1

The Application of a Silk Fibroin Scaffold to Otology

By

Dr Brett Levin

B Med Sci (Macquarie University), MBBS (Hons) (University of Sydney)

This thesis is presented for the degree of Masters of Medical Science at the University of Western Australia

School of Surgery

2010

  2

CONTENTS

• Declaration

• Acknowledgments

• Publications Arising from Thesis

• Presentations Arising from Thesis

• Abstract

• Introduction

• CHAPTER 1 (LITERATURE REVIEW) - Utilising a Silk Fibroin Scaffold

as a Novel Device

• CHAPTER 2 - Phenotypic and Genotypic Profile of Human Tympanic

Membrane Derived Cultured Cells

• CHAPTER 3 - Structure and Properties of Biomedical Films Prepared from

Aqueous and Acidic Silk Fibroin Solutions

• CHAPTER 4 - Preliminary Results of the Application of a Silk Fibroin

Scaffold to Otology

• CHAPTER 5 - Utilising Silk Fibroin Membranes as Scaffolds for the Growth

of Tympanic Membrane Keratinocytes and their Application to

Myringoplasty Surgery

• Conclusion

  3

DECLARATION

I declare that the substance of this thesis has not been previously submitted for a

degree at this or any other University.

I declare that all assistance received during the preparation of this thesis has been

acknowledged. The chapters in this thesis that have been accepted for publication or

submitted for publication have the permission of all co-authors to be included in this

thesis.

------------------ ----------

Dr Brett Levin Date

  4

ACKNOWLEDGEMENTS

This thesis would not have been possible without the financial support of the Garnett

Passe and Rodney Williams Memorial Foundation during my second year, for which

I am most grateful. I thank the University of Western Australia for its financial

support in granting me a University Postgraduate Award Scholarship in my first year

and also for assistance with interstate relocation expenses and travelling to

conferences to present my research.

My supervisors, Dr Robert Marano, Adj Prof Robert Eikelboom and Winthrop Prof

Marcus Atlas have provided invaluable support and assistance with this thesis. I

thank them for their help with the challenges associated with scientific research and

for showing me that it is possible to work as a clinician whilst maintaining a strong

research focus.

I thank my research assistant, Sharon Redmond, for all her help with laboratory

experiments. Her work ethic and efforts were truly remarkable. Members of the

Department of Surgery at The University of Western Australia with whom I worked

were always willing to provide guidance and assistance whenever required. I

acknowledge the facilities, scientific and technical assistance of the Australian

Microscopy & Microanalysis Research Facility at the Centre for Microscopy,

Characterisation & Analysis at The University of Western Australia.

  5

The collaboration we formed with the Centre for Material and Fibre Innovation at

Deakin University, Victoria, has been excellent. In particular, I thank Mr Rangam

Rajkhowa for his assistance in manufacturing the silk fibroin membranes and for

always being willing to provide additional materials for experimentation as required.

The tremendous support from my wife, Cara, did not go unnoticed. The relocation

interstate and the journey from clinical medicine into the scientific world would not

have been as exciting without her by my side.

  6

PUBLICATIONS ARISING FROM THESIS

• Levin B, Redmond SL, Rajkhowa R, Eikelboom RH, Marano RJ, Atlas MD.

Preliminary Results of the Application of a Silk Fibroin Scaffold to Otology.

Otolaryngology - Head and Neck Surgery 2010; 142, Issue 3, S33-S35.

• Levin B, Rajkhowa R, Redmond SL, Atlas MD. Grafts in

Myringoplasty: Utilising a Silk Fibroin Scaffold as a Novel Device [Review

Article]. Expert Review of Medical Devices 2009; 6(6): 653-664.

• Redmond SL, Levin B, Heel KA, Atlas MD, Marano RJ. Phenotypic and

Genotypic Profile of Human Tympanic Membrane Derived Cultured Cells.

Accepted for publication in Journal of Molecular Histology on 12 October

2010.

• Rajkhowa R, Levin B, Redmond S, Li L, Wang L, Kanwar J, Atlas M, Wang

X. Structure and Properties of Biomedical Films Prepared from Aqueous and

Acidic Silk Fibroin Solutions. Accepted for publication in Journal of

Biomedical Materials Research: Part A on 3 November 2010.

• Levin B, Redmond SL, Rajkhowa R, Eikelboom RH, Marano RJ, Atlas MD.

Utilising Silk Fibroin Membranes as Scaffolds for the Growth of Tympanic

Membrane Keratinocytes and their Application to Myringoplasty Surgery.

Under review by Biomedical Materials.

  7

PRESENTATIONS ARISING FROM THESIS

• “Surgeon Scientist: Research and Clinical Experiences”. Oral presentation at

the Western Australian Society of Otolaryngology Head and Neck Surgery

Meeting, Bunker Bay, October 2009.

• “Tissue Engineering an Artificial Tympanic Membrane”. Oral presentation at

the Translational Medicine Symposium, Yallingup, July 2009.

• “The Use of a Silk Fibroin Scaffold to Support the Growth of Human

Tympanic Membrane Keratinocytes in vitro”. Oral presentation at the

Australian Society for Medical Research Symposium, Perth, June 2009.

• “The Application of a Silk Fibroin Scaffold to Otology”. Oral presentation at

the Australian Society of Otolaryngology Head and Neck Surgery Meeting,

Gold Coast, June 2009.

• “Animal and Human Trials of an Artificial Tympanic Membrane Scaffold”.

Poster presentation at Frontiers in Otorhinolaryngology, Noosa, July 2008.

  8

ABSTRACT

Chronic tympanic membrane perforations represent a significant source of

morbidity. Myringoplasty is the surgical procedure that repairs these perforations

using a graft material. Although a variety of materials have been successfully used

for this purpose, all have limitations. In recent years, silk fibroin has been studied

with biomedical applications in mind due to its biocompatibility, biodegradability

and diverse mechanical properties. This research project examines the phenotype and

genotype of cultured human tympanic membrane keratinocytes as well as the

structure and properties of silk fibroin films and factors affecting their strength and

degradation. Primarily, it examines the growth of human tympanic membrane

keratinocytes on silk fibroin scaffolds. Light microscopy, immunofluorescent

staining, confocal imaging and scanning electron microscopy all demonstrate the

proliferation and adhesion of keratinocytes on the fibroin scaffolds. The scaffolds

successfully supported the keratinocytes, which continued to maintain their cell

lineage. Integral cellular proteins that function in proliferation, differentiation and

adhesion were also maintained. The biocompatibility, biodegradability, transparency

and tensile strength of silk fibroin make it an attractive option to study as an

alternative graft in myringoplasty surgery.

  9

INTRODUCTION

This thesis examines the interaction between human tympanic membrane

keratinocytes and the protein, silk fibroin. It describes a project within a long-

term research programme to determine whether silk fibroin scaffolds can be

used as alternative graft materials to repair chronic tympanic membrane

perforations in myringoplasty surgery. In addition it examined separately the

phenotypic and genotypic profile of cultured tympanic membrane cells and

the structure and properties of silk fibroin films.

Chapter 1 initially reviews the current literature relating to myringoplasty.

The morbidity associated with tympanic membrane perforations is examined

and the evolution of myringoplasty is then discussed. The properties of an

ideal graft, the various graft materials currently used and their limitations are

described.

Chapter 1 then proceeds to review the current literature relating to silk

fibroin, bearing in mind its potential use in otology. The properties of silk

fibroin and its biocompatibility are discussed. Various in vitro and in vivo

studies relevant to this thesis are summarised. Studies looking at the growth

of keratinocytes (not from the tympanic membrane) are considered. The

wound healing effects of silk fibroin and its biodegradation (both particularly

relevant to myringoplasty) are reviewed.

  10

Chapter 2 examines the genotype, phenotype and basic cell morphology of

the cultured human tympanic membrane cells used in this research. These

cells are compared to epithelial (HaCaT cells) and mesenchymal (dermal

fibroblasts) reference cells. This is achieved using immunofluorescent

staining, flow cytometry and Polymerase Chain Reaction (PCR). The chapter

reiterates that keratinocytes in culture can undergo phenotypic transformation

and that the cell passage used in experimentation needs consideration. As

second author, I contributed to drafting the original article, designing the

experiments and collecting data as well as reviewing the final version prior to

submission.

Chapter 3 studies the structure and properties of the silk fibroin membranes

that were used as scaffolds for tympanic membrane keratinocyte growth. The

mechanical properties, factors affecting tensile strength as well as the control

of biodegradation of silk fibroin are examined. These characteristics are of

particular importance when considering graft materials to be used in

myringoplasty. As second author, I contributed to drafting the original article,

advising the first author on clinical application of the research, grammatical

assistance and reviewing the final version prior to submission.

Chapters 4 and 5 study the growth of human tympanic membrane

keratinocytes on the silk fibroin scaffolds. The need for a novel graft is

  11

reiterated and the biomedical applications of silk fibroin are examined. The

preparation of the fibroin membranes and the culture of human keratinocytes

are detailed. The seeding of keratinocytes onto the fibroin membranes and the

immunostaining for various protein markers is described. The experimental

results are discussed and the benefit of further research pursuing the use of

silk fibroin in myringoplasty is expressed.

This thesis is presented as a series of five scientific papers, two of which

have been published, two accepted for publication and one currently under

review. These chapters can be read as part of the whole thesis or as separate

entities. Each chapter contains an introduction, literature review, materials

and methods section, results, discussion and references. There is some

unavoidable overlap within these subsections since chapters are concerned

with similar or related research questions. A general concluding chapter

closes the thesis providing an overall discussion of the issues raised in the

thesis, concluding remarks and suggestions for future research as a result of

this thesis.

Chapter 1

Utilising a Silk Fibroin Scaffold as a Novel Device

Levin B, Rajkhowa R, Redmond SL, Atlas MD. Grafts in Myringoplasty: Utilising a Silk Fibroin Scaffold as a Novel Device [Review Article]. Expert Review of Medical Devices 2009; 6(6): 653-664.

653

Review

www.expert-reviews.com ISSN 1743-4440© 2009 Expert Reviews Ltd10.1586/ERD.09.47

Myringoplasty backgroundEpidemiology of tympanic membrane perforationsAcute perforations of the tympanic membrane (TM) usually heal without treatment, with up to 80% undergoing spontaneous closure [1]. Those that persist and become chronic TM perforations usually result from otitis media (OM) or traumatic injury [2]. OM is defined as infection of the middle ear cleft, and, apart from the common cold, is the most common disorder for which children and their families seek pediatric care [3]. It is the most frequent bacterial pediatric infection and the most com-mon indication for the prescription of antibi-otic therapy in children [4]. Acute OM usually resolves spontaneously or with medical treat-ment. If it results in a TM perforation, this will usually resolve spontaneously. However, recur-rent OM may lead to chronic TM perforations (FIGURE 1). Chronic suppurative OM (CSOM) describes a chronic infection of the middle ear cleft in which a nonintact TM (i.e., perforation)

and discharge (otorrhea) are present, whereas chronic OM with a perforation refers to a nondischarging perforation [201].

Regardless of etiology, chronic perforations and CSOM represent a significant cause of morbidity worldwide. They are particularly prevalent among Australian Aborigines, the Inuits of Alaska, Canada and Greenland, and certain Native Americans [5]. Risk factors for ear disease and TM perforations in these popu-lations include overcrowding, poor hygiene, poor nutrition, passive smoking, high rates of bacterial colonization of the nasopharynx and limited access to healthcare [5].

Morbidity of TM perforationsThe complications associated with chronic TM perforations are vast, with hearing loss listed as the most common sequela [6]. The WHO describes the global burden of hearing impairment as the most frequent sensory defi-cit in human populations, affecting more than 250 million people worldwide [202]. Importantly,

Brett Levin†, Rangam Rajkhowa, Sharon Leanne Redmond and Marcus David Atlas†Author for correspondenceEar Sciences Centre, School of Surgery, 2nd Floor M-Block, The University of Western Australia, Queen Elizabeth II Medical Centre, Nedlands, WA 6009, AustraliaTel.: +61 (08) 9346 4153Fax: +61 (02) 8088 3820blevin@med.usyd.edu.au

Chronic perforations of the eardrum or tympanic membrane represent a significant source of morbidity worldwide. Myringoplasty is the operative repair of a perforated tympanic membrane and is a procedure commonly performed by otolaryngologists. Its purpose is to close the tympanic membrane, improve hearing and limit patient susceptibility to middle ear infections. The success rates of the different surgical techniques used to perform a myringoplasty, and the optimal graft materials to achieve complete closure and restore hearing, vary significantly in the literature. A number of autologous tissues, homografts and synthetic materials are described as graft options. With the advent and development of tissue engineering in the last decade, a number of biomaterials have been studied and attempts have been made to mimic biological functions with these materials. Fibroin, a core structural protein in silk from silkworms, has been widely studied with biomedical applications in mind. Several cell types, including keratinocytes, have grown on silk biomaterials, and scaffolds manufactured from silk have successfully been used in wound healing and for tissue engineering purposes. This review focuses on the current available grafts for myringoplasty and their limitations, and examines the biomechanical properties of silk, assessing the potential benefits of a silk fibroin scaffold as a novel device for use as a graft in myringoplasty surgery.

KEYWORDS: biomaterial • fibroin • graft • myringoplasty • perforation • silk • tissue engineering • tympanic membrane

Grafts in myringoplasty: utilizing a silk fibroin scaffold as a novel deviceExpert Rev. Med. Devices 6(6), 653–664 (2009)

For reprint orders, please contact reprints@expert-reviews.com

Expert Rev. Med. Devices 6(6), (2009)654

Review

the WHO reports that CSOM in children is one of the most common causes of preventable hearing loss, particularly in the developing world, and accounts for up to 80% of the worldwide burden of hearing impairment [201]. If this hearing deficit occurs during a period of critical development, it can give rise to speech, language and developmental problems in children [7]. Other extracranial complications of TM perforations include acquired cholesteatoma, facial paralysis, subperiosteal abscess formation and labyrinthitis [8–11]. Significant intracranial complications that

can result from CSOM include meningitis, lateral sinus throm-bosis and cerebral and extradural abscess formation [6,201], all of which are potentially life-threatening if left untreated (FIGURE 2).

Historical perspectiveThe history of the management of TM perforations dates back to 1640, when Banzer attempted closure using a tube of elk’s claw covered with pig bladder. This was followed by a number of organic and nonorganic devices trialed by various inventors and authors over the next 200 years [12]. These included devices such as lead tubing, cotton wool, wood, paper, rubber, skin, powder, fish bones and silver foil. However, the first surgical closure of the TM was performed by Berthold in 1878 using a full-thickness skin graft [13].

Today, the treatment for chronic TM perforations is the sur-gical procedure myringoplasty (also known as type 1 tympano-plasty). With the advent of the operating microscope, advances in anesthesia and antibiotic therapy, as well as the availability of a variety of graft materials, myringoplasty is one of the most commonly performed otological procedures in both adults and children. Despite this, there is marked diversity in the literature of reported success rates, and factors that are thought to influ-ence this remain controversial [A!"#$ MD, A%#%&#'(()#*#$+#), KS,

W#!!$ GD, U%-./"0$12& D#!#] [14].

Incisions, operative techniques & approachesThere are three main incisions that can be used to access the TM to attempt closure – endaural (incision through the ear canal), postauricular (incision behind the ear), and transcanal or transmeatal (across the ear canal). The choice of incision is often dependent on the location of the perforation and on the surgeon’s preference [15], and no single incision is optimal for all types of perforation [1]. Endaural incisions are often used for posterior perforations, while the postauricular approach gives access to the anterior part of the TM with preservation of the anterior canal wall skin. The transcanal approach is used frequently for small, central perforations, in which the canal is of adequate width [16].

Once the incision has been selected, the surgeon can either place the graft material lateral to the perforation (overlay or onlay) or medial to it (underlay). Multiple studies, includ-ing a randomized prospective study [17], have shown the two techniques to be comparable [17–19]. The over–under technique combines the two techniques but requires the handle of the malleus to be present. It is reportedly useful for perforations in any of the four TM quadrants, allowing precise graft placement unobscured by the malleus, and is advantageous in cases where ossicular reconstruction is a possibility [20]. In recent years, an additional technique known as inlay myringoplasty has been described and uses specifically designed cartilage inserted via a transcanal butterfly technique [21,22]. This has been shown to be useful for the repair of small, moderate and large perfora-tions, although many of these patients underwent concurrent mastoidectomy [22]. Other studies have shown inconsistent long-term results with this technique when compared with the underlay technique [23].

Figure 1. Normal and perforated tympanic membrane. (A) Normal right tympanic membrane (dotted line). The remainder of the image displays part of the external auditory canal. The diameter of the tympanic membrane is 1 cm. (B) Perforated left tympanic membrane (dotted line) in the same patient.

