Skip to main content

Advertisement

ADVERTISEMENT

Original Research

Evaluating the Effect of 3% Papain Gel Application in Cutaneous Wound Healing in Mice

April 2017
1044-7946
Wounds 2017;29(4):96–101. Epub 2017 January 23

Abstract

While the US Food and Drug Administration has not approved the use of 3% papain gel in the United States, the authors feel this study adds to the literature regarding its use. Introduction. The aim of this study was to evaluate the effect of 3% papain gel on wounds in mice.Materials and Methods. Thirty healthy C57BL mice (25–30 g) aged 10 weeks were randomly divided into 2 groups: mice treated with 3% papain gel and mice treated with placebo gel. Skin incisions were performed with a 6-mm metallic punch with a cutting blade edge. On days 3 and 7 after creating the lesion, the mice were euthanized and lesion samples were collected. The lesion samples were processed and sectioned into 3 fragments of skin to be stained with 3 types of dye: hematoxylin and eosin, Picrosirius red, and Weigert. In addition, immunohistochemical analysis (α-SM actin and Ki67) followed by real-time polymerase chain reaction (PCR) protocol was performed on the samples. Results. On gross examination, the 3% papain-treated group took less time to heal the wounds compared with the control. On day 7, microscopic examination showed the 3% papain-treated group had lower numbers of inflammatory cells, increased neovascularization, and improved organization of collagen and elastic fibers. Using PCR analysis, the 3% papain-treated group showed a significant increase in transforming growth factor beta and interleukin-6 expression compared with the control (P < .05). Conclusion. Due to a reduced local inflammatory response, increased angiogenesis, and improved organization of collagen deposition, these findings demonstrate 3% papain gel can improve cutaneous wound healing in mice.

Introduction

Papain is a complex mixture of proteolytic enzymes and peroxidases present in Carica papaya L. (Caricaceae). Papain has been largely used in the treatment of injuries to accelerate the healing process and can be used in various stages of tissue repair and at different concentrations (2%–10%).1,2 

Papain reduces pH in the wound bed, which stimulates the production of cytokines that promote cellular reproduction and leaves the area unfavorable for the emergence of pathogenic microorganisms.3 In addition to its debriding action and accelerating the tissue repair process, papain has anti-inflammatory,4 bacteriostatic, and bactericide actions.5

Due to its dynamic nature, wound care is dependent on the healing stage. There are several types of coverage available in Brazil for the treatment of injuries. There are several coverages (eg, alginate, hydrogel, growth factors), but the choice depends on the intrinsic and extrinsic factors related to the injury. The high cost of some healing modalities limits access for most of the population. In this context, papain-based dressings have been used in Brazil since 1983 because of its low cost and accessibility.6

Within the past 5 years, a systematic review7 was performed to examine evidence on the use of papain in the wound healing process. The review revealed low-quality studies on prevalence in accordance with the international scales assessment, indicating the need for further research with greater methodological rigor to produce a more accurate assessment of the effectiveness of papain in the process of tissue repair.7 However, the aim of this study was to evaluate the effect of 3% papain gel in cutaneous wound healing.

Materials and Methods

Animals. Thirty C57BL male mice (25–30 g) were used in this study and were obtained from the University
of Campinas Central Breeding Centre in Campinas, São Paulo, Brazil. Animals were individually housed in cages with
water and food ad libitum under standard experimental conditions of relative humidity and temperature (23–25°C), with a 12-hour light/dark cycle. All experimental animal work was carried out in accordance with the Brazilian Legislation (no. 11.794, from October 8, 2008) and approved by the Ethical Committee for Animal Use at the University of Campinas (ID protocol: 3287-1). 

Experimental protocol. The surgical procedure was performed on the animals at 10 weeks of age. They were anesthetized using intraperitoneal injection of a combination of 80 mg/kg of ketamine and 8 mg/kg of xylazine.8 

Before the wound was made, trichotomy of the dorsal region was performed with a razor blade and antiseptic treatment (70% alcohol). Skin incisions were performed with a 6-mm metallic punch with a cutting blade edge.
A 6-mm wound was made in the dorsal region of the mice; the skin and subcutaneous tissue were removed to expose muscle fascia. Immediately after the wound creation, the treatments were applied to the respective animals in the control group (group C) and experimental group (group P) group. After surgery, the animals received an analgesic of oral sodium dipyrone (6.2 mg/kg), which was added to their drinking water for 5 days; the mice were placed in individual cages and maintained in a room with a 12-hour day/night cycle with a controlled temperature of 23–25°C (under light with eye protection) to prevent traumatic damage to wounds by other mice and hypothermia. The treatments were applied once daily during the entire treatment period. In all groups, the wound beds locally were treated topically with ~0.15 g of the control gel (ie, without papain) or papain gel to cover the injury immediately after wounding (Day 0). The gel was reapplied daily for the duration of treatment. The wounds were not sutured or covered, but healed by secondary intention with no signs of infection. 

Treatments preparation. Powdered plant material (Lote: B-15/SEPT/2012#8) was supplied by Prozyn Biosolutions Corporation (São Paulo, Brazil). The 3% papain gel was prepared in the University Hospital Pharmacy at the University of Campinas in São Paulo, Brazil, using 3% papain (wt/vol) powder solubilized in hydroxypropyl methylcellulose (HPMC), a cellulose-derived polymer with high moisturizing power and a hydrostatic aspect with adhesive mucus characteristics. The placebo gel was composed of only a HPMC base. The gels (3% papain and placebo) were refrigerated at 4°C during treatment to maintain stability. 

Wound measurement. Digital images and wound area measurements were recorded to visualize and evaluate the wound healing process, which was estimated on the remaining wound area. The authors photographed the wounds on days 0, 2, 4, 8, 10, and 13 post wounding to measure the wound area using digital images captured with Sony Cyber Shot1 (model W510 12.1 Megapixel; Tokyo, Japan) and measured with Adobe Photoshop CS5 Extended (Adobe Systems Software Ireland Ltd, Dublin, Ireland). The same author captured all images, and the photos were taken from the same distance and aperture settings and analyzed for wound surface area (cm2). The authors estimated the percentage of wound area as the difference between the total lesion area immediately after wounding (d0) and the wound area still uncovered with epidermis/ lesion area (d0). 

Sample collection and tissue processing and staining. To obtain samples for morphological analysis, the animals were euthanized with an excessive dose of anesthesia. The tissue samples were carefully excised using scissors to remove a sufficient and constant amount of the surrounding wound margin from the healing skin of mice from all groups on days 3 and 7 post wounding. Fragments of lesions and adjacent normal skin were fixed in 4% paraformaldehyde (Pharma Synth Formulations, Diadema, São Paulo, Brazil) and then embedded in paraffin. 

The 5-µm sectioned fragments of skin from animals of all groups were incubated with the primary antibodies of anti-alpha smooth muscle actin (α-SMA; 1:100; Abcam, Cambridge, MA) and cell proliferation marker anti-Ki67 (1:100; Spring BioScience, Abingdon, UK). The samples were kept in an incubator at 37°C overnight, according to protocol, and were then deparaffinized. Subsequently, the slides were hydrated with 0.05 M of phosphate buffered saline (PBS) for 5 minutes at room temperature. After this phase, sections were incubated under light for 15 minutes with 50 mL of peroxidase. The samples passed the antigen retrieval procedure and were placed in 10 mM citrate buffer (pH 6.0) at 95°C for 40 minutes. After cooling the slides, blocking was performed with cuts of 50 mL per cut of the Invitrogen reagent kit (Thermo Fisher Scientific Inc, Carlsbad, CA) for 10 minutes. With the exception of the blades used as a negative control reaction, each primary cutting antibody of interest was diluted in 0.05 M of PBS and added overnight at 4°C. The sections were stained with Mayer’s hematoxylin for 5 minutes and passed through the assembly step of the blades for analysis under a Nikon Eclipse Ti-U microscope with a 20x10 objective; images were taken with a Nikon DS-Ri1 camera and analyzed using the NIS - Elements BR 3.1 software (Nikon Corporation, Tokyo, Japan). The subsequent images evaluated for general aspects of the wound and a qualitative analysis of the inflammatory infiltrate and blood vessels. Ten randomly selected fields were used per animal in each analysis. 

Vessels were identified by the presence of blood cells in the lumens or positive staining for α-SMA in the wall; cell proliferation for Ki67 was analyzed qualitatively.

Analysis of collagen and elastic fibers. Using a binocular microscope (Zeiss, Oberkochen, Germany) under polarized light, type I (red, orange, yellow) and type III (green) collagen fibers were analyzed in histological sections stained with picrosirius. In some nonstained histological sections and with a filter of quartz (collagen fibers were not quantitatively evaluated), the organization of collagen fibers was also evaluated. The elastic fibers were analyzed in histological sections stained with Weigert’s resorcin-fuchsin.

Real-time polymerase chain reaction (PCR) analysis. The sectioned fragments of skin from animals of all groups (10 mice, 5/group) were used for messenger ribonucleic acid (mRNA) expression studies. Total RNA was isolated from granulation and/or healing tissue for complementary deoxyribonucleic acid (DNA) synthesis. The healing wounds of the sectioned fragments of skin from animals of all groups were collected on days 3 and 7 post wounding.

Expression of the mRNA of the following cytokines was investigated: tumor necrosis factor α (TNFα; Mm00443258_m1), interleukin 6 (IL-6; Mm00446190_m1), IL-1β (Mm00434228_m1), vascular endothelial growth factor (VEGF; Mm01281449_m1), and transforming growth factor β (TGFβ; Mm01227699_m1).  All Taq
Man assays for gene expression of the genes corresponding to specific cytokines studied, as well as the endogenous control, were purchased from Applied Biosystems (Foster City, CA). For endogenous controls, beta-actin and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) were used.

The reactions were performed in duplicate using 96-well plates with 1 µL of the product in each assay reaction; there was a specific assay for each gene and the endogenous control, which used 10 µL of Master Mix (Thermo Fisher Scientific Inc, Carlsbad, CA), 100 µg of complementary DNA, and 5 µL of distilled water. Fluorescence detection was performed during each cycle, and a fluorescence variation graph was obtained at the end of the run, according to the number of cycles (Cycle Number).

Statistical analysis.  To calculate the size of the wounds and their evolution during the assay period, GraphPad Prism, version 6.0 (GraphPad Software, San Diego, CA), was used; the groups were compared using two-way analysis of variance (ANOVA) followed by Bonferroni post-hoc test. For analysis of gene expression of the mRNA of interest as investigated by PCR, the results were evaluated with paired one-way ANOVA with multiple comparisons and the Tukey post-hoc test. The significance level was established as P < .05.

Results

Morphological evaluation. The animals were randomly divided before surgery into group C and group P, and the effect of 3% papain gel treatment on the healing rate of the cutaneous wound in the mice was evaluated. Topical application of 3% papain gel improved wound closure when compared to the C group throughout the 13-day observation period (eFigure 1A). In the daily monitoring of the wounds of the P group, a perilesional area was observed with the presence of regular edges compared with group C. Macroscopic analysis showed the persistence of dry scabs in group C until day 10 of treatment. 

The morphometric analysis (eFigure 1B) using the flatness of the images of days 0, 2, 4, 8, 10, and 13 postinjury showed no statistical difference in wound resolution time between the 2 groups. 

Histological analysis. The representative images of hematoxylin and eosin (H&E)-stained wound sections showed a more rapid reinstatement of normal tissue structure following the application of 3% papain gel, based on comparison of group C (eFigure 2). On day 3 postinjury, the authors observed a sharp drop in the number of inflammatory cells in group P compared to group C, which experienced a sharp rise in the number of inflammatory cells. Otherwise, group P was marked by a higher number of newly formed blood vessels compared with group C (eFigure 2).

On day 7 postinjury, group P showed a reduction of the inflammatory infiltrate and thick regenerated epithelial layer covering the healed epithelium compared with group C. The wound area of group C still showed high inflammation, because the wounds were not completely closed and extensive scabs and a disorganized epithelium were visible (eFigure 2).

Immunohistochemistry. On day 3 postinjury, the newly formed blood vessels were uniformly distributed within the granulation tissue, with clearly marked lumen of the vessels and a large perimeter in group P when compared with group C. In contrast, on day 7 postinjury, there was no difference between the 2 groups (eFigure 3). 

The representative images of Ki67-positive antibody are also presented (eFigure 4). On day 3 postinjury, the appearance of the proliferative cells in group P had accelerated more and were more organized than the appearance of the proliferative cells of the animal that did not receive the treatment with papain. However, on day 7 postinjury, cell proliferation was seen in group C. 

Once the mice were divided into the 2 groups — control and experimental — the authors then assessed the wounds at 2 time points: days 3 (C3 and P3) and 7 (C7 and P7) post wounding.

Collagen and elastic fibers: qualitative aspects. In specimens of all groups (C3, C7, P3, and P7), the dense type I collagen fibers predominate (eFigure 5A). Bundles of horizontal and vertical collagen fibers determined a mesh with a pantographic aspect in C3, C7, and P7. This characteristic was not observed in the P3 group, where the fibers were for the most part horizontally arranged (eFigure 5B).

The elastic fibers were detected in all groups (C3, C7, P3, and P7). However, the appearance of elastic fibers was thicker in P3 and C7 (particularly, when the groups were compared at 7 days postinjury [C7 and P7]). Although the elastic fibers were thinner in P7, their appearance was similar to the intact skin (eFigure 6). 

Gene expression analysis (RT-PCR). The analysis of gene expression of VEGF, TGFβ, TNFα, IL-6, and IL-1β
is shown in eFigure 7. On the third day, there was a slight increase in IL-1β expression in group P compared with group C. On the other hand, on day 7 showed a significant decrease in IL-1β in group P when compared with other groups and times (eFigure 7A). 

In relation to TNFα, lower gene expression was observed in animals treated with papain on day 7 post wounding when compared with the other groups on day 3 (P3 vs P7; P < .05; eFigure 7B). There was also a decrease in group C (C3 vs C7; P < .05). There is no statistical difference between groups C3 and P3, or C7 and P7. The IL-6 level was increased in group P on day 3 postinjury, compared with other groups (P < .05, eFigure 7C). Expression of VEGF mRNA was slightly increased in group P when compared with other groups and times on day 7 (eFigure 7D). However, mRNA expression of TGFβ significantly increased in group P on day 7, compared with other groups and times (P < .05; eFigure 7E).

Discussion

The use of papain in the treatment of wounds has been known worldwide since the 1950s7 due to its anti-inflammatory, antibacterial, antioxidant, and immunomodulatory properties.3,9,10 However, it is not approved for use by the US Food and Drug Administration. 

The results of this study demonstrate treatment with 3% papain gel decreases the inflammatory infiltrate on day 3 after injury and reepithelialization of the area occurs around day 7 after injury, when compared to the control by qualitative analysis. On the other hand, expression of the genes for pro-inflammatory cytokines IL-1β and TNFα showed a slight increase on day 3 after injury; for IL-6, there was a significant increase. The increase of IL-6 in preference of other pro-inflammatory cytokines may indicate a specific role for 3% papain gel. The real-time PCR also showed on day 7 after injury, there was a reduction of cytokines of these genes, which revealed that the inflammatory phase has been resolved. Previous studies showed IL-1β, IL-6, and TNFα play important roles in the healing process11 and also in the induction and maintenance of inflammation.12

Studies have reinforced the beneficial action of papain in different formulations to treat injuries. In 2 studies13,14 investigating lesions in diabetic animals treated with papain powder, morphological evaluation showed contraction of the lesion area, granulation tissue, and hydroxyproline levels were significantly improved.

Collard and Roy15 investigated tissue repair in diabetic animals undergoing oral supplementation with papain and observed an increased expression of CD68 and CD31 in animals with injuries that were treated with papain, which suggests an effective recruitment of macrophages and the subsequent improvement in angiogenic response. 

During tissue repair, angiogenesis is responsible for supplying the nutrients and oxygen necessary for the wound and accelerating the formation of granulation tissue.16 

In the present study, the authors observed the number of new capillaries, assessed by histology, were most intense in those animals that received topical treatment with 3% papain on day 3 postinjury. Furthermore, 3% papain significantly modulated the immunostaining of α-SMA, evaluated by immunohistochemistry, and upregulation of VEGF by mRNA. Inhibition of angiogenesis impairs wound healing, and VEGF and TGF-β1 are involved in the stimulation, promotion, and stabilization of new blood vessels.17 Angiogenesis is induced by VEGF endothelial cell proliferation and the prevention of apoptosis.18 Confirmed by the immunostaining of α-SMA, VEGF mRNA levels support an appropriate response at the site of injury in group P.

Not only does VEGF promote angiogenesis, but it can also increase collagen fibers and formation of granulation tissue and promote reepithelization.18 In this study, there was a significant increase in the immunostaining of Ki67, a cell proliferation marker, on day 3 after injury in animals treated with papain. In addition, the prevalence of collagen type I fibers with transverse and longitudinal characteristics composing the aspect mesh pantograph was increased.

These data suggest 3% papain shortens or reduces the phase of inflammatory-treated animals, resulting in a better quality of granulation tissue by improving cell proliferation and the deposition of collagen. The proliferation and migration of endothelial cells is essential for the formation of granulation tissue, which is vital to the success of healing.19

There also was an increase in TGFβ mRNA expression. This plays an important role in the proliferative phase by stimulating the proliferation and differentiation of fibroblasts, collagen production, and wound contraction.20,21 Collagen is a component of the dermis, which functions to resist mechanical abrasions and maintains the structural integrity of the fabric22; its orientation is important in developing tissues.23

The elastic fibers correspond with the stages of normal elastogenesis, and these elastic fibers are necessary to provide strength and elasticity to connective tissues of the skin, similar to that in the lungs and blood vessels.24 The organization of elastin and laminae strongly provides the resilience of tissues for their mechanical properties, but elastic fibers are rarely considered in the wound healing process.25 In the present study, the papain treatment improved the early emergence of elastogenesis. 

By means of 3 different analyses (classification, organization, and direction), this study showed papain organized collagen fibers in a complete manner (mesh pantograph), because they resembled collagen organization in intact skin. It is predominantly the presence of collagen type I that indicates the maturity of collagen fibers. The absence of type III collagen can be justified since the deficiency in collagen type III in wounds may indicate increased wound contraction, which results in a shorter wound closure time.26 These findings are in agreement with the study13 using diabetic animals subjected to treatment with papain extract; therefore, increasing hydroxyproline levels were observed in the lesion area, which is an indicator of collagen production.

These data suggest 3% papain shortens or reduces inflammation in treated animals and improves cell proliferation and the deposition of collagen, thus resulting in better quality granulation tissue. The proliferation and migration of endothelial cells is essential for the formation of granulation tissue, which is vital to the success of wound healing.27,28

Conclusions

Topical 3% papain gel treatment showed more organized healing when compared to group C. This can be attributed to a reduction of the local inflammatory response, up-regulation of angiogenic genes such as VEGF and TGFβ, stimulation of the proliferative phase, increased collagen deposition, and improved organization of the elastic fibers. Moreover, the treatment with 3% papain gel may have beneficial effects on wound healing. 

References

1. Ramundo J, Gray M. Enzymatic wound debridement. J Wound Ostomy Continence Nurs. 2008;35(3):273–280. 2. Guzman AV, Stein de Guzman MG. The enzymatic debridement of suppurations, necrotic lesions and burns with papain. J Int Coll Surg. 1953;20(6):695–702. 3. Falanga V.  Wound bed preparation and the role of enzymes: a case for multiple actions of therapeutic agents. Wounds. 2002;14(2):47–57. 4. Owoyele BV, Adebukola OM, Funmilayo AA, Soladoye AO. Anti-inflammatory activities of ethanolic extract of Carica papaya leaves. Inflammopharmacology. 2008;16(4):168–173. 5. Osato JA, Santiago LA, Remo GM, Cuadra MS, Mori A. Anti-microbial and antioxidant activities of unripe papaya. Life Sci. 1993;53(17):1383–1389. 6. Monetta L. Uso da papaína nos curativos feitos pela enfermagem [in Portugese]. Rev Bras Enf Brasília. 1987;40(1):66–73. 7. Leite AP, de Oliveira BGRP, Soares MF, Barrocas DLR. Uso e efetividade da papaína no processo de cicatrização de feridas: uma revisão sistemática [in Portugese]. Rev Gaúcha Enferm. 2012;33(3):198–207. 8. Damy SB, Camargo RS, Chammas R, Figueiredo LF. Fundamental aspects on animal research as applied to experimental surgery [in Portuguese]. Rev Assoc Med Bras. 2010;56(1):103–111. 9. Ferreira AM, Watanabe E, Nascimento AP, de Andrade D, Ito IY. Atividade anti-bacteriana in vitro de géis com diferentes concentrações de papaína. Rev Eletr Enf. 2008; 10(4):1035–1040. 10. Nafiu AB, Rahman MT. Anti-inflammatory and antioxidant properties of unripe papaya extract in an excision wound model. Pharm Biol. 2015;53(5):662–671. 11. Chen L, Arbieva ZH, Guo S, Marucha PT, Mustoe TA, Di- Pietro LA. Positional differences in the wound transcriptome of skin and oral mucosa. BMC Genomics. 2010;11:471. doi: 10.1186/1471-2164-11-471. 12. Furuichi K, Wada T, Iwata Y, et al. Interleukin-1-dependent sequential chemokine expression and inflammatory cell infiltration in ischemia-reperfusion injury. Crit Care Med. 2006;34(9):2447–2455. 13. Nayak SB, Pinto Pereira L, Maharaj D. Wound healing act- ivity of Carica papaya L. in experimentally induced diabetic rats. Indian J Exp Biol. 2007;45(8):739–743. 14. Mikhal’chik EV, Ivanova AV, Anurov MV, et al. Wound-heal- ing effect of papaya-based preparation in experimental thermal trauma. Bull Exp Biol Med. 2004;137(6):560–562. 15. Collard E, Roy S. Improved function of diabetic wound-site macrophages and accelerated wound closure in response to oral supplementation of a fermented papaya preparation. Antioxid Redox Signal. 2010;13(5):599–606. 16. Brem H, Tomic-Canic M. Cellular and molecular basis of wound healing in diabetes. J Clin Invest. 2007;117(5):1219–1222. 17. Johnson KE, Wilgus TA. Multiple roles for VEGF in non-melanoma skin cancer: angiogenesis and beyond [published online ahead of print October 17, 2012]. J Skin Cancer. 2012;2012:483439. 18. Greaves NS, Ashcroft KJ, Baguneid M, Bayat A. Current understanding of molecular and cellular mechanisms in fibroplasia and angiogenesis during acute wound healing [published online ahead of print July 30, 2013]. J Dermatol Sci. 2013;72(3):206–217. 19. Kant V, Gopal A, Kumar D, et al. Curcumin-induced angiogenesis hastens wound healing in diabetic rats [published online ahead of print October 18, 2014]. J Surg Res. 2015;193(2):978–988. 20. Behm B, Babilas P, Landthaler M, Schreml S. Cytokines, chemokines and growth factors in wound healing. J Eur Acad Dermatol Venereol. 2012;26(7):812–820. 21. Hajiaghaalipour F, Kanthimathi MS, Abdulla MA, Sanusi J. The effect of Camellia sinensis on wound healing potential in an animal model. Evid Based Complement Alternat Med. 2013;(2013):386734. 22. Ajlia SA, Majid FA, Suvik A, Effendy MA, Nouri HS. Efficacy of papain-based wound cleanser in promoting wound regeneration. Pak J Biol Sci. 2010;13(12):596–603. 23. Valenti DMZ, Silva JA, Teodoro WR, Velosa PA, Mello SBV. Avaliação da histoarquitetura do colágeno no tecido cutâneo após a utilização tópica da argila em ratos [in Portugese]. Rev Bras de Ciênc Saúde. 2008;8(23):22–31. 24. Kielty CM, Sherratt MJ, Shuttleworth CA. Elastic fibres. J Cell Sci. 2002;115(Pt 14):2817–2828. 25. Amadeu TP, Braune AS, Porto LC, Desmoulière A, Costa AM. Fibrillin-1 and elastin are differentially expressed in hypertrophic scars and keloids. Wound Repair Regen. 2004;12(2):169–174. 26. Volk SW, Wang Y, Mauldin EA, Liechty KW, Adams SL. Diminished type III collagen promotes myofibroblast differentiation and increases scar deposition in cutaneous wound healing [published online ahead of print January 19, 2011]. Cells Tissues Organs. 2011;194(1):25–37. 27. Mandelbaum SH, Di Santis EP, Mandelbaum MHS. Cicatrização: Conceitos atuais e recursos auxiliares - Parte I e Parte II [in Portugese]. Anais Bras Derm. 2003;78(5):525–542. 28. Krishnaswamy KR, Korrapati PS. Role of dermatopontin in re-epithelialization: implications on keratinocyte migration and proliferation. Sci Rep. 2014;4:7385. doi: 10.1038/srep07385.

Advertisement

Advertisement

Advertisement