-interacting Harb Perspect Biol 2:a000935 protein that balances NFkappaB and JNK signalling to promote apoptosis. Chen C-Y, Oliner JD, Zhan Q et al. (1994) Interactions between p53 and JCellSci122:2651–61 MDM2 in a mammalian cell cycle checkpoint pathway. Proc Nadl Acad Sebastian A, Syed F, McGrouther DA et al. (2011a) A novel in vitro assay for Sci 91:2684–8 electrophysiological research on human skin fibroblasts: degenerate Chu F, Borthakur A, Sun X et al. (2004) The Siva-1 putative amphipathic helical electrical waves downregulate collagen I expression in keloid fibroblasts. region (SAH) is sufficient to bind to BCL-XL and sensitize cells to UV Exp Dermatol 20:64–8 radiation induced apoptosis. Apoptosis 9:83–95 Sebastian A, Syed F, Perry D et al. (2011b) Acceleration of cutaneous healing Dazard JE, Piette J, Basset-Seguin N et al. (2000) Switch from p53 to MDM2 as by electrical stimulation: degenerate electrical waveform down-regulates differentiating human keratinocytes lose their proliferative potential and inflammation, up-regulates angiogenesis and advances remodeling in increase in cellular size. Oncogene 19:3693–705 temporal punch biopsies in a human volunteer study. Wound Repair Du W, Jiang P, Li N et al. (2010) Suppression of p53 activity by Siva1. Cell Regen 19:693–708 Death Differ 16:1493–504 Vogelstein B, Lane D, Levine AJ (2000) Surfing the p53 network. Nature Flatt PM, Price JO, Shaw A et al. (1998) Differential cell cycle checkpoint 408:307–10 response in normal human keratinocytes and fibroblasts. Cell Growth Vousden KH, Lane DP (2007) p53 in health and disease. Nat Rev Mol Cell Biol Differ 9:535–43 8:275–83

1174 Journal of Investigative Dermatology (2015), Volume 135