A

B

Levin, Rajkhowa, Redmond & Atlas

www.expert-reviews.com 655

Review

Myringoplasty graft materials & devicesProperties of an ideal graftThe ideal graft material to reconstruct the TM is yet to be identified but, according to Schulte et al., would satisfy the following criteria [24]:• Easy to obtain and fashion

• Available in unlimited quantities

• Cause no cosmetic changes

• Be inexpensive

• Safe

• Cause no interference in postoperative follow-up

• Resist negative ear pressures

• Cause no interference with the ear conduction component of hearing

In addition, evidence indicates that the biomechanical and material properties of grafts are especially important for hearing outcomes [25]. Ideally, a graft should also be transparent because lifelong examination of the middle ear is vital in patients with a history of recurrent OM.

Tympanic membrane closure & hearing outcomesThe biomechanical properties of graft materials are of crucial importance, as they are not only required to close perforations but also restore hearing. The TM is only 30–90 µm thick and the macro- and micro-anatomical structure is largely responsible for determining its function [26]. This implies that the more closely a graft replicates the native TM structurally, the more likely it is to do so functionally. Although clinical reviews report high closure rates [14], other reviews suggest that actual success rates may be lower than reported, as only a minority of papers record hearing outcomes [A!"#$ MD, A%#%&#'(()#*#$+#), KS, W#!!$ GD, U%-./"0$12&

D#!#] [27]. Studies that have considered hearing outcomes report that normal hearing is only achieved in 43–80% of cases [28–30]. Moreover, it has been reported that as patients are followed-up for greater periods of time, the reported success rate of myringo-plasty decreases, suggesting the importance of long-term follow-up [A!"#$3MD, A%#%&#'(()#*#$+#), KS, W#!!$ GD, U%-./"0$12& D#!#].

The device or graft used to repair TM perforations has under-gone considerable evolution since Berthold’s full-thickness skin graft in 1878 [13]. Today, multiple materials are part of the oto-laryngologist’s armamentarium when considering a myringo-plasty operation. The reported success rates vary significantly in the literature, with closure rates between 60 and 99% in adults and 35 and 94% in children [14]. Apart from TM closure, improvements in hearing also vary with graft materials and have been reported as an independent indication for myringoplasty, regardless of the site and size of the perforation, or gender and age of the patient [31].

Common autologous grafts Although autologous temporalis fascia is cited as the most commonly used graft material for all types of perforation [14], other commonly harvested materials used successfully today

include cartilage [32–34], perichondrium [35–37] and fat [35,38–40], while common synthetic graft materials include paper [35,41] and alloderm [42,43].

Soft tissue grafts, such as temporalis fascia or perichondrium, are usually air dried in the operating theatre for a number of minutes prior to insertion. Subsequently, placement needs to be performed promptly and accurately as the grafts quickly lose gross structural stability and become technically difficult to manage when they rehydrate [44]. Cartilage grafts have received renewed interest over the last two decades as an alternative to perichondrium or fascia owing to excellent results achieved when reconstructing the TM in cases of advanced middle ear disease or eustachian tube dysfunction [45]. However, for a num-ber of years, cartilage was thought to have possible deleterious effects on postoperative hearing when compared with tempora-lis fascia [46]. Cartilage grafts that are cut too thickly have been shown to impair optimal acoustic transmission, while those that are too thin will sacrifice stability [25]. Moreover, in experi-mental studies, cartilage grafts tend to vibrate at significantly lower amplitudes than that of the adjacent tympanic mem-brane, which also has acoustic implications [25]. Limitations of temporalis fascia, cartilage and perichondrial grafts include possible donor site morbidity, their lack of transparency, and the differing biomechanical and material properties of these grafts when compared with intact TM, which may interfere with the conductive pathway of hearing [25].

Combinations of the aforementioned grafts have also been stud-ied, including cartilage–perichondrium composite grafts [44,47] citing high closure success rates and also the ability to use these

Figure 2. Subtotal perforation of tympanic membrane (dotted line) exposing the middle ear. This patient would require closure of the perforation with a graft and meticulous follow-up to examine for potential complications.

A

Silk fibroin device for myringoplasty

Expert Rev. Med. Devices 6(6), (2009)656

Review

grafts in repairing larger perforations, anterior perforations and perforations in wet or discharging ears, which can all be challeng-ing with other graft materials [47]. The cartilage-perichondrium composite graft is also rigid and thick compared with tempora-lis fascia or perichondrium [44]. However, it is nontransparent and slippery, and delays in hearing improvement up to 6 months postoperatively can occur as the graft takes time to soften [44,47]. Fat is another autologous graft that is frequently used to repair perforations but has the same limitations as the aforementioned grafts, and its success is largely dependent on the perforations being small and central [38–40].

Common synthetic & processed graft devicesThere is little evidence to suggest that any one of these grafts or devices is superior for all perforation types [A!"#$ MD,

A%#%&#'(()#*#$+#), KS, W#!!$ GD, U%-./"0$12& D#!#]. As with fat grafts, paper and alloderm are generally used for smaller, cen-tral perforations [38–43] and can be used in an outpatient or office setting without general anesthesia [40–42]. Alloderm is an allograft from human skin, which is processed to render the graft acellular, thereby reducing its immunogenicity. Some authors have reported advantages of this material that include a reduction in operative time, preservation of native tissues and the avoidance of donor site morbidity from graft harvesting [48]; however, their lack of transparency and their biomechanical and material properties differ from normal TM.

Other autologous grafts, allografts & synthetic devicesDespite the multitude of commonly used grafts available, a num-ber of other autologous and synthetic grafts have been described over the years and have been reported as being useful in myrin-goplasty. Dura has been tested with high closure rates, but it has not been shown to be as successful as temporalis fascia for all perforation types and has been recommended in situations where there may be insufficient fascia [49,50]. Pedicled meatal skin grafts have also been described to support fascia or cartilage [51,52], and

free skin grafts have been used successfully, obviating the need to raise a tympanomeatal flap [53]; However, these are associ-ated with donor site morbidity. Synthetic biomaterials, such as Carbylan™ (hyaluranan plus additional carboxyl groups), have been tested in animal models with acute perforations, showing good biocompatibility when compared with other biomaterials, but still require further investigation [54].

After acellular human dermis (alloderm) showed promise in the repair of chronic TM perforations [42,43,48], Spiegel et al. tested acellular porcine mucosa to repair perforations created in the drums of chinchillas [55], with closure results comparable to those of cartilage. Acellular intestinal submucosa has many of the ideal qualities sought in a good graft material: it is inex-pensive, readily available, does not require the harvesting of an autologous graft, thus reducing operative time and donor site morbidity, and it is easy to work with. It is not antigenic once it is manipulated to become acellular and does not carry a risk of human disease transmission [55]. However, it still pos-sesses different bio mechanical properties to native TM, and further studies in humans may be warranted to evaluate its clinical implications.

Autologous nasal mucosa has been tested as a graft in humans, with a closure rate of 91% [56]. It is described as being easy to harvest and prepare, but the authors note that it was only inves-tigated for its use in repairing small- and medium-sized defects. Other tissues used over the years as graft materials include peri-umbilical superficial fascia [57], vein grafts [58–63], fresh and pre-served amniotic membrane [64], and heart valves [65], all of which have variable success rates and donor site morbidity. Irradiated human rib cartilage has also been tested as a graft material [24], achieving a reduction in the air–bone gap of 20 dB or less in 70% of patients; however, this requires autopsy tissue donation and transplantation and, therefore, necessitates following rigorous screening guidelines for various diseases.

It is clear from the reports of the multitude of autologous grafts, homografts (allografts) and synthetic devices that have been tested in myringoplasty surgery that no single graft or device provides the optimal solution for all perforation types. The fact that devices are still being investigated for this purpose reiterates the fact that otolaryngologists are searching for a graft that will act as a true artificial eardrum, closing the per-foration and restoring premorbid hearing. Perhaps this is only feasible with a graft that is biodegradable. This would require the graft or device to be in the form of a scaffold that supports the overgrowth of new, native TM cells over a period of time. Ideally, this scaffold would biodegrade at a rate that matches the formation of neotissue, allowing new cells to function similarly to the original cells constituting that tissue.

Silk fibroin backgroundProperties of silk as a biomaterialWith the advent of tissue engineering, the development of an artificial TM from other biomaterials may become feasible. Silks have a long history of use in the human body in the form of suture materials owing to their mechanical properties, which

Figure 3. Silkworm cocoons cut open with pupa removed prior to any processing.

A

3.5 cm

Levin, Rajkhowa, Redmond & Atlas

www.expert-reviews.com 657

Review

include high tensile strength and flexibility (FIGURE 3) [66]. They are biomaterials that may possess many of the desirable quali-ties of an ideal graft [24], with biodegradability as an additional attractive property. Fibroin is the core structural protein in the filament extruded by silkworms and is encased by sericin, a glue-like protein. Despite their longstanding use as suture mate-rials, including the successful embolization of cerebral arterio-venous malformations [67–69], adverse immunological events have been associated with silk proteins, affecting their clinical applications [70,71]. However, the identification of sericin as the potential immunogenic protein and the ability to isolate fibroin from sericin by a thermochemical process termed degumming has sparked new interest in the use of fibroin for biomedical applications (FIGURE 4) [72–75]. Silk fibroin (SF) can also be derived from spiders (termed spidroin), but SF from silkworms (in par-ticular the Bombyx mori species) has been more widely studied owing to the higher levels of silk production by silkworms and the ease of obtaining it from nonpredatory animals.

Silk fibroin has several attractive physicochemical properties that have prompted investigation for its biomedical use. These include biocompatibility, good oxygen and water vapor perme-ability, and biodegradability [76,77]. SF can also be processed to become highly transparent [78]. Its fibers provide a remarkable combination of strength and toughness, bestowing excellent stability owing to extensive hydrogen bonding, the hydrophobic nature of SF and its significant crystallinity [77,79]. The ability to modify the morphology and structure of SF during processing has vast implications for biomedical use. It can be processed to form films (membranes), fibers, foams, hydrogels, mats and meshes, depending on experimental or clinical requirements (FIGURE 5) [79].

Immunological responses to silk biomaterialsPrior to clinical application, the immunogenic responses to SF required detailed study. All nonautologous biomaterials will illicit some foreign body response when interacting with living tissue; however, the biocompatibility of SF is well demonstrated [80–85], and SF from B. mori has been shown to be largely immuno-logically inert, invoking minimal immune response [85]. Silk-induced asthma and hypersensitivity responses have been reported in the literature [86–88], but the immuno globulin E responses elic-but the immuno globulin E responses elic-ited were thought to be due to the sericin component of silk [89]. There has now been widespread acceptance that isolated fibroin elicits minimal immune response, and differences have also been found in hypersensitivity responses between traditional silk sutures and sutures containing degummed silk [79]. Moreover, it has been shown that silk can be sterilized, either by means of autoclaving [90] or ethanol immersion [91]. As a result, interest in studying the biomedical, clinical and surgical applications of SF has been rekindled.

In vitro researchMultiple in vitro studies have shown that SF has potential for biomedical applications. It has been shown to be bio compatible with dorsal root ganglia [81] and to be beneficial to the survival

of Schwann cells without exerting any cytotoxic effects on these cells [81,92]. SF extract has been used to successfully culture hippo-campal neurons as well as support their survival and growth [93]. SF in the membranous form has been tested as a substrate for corneal limbal epithelial cell growth and was found to be com-parable to tissue culture plastic [94]. Fibroin has been extensively studied as a scaffold for musculoskeletal cell growth [91,95–98], with results yielding hope for an array of tissue engineering appli-cations, including the repair of musculo skeletal defects. Fibroin in various forms has been shown to support the adhesion, pro-liferation and differentiation of stem cells in vitro, which has allowed stem cell-based tissue engineering using 3D scaffolds to further expand silk’s use as a biomaterial, particularly relating to skeletal and connective tissues [71,99].

A SF scaffold has been shown to support the growth and angiogenesis potential of human endothelial cells [74,83], as well as the adherence and growth of epithelial, fibroblast, glial and osteoblast cells [100]. It has been tested as a scaffold for sup-porting rat hepatocytes [101], and was shown to be a suitable substratum for hepatocyte attachment and culture.

Silk fibroin has also been studied with applications in pharma-ceutical fields with potential therapeutic relevance [102,103]. Uebersax et al. found that silk matrices were suitable for design-ing adenosine-releasing bioincubators that may be useful in the management of epilepsy [102], while Bayraktar et al. isolated SF and explored its potential use as an aqueous coating material for theophylline tablets [103]. Their results demonstrated that SF possessed good film-forming and coating properties, with the potential for sustaining controlled drug release.

In vivo researchPrior to considering the use of SF as a graft biomaterial, not only are in vivo studies required to verify the biocompatibility established in the laboratory but also to evaluate its safety when used in animals and humans. SF has been tested in vivo in a

Figure 4. Silk fibroin after degumming (sericin protein has been removed).

5 cm

Silk fibroin device for myringoplasty

Expert Rev. Med. Devices 6(6), (2009)658

Review

number of organ systems. Dal Pra et al. implanted SF sheets into the subcutaneous tissues of mice and examined the in vivo bio compatibility. A mild foreign body response was elicited, but there was no significant macrophage or lymphocyte infiltration, and no hyperkeratosis or fibrosis formation [104]. SF films seeded with rat mesenchymal stem cells have also been implanted intra muscularly, with histological and immunohistochemical analysis revealing the presence of fibroblasts, few macrophages, the absence of giant cells and a smaller inflammatory reaction when compared with collagen films [99]. Kim et al. created bony defects in the calvaria of rabbits and implanted SF membranes into these defects, examining cell numbers and calcification at the surgical sites [77]. A bony union was observed after 8 weeks, and at 12 weeks, the defects had completely healed with new bone. Meinel et al. also examined the bone regeneration prop-erties of silk and extended their investigation to larger defects in load-bearing long bones [105]. They evaluated SF scaffolds as osteopromotive matrices in femoral defects created in nude rats. The SF scaffolds seeded with human mesenchymal stem cells exhibited good osteogenic potential, almost bridging the defects with new bone after 8 weeks and exhibited good load-bearing capabilities and torque when compared with their other experimental and control groups [105].

Silk fibroin has also been investigated for its use in ligament tissue engineering, where anterior cruciate ligament regeneration was examined using mesenchymal stem cells and silk compos-ite scaffolds (combined silk netted meshes and sponges), which mimicked the structure of ligament extracellular matrix [106]. In this study, Fan et al. found that at 24 weeks postimplantation, histo pathological examination confirmed that the stem cells were distributed throughout the regenerated ligament and that they had differentiated to exhibit the appropriate fibroblastic mor-phology. Key ligament extracellular matrix components were produced, including collagen I and III, as well as tenascin-C. Moreover, direct ligament–bone insertion was observed, which resembled native anterior cruciate ligament–bone insertion [106].

SF has also been examined for potential use as a nerve graft to assist peripheral nerve regeneration [107]. Yang et al. developed a graft composed of a SF-nerve guidance conduit with mechanical properties beneficial for nerve regeneration. The graft was tested in rat sciatic nerve defects, achieving results similar to those attained with nerve autografts [107].

Silk fibroin & otolaryngology: head & neck surgeryOf particular interest, SF has recently been tested in vivo in the field of otolaryngology [108]. Ni et al. examined whether SF could be used to reconstruct tracheal defects by implant-ing titanium meshes coated with SF inside these defects. Histopathological examination of the specimens did not reveal any signs of infectious complications nor excessive granulation tissue formation, and the skin covering the embedded SF was normal. Thin layers of fibroblasts (as well as microvessels) were observed growing in an orderly fashion covering the SF films. Importantly, CT scans showed no obvious tracheal stenosis, leading the authors to conclude that SF may be used as a coat-ing material for artificial tracheas and may have future merit in the challenging task of human tracheal reconstruction [108].

Application of silk fibroin to myringoplastyKeratinocyte growth on silk fibroin scaffoldsTo extend the use of SF in otolaryngology to repairing TM per-forations, the biocompatibility of SF with keratinocytes needs to be established as these cells compose the outer epidermal layer of the TM. Unger et al. examined the growth of human cells from multiple tissue origins and demonstrated that human kera-tinocytes adhere to one another, grow and spread over SF nets, which acted as supporting structures, allowing the keratinocytes to form tight cellular contacts on and between individual fibroin fibers [100]. Park et al. also seeded human epidermal keratino-cytes on chitin/SF scaffold blends, and examined their bio-compatibility and cellular behaviour [109]. Cell attachment and spreading assays confirmed that the keratinocytes proliferated, adhered to the blends and then spread over the substrates [109]. Previous studies have obtained similar results when keratinocyte growth has been examined on SF scaffolds [110,111]; however, the biocompatibility and cell growth of specific TM keratinocytes on a SF scaffold needs to be investigated [112].

Recently, cultured human TM cells were seeded onto SF membranes and cell growth, proliferation, and adhesion were compared with tissue culture plastic (polyethylene terepthlate) membranes [113]. The authors found that SF was more suitable as a substrate than standard tissue culture plastic for the growth and adhesion of TM cells. The biocompatibility of SF with human TM keratinocytes reiterates the interest of considering SF as graft material for use in myringoplasty.

Wound-healing effects of silk fibroinA scaffold that is not only biocompatible with keratinocytes, but also conducts wound healing would be ideal if it is to be considered for repairing TM perforations. For functional tissue repair, the ideal scaffold should:

Figure 5. Silk fibroin in the membranous (film) form after processing.

A

5 cm

1 2 3 4 5 cm

Levin, Rajkhowa, Redmond & Atlas

www.expert-reviews.com 659

Review

• Support cell attachment, migration, interaction, proliferation and differentiation;

• Exhibit biocompatibility;

• Biodegrade at a rate that matches neotissue growth and facilitate the integration of engineered tissue into the host tissue;

• Provide structural support for cells and neotissue postimplantation;

• Have versatile processing options to alter structure and morphology related to tissue-specific needs [71].

Although the effect of SF on TM defects requires investiga-tion, several studies have shown that SF may be efficacious in the promotion of wound healing in general [77,108,114–116].

Kweon et al. compared sponges containing SF with Vaseline® gauze as wound dressings. They reported faster wound contrac-tion with the SF sponges than with the gauze and histologically confirmed that there was organized deposition of collagen in the dermis covering the wounds dressed with SF [114]. Roh et al. also examined the wound healing effects of a SF/alginate-blended sponge in full-thickness skin defects in rats [115]. These skin defects were dressed with the SF/alginate blended sponge and compared with commercially available NuGauze™ at various postoperative time points, calculating residual wound area, regenerated epithelium and collagen deposition. They observed healing time to be signifi-cantly reduced with the SF-blended sponges and that the area of re-epithelialization was also significantly greater with these sponges, indicating their superiority over NuGauze with respect to wound healing. Collagen deposition was not significantly different between experimental and control groups [115]. Sugihara et al. created full-thickness dermatotomies on the dorsal walls of mice and examined the efficacy of a transparent SF film used as a biological wound dressing [116]. These dressings were compared with conventional, commercially available hydrocolloid dressings. The authors reported that the area of wounds dressed with SF was reduced to 10% of the original size by day 14 and that the wounds were covered with regenerated epidermis by day 21. By contrast, there was less reduc-tion in wound size and less epidermal regeneration in the wounds dressed with hydrocolloid dressings. Healing times of SF-dressed wounds were reported as being 7 days quicker than those with con-trol dressings, and histological findings revealed greater collagen regeneration and reduced inflammatory responses in the SF-dressed wounds [116]. Advantages from a surgical point of view include the ease of obtaining SF, its ability to be sterilized and its transparency, allowing for careful postoperative wound observation.

Silk fibroin degradationBiodegradation is a highly desired property when investigating materials with biomedical applications. This property would be useful when considering the repair of TMs as this could allow for the restoration of normal structure and function after biodegradation.

It is well established that SF is biodegradable [79,117–119]. The rate at which this occurs may be affected by processing meth-ods that cause conformational changes in its protein structure.

However, biodegradation of SF fibers ultimately occurs through proteolytic degradation, with resorption typically occurring within a year [79]. Moreover, the rate at which enzymatic pro-teolysis occurs can be predictably controlled in vitro to meet clinical needs [119]. Wang et al. studied the in vivo degradation of implanted SF scaffolds in rats and found that SF scaffolds prepared from all-aqueous processing completely degraded within 2–6 months, whereas those prepared via organic sol-vent (hexafluoroisopropanolol) processing persisted for more than 1 year [117]. These in vitro and in vivo insights have implications in tissue engineering, because processing meth-ods and SF morphological manipulation can be guided by tissue-specific requirements.

Expert commentary & five-year viewChronic TM perforations continue to represent a significant source of morbidity worldwide. The evolution of myringoplasty surgery has allowed for the successful closure of these perfora-tions, improvements in hearing outcomes and a reduction in

Figure 6. Light microscopy photos of human tympanic membrane keratinocytes. (A) Human tympanic membrane keratinocytes in culture; !10 magnification. (B) Human tympanic membrane keratinocyte growth on silk fibroin membrane, day 18; !4 magnification.

5 cm

A

B

20 µm

20 µm

Silk fibroin device for myringoplasty

Expert Rev. Med. Devices 6(6), (2009)660

Review

the associated long-term sequelae. Regardless of the incision and operative technique employed, no single graft material or device has been unequivocally established for use in all perfora-tion types, despite the variety of autologous grafts, allografts and synthetic devices available for selection.

Patients with TM perforations that are closed successfully often do not exhibit completely normal hearing postoperatively. Autologous and synthetic grafts do not biodegrade, which may contribute to this, as the TM has a very specific multilayered microstructure that is different to any graft materials described. However, a material that biodegrades over time and allows the simultaneous formation of neotissue may provide a solution to this problem by facilitating the formation of a neotympanum (new TM) with no residual graft material, thereby potentially improving both surgical and hearing outcomes for patients.

With the advent and development of tissue engineering, bio-materials continue to be studied with clinical applications in mind. Degummed silk possesses remarkable properties, includ-ing biocompatibility, high tensile strength, flexibility and bio-degradability. The ability to manipulate these properties and control SF morphology during manufacturing has proven to be invaluable as tissue-specific or clinical requirements can guide the processing of SF [71].

It is likely that these properties, combined with tissue-engi-neering expertise, will allow scaffolds to be processed that will have many diverse functions, including wound healing and guided tissue repair, as well as regeneration to restore function. This is already evident, with a number of studies demonstrating the repair of musculoskeletal defects. The use of SF to repair skin defects may also lead to an improvement in current biological wound dressings.

Tissue engineering using stem cells and silk biomaterials is of particular interest, and SF has been shown to support stem cell adhesion, growth and differentiation, as well as promote tissue

repair [71]. 3D SF scaffolds supporting stem cells continue to show tremendous promise in engineering tissues, for example bone, ligaments and skin. The ability to sterilize SF will con-tinue to allow in vivo research in a wide range of organ systems. Sterilization will also allow surgeons to consider its potential utility in repairing defects, including TM perforations.

The application of SF in the field of otolaryngology has only recently been employed. SF has demonstrated potential in the repair of tracheal defects, which have presented challenges to surgeons for decades and continue to do so [108]. The promo-tion of fibroblast growth at the defect site and the prevention of tracheal stenosis are major advantages of this material. Future in vivo studies examining the time until degradation will be helpful and may see this material used in the reconstruction of human tracheal defects.

As otolaryngology is such a diverse specialty, the use of SF within the specialty is likely to expand. Our group has success-fully cultured human autologous TM cells (keratinocytes) in our laboratory and seeded them onto SF films, with promising preliminary results (FIGURE 6) [112]. The films have successfully supported the growth, proliferation and adhesion of human TM cells in vitro and immunofluorescent staining for these properties as well as the maintenance of keratinocyte lineage have yielded encouraging results (FIGURE 7) [112].

Silk fibroin in a thin membranous form appears appropriate to mimic the natural TM. However, as the TM is a multilayered structure, strategies will be required to develop a scaffold to guide multi layered TM regeneration [112]. The scaffold design needs to consider the mechanical properties of the human TM and also mechanical strains imposed on it. Aqueous-based processing may be advantageous over organic and ionic solu-tions, as this avoids harmful chemicals and reduces the risks of residual solvents in nonaqueous processing.

Future in vivo research using animal models to examine the interaction of SF with TM keratinocytes may further develop its application for use in myringoplasty surgery. Certainly, its strength, flexibility, transparency, wound-healing ability and biodegradability make this an attractive option to pursue. A SF scaffold may permit the overgrowth of native TM cells while slowly biodegrading and could, therefore, result in the first true artificial TM, in terms of both structure and function.

AcknowledgementsThe authors thank Robert Eikelboom and Robert Marano for their assistance in reading the manuscript.

Financial & competing interests disclosureBrett Levin is supported by a Garnett Passe and Rodney Williams Memorial Foundation Surgeon Scientist Scholarship. The authors have no other relevant affiliations or financial involvement with any organiza-tion or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.

Figure 7. Keratinocytes on silk fibroin membrane immunofluorescently stained with ESE-1. This stains specifically for terminally differentiated keratinocytes, verifying the cell lineage of human tympanic membrane cells.

A

20 µm20 µm

Levin, Rajkhowa, Redmond & Atlas

www.expert-reviews.com 661

Review

Key issues

• Chronic tympanic membrane perforations represent a significant source of morbidity worldwide.• Myringoplasty is a successful operation that aims to repair perforations, improve hearing and prevent patient susceptibility to middle

ear infections.• Multiple autologous grafts, allografts and synthetic grafts are currently used for myringoplasty, all of which have limitations.• Silk fibroin (SF) has remarkable material properties, including stability, high tensile strength and the ability to form diverse morphologies.• The biocompatibility and slow biodegradability of SF make it an attractive biomaterial for tissue engineering purposes.• Several in vitro and in vivo studies have considered the clinical application of SF across a range of medical specialities, but it remains

relatively novel in the field of otolaryngology.• The biocompatibility of SF with keratinocytes and its potential for exerting wound healing effects make it an attractive option to pursue

as a graft for myringoplasty.• In vivo studies investigating the interaction of SF with normal and perforated tympanic membranes need to be undertaken to ensure

safety for use in the middle ear.• A SF graft may improve long-term hearing outcomes with the overgrowth of new tympanic membrane cells and the biodegradation of

SF allowing native tympanic membrane to replace the perforation.

ReferencesPapers of special note have been highlighted as:• of interest•• of considerable interest

1 Farrior JB. Incisions in tympanoplasty: anatomic considerations and indications. Laryngoscope 93(1), 75–86 (1983).

2 Darrow DH, Dash N, Derkay CS. Otitis media: concepts and controversies. Curr. Opin. Otolaryngol. Head Neck Surg. 11(6), 416–423 (2003).

3 Freid VM, Makuc DM, Rooks RN. Ambulatory health care visits by children: principal diagnosis and place of visit. Vital Health Stat. 13(137), 1–23(1998).

4 Nash DR, Harman J, Wald ER et al. Antibiotic prescribing by primary care physicians for children with upper respiratory tract infections. Arch. Pediatr. Adolesc. Med. 156(11), 1114–1119 (2002).

5 Bluestone CD. Epidemiology and pathogenesis of chronic suppurative otitis media: implications for prevention and treatment. Int. J. Pediatr. Otorhinolaryngol. 42(3), 207–223(1998).

6 Verhoeff M, van der Veen EL, Rovers MM et al. Chronic suppurative otitis media: a review. Int. J. Pediatr. Otorhinolaryngol. 70(1), 1–12 (2006).

7 Nittrouer S, Burton LT, Nittrouer S, Burton LT. The role of early language experience in the development of speech perception and phonological processing abilities: evidence from 5-year-olds with histories of otitis media with effusion and low socioeconomic status. J. Comm. Disorders 38(1), 29–63 (2005).

8 Osma U, Cureoglu S, Hosoglu S. The complications of chronic otitis media: report of 93 cases. J. Laryngol. Otol. 114(2), 97–100 (2000).

9 Kangsanarak J, Fooanant S, Ruckphaopunt K, Navacharoen N, Teotrakul S. Extracranial and intracranial complications of suppurative otitis media. Report of 102 cases. J. Laryngol. Otol. 107(11), 999–1004 (1993).

10 Trimis G, Mostrou G, Lourida A, Prodromou F, Syriopoulou V, Theodoridou M. Petrositis and cerebellar abscess complicating chronic otitis media. J. Paediatr. Child Health 39(8), 635–636 (2003); see comment.

11 Panda NK, Sreedharan S, Mann SB, Sharma SC. Prognostic factors in complicated and uncomplicated chronic otitis media. Am. J. Otolaryngol. 17(6), 391–396 (1996).

12 Chu EA, Jackler RK, Chu EA, Jackler RK. The artificial tympanic membrane (1840–1910): from brilliant innovation to quack device. Otol. Neurotol. 24(3), 507–518 (2003).

•• Interesting study examining the repair of tympanic membranes over centuries (prior to surgery) and the evolution of operative repair.

13 Ringenberg JC. Closure of tympanic membrane perforations by the use of fat. Laryngoscope 88(6), 982–993 (1978).

14 Aggarwal R, Saeed SR, Green KJ, Green KJM. Myringoplasty. J. Laryngol. Otol. 120(6), 429–432 (2006).

• Excellent clinical review of myringoplasty, with particular reference to prognostic factors.

15 Inwood JL, Wallace HC, Clarke SE. Endaural or postaural incision for myringoplasty: does it make a difference to the patient? Clin. Otolaryngol. Allied Sci. 28(5), 396–398 (2003); see comment.

16 Gerard JM, Gersdorff M. The Tutopatch graft for transcanal myringoplasty. B-ENT 2(4), 177–179 (2006).

17 Singh M, Rai A, Bandyopadhyay S et al. Comparative study of the underlay and overlay techniques of myringoplasty in large and subtotal perforations of the tympanic membrane. J. Laryngol. Otol. 117(6), 444–448 (2003).

18 Crovetto De La Torre M, Fiz Melsio L, Escobar Martinez A. [Myringoplasty in chronic simple otitis media. Comparative analysis of underlay and overlay techniques]. Acta Otorrinolaringologica Espanola 51(2), 101–104 (2000).

19 Vartiainen E, Nuutinen J. Success and pitfalls in myringoplasty: follow-up study of 404 cases. Am. J. Otol. 14(3), 301–305 (1993).

• Large retrospective study examining factors thought to influence the success of myringoplasty.

20 Kartush JM, Michaelides EM, Becvarovski Z et al. Over–under tympanoplasty. Laryngoscope 112(5), 802–807 (2002).

21 Eavey RD. Inlay tympanoplasty: cartilage butterfly technique. Laryngoscope 108(5), 657–661 (1998).

22 Ghanem MA, Monroy A, Alizade FS et al. Butterfly cartilage graft inlay tympanoplasty for large perforations. Laryngoscope 116(10), 1813–1816(2006).

Silk fibroin device for myringoplasty

Expert Rev. Med. Devices 6(6), (2009)662

Review

23 Effat KG, Effat KG. Results of inlay cartilage myringoplasty in terms of closure of central tympanic membrane perforations. J. Laryngol. Otol. 119(8), 611–613 (2005).

24 Schulte DL, Driscoll CL, McDonald TJ, Facer GW, Beatty CW. Irradiated rib cartilage graft for reconstruction of the tympanic membrane: preliminary results. Am. J. Otol. 19(2), 141–144 (1998).

25 Huttenbrink K. Current Topics in Otolaryngology – Head and Neck Surgery. Middle Ear Surgery – Recent Advances and Future Directions. Jahnke K (Ed.). Georg Thieme Verlag Publishers, Stuttgart, Germany (2004).

26 Vorwerk U, Hey M, Steinicke G, Begall K. High-speed digital videoimaging of fast eardrum motions during Valsalva maneuver. ORL J. Otorhinolaryngol. Relat. Spec. 60(3), 138–142 (1998).

27 Gerber MJ, Mason JC, Lambert PR. Hearing results after primary cartilage tympanoplasty. Laryngoscope 110(12), 1994–1999 (2000).

28 Dirckx JJ, Decraemer WF, von Unge M, Larsson C. Volume displacement of the gerbil eardrum pars flaccida as a function of middle ear pressure. Hearing Res. 118(1–2), 35–46 (1998).

29 Nagai T, Tono T. Encapsulated nerve corpuscles in the human tympanic membrane. Arch. Oto-Rhino-Laryngol. 246(3), 169–172 (1989).

30 Teoh SW, Flandermeyer DT, Rosowski JJ. Effects of pars flaccida on sound conduction in ears of Mongolian gerbil: acoustic and anatomical measurements. Hearing Res. 106(1–2), 39–65 (1997).

31 Karela M, Berry S, Watkins A, Phillipps JJ, Berry S. Myringoplasty: surgical outcomes and hearing improvement: is it worth performing to improve hearing? Probl. Actuels Otorhinolaryngol. 265(9), 1039–1042 (2008).

32 Dornhoffer JL, Dornhoffer JL. Cartilage tympanoplasty. Otolaryngol. Clin. N. Am. 39(6), 1161–1176 (2006).

33 Tos M, Tos M. Cartilage tympanoplasty methods: proposal of a classification. Otolaryngol. Head Neck Surg. 139(6), 747–758 (2008).

34 Yung M. Cartilage tympanoplasty: literature review. J. Laryngol. Otol. 122(7), 663–672 (2008).

35 Dursun E, Dogru S, Gungor A et al. Comparison of paper-patch, fat, and perichondrium myringoplasty in repair of

small tympanic membrane perforations. Otolaryngol. Head Neck Surg. 138(3), 353–356 (2008).

36 Gierek T, Slaska-Kaspera A, Majzel K et al. [Results of myringoplasty and type I tympanoplasty with the use of fascia, cartilage and perichondrium grafts]. Otolaryngol. Polska 58(3), 529–533 (2004).

37 Williamson PA, Thomas DM, Beasley P. Posterior tragal perichondrium harvesting for myringoplasty. Clin. Otolaryngol. Allied Sci. 24(4), 252–254 (1999).

38 Fiorino F, Barbieri F. Fat graft myringoplasty after unsuccessful tympanic membrane repair. Eur. Arch. Oto-Rhino-Laryngol. 264(10), 1125–1128 (2007).

39 Landsberg R, Fishman G, DeRowe A et al. Fat graft myringoplasty: results of a long-term follow-up. J. Otolaryngol. 35(1), 44–47 (2006).

40 Ozgursoy OB, Yorulmaz I. Fat graft myringoplasty: a cost-effective but underused procedure. J. Laryngol. Otol. 119(4), 277–279 (2005).

41 Golz A, Goldenberg D, Netzer A et al. Paper patching for chronic tympanic membrane perforations. Otolaryngol. Head Neck Surg. 128(4), 565–570 (2003).

42 Saadat D, Ng M, Vadapalli S, Sinha UK. Office myringoplasty with alloderm. Laryngoscope 111(1), 181–184 (2001).

43 Youssef AM. Use of acellular human dermal allograft in tympanoplasty. Laryngoscope 109(11), 1832–1833 (1999).

44 Mansour MH, Askar MH, Albirmawy OA. Repair of tympanic membrane perforation using a modified cartilage-perichondrium composite ring graft. J. Laryngol. Otol. 120(11), 952–954 (2006).

45 Dornhoffer J, Dornhoffer J. Cartilage tympanoplasty: indications, techniques, and outcomes in a 1,000-patient series. Laryngoscope 113(11), 1844–1856 (2003).

46 Puls T. Tympanoplasty using conchal cartilage graft. Acta Otorhinolaryngol. Belg. 57(3), 187–191 (2003).

47 El-Hennawi DM. Cartilage perichondrium composite graft (CPCG) in pediatric tympanoplasty. Int. J. Pediat. Otorhinolaryngol. 59(1), 1–5 (2001).

48 Vos JD, Latev MD, Labadie RF et al. Use of AlloDerm in type I tympanoplasty: a comparison with native tissue grafts. Laryngoscope 115(9), 1599–1602 (2005).

49 Yetiser S, Tosun F, Satar B. Revision myringoplasty with solvent-dehydrated human dura mater (Tutoplast). Otolaryngolo. Head Neck Surg. 124(5), 518–521 (2001).

50 Packer P, Mackendrick A, Solar M. What’s best in myringoplasty: underly or overlay, dura or fascia? J. Laryngol. Otol. 96(1), 25–41 (1982).

51 Raghavan U, Malki DS, Mahmoud NA. Myringoplasty: update on onlay pedicle skin flap and temporalis fascia sandwich graft. J. Laryngol. Otol. 114(3), 174–177 (2000).

52 Henein RR, Gandhi AG. Myringoplasty using free tragal cartilage and pedicled meatal skin flap. J. Laryngol. Otol. 94(9), 1005–1008 (1980).

53 Chanvimalueng W. A clinical comparison of outpatient and standard myringoplasty. Ear Nose Throat J. 79(2), 113–114 (2000).

54 Park AH, Hughes CW, Jackson A et al. Crosslinked hydrogels for tympanic membrane repair. Otolaryngol. Head Neck Surg. 135(6), 877–883 (2006).

55 Spiegel JH, Kessler JL, Spiegel JH, Kessler JL. Tympanic membrane perforation repair with acellular porcine submucosa. Otol. Neurotol. 26(4), 563–566 (2005).

56 Strasser GM, Schratzenstaller B. [Myringoplasty with autologous mucosa-graft]. Laryngorhinootologie 87(2), 93–95 (2008).

57 Ajulo SO, Myatt HM, Alusi G. Peri-umbilical superficial fascial graft myringoplasty – a simple alternative. Clin. Otolaryngol. Allied Sci. 18(5), 433–435 (1993).

58 Miyoshi T, Watanabe T, Yokose H, Okitsu T. [Myringoplasty by means of a vein graft]. Jibiinkoka 43(1), 25–30 (1971).

59 Jelinek R, Herodek F. [Myringoplasty using a venous graft according the Shea]. Ceskoslovenska Otolaryngol. 16(6), 362–364 (1967).

60 Valenzuela C. Myringoplasty. Experience with vein graft technics. Int. Surg. 47(1), 87–96 (1967).

61 Richards SH, McGee TM. Experimental observations on vein graft myringoplasty. J. Laryngol. Otol. 79(11), 952–963 (1965).

62 Mitchell JF. Myringoplasty by homogenous vein graft. J. Laryngol. Otol. 81(3), 339–346 (1967).

63 Heck WE. Vein graft myringoplasty. Laryngoscope 71, 847–853 (1961).

64 Lyons GD, Dyer RF, Ruby JR. Morphologic analysis of tympanic membrane grafts. Laryngoscope 87(10 Pt 1), 1705–1709 (1977).

65 Cornish CB, Scott PJ. Freeze dried heart valves as tympanic grafts. Arch. Otolaryngol. 88(4), 350–356 (1968).

Levin, Rajkhowa, Redmond & Atlas

www.expert-reviews.com 663

Review

66 Moy RL, Lee A, Zalka A. Commonly used suture materials in skin surgery. Am. Fam. Physician 44(6), 2123–2128 (1991).

67 Schmutz F, McAuliffe W, Anderson DM, Elliott JP, Eskridge JM, Winn HR. Embolization of cerebral arteriovenous malformations with silk: histopathologic changes and hemorrhagic complications. Am. J. Neuroradiol. 18(7), 1233–1237 (1997).

68 Song JK, Eskridge JM, Chung EC et al. Preoperative embolization of cerebral arteriovenous malformations with silk sutures: analysis and clinical correlation of complications revealed on computerized tomography scanning. J. Neurosurg. 92(6), 955–960 (2000).

69 Dehdashti AR, Muster M, Reverdin A, de Tribolet N, Ruefenacht DA. Preoperative silk suture embolization of cerebral and dural arteriovenous malformations. Neurosurg. Focus 11(5), E6 (2001).

70 Salthouse TN. Biologic response to sutures. Otolaryngol. Head Neck Surg. 88(6), 658–664 (1980).

71 Wang Y, Kim HJ, Vunjak-Novakovic G et al. Stem cell-based tissue engineering with silk biomaterials. Biomaterials 27(36), 6064–6082 (2006).

•• Superb review of stem cell-based tissue engineering and the application of silk fibroin to this arena. Future directions and uses of silk biomaterials are discussed.

72 Liu H, Ge Z, Wang Y et al. Modification of sericin-free silk fibers for ligament tissue engineering application. J. Biomed. Mater. Res. Part B Appl. Biomater. 82(1), 129–138 (2007).

73 Yamada H, Igarashi Y, Takasu Y et al. Identification of fibroin-derived peptides enhancing the proliferation of cultured human skin fibroblasts. Biomaterials 25(3), 467–472 (2004).

74 Unger RE, Peters K, Wolf M, Motta A, Migliaresi C, Kirkpatrick CJ. Endothelialization of a non-woven silk fibroin net for use in tissue engineering: growth and gene regulation of human endothelial cells. Biomaterials 25(21), 5137–5146 (2004).

75 Gotoh Y, Niimi S, Hayakawa T et al. Preparation of lactose-silk fibroin conjugates and their application as a scaffold for hepatocyte attachment. Biomaterials 25(6), 1131–1140 (2004).

76 Santin M, Motta A, Freddi G, Cannas M. In vitro evaluation of the inflammatory potential of the silk fibroin. J. Biomed. Mater. Res. 46(3), 382–389 (1999).

77 Kim KH, Jeong L, Park HN et al. Biological efficacy of silk fibroin nanofiber membranes for guided bone regeneration. J. Biotechnol. 120(3), 327–339 (2005).

78 Minoura N, Tsukada M, Nagura M. Physico–chemical properties of silk fibroin membrane as a biomaterial. Biomaterials 11(6), 430–434 (1990).

79 Altman GH, Diaz F, Jakuba C et al. Silk-based biomaterials. Biomaterials 24(3), 401–416 (2003).

•• Excellent review of silk biomaterials from silk worms and spiders. Discusses properties of silk, biomedical utility, biocompatibility and degradation.

80 Gellynck K, Verdonk P, Forsyth R et al. Biocompatibility and biodegradability of spider egg sac silk. J. Mater. Sci. Mater. Med. 19(8), 2963–2970 (2008).

81 Yang Y, Chen X, Ding F et al. Biocompatibility evaluation of silk fibroin with peripheral nerve tissues and cells in vitro. Biomaterials 28(9), 1643–1652 (2007).

82 Mei N, Zhou P, Pan LF et al. Biocompatibility of poly (3-hydroxybutyrate-co-3-hydroxyhexanoate) modified by silk fibroin. J. Mater. Sci. Mater. Med. 17(8), 749–758 (2006).

83 Ma X, Cao C, Zhu H, Ma X, Cao C, Zhu H. The biocompatibility of silk fibroin films containing sulfonated silk fibroin. J. Biomed. Mater. Res. Part B Appl. Biomater. 78(1), 89–96 (2006).

84 Jiang Y, Chen H, Zhou W et al. [Research on preparation of silk fibroin and its biocompatibility with rat bone marrow mesenchymal stem cells]. Sheng Wu Yi Xue Gong Cheng Xue Za Zhi 23(3), 560–564 (2006).

85 Acharya C, Ghosh SK, Kundu SC. Silk fibroin protein from mulberry and non-mulberry silkworms: cytotoxicity, biocompatibility and kinetics of L929 murine fibroblast adhesion. J. Mater. Sci. Mater. Med. 19(8), 2827–2836 (2008).

86 Uragoda CG, Wijekoon PN. Asthma in silk workers. J. Soc. Occupation. Med. 41(3), 140–142 (1991).

87 Vovolis V, Galatas I. Silk-induced asthma. Allergy Asthma Proc. 20(2), 107–108 (1999).

88 Wen CM, Ye ST, Zhou LX, Yu Y. Silk-induced asthma in children: a report of 64 cases. Ann. Allergy 65(5), 375–378 (1990).

89 Zaoming W, Codina R, Fernandez-Caldas E, Lockey RF. Partial characterization of the silk allergens in mulberry silk extract. J. Invest. Allergol. Clin. Immunol. 6(4), 237–241 (1996).

90 Meinel L, Hofmann S, Karageorgiou V et al. Engineering cartilage-like tissue using human mesenchymal stem cells and silk protein scaffolds. Biotechnol. Bioeng. 88(3), 379–391 (2004).

91 Li C, Vepari C, Jin HJ et al. Electrospun silk-BMP-2 scaffolds for bone tissue engineering. Biomaterials 27(16), 3115–3124 (2006).

92 Allmeling C, Jokuszies A, Reimers K et al. Use of spider silk fibres as an innovative material in a biocompatible artificial nerve conduit. J. Cell. Mol. Med. 10(3), 770–777 (2006).

93 Tang X, Ding F, Yang Y, Hu N, Wu H, Gu X. Evaluation on in vitro biocompatibility of silk fibroin-based biomaterials with primarily cultured hippocampal neurons. J. Biomed. Mater. Res. A. 91(1), 166–174. (2009).

94 Chirila T, Barnard Z, Zainuddin Z et al. Bombyx mori silk fibroin membranes as potential substrata for epithelial constructs used in the management of ocular surface disorders. Tissue engineering Part A 14(7), 1203–1211 (2008).

95 Marolt D, Augst A, Freed LE et al. Bone and cartilage tissue constructs grown using human bone marrow stromal cells, silk scaffolds and rotating bioreactors. Biomaterials 27(36), 6138–6149 (2006).

96 Wang Y, Blasioli DJ, Kim HJ et al. Cartilage tissue engineering with silk scaffolds and human articular chondrocytes. Biomaterials 27(25), 4434–4442 (2006).

97 Meinel L, Karageorgiou V, Hofmann S et al. Engineering bone-like tissue in vitro using human bone marrow stem cells and silk scaffolds. J. Biomed. Mater. Res. A 71(1), 25–34 (2004).

98 Acharya C, Ghosh SK, Kundu SC, Acharya C, Ghosh SK. Silk fibroin film from non-mulberry tropical tasar silkworms: a novel substrate for in vitro fibroblast culture. Acta Biomaterialia 5(1), 429–437 (2009).

99 Meinel L, Hofmann S, Karageorgiou V et al. The inflammatory responses to silk films in vitro and in vivo. Biomaterials 26(2), 147–155 (2005).

100 Unger RE, Wolf M, Peters K et al. Growth of human cells on a non-woven silk fibroin net: a potential for use in tissue engineering. Biomaterials 25(6), 1069–1075 (2004).

Silk fibroin device for myringoplasty

Expert Rev. Med. Devices 6(6), (2009)664

Review

101 Cirillo B, Morra M, Catapano G. Adhesion and function of rat liver cells adherent to silk fibroin/collagen blend films. Int. J. Artif. Organs 27(1), 60–68 (2004).

102 Uebersax L, Fedele DE, Schumacher C et al. The support of adenosine release from adenosine kinase deficient ES cells by silk substrates. Biomaterials 27(26), 4599–4607 (2006).

103 Bayraktar O, Malay O, Ozgarip Y et al. Silk fibroin as a novel coating material for controlled release of theophylline. Eur. J. Pharma. Biopharma. 60(3), 373–381 (2005).

104 Dal Pra I, Freddi G, Minic J et al. De novo engineering of reticular connective tissue in vivo by silk fibroin nonwoven materials. Biomaterials 26(14), 1987–1999 (2005).

105 Meinel L, Betz O, Fajardo R et al. Silk based biomaterials to heal critical sized femur defects. Bone 39(4), 922–931 (2006); erratum: 43(6), 1123 (2008).

106 Fan H, Liu H, Wong EJ et al. In vivo study of anterior cruciate ligament regeneration using mesenchymal stem cells and silk scaffold. Biomaterials 29(23), 3324–3337 (2008).

107 Yang Y, Ding F, Wu J et al. Development and evaluation of silk fibroin-based nerve grafts used for peripheral nerve regeneration. Biomaterials 28(36), 5526–5535 (2007).

108 Ni Y, Zhao X, Zhou L et al. Radiologic and histologic characterization of silk fibroin as scaffold coating for rabbit tracheal defect repair. Otolaryngol. Head Neck Surg. 139(2), 256–261 (2008).

109 Park KE, Jung SY, Lee SJ et al. Biomimetic nanofibrous scaffolds: preparation and characterization of chitin/silk fibroin blend nanofibers. Int. J. Biolog. Macromol. 38(3–5), 165–173 (2006).

110 Min BM, Lee G, Kim SH et al. Electrospinning of silk fibroin nanofibers and its effect on the adhesion and spreading of normal human keratinocytes and fibroblasts in vitro. Biomaterials 25(7–8), 1289–1297 (2004).

111 Min BM, Jeong L, Nam YS et al. Formation of silk fibroin matrices with different texture and its cellular response to normal human keratinocytes. Int. J. Biolog. Macromol. 34(5), 281–288 (2004).

112 Levin B, Redmond SL, Rajkhowa R et al. Preliminary results of the application of a silk fibroin scaffold to otology. Otolaryngol. Head Neck Surg. DOI:10.1016/j.otohns.2009.06.746 (2009) (Epub ahead of print).

113 Ghassemifar R, Redmond SL, Zainuddin Z et al. Advancing towards a tissue-engineered tympanic membrane: silk fibroin as a substratum for growing human eardrum keratinocytes. J. Biomater. Appl. DOI: 10.1177/0885328209104289 (2009) (Epub ahead of print).

114 Kweon H, Yeo JH, Lee KG et al. Semi-interpenetrating polymer networks composed of silk fibroin and poly(ethylene glycol) for wound dressing. Biomed. Mater. 3(3), 034115 (2008).

115 Roh DH, Kang SY, Kim JY et al. Wound healing effect of silk fibroin/alginate-blended sponge in full thickness skin defect of rat. J. Mater. Sci. Mater. Med. 17(6), 547–552 (2006).

116 Sugihara A, Sugiura K, Morita H et al. Promotive effects of a silk film on epidermal recovery from full-thickness skin wounds. Proc. Soc. Exp. Biol. Med. 225(1), 58–64 (2000).

117 Wang Y, Rudym DD, Walsh A et al. In vivo degradation of three-dimensional silk fibroin scaffolds. Biomaterials 29(24–25), 3415–3428 (2008).

• Studies in vivo degradation of silk fibroin and examines the various factors that influence this process, discussing how these can be manipulated for clinical purposes.

118 Kim UJ, Park J, Kim HJ et al. Three-dimensional aqueous-derived biomaterial scaffolds from silk fibroin. Biomaterials 26(15), 2775–2785 (2005).

119 Horan RL, Antle K, Collette AL et al. In vitro degradation of silk fibroin. Biomaterials 26(17), 3385–3393 (2005).

Websites201 WHO. Chronic Suppurative Otitis Media.

Burden of Illness and Management Options (2004) www.who.int/pbd/deafness/activities/hearing_care/otitis_media.pdf

202 WHO. Global Burden of Hearing Loss in the Year 2000 www.who.int/healthinfo/statistics/bod_hearingloss.pdf

Affiliations• Brett Levin, B Med Sci, MBBS (Hons)

Ear Science Institute Australia, Ear Sciences Centre, School of Surgery, The University of Western Australia, Sir Charles Gairdner Hospital, Perth, WA, Australia Tel.: +61 (08) 9346 4153 Fax: +61 (02) 8088 3820 blevin@med.usyd.edu.au

• Rangam Rajkhowa, MTech Centre for Material and Fibre Innovation (CMFI), Deakin University, Geelong, VIC, Australia

• Sharon Leanne Redmond, Assoc Dip Appl Sci (Biol) Ear Science Institute Australia and Ear Sciences Centre, School of Surgery, The University of Western Australia, Perth, WA, Australia

• Marcus David Atlas, MBBS, FRACS Ear Science Institute Australia, Ear Sciences Centre, School of Surgery, The University of Western Australia, Perth, WA, Australia and St John of God Hospital, Perth, WA Australia and Sir Charles Gairdner Hospital, Perth, WA, Australia

Levin, Rajkhowa, Redmond & Atlas

Chapter 2

Phenotypic and Genotypic Profile of Human Tympanic Membrane Derived Cultured Cells

Redmond SL, Levin B, Heel KA, Atlas MD, Marano RJ. Phenotypic and Genotpyic Progile of Human Tympanic Membrane Derived Cultured Cells. Accepted for publication in Journal of Molecular Histology on 12 October 2010.

ORIGINAL PAPER

Phenotypic and genotypic profile of human tympanic membranederived cultured cells

Sharon L. Redmond • Brett Levin • Kathryn A. Heel •

Marcus D. Atlas • Robert J. Marano

Received: 20 September 2010 / Accepted: 2 November 2010

� Springer Science+Business Media B.V. 2010

Abstract The human tympanic membrane (hTM), known

more commonly as the eardrum, is a thin, multi-layered

membrane that is unique in the body as it is suspended in

air. When perforated, the hTM’s primary function of

sound-pressure transmission is compromised. For the pur-

poses of TM reconstruction, we investigated the phenotype

and genotype of cultured primary cells derived from hTM

tissue explants, compared to epithelial (HaCaT cells) and

mesenchymal (human dermal fibroblasts (HDF)) reference

cells. Epithelium-specific ets-1 (ESE-1), E-cadherin,

keratinocyte growth factor-1 (KGF-1/FGF-7), keratinocyte

growth factor-2 (KGF-2/FGF10), fibroblast growth factor

receptor 1 (FGFR1), variants of fibroblast growth factor

receptor 2 (FGFR2), fibroblast surface protein (FSP), and

vimentin proteins were used to assess the phenotypes of all

cultured cells. Wholemount and paraffin-embedded hTM

tissues were stained with ESE-1 and E-cadherin proteins to

establish normal epithelial-specific expression patterns

within the epithelial layers. Immunofluorescent (IF) cell

staining of hTM epithelial cells (hTMk) demonstrated

co-expression of both epithelial- and mesenchymal-specific

proteins. Flow cytometry (FCM) analysis further demon-

strated co-expression of these epithelial and mesenchymal-

specific proteins, indicating the subcultured hTMk cells

possessed a transitional phenotype. Gene transcript analy-

sis of hTMk cells by reverse transcriptase polymerase

chain reaction (RT-PCR) revealed a down regulation of

ESE-1, E-cadherin, FGFR2, variant 1 and variant 2

(FGFR2v1 and FGFR2v2) between low and high passages,

and up-regulation of KGF-1, KGF-2, and FGFR1. All

results indicate a gradual shift in cell phenotype of hTMk-

derived cells from epithelial to mesenchymal.

Keywords Eardrum � Primary cells � Epithelial �Flow cytometry � RT-PCR � Keratinocytes �Confocal microscopy

Introduction

Cell culture is an important tool for studying cells in detail

as well as the structural characteristics of their tissue of

origin. By utilizing in vitro models that are biologically

relevant to the tissue and parental cell type, important

information can be revealed with a view towards recon-

struction and replacement of diseased or damaged struc-

tures (Atala 2006), in this case chronic perforations of the

hTM. We have reported the culture of hTM epithelial cells

(hTMk) on silk substrates (Levin et al. 2010) with the aim

of testing alternative grafts for repairing perforated human

eardrums. However, if cultured primary cells are to be used

as the progenitors of an engineered tissue or organ, their

phenotypic and genotypic integrity must be maintained

throughout the culture process.

Structurally, the whole TM is composed of three layers,

which have been widely discussed (Broekaert 1995; Lim

S. L. Redmond � M. D. Atlas � R. J. Marano (&)

Molecular and Cellular Otolaryngology Research Laboratory,

Ear Science Institute Australia, 2nd Floor, M Block, Room 2.27

(M507), QEII Medical Centre, Nedlands, WA 6009, Australia

e-mail: rob.marano@earscience.org.au

S. L. Redmond � B. Levin � M. D. Atlas � R. J. Marano

Ear Sciences Centre, School of Surgery, The University

of Western Australia, Nedlands, WA 6009, Australia

K. A. Heel

Centre for Microscopy, Characterisation and Analysis (M510),

The University of Western Australia, 35 Stirling Highway,

Crawley, WA 6009, Australia

123

J Mol Hist

DOI 10.1007/s10735-010-9303-5

1995; Stenfors 1984). The middle connective/collagenous

layers of the TM (lamina propria) support the growth of

both an epithelial and mucosal layer on its outer and inner

surfaces, respectively. This demonstrates that TM tissue as

a whole consists of a mixture of cell layers and therefore,

cell types, typically fibroblasts, epithelial, endothelial,

nerve, mucosal, and mast cells, along with a variety of

leukocyte and lymphocytic cells (Stenfeldt et al. 2006).

The epithelial cell layers are of paramount importance in

cases of TM perforations and wound healing, as they are

the first cell type involved in wound closure (Santa Maria

et al. 2010a; Taylor and McMinn 1967).

The isolation and confirmation of the cells’ epithelial

origin may provide a foundation for ongoing in vitro work

in tissue engineering scaffolds to be used in TM recon-

struction (Levin et al. 2009). The aim of this study is to

characterize primary hTMk cells derived from normal hTM

tissue explants and to compare them to the parent tissue

and reference cell types. The isolation and characterization

of hTMk cells for use in vitro will enable detailed cyto-

toxicity and wound healing studies (Santa Maria et al.

2010b) to be performed, which will lead to a better

understanding of diseased states of the hTM and improved

therapeutics.

Materials and methods

Culture of primary hTM cells

Fresh intact hTM tissue (n = 6) was obtained from con-

senting patients who underwent otological procedures

requiring the permanent surgical removal of the tympanic

membrane. All hTM tissue received was disease-free and

considered morphologically normal. Small pieces of hTM

tissue, approximately 0.25 mm2, were placed into BD

Falcon 6-well culture plates (BD Biosciences, Australia).

DMEM 4,500 mg/l D-glucose was supplemented with

100 U/ml penicillin, 100 lg/ml streptomycin (Invitrogen,

Australia), and 10% FBS (Invitrogen). The derived primary

cells were incubated for 20 days in a humidified cell cul-

ture incubator at 37�C, with 5% carbon dioxide (CO2).

Subsequent cell passages were incubated for 5 days in BD

Falcon T-75 culture flasks (BD Biosciences).

Culture of reference cells

HaCaT keratinocyte cells were maintained in DMEM high

glucose as per primary hTM cells.

Primary human dermal fibroblast (HDF) cells were

cultured from frozen cell stocks originally derived from

burn-injured patients. Cells were maintained in DMEM/

F12?GLUTAMAX (Invitrogen) culture media supplemented

with 100 U/ml penicillin, 100 lg/ml streptomycin (Invitro-

gen), and 10% FBS (Invitrogen) for 5 days in BD Falcon

T-75 culture flasks (BD Biosciences). Cells were incubated in

a humidified cell culture incubator at 37�C, with 5% CO2.

Antibodies

Primary antibodies: rabbit polyclonal anti-human ESE-1

(NovusBiologicals, USA), mouse monoclonal anti-human

vimentin, and mouse monoclonal anti-human E-cadherin

(CDH1) (Zymed, USA), and mouse monoclonal anti-

human fibroblast surface protein (AbCam, UK). Donkey

polyclonal anti-goat FGF-7 (KGF-1) and goat polyclonal

anti-rabbit FGF-10 (KGF-2) (Santa Cruz Biotechnology,

USA), rabbit polyclonal anti-human FGFR1 and rabbit

polyclonal anti-human FGFR2 (AbCam, UK). Secondary

antibodies: Alexa Fluor-488 conjugated goat anti-mouse,

Alexa Fluor-546 conjugated goat anti-rabbit, Alexa Fluor-

488 conjugated donkey anti-goat (Molecular Probes, USA).

Nuclei were counterstained with 40,6-diamidino-2-pheny-

lindole (DAPI) (Molecular Probes).

Cultured cell IF staining

Cultured cells were seeded into BD Falcon 8-well chamber

slides (BD Biosciences) and incubated for 7 days at 37�C

in a humidified tissue culture incubator. Prior to staining,

cells were washed briefly in PBS, pH7.2 (Invitrogen), then

incubated in ice-cold methanol at -20�C for 10 min,

followed by three washes with PBS, pH7.2 containing

0.1% Tween 20 (PBS-Tw). Incubation with primary anti-

bodies was performed at room temperature (RT) for

60 min, followed by three washes with PBS-Tw. Second-

ary antibody incubations were performed at RT for 60 min

followed by three washes in PBS-Tw. Cell nuclei were

counterstained with DAPI for 20 min, washed with PBS-

Tw, cover slipped in anti-fade mounting medium and

sealed with nail varnish. Controls were performed for each

cell type by omitting the primary antibody and incubating

in the appropriate conjugated secondary only.

Paraffin section IF staining

Whole hTM tissue was fixed in 4% paraformaldehyde

(Scharlau Chemie, Spain) in 19 PBS, pH 8.0, overnight.

Following fixation, tissue were transferred to 70% ethanol

and processed in a Leica tissue processor, model TP1020

(Wetzlar, Germany). Tissues were embedded on edge to

permit the examination of all cell layers. Briefly, 4 lm

paraffin sections were deparaffinized through a series of

solvents to water and rinsed in 19 Tris buffered saline

(TBS), pH 7.4 containing 0.1% Triton X-100 (TBS-Tx).

J Mol Hist

123

Sections were incubated in 3% hydrogen peroxide (H2O2)

in methanol for 5 min, washed three times in TBS-Tx, and

then blocked with 20% heat-inactivated serum for 20 min.

Primary antibodies were incubated for 60 min followed by

three washes of TBS-Tx. Subsequent incubation with the

corresponding secondary antibody (1:500) diluted in PBS-

Tw, was for 60 min followed by three washes in TBS-Tx.

Finally, cell nuclei were counterstained with DAPI for

20 min, washed in three changes of TBS-Tx, cover slipped

in an anti-fade mounting medium. All staining procedures

were performed at RT in a light-tight humidified staining

box under low light conditions unless otherwise stated.

Wholemount IF staining

Wholemount tissues were fixed with 4% paraformalde-

hyde (Scharlau Chemie) in 19 PBS, pH 8.0, overnight.

Following fixation, tissues were washed with PBS-Tw, on a

shaker for 8 h with hourly PBS-Tw buffer changes. Tissues

were then incubated with E-cadherin and ESE-1 primary

antibodies (double-labeled) diluted in PBS-Tw for 12 h in

a humidified box at 4�C. Tissues were then washed with the

PBS-Tw buffer for 8 h with hourly buffer changes. Incu-

bation with goat anti-mouse-AF488 (E-cadherin) and goat

anti-rabbit-AF546 (ESE-1) secondary antibodies diluted in

PBS-Tw (1:500) was performed overnight. Tissues were

washed with PBS-Tw buffer with hourly changes for 8 h.

Incubation with DAPI nuclear counterstain was performed

overnight at 4�C followed by hourly changes of PBS-Tw

buffer for 8 h. Tissue was mounted with 50:50 glycerol/

19 PBS mounting media. Staining procedures were per-

formed at RT in a light-tight humidified staining box under

low light conditions unless stated otherwise.

Confocal laser-scanning microscopy and imaging

Images were captured sequentially using a MRC 1000/

1024 UV confocal laser-scanning microscope (BioRad,

UK) on a Nikon Diaphot 300 using a glycerol

209 objective (Nikon Fluor, NA 0.75) for wholemount

tissues and a 409 oil immersion objective (Nikon Fluor,

NA 1.3) for detailed paraffin section and cell IF imaging. A

488 nm laser was used to excite AF-488 secondary anti-

bodies and fluorescence emission detected through a 522/

35 nm bandpass filter. A 543 nm laser was used to excite

AF546 secondary antibodies and fluorescence emission

detected through a 580/32 nm bandpass emission filter. A

351 nm laser was used for the excitation of DAPI and

fluorescence emission detected through a 455/30 nm

bandpass filter. Images were saved as 8-bit greyscale

images and were subsequently merged using ImageJ

(1.42e) (Wayne Rasband, http://rsb.info.nih.gov/ij/).

Flow cytometry of cultured cells

Cells were detached and washed in PBS, pH 7.4 containing

1% BSA. Permeabilization of cell membranes was per-

formed using the BD Cytofix/Cytoperm fixation/permea-

bilization kit (BD Biosciences) for 20 min at RT.

Following this step, cells were washed twice in a 1:10

dilution of 109 stock of BD Perm/Wash buffer. Cell pel-

lets were re-suspended with primary antibodies diluted in

wash buffer and incubated for 30 min at RT, washed twice,

and subsequently incubated with secondary antibodies for

30 min at RT. An unstained cell population was processed

concurrently for each cell type. Finally, after washing twice

more, cells were re-suspended and analyzed on a BD

FACSCalibur flow cytometer (BD Biosciences) with

488 nm excitation. All cells were gated in a forward/side

scatter plot (FSC vs SSC). AF488 fluorescence emission

was detected with a 530/30 bandpass filter and acquired in

log. Data was collected for 10,000 cells and an AF488

fluorescence histogram was used to compare the unstained

and stained samples. Sample data were analyzed using

FlowJo v8.8.6 (Treestar Inc, USA).

The MFI values of protein expression were compared

using ANOVA with a posthoc Fishers LSD (protected

t tests) using GB-STAT v9.0 (Dynamic Microsystems Inc.,

USA). Differences were considered significant at P \ 0.05.

Total RNA isolation and cDNA synthesis

Total RNA was extracted from fresh hTM explants and

cultured cells using the All-Prep DNA/RNA/protein kitTM

(Qiagen, Australia). One quarter of a whole hTM or pel-

leted cells from confluent wells of a 6-well plate were used

for RNA extraction. Total RNA was eluted from the spin

columns with 40 ll of RNase free water, with a 5 ll ali-

quot taken from each and diluted 1 in 2 with Tris/EDTA

(pH 7.5) to measure the concentration using a NanoPho-

tometerTM (IMPLEN, Germany).

Synthesis of first strand cDNA was performed using the

SuperScriptTM III Reverse Transcriptase (RT) kit (Invit-

rogen) with 1 lg of total RNA as template in a final

reaction volume of 20 ll for 60 min. Samples were cleaned

using Centri–SpinTM-40 columns (Princeton Separations

Inc., USA). cDNA concentration was measured using the

NanoPhotometerTM (IMPLEN), adjusted to 1 ng/ll and

stored at -20�C.

Polymerase chain reaction

Genes and their corresponding primers are listed in

Table 1. Reaction components per tube were as follows:

12.5 ll of iQTM iSYBR Green� supermix (Bio-Rad),

10 pmol each forward and reverse primer and 5 ng of

J Mol Hist

123

template cDNA with the final volume adjusted to 25 ll

with ddH2O. PCR was performed using an iCyclerTM (Bio-

Rad) with the following cycling conditions, 1 9 95�C for

3 min followed by 35 cycles of 95�C—10 s/60�C—15

s/72�C—30 s with a final hold of 72�C for 5 min. Ampli-

cons were separated in a 1.5% agarose gel in 19 TAE

(75 min at 90 V) and visualized using a Geliance 600 Gel

Doc (Perkin-Elmer, Australia) with GeneSnap Software

(Perkin-Elmer). All samples were electrophoresed on a

single gel to ensure uniformity. TrackItTM 100 bp DNA ladder

(Invitrogen) was used to verify the correct product sizes.

Results

Initially, cell morphology was assessed using light

microscopy. Following this the localization and expression

by each cell type and hTMk passage (high and low) was

established by IF staining and FCM analysis. Lastly,

RT-PCR was performed to establish the gene expression

pattern of all cell types.

Basic cell morphology of cultured cells

Explanted tissue and primary hTMk cells

Cells emerged from primary tissue explants after 24–36 h

in culture. The cells grew as a monolayer, initially closely

surrounding each tissue explant (Fig. 1, panel a) until

complete well coverage was achieved. Microscopically, the

cell populations were homogenous, tightly packed, and

cobblestone-like in appearance, characteristic of keratino-

cytes in vitro. No fibroblasts were visible in these cultures.

Changes in primary hTMk cells following subculture

(passage)

The hTMk cells at both low (P1) (Fig. 1, panel b) and high

(P7) (Fig. 1, panel c) passage numbers exhibited epithelial-

like morphology where cells appeared slightly elongated.

In all passages, the cells were observed to stream together

en masse to form tightly packed, spiral patterns throughout

the cell population. The hTMk cells did not overlap each

other at any point and remained as a homogenous popu-

lation. Low passage hTMk P1 cells appeared more tightly

packed together than cells at higher passage P7.

Morphology of HaCaT cell line

HaCaT cells consisted of a predominantly homogenous

population of keratinocytes (Boukamp et al. 1988). These

cells exhibited typical keratinocyte morphology (Fig. 1,

panel d) in vitro, displaying a tightly packed cobblestone-

like appearance. Immortilised HaCaT cells were chosen, as

they remained stable and unchanged despite multiple

passages.

Morphology of primary dermal fibroblasts

Human dermal fibroblast (HDF) cells, originally derived from

post-burn dermis, were chosen because they are a transitional

mesenchymal cell population. The cells exhibited a more

Table 1 Nucleotide sequence

of primers and expected size of

PCR products used for gene

transcript analysis

Gene Primer sequences (50–30) Size

(bp)

Accession

number

KGF-1 (FGF7) F: GACATGGATCCTGCCAACTT 319 NM_002009.3

R: CCCTTTGATTGCCACAATTC

KGF-2 (FGF-10) F: ATGTCCGCTGGAGAAAGCTA 295 NM_004465.1

R: CTCCCATTATGCTGCCAGTT

FGFR2v1 (variant 1) F: GGATCAAGCACGTGGAAAAG 154 NM_000141.4

R: TAGAATTACCCGCCAAGCAC

FGFR2v2 (variant 2) F: GGATCAAGCACGTGGAAAAG 204 NM_022970.3

R: TGTTTTGGCAGGACAGTGAG

ESE-1 F: CTGAGCAAAGAGTACTGGGACTGTC 189 NM_004433.4

R: CCATAGTTGGGCCACAGCCTCGGAGC

FGFR1 (variant 1) F: ACGGCCGACTGCCTGTGAAG 381 NM_023110.2

R: AGCTCCGGGTGTCGGGAAAG

E-cadherin F: GGCTGATACTGACCCCACAG 179 NM_004360.3

R: CGTACATGTCAGCCAGCTTC

GAPDH F: GAAGGTGAAGGTCGGAGTC 226 NM_002046.3

R: GAAGATGGTGATGGGATTTC

J Mol Hist

123

elongated spindle-shaped cell (Fig. 1, panel e) relative to the

passaged hTMk cells. The HDF cells were not uniform in

composition, appearing irregular in shape and moved in a

more disorganized manner as individuals rather than

en masse, typical of fibroblast cells in culture.

IF staining of epithelial- and mesenchymal-specific

protein expression

Wholemount and paraffin-embedded hTM tissue

Strong IF staining for the epithelial specific proteins of

ESE-1 and E-cadherin were seen in freshly prepared

wholemount hTM tissue (Fig. 2a, panel a and c respec-

tively), which established the normal staining pattern in the

tissue of origin. Staining was visualized by confocal laser-

scanning microscopy and seen exclusively in the epithelial

layer where cells exhibited strong cytoplasmic membrane

(green fluorescence) expression. DAPI staining (red fluo-

rescence) identified the cell nuclei. Additionally, IF stain-

ing of ESE-1 and E-cadherin on paraffin-embedded hTM

tissue sections was visualized exclusively in the epithelial

layers (Fig. 2a, panel b and d respectively). The white

arrows in Fig. 2a, panels c and d, indicate strong cyto-

plasmic membrane expression of ESE-1 and E-cadherin

proteins, respectively. No expression of ESE-1 or E-cad-

herin proteins was seen in the inner mucosal or the middle

collagenous layers.

Cultured cells

Epithelial-specific proteins Strong cytoplasmic mem-

brane-specific ESE-1 protein expression was seen in

HaCaT cells (Fig. 2b). The expression in HDF cells was a

combination of cytoplasmic and punctate cytoplasmic

membrane staining (Fig. 2b). The expression in hTMk P2

cells was located primarily on the cytoplasmic membrane.

However, some cytoplasmic staining was observed. The

expression in hTMk P7 cells was exclusively cytoplasmic

throughout the entire cell population (Fig. 2b).

E-cadherin expression was very strong in HaCaT cells

with the protein specifically localized to the cytoplasmic

membrane indicating strong cell–cell adhesion (Fig. 2b).

However, expression of E-cadherin in hTMk P2, P7, and

HDF cells was internalized and found primarily within the

cell cytoplasm and nuclei.

Keratinocyte growth factors and associated recep-

tors Low KGF-1 (FGF-7) IF staining was seen in HaCaT

cells with expression primarily localized to cell nuclei,

Fig. 1 Cell culture of hTMk and reference cells (panels a–e). hTM

exhibited cobblestone-like morphology in initial culture (panel a). A

white asterick pinpoints the tissue explant. Changes in cell morphol-

ogy following subculture of hTMk are evident at low passage (P1)

(panel b) and high passage (P7) (panel c), now appearing slightly

more elongated. HaCaT keratinocytes (panel d) appear typically

cobblestone-like in appearance. Transitional HDF cells (panel e)

appear even more loosely packed, excessively elongated, and spindle-

shaped. Scale Bar is 20 lm for all panels

J Mol Hist

123

Fig. 2 a Wholemount hTM

tissue: Expression of ESE-1

(panel a), and E-cadherin (panel

c) proteins are seen exclusively

on the cytoplasmic membranes

of epithelial cells. Paraffin-

embedded tissue: Distribution

of ESE-1 (panel b) and

E-cadherin (panel d) proteins is

entirely within the epithelial

layers. ESE-1 expression (whitearrow heads, panel b) is

primarily expressed in the basal

layer of the epithelium.

However, positive expression

was also seen in the middle

epithelial layer. Strong E-

cadherin (white arrow heads,

panel d) staining was observed

within all layers. b Cultured

cells were stained for ESE-1,

E-cadherin, KGF-1, KGFR,

KGF-2, FGFR2, FSP and

vimentin, and compared to

HaCaT and primary human

HDF reference cells. Transition

of hTMk cells away from a

purely epithelial to a transitional

mesenchymal phenotype was

observed as passage numbers

increase. HaCaT cells show

strong epithelial-specific

staining. HDF cells show a

transitional cell type with strong

expression of mesenchymal-

specific proteins. Scale Bar is

20 lm for all panels

J Mol Hist

123

although low-level expression was also seen in the cyto-

plasm (Fig. 2b). The localization of KGF-1 in hTMk P2

cells was punctate and cytoplasmic membrane-bound, with

negligible staining observed in cell nuclei (Fig. 2b). The

nuclei of HDF cells exhibited strong, punctate staining for

KGF-1. Additionally, moderate staining on the cytoplasmic

membranes of HDF cells was visible.

KGF-2 (FGF-10) expression in HaCaT cells was local-

ized to the cytoplasm (Fig. 2b). Low-level cytoplasmic

staining was seen in hTMk P2 cells (Fig. 2b), compared to

strong expression observed in hTMk P7, which was

localized to the cytoplasmic membrane and punctate in

appearance. Similar cytoplasmic localization was seen in

HDF cells however, staining patterns appeared very diffuse

and were not apparent in all cells.

FGFR1 expression in HaCaT cells appears punctate and

was localized to cell nuclei (Fig. 2b). A combination of

nuclear, peri-nuclear, and cytoplasmic staining was seen in

hTMk P2 cells (Fig. 2b). Similar patterns were evident in

hTMk P7 cells with increased staining seen within cell

nuclei (Fig. 2b). The expression of FGFR1 in HDF cells

was strongly visible within cell nuclei with weaker

expression appearing on the cytoplasmic membrane

(Fig. 2b).

FGFR2 is strongly expressed in HaCaT cells, and was

localized to the cytoplasm and cytoplasmic membrane

(Fig. 2b). Diffuse cytoplasmic staining for FGFR2 was

seen in hTMk P2 cells with decreased staining seen within

cell nuclei when compared to HaCaTs (Fig. 2b). Locali-

zation of moderate FGFR2 expression was seen within the

cell cytoplasm of hTMk P7 cells, with minimal expression

seen in cell nuclei. Expression within HDF cells was

nuclei-specific with no staining visible in either the cyto-

plasm or cytoplasmic membranes of cells.

Mesenchymal-specific proteins Low expression of FSP

was localized to the cytoplasmic membrane in HaCaT

cells, whereas vimentin expression was negligible

(Fig. 2b). Low FSP expression was seen throughout the

cytoplasm of all hTMk P2 cells, with stronger cytoplasmic

staining seen within hTMk P7 cells (Fig. 2b). Increased

cytoplasmic expression in hTMk P2 and P7 cells when

compare to HaCaTs was seen for vimentin (Fig. 2b). In

contrast to hTMk cells, HDFs expressed a lower level of

FSP, which was localized to the cytoplasm of all cells.

Vimentin expression in HDF cells was strong and localized

to the cytoplasm (Fig. 2b).

FCM analysis of cultured cells

HaCaT, hTMk P3 and P7, and HDF cells were analysed for

protein expression of ESE-1, KGF-1 (FGF-7), KGF-2

(FGF-10), FGFR1, FGFR2, FSP, and vimentin. The his-

togram for each protein was colour-coded for comparison

between cell types and passages (Fig. 3), in relation to

unstained cell populations (grey histograms). The histo-

grams display the frequency distribution of the fluores-

cence intensity of each protein for each cell type. Table 2

outlines the mean fluorescence intensity (MFI) values

(±SEM) for cultured cell populations for each of the pro-

teins listed above. Subpopulations in HDF and hTMk P3

cells are represented by G1 (low expressing) and G2

(highly expressing).

Fig. 3 Cultured cells were analysed for protein expression of ESE-1

(green histogram), KGF-1 (dark blue), KGF-2 (light blue), FGFR1

(yellow), FGFR2 (purple), FSP (orange), and vimentin (red). Higher

expression of ESE-1 and FGFR2 by cells indicated epithelial cell

attributes. Higher expression of KGF-1, KGF-2, FGFR1, along with

FSP and vimentin indicated that cells exhibited mesenchymal

attributes. Transitioning proteins for HDF and hTMk P3 cells were

gated as high and low MFIs and are represented by G1 (asterisk) and

G2 (plus)

J Mol Hist

123

ESE-1 protein expression

All cell types exhibited strong expression for ESE-1 (green

histogram) relative to the respective unstained cell popu-

lations (grey histogram). HaCaT and hTMk P7 cells had

very similar MFI values. Although HaCaT cell mean

expression was 2.3 times greater than hTMk P3 cells, it

was not statistically significant (P [ 0.05). HDF cell

expression fell half way between that of hTMk P3 and P7

cell expression. No significant differences were found

between the cell types.

Keratinocyte growth factors and associated receptors

KGF-1 expression for all cell types was low (dark blue

histogram). HaCaT and hTMk P3 cells had similar protein

expression. The hTMk P7 cells produced the highest MFI

value, 1.3 times higher than HDF cells. There was a sig-

nificant difference between the MFI values of HaCaT and

hTMk P3 and similarly between hTMk P3 and hTMk P7

(P \ 0.05).

KGF-2 expression in all cell types was relatively low

(light blue histogram). hTMk P3 cells produced the lowest

MFI of all cell types. HaCaT, hTMk P7, and HDF cells

produced very similar MFI values compared to the

unstained cell populations. There was a significant differ-

ence (P \ 0.05) between the MFI value of hTMk P3

compared to both HaCaT and hTMk P7 cells.

FGFR1 was moderately expressed in most cell types

(yellow histogram). HaCaT cells had the lowest MFI of all

cell types. The hTMk P3 and P7 cells had similar FGFR1

expression levels, 2.3 and 2.1 times higher than HaCaTs

respectively. This result was statistically significant

(P \ 0.05). The HDF cell population exhibited an asym-

metrical peak and was therefore, gated as two subpopula-

tions (G1 and G2), as illustrated in Fig. 3. The G1 (*)

subpopulation contained 32.5% of cells and G2 (?) sub-

population, contained 67.4%. The G2 (?) subpopulation

had a MFI 3.2 times greater than the cell population in G1

(*). Additionally, the G2 subpopulation of HDF cells was

2.5 times greater than hTMk P3 and P7 cells, and 5.8 times

that of HaCaT cells, which was statistically significant

(P \ 0.05).

FGFR2 expression in all cell types varied between

low to moderate (purple histogram). The hTMk P3, P7

and HDF cells had very similar MFI values compared

to HaCaT cells. This difference between the MFI values

for HaCaT and HDF cells was statistically significant

(P \ 0.05).

Mesenchymal-specific proteins

FSP expression was low to moderate (orange histogram).

HaCaT cell expression of FSP produced the lowest MFI,

followed by hTMk P3 cells with similar expression. The

HDF cells produced the highest MFI along with hTMk P7

cells. The difference between the similarly expressing MFI

levels by HaCaT and hTMk P3 to that of both hTMk P7

and HDF cells was statistically significant (P \ 0.05).

Vimentin expression (red histogram) was low in HaCaT

and hTMk P3 cells, very high in hTMk P7 and HDF cells.

HDF cells produced the highest MFI value, 125 times

higher than HaCaT cells. The hTMk P3 cell population

exhibited a skewed peak, and was gated as two subpopu-

lations (G1 and G2) (Fig. 3). The G1 (*) subpopulation

contained 12.3% and the G2 (?) subpopulation contained

87.5% of cells. The G2 (?) subpopulation produced a MFI

6.3 times greater than the cell population in G1 (*). Sta-

tistically significant differences (P \ 0.05) were seen

between the very low expressing HaCaT cells, the highly

expressive hTMk P7 and HDF cells, along with both hTMk

P3 subpopulations.

Table 2 FCM analysis of the MFI of protein expression by cultured cell populations (mean ± SEM, n = 4)

HaCaT hTMk, P3 hTMk, P7 HDF cells

Unstained 3.15 ± 1.27 2.88 ± 1.2 6.95 ± 1.88 3.36 ± 1.31

ESE-1 507.25 ± 108.10 215.5 ± 10.48 440.5 ± 96.20 333.75 ± 30.50

KGF-1 29.68 ± 3.70 * 24.15 ± 3.47* 53.98 ± 11.28rj 41.65 ± 4.28

KGF-2 40.6 ± 0.87j 8.44 ± 2.06r* 45.25 ± 8.19j 34.25 ± 12.59

FGFR1 237 ± 12.10js* 517.5 ± 16.08rs 548.25 ± 19.70rs G1: 425 ± 14.72

G2: 1,374 ± 26.48rj*

FGFR2 47.55 ± 0.40s 69.75 ± 5.90 61.85 ± 5.50 79.2 ± 10.60r

FSP 33.78 ± 0.48*s 76.4 ± 6.18s 153.75 ± 9.10r 179.5 ± 36.90rj

Vimentin 14.85 ± 0.09 *js G1: 58.9 ± 5.48

G2: 370 ± 13.74 *sr

1,623 ± 93.60rj 1,856.25 ± 10.80rj

Symbols denote a significant (P value \ 0.05) difference between HaCaT (r), hTMk, P3 (j), hTMk, P7 (*), or HDF (s)

J Mol Hist

123

Gene transcript analysis

PCR and gel electrophoresis (Fig. 4) was used to examine

the relative expression patterns of various genes within

whole hTM tissue (panel a), hTMk P2 cultured cells (panel

b), hTMk P7 cells (panel c), HaCaTs (panel d), and cul-

tured HDF cells (panel e). Lane 1 was the reference gene

GAPDH, which shows very strong expression throughout

all samples. Lane 2 represents KGF-1, which expressed

strongly in hTMk P2, P7 and HDF cells, moderately in

hTM tissue and weakly in HaCaTs. Lane 3 is KGF-2 and

was expressed weakly in hTM tissue, moderately in hTMk

P2 and P7, and was absent in both HaCaT and HDF cells.

FGFR2v1 (lane 4) was very weakly expressed in hTM

tissue, moderately expressed in hTMk P2, absent from

hTMk P7 and HaCaTs, and very strongly expressed in

HDF cells. The faint band seen in panel d, lane 4, for

HaCaT cells was considered to be non-specific binding, as

the product size was incorrect. FGFR2v2 (lane 5) was

strongly expressed in hTM tissue, very strongly expressed

in HaCaTs, weak in HDF cells and absent from both hTMk

P2 and P7 cells. ESE-1 (lane 6), an epithelial-specific

marker, was very strong in HaCaTs, strong in hTMk P2,

moderate in hTMk P7, weak in hTM tissue, and absent

from HDF cells. FGFR1 (lane 7) expressed weakly in both

hTM tissue and HaCaTs, while expressing strongly in

hTMk P2, P7, and HDF cells. E-cadherin (lane 8) exhibited

strong expression in hTM tissue, weak expression in hTMk

P2, was absent from hTMk P7, very strongly expressed in

HaCaTs and moderately expressed in HDF cells.

Discussion

Historically, the distinction between epithelial and mes-

enchymal cell populations in vitro was based solely on cell

morphology (Karasek 1975; Shook 2003). Initial observa-

tions of the derived hTMk primary cells, before subculture,

were reminiscent of classical keratinocyte architecture in

vitro. By utilizing normal hTM wholemount and paraf-

fin-embedded tissue sections, we have determined the

localization of the epithelial-specific proteins, ESE-1 and

E-cadherin, to be exclusively within the epithelial layers of

the hTM. This was a logical first step in linking the derived

hTMk cells to the epithelial layers of the hTM tissue of

origin.

Using multiple techniques, we have shown the

co-expression of epithelial- and mesenchymal-specific

characteristics exhibited by hTMk cells. Following sub-

culture of hTMk cells, experimental results indicated an

early transition with respect to both phenotypic and geno-

typic characteristics, which progressed with increasing

passage number. The phenotypic drift observed in these

cells may be due to a phenomenon known as epithelial-

mesenchymal transition (EMT) (Hay 2005; Lee et al. 2006;

Radisky 2005; Thiery and Sleeman 2006), a process that

has been described as a gradual transition of an epithelial

cell type to a mesenchymal cell type. EMT is initially

Fig. 4 Relative gene expression patterns for GAPDH (lane 1), KGF-

1 (lane 2), KGF-2 (lane 3), FGFR2v1 (lane 4), FGFR2v2 (lane 5),

ESE-1 (lane 6), FGFR1 (lane 7), and E-cadherin (lane 8), performed

on whole hTM tissue (panel a) and cultured cells by RT-PCR. The

genotypic profile for both hTMk P1 (panel b) and P7 (panel c) cells

indicated co-expression of both epithelial- and mesenchymal-specific

gene expressions, indicative of cells in transition. HaCaT cells (panel

d) have a genotypic profile indicative of an epithelial phenotype,

while HDF cells (panel e) exhibit primarily mesenchymal traits

J Mol Hist

123

demonstrated by a loss of apico-basal polarity in epithelial

cells due to a decrease in adhesion-dependent proteins such

as E-cadherin (Thiery 2003). The loss and reorganization

of the cytoskeleton of epithelial cells undergoing EMT

coincides with an increased cell expression of vimentin, a

widely accepted and ubiquitous mesenchymal cell marker.

This transitional process necessitates a change in cell

morphology where cells shift away from their classical

keratinocyte architecture to become elongated and fibro-

blast-like in morphology.

Investigations by means of FCM into the phenotypic

changes occurring in passaged hTMk cells revealed mod-

erate to strong expression of the mesenchymal-specific

proteins vimentin, FSP, and FGFR1. In contrast, HaCaT

cells exhibited low expression of mesenchymal markers,

confirming the stability of this cell lines’ epithelial phe-

notype. FCM results were confirmed by RT-PCR gene

transcript analysis (Fig. 4). The strong gene expression of

FGFR1 seen in both hTMk cell passages (Fig. 4) further

emphasizes the phenotypic shift in these cells from epi-

thelial to mesenchymal. Two subpopulations of HDF cells

expressing the FGFR1 protein, highlighted in Fig. 3, show

their transitional nature.

IF staining (Fig. 2b) revealed that hTMk P2 cells pos-

sessed a dual phenotype. Strong protein expression was

seen for both FSP and vimentin, both mesenchymal-

specific cell markers and not normally expressed in non-

tumorigenic epithelial cells. However, strong expression of

ESE-1 was visualized within the cytoplasm of hTMk P2

cells. The IF staining results, further validated by FCM

(Fig. 3), clearly show the co-expression of both epithelial

and mesenchymal proteins in all cell types with the

exception of the stable HaCaT cells (Ronnov-Jessen et al.

1992). Gene transcript analysis (Fig. 4) revealed very

strong expression of both ESE-1 and E-cadherin in HaCaTs

and very low expression of the fibroblast cell marker

FGFR1. In contrast to HaCaTs, FCM analysis of HDF cells

showed similar protein expression patterns to hTMk cells,

exhibiting co-expression of both epithelial- and mesen-

chymal-specific markers, indicating that HDF cells are also

transitional. This result was expected as the cells were

derived from an area of regenerating skin.

In this study KGF-1 (FGF-7) and KGF-2 (FGF-10),

along with their receptors were investigated as they are

widely accepted as important paracrine mediators of

keratinocyte differentiation and proliferation (Finch et al.

1989; Rubin 1989). KGF-1 and KGF-2 are similar in both

sequence and functional properties, and are expressed

exclusively by cells of mesenchymal origin. Thus, it was

not surprising that HDF cells were strongly positive for

KGF-1. Following any injury and subsequent re-epitheli-

alization, KGF-1 is strongly upregulated in mesenchymal

cells (Werner et al. 1992). Conversely, the HaCaT

keratinocyte cell line exhibited very weak gene expression

for KGF-1. We did not anticipate strong gene expression

for hTMk P1 and P7 cells, as it is widely reported in the

literature that KGF-1 is not expressed by normal cells of

epithelial origin (Finch et al. 1989; Werner 1998; Pereira

et al. 2007). However, Parrott et al. (2000) reported

observing high levels of protein and gene expression of

KGF-1 in normal human and bovine ovarian surface epi-

thelium and ovarian cancer tissues. It is important to note

that ovarian surface epithelium are derived from the

mesoderm and are modified peritoneal mesothelial cells.

Moderate expression of KGF-1 was seen in hTM whole

tissue by RT-PCR and was likely derived from fibroblasts

residing in the middle connective tissue layer. Studies have

revealed KGF-1 to be located exclusively in the dermis and

not the epidermis (Finch et al. 1989).

KGF-2 (FGF-10) expression, unlike KGF-1 (FGF-7), is

not induced during wound healing (Werner et al. 2007). As

the HDF cells were derived from an area undergoing active

wound healing and regeneration, it was expected that no

gene expression of KGF-2 would be evident by RT-PCR.

However, the expression observed in both hTMk cell

passages by FCM and RT-PCR again was indicative of

cells with a transitioning phenotype and therefore, further

evidence of EMT.

FGFR2b is the main receptor for KGFs and binds with

high affinity. By utilizing two variants of FGFR2, v1 and

v2, to distinguish between mesenchymal (v1) and epithelial

(v2) cell types, RT-PCR analysis for FGFR2v1 revealed

very strong expression in HDF cells and no expression in

HaCaT cells. In keeping with a transitional cell type,

RT-PCR analysis of hTMk cells revealed a decrease in

gene expression of FGFR2v2, and an increase in expres-

sion of FGFR2v1, with increasing passage, which is in

keeping with a transitioning cell population.

In this paper, we have shown, using several techniques,

that keratinocytes cultured from the human tympanic

membrane undergo an epithelial to mesenchymal transition

during continued passage. This research highlights the need

for the evaluation and monitoring the phenotype of

experimental cells at both low and high passage numbers to

maintain experimental reliability and validity. A stable and

committed keratinocyte cell population (in vitro) derived

from the hTM is essential in our current goal of developing

a biocompatible scaffold tailored specifically to enhancing

keratinocyte growth with the long-term aim of utilization in

myringoplasty (Ghassemifar et al. 2010; Levin et al. 2010).

Acknowledgments This study has full human ethics approval from

the St John of God Health Care Ethics Committee (St John of God

Hospital, Subiaco, Western Australia). The Garnett Passe and Rodney

Williams Memorial Foundation, and the Medical and Health Research

Infrastructure Fund, Western Australian Department of Health, fun-

ded this study. We gratefully acknowledge Winthrop Professor Fiona

J Mol Hist

123

Wood (Burn Injury Research Unit, UWA), for supplying the human

HDF cells and Winthrop Professor Marcus Atlas for the supply of

valuable hTM whole tissue samples for this study. Adjunct Professor

Robert Eikelboom and Dr Bing Teh are thanked for valuable com-

ments relating to the manuscript. The authors acknowledge the

facilities, scientific and technical assistance of the Australian

Microscopy & Microanalysis Research Facility at the Centre for

Microscopy, Characterisation & Analysis, The University of Western

Australia, a facility funded by The University, State, and Common-

wealth Governments.

References

Atala A (2006) Tissue-engineered autologous bladders for patients

needing cystoplasty. Lancet 367:1241–1246

Boukamp P, Petrussevska R, Breitkreutz D, Hornung J, Markham A,

Fusenig N (1988) Normal keratinization in a spontaneously

immortalized aneuploid human keratinocyte cell line. J Cell Biol

106:761–771

Broekaert D (1995) The tympanic membrane: a biochemical updating

of structural components. Acta Otorhinolaryngol Belg

49:127–137

Finch P, Rubin J, Miki T, Ron D, Aaronson S (1989) Human KGF is

FGF-related with properties of a paracrine effector of epithelial

cell growth. Science 245:752–755

Ghassemifar R, Redmond SL, Zainuddin, Chirila TV (2010)

Advancing towards a tissue-engineered tympanic membrane:

silk fibroin as a substratum for growing human eardrum

keratinocytes. J Biomater Appl 24:591–606

Hay ED (2005) The mesenchymal cell, its role in the embryo, and the

remarkable signaling mechanisms that create it. Dev Dyn

233:706–720

Karasek MA (1975) In vitro growth and maturation of epithelial cells

from postembryonic skin. J Invest Dermatol 65:60–66

Lee J, Dedhar S, Kalluri R, Thompson E (2006) The epithelial-

mesenchymal transition: new insights in signaling, development,

and disease. J Cell Biol 172:973–981

Levin B, Rajkhowa R, Redmond SL, Atlas MD (2009) Grafts in

myringoplasty: utilizing a silk fibroin scaffold as a novel device.

Exp Rev Med Dev 6:653–664

Levin B, Redmond SL, Rajkhowa R, Eikelboom RH, Marano RJ,

Atlas MD (2010) Preliminary results of the application of a silk

fibroin scaffold to otology. Otolaryngol Head Neck Surg 142:

S33–S35

Lim D (1995) Structure and function of the tympanic membrane: a

review. Acta Otorhinolaryngol Belg 49:101–115

Parrott JA, Kim G, Mosher R, Skinner MK (2000) Expression and

action of keratinocyte growth factor (KGF) in normal ovarian

surface epithelium and ovarian cancer. Mol Cell Endocrinol

167:77–87

Pereira CT, Herndon DN, Rocker R, Jeschke MG (2007) LIposomal

gene transfer of keratinocyte growth factor improves wound

healing by altering growth factor and collagen expression. J Surg

Res 139:222–228

Radisky DC (2005) Epithelial-mesenchymal transition. J Cell Sci

118:4325–4326

Ronnov-Jessen L, Celis J, Van Deurs B, Petersen O (1992) A

fibroblast-associated antigen: characterization in fibroblasts and

immunoreactivity in smooth muscle differentiated stromal cells.

J Histochem Cytochem 40:475–486

Rubin J (1989) Purification and characterization of a newly identified

growth factor specific for epithelial cells. Proc Natl Acad Sci

USA 86:802

Santa Maria PL, Redmond SL, Atlas MD, Ghassemifar R (2010a)

Histology of the healing tympanic membrane following perfo-

ration in rats. Laryngoscope 120:2061–2070

Santa Maria PL, Redmond SL, Atlas MD and Ghassemifar R (2010b)

The role of epidermal growth factor in the healing tympanic

membrane following perforation in rats. J Mol Histol (in press)

Shook D (2003) Mechanisms, mechanics and function of epithelial-

mesenchymal transitions in early development. Mech Dev

120:1351–1383

Stenfeldt K, Johansson C, Hellstrom S (2006) The collagen structure

of the tympanic membrane: collagen types I, II, and III in the

healthy tympanic membrane, during healing of a perforation, and

during infection. Arch Otolaryngol Head Neck Surg 132:

293–298

Stenfors L (1984) The tympanic membrane. Acta Otolaryngol 98:28

Taylor M, McMinn R (1967) Cytology of repair in experimental

perforation of the tympanic membrane and its relationship to

chronic perforations in man. Trans Am Acad Ophthalmol

Otolaryngol 71:802–812

Thiery J (2003) Epithelial-mesenchymal transitions in development

and pathologies. Curr Opin Cell Biol 15:740–746

Thiery J, Sleeman J (2006) Complex networks orchestrate epithelial-

mesenchymal transitions. Nat Rev Mol Cell Biol 7:131–142

Werner S (1998) Keratinocyte growth factor: a unique player in

epithelial repair processes. Cytokine Growth Factor Rev

9:153–165

Werner S, Peters K, Longaker M, Fuller-Pace F, Banda M, Williams L

(1992) Large induction of keratinocyte growth factor expression

in the dermis during wound healing. Proc Natl Acad Sci USA

89:6896–6900

Werner S, Krieg T, Smola H (2007) Keratinocyte-fibroblast interac-

tions in wound healing. J Invest Dermatol 127:998–1008

J Mol Hist

123

Chapter 3

Structure and properties of biomedical films prepared from aqueous and acidic silk fibroin

solutions

Rajkhowa R, Levin B, Redmond S, Li L, Wang L, Kanwar J, Atlas M, Wang X. Structure and Properties of Biomedical Films Prepared from Aqueous and Acidic Silk Fibroin Solutions. Journal of Biomedical Materials Research: Part A. 97A: 37-45; 2011.

Structure and properties of biomedical films prepared from aqueousand acidic silk fibroin solutions

Rangam Rajkhowa,1 Brett Levin,2 Sharon L. Redmond,2,3 Lu Hua Li,1 Lijing Wang,4

Jagat R. Kanwar,1 Marcus D. Atlas,2,3 Xungai Wang1

1Institute for Technology Research and Innovation, Deakin University, Geelong, Victoria, Australia2Ear Sciences Centre, School of Surgery, The University of Western Australia, Perth, Western Australia, Australia3Ear Science Institute Australia, Perth, Western Australia, Australia4School of Fashion and Textiles, RMIT University, 25 Dawson Street, Brunswick, Vic. 3056, Australia

Received 24 September 2010; revised 29 October 2010; accepted 3 November 2010

Published online 9 February 2011 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/jbm.a.33021

Abstract: Silk fibroin films are promising materials for a

range of biomedical applications. To understand the effects of

casting solvents on film properties, we used water (W), formic

acid (FA), and trifluoroacetic acid (TFA) as solvents. We char-

acterized molecular weight, secondary structure, mechanical

properties, and degradation behavior of cast films. Significant

degradation of fibroin was observed for TFA-based film com-

pared to W and TA-based films when analyzed by SDS–PAGE.

Fibroin degradation resulted in a significant reduction in ten-

sile strength and modulus of TFA-based films. Compared to

water, TFA-based films demonstrated lower water solubility

(19.6% vs. 62.5% in 12 h) despite having only a marginal

increase in their b-sheet content (26.9% vs. 23.7%). On the

other hand, FA-based films with 34.3% b-sheet were virtually

water insoluble. Following solubility treatment, b-sheet con-

tent in FA-based films increased to 50.9%. On exposure to

protease XIV, water-annealed FA-based films lost 74% mass

in 22 days compared to only 30% mass loss by ethanol

annealed FA films. This study demonstrated that a small vari-

ation in the b-sheet percentage and random coil conforma-

tions resulted in a significant change in the rates of enzymatic

degradation without alteration to their tensile properties. The

film surface roughness changed with the extent of enzymatic

hydrolysis. VC 2011 Wiley Periodicals, Inc. J Biomed Mater Res Part A:

97A: 37–45, 2011.

Key Words: silk fibroin film, tensile properties, degradation,

biomaterial

INTRODUCTION

Silk from domesticated silkworms Bombyx mori (B. mori)contains a fiber forming protein, fibroin, encased by silkgum, and other protective coatings.1 Silk fibroin is biocom-patible, particularly after removing the potentially immuno-genic nonfibroin proteins.2 Therefore, silk proteins havebeen engineered into a wide range of fibroin-based materi-als with diverse morphologies and mechanical properties.3,4

Furthermore, it is possible to regenerate fibroin with thedesired amount of crystallinity, resulting in control over therate of biodegradation.4,5 These advantages have sparkednew interest in the use of silk fibroin for biomedical appli-cations, including tissue engineering.3,6–8 Silk has beenextensively studied with respect to supporting a number ofdifferent animal and mammalian cells as well as for regen-eration of specific tissues.3,6,7

Silk film is a type of regenerated silk material. It hasoxygen and water vapor permeability9,10 and can be used asa graft for wound healing, tissue engineering, and degradablebiosensing applications.3,6,8,11–14 Animal studies have shownthat the performance of silk film was superior to some com-mercial wound-healing products.11,12 It also has the desiredcharacteristics for use as a substrate for engineering skin6

and corneas.15–18 We have recently examined silk fibroin filmsas a potential scaffold for ear drum [tympanic membrane(TM)] keratinocyte growth with the long-term aim of using itas a graft to repair chronic ear drum perforations.19–21

Silk films can be regenerated from aqueous,22,23 acidic,22,24,25

organic,26,27 and ionic28 solutions. Film-forming conditionssuch as the concentration of silk solution,29 tempera-ture,29,30 drying rate,31,32 and annealing methods4,33,34 playan important role in determining the structure and proper-ties of films. Published data indicate that the molecularweight distribution of regenerated silk could affect the pro-liferation of fibroblasts.35 Similarly, b-sheet content deter-mines the degradation rate and may influence cell-support-ing properties.4,5 Despite some studies on the effect ofsolvents used for dissolving native and reconstituted silk onfibroin secondary structure,25,36 further work is required tounderstand the implications of solvent systems on the struc-ture and functional properties of films and in turn their bio-logical applications. Such studies will help to establish theprocess-induced optimal conditions for tailoring the silkfilms for specific application requirements. This work exam-ined differences in molecular weight, secondary structure,mechanical properties, and degradation behavior of silk

Correspondence to: X. Wang; e-mail: xwang@deakin.edu.au

VC 2011 WILEY PERIODICALS, INC. 37

films cast under similar conditions from water (W), formicacid (FA), and trifluoroacetic acid (TFA).

MATERIALS AND METHODS

Silk fiber degummingMultivoltine B. mori silk cocoons with pupae removed weresourced from northeast India. Degumming was performedto remove sericin with a laboratory dyeing machine (Thies,USA) using 2 g/L sodium carbonate and 0.6 g/L sodiumdodecyl sulphate (Sigma-Aldrich, Australia) at 100�C with amaterial mass (g) to liquor volume (mL) ratio of 1:25. After20 min of degumming, the cocoons were washed thoroughlywith warm distilled water (dH2O) followed by cold dH2O.

Preparation of silk solutions and regenerated silk filmsDegummed silk fibers were dissolved in 10M lithium thio-cyanate (LiSCN) (Sigma-Aldrich) at room temperature (RT)for 2 h with a fiber mass (g) to liquor volume (mL) ratio of1:10. Undissolved matters in the solution were removed bycentrifugation (model CT15RT) (Techcomp, China) at 4800rpm (4170 � g rcf) for 10 min. The clear solution was dia-lyzed for 4 days at 4�C using dialysis sacks (molecularweight cut-off 12 kDa, from Sigma-Aldrich) against milli-Qwater, which was changed every 6–12 h. Following dialysis,silk solutions were again centrifuged to remove any undis-solved material. The final concentration of silk solutionswas determined by weighing the remaining solid after over-night drying at 60�C. The concentration of silk solution wasadjusted to 3% (w/v) by diluting with milli-Q water. Aportion of silk solution was kept for the preparation ofwater-based films, whilst the remaining solution was frozenat �80�C overnight and lyophilized for dry silk fibroin foamproduction. The foams were dissolved separately in 98% FA(Sigma-Aldrich) and �98% TFA (Sigma-Aldrich) to obtain3% fibroin solutions (w/v). Water (W), FA, and TFA-basedsilk solutions were poured into polyethylene discs (1 mL ina 25-mm diameter disc) and dried inside a chemical flowfume hood (covered to allow slow drying) until dry filmswere obtained. The films, �33 lm in thickness, were keptinside the fume hood for further 24 h for residual solventsto evaporate.

Dry films were divided into two parts: one part wasstored at RT under desiccation and the second annealed bysubmerging in 75% ethanol for 6 h to induce b-sheets torender it water insoluble. The treated films were washedthoroughly with milli-Q water to remove residual chemicals(ethanol/FA/TFA), dried at RT, and stored under desiccation.The nomenclatures used to define each type of film in thisstudy are presented in Table I.

Molecular weight distribution by SDS–PAGETo evaluate the changes in molecular weight distribution ofregenerated silk fibroin films due to solvents, SDS–PAGEmeasurements were performed on ethanol annealed films(W-E, FA-E, and TFA-E) having no residual solvents. Proteinsamples for SDS–PAGE analysis were prepared by dissolvingfilms in 10M LiSCN (Sigma-Aldrich) at RT for 2 h, followedby dialysis, and finally a dilution in milli-Q water to obtain a

1% (w/v) water-based silk solution. Fifty microliters of testsample were loaded onto the gel (NH11-420, Nu Sep,Australia), along with 5-lL broad range prestained mole-cular weight standard (Bio-Rad Laboratories, Australia), forthe estimation of protein molecular weights. BG-165 (NuSep, Australia) and 1� Tris–HEPES–SDS were used as thesample buffer and running buffer, respectively, for SDS–PAGE analysis. A constant 150 V and a starting current of120 mA were applied across the gel for 45 min to allowadequate protein band separation. Gels were subsequentlystained with Commassie Blue R-250 (Sigma-Aldrich) for12 h at RT followed by destaining (40% ethanol and 10%acetic acid) for 12 h. Gel images were captured in a BioRadUniversal Hood II (BioRad Laboratories).

Dissolution of silk filmsThe desiccated silk films (�50 mg each) were accuratelyweighed, and each film was immersed in 5-mL milli-Q waterand stored in a 37�C incubator (T1-150 G, ThermolineScientific, Australia) for 12 h. Films were gently washed inmilli-Q water and dried at RT. Following desiccation for 24 h,films were weighed to determine percentage loss in weight.Four specimens were measured for each type of silk film,and their mean and standard deviations were calculated.

In vitro enzymatic degradationDegradation of the silk fibroin films was evaluated usingprotease XIV from Streptomyces griseus (Sigma-Aldrich) withan activity of 4.5 U/mg. Each desiccated silk film sample of�50 mg was weighed and treated with 1 mg of proteasedissolved in 1 mL of 0.1M phosphate buffer (pH 7.4). Phos-phate buffer (pH 7.4) was used as the control solution. Thefilm (g) to liquid (mL) ratio was 1:150, and the sampleswere incubated at 37�C at 50 rpm in an orbital shaker(Thermoline Scientific). Fresh enzyme solutions were usedevery 2 days to maintain optimal enzymatic activity. Depre-dated films were collected at 2, 8, 16, and 22 days, washedthoroughly with milli-Q water, dried at RT, desiccated for 24h, and weighed. The percentage loss in mass of each silkfilm type (n ¼ 4) was calculated.

Tensile property measurementsTensile tests were performed on a Universal Testing Instru-ment (Lloyd Instruments, UK) equipped with 100 N capacityload cell and analysis software, LrLrxCon (n ¼ 15). Drytests were performed after conditioning the specimens at

TABLE I. Nomenclature of Silk Films Outlining the Solvents

and Annealing Agents Used in Their Preparation

Name of Films Solvent Used Annealed by

W Water –W-E Water EthanolTFA Trifluoroacetic acid –TFA-E Trifluoroacetic acid EthanolFA Formic acid –FA-E Formic acid EthanolFA-W Formic acid Water

38 RAJKHOWA ET AL. STRUCTURE AND PROPERTIES OF BIOMEDICAL FILMS

20�C 6 2�C and 65% 6 2% relative humidity for 48 h. Wettests were conducted after the specimens were immersed inmilli-Q water for 24 h in the same conditioning room usedfor conditioning dry test specimens. All tests were per-formed using a displacement rate of 15 mm/min and agauge length of 15 mm. Breaking stress and Young’s modu-lus were calculated by the LrLrxCon and was based on theload elongation data generated during testing. Additionally,the cross-sectional area was calculated based on specimenwidth (5 mm) and nominal thickness, which was �33 lm.The thickness of each film type was determined by SEMobservations.

FTIR analysisInfrared spectra of desiccated silk films were recorded withan attenuated total reflectance Fourier transform infrared(FTIR) spectrophotometer (Vertex 70) (Bruker BiosciencesPty, Australia). Each spectrum was obtained in absorbancemode in the range of 4000–600 cm�1. Both sides werescanned twice on each silk film, hence an average of fourscans was used in each analysis. Fourier self-deconvolution(FSD) spectrum provides better resolution than originalFTIR peaks and results in minimum uncertainty duringcurve fitting.37 Hence to measure different conformations,average spectrum in the amide I mode (1595–1705 cm�1)was deconvoluted, and curve was fitted using OPUS 5.5 soft-ware adapting the procedure used by Hu et al.37 with slightmodifications. Briefly, a baseline correction was initiallyperformed for the amide I region. Deconvolution was thencarried out adapting a Lorentzian model using a bandwidthof 25 cm�1 and a noise reduction factor of 0.3. A straightbase line correction of FSD spectrum was performed againfollowed by curve fitting using a Gaussian model. Duringcurve fitting, 11 fixed band positions (1611, 1619, 1624,1630, 1640, 1650, 1659, 1666, 1680, 1691, and 1698 cm�1)were initially used. Band width and positions were thenautomatically adjusted by the autofit program using aLevenberg–Marquardt algorithm. Finally, each individualspectrum was area normalized to obtain percentage confor-mations within the amide I region. The procedure wasrepeated four times for each spectrum.

Atomic Force Microscopy ImagingA Cypher atomic force microscopy (AFM; Asylum Research,USA) was used to obtain height images in tapping mode inair with a silicon cantilever, 125 lm in length, and a forceconstant of 40 N/m. The scan size was 1 lm � 1 lm, andthe scan rate was 0.78 Hz.

RESULTS AND DISCUSSION

SDS–PAGE analysisSDS–PAGE analysis was useful for revealing qualitativechanges in molecular weights during silk film processing. Itwas expected that fibroin with high alanine and glycine con-tents would show some migration during electrophoresiswhen compared with normal globular proteins. The electro-phoretic pattern for the W-E film appears to show somesmearing down to �110 kDa (Fig. 1). In the case of TFA-E

films, molecular weight distribution widened significantlywith the pattern migrating to �70 kDa. Despite visiblesmearing of FA-E films when compared with W-E, the elec-trophoretic pattern was still appreciably less than TFA-E.The estimated molecular weight of the silk fibroin heavychain in the silkworm’s gland is 391 kDa.38–41 However, thedegumming and dissolution processes are responsible forcreating polydispersion and associated smearing in electro-phoretic patterns. LiSCN used in this study is known tocause very low degradation in silk compared to other chaot-ropic solvents.42 Therefore, the smearing in W-E is largelyattributed to degumming conditions. We recently observedsuch smearing in the case of aqueous silk solutions beforefilm formation.43 Thus, the additional increase in molecularweight distribution in TFA-E, compared to W-E, can beattributed to degradation of silk fibroin by TFA. Comparedto W-E, FA did not cause additional change in molecularweight distribution of the fibroin heavy chain. Figure 1 dem-onstrates that apart from the smeared band of the fibroinheavy chains, a �30 kDa band, which is most likely the lightchain, appears noticeable in W-E and FA-E but very weak inTFA-E. A reduction in the intensity of the �30 kDa band inthe TFA lane indicates rupture of the light chains due toTFA treatment. Further studies are warranted to shed morelight on the mechanisms of hydrolysis and degradation offibroin due to different types of acidic solvents.

Conformational change in silk filmsFTIR amide I mode peaks (vibration due to C¼¼O stretching)are good indicators of the conformation of proteins. Thisoccurs because the frequency of vibration of C¼¼O in theprotein backbone changes due to the influence of hydrogenbonds between NAH and C¼¼O, which are dependant on

FIGURE 1. SDS–PAGE analysis of silk films.

ORIGINAL ARTICLE

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | APR 2011 VOL 97A, ISSUE 1 39

conformations adopted by protein chains. Moreover, spectrain this region are not significantly influenced by the proteinside chains and are hence suitable for conformationalassignments.37 Specifically, for B. mori fibroin, Tretinnikovand Tamada attributed FTIR peaks at 1619, 1626, and 1631cm�1 to b-sheets, 1641 and 1649 cm�1 to random coil, andthe peak at 1659 cm�1 to a-helix conformations.23 Chenet al.44,45 used peaks around 1620 and 1691 cm�1 forb-sheets and b-turns, respectively, and 1652 cm�1 for ran-dom coil/a-helix conformations for measuring ethanol-induced conformational changes of B. mori films. On the basisof the second derivative plots, they attributed a peak around1968 cm�1 also to a-helix conformation.46 Reviewing litera-ture on protein structures, Hu et al.37 used bands between1616–1637 cm�1 and 1697–1703 cm�1 for b-sheets, 1638–1655 cm�1 for random coil, 1656–1662 cm�1 for a-helix,and 1663–1696 cm�1 for b-turn conformations.

Amide I peaks of silk films before and after ethanoltreatment are presented in Figure 2 with the originalabsorption plots being displaced vertically without changingtheir scales. The second derivative spectra for W, FA, andW-E films are presented in Figure 3. Plots of other ethanol-treated films are similar to W-E; hence, are not shown inFigure 3. When comparing the plots ‘‘W’’ and ‘‘W-E’’ (Fig. 3),it is evident that after ethanol treatment, the component forthe water-based film at 1640 cm�1 disappeared and that at1648 cm�1 was reduced significantly (both assigned torandom coil). Conversely, the component at 1693 cm�1

(b-turn) shifted to 1699 cm�1 (b-sheet). Ethanol treatmentalso resulted in a prominent second derivative componentat 1623 cm�1 (b-sheet). On the other hand, Figure 3 alsoshows that FA films had an intermediate structure withcomponents at 1648 and 1693 cm�1 along with 1623 and1699 cm�1, respectively. There was no difference in spectraat 1663 and 1680 cm�1 for all films.

Second derivative plots in Figure 3 were obtained usingnine-point smoothening and corresponded to the conforma-tion-based peak assignments used by Hu et al.37 duringcurve fitting. Hence, we followed their procedure for curvefitting and estimation of percentage conformations in this

study. The curve-fitted amide I FSD spectra are presented inFigure 4. Table II lists the percentage of different conforma-tions derived from area normalized data. The b-sheetcontents of as cast W, TFA, and FA films are 23.7, 26.9, and34.3%, respectively. This suggests a significantly high contentof b-sheets (silk II) in FA-based films even before ethanoltreatment. As expected, irrespective of the solvents used forsilk solution preparation, all films annealed by ethanol showmore than 50% b-sheet conformation. It is also evident fromTable II that on annealing, a-helix (a hydrated crystal struc-ture, also known as silk I) in all annealed films reduced to8–9%. The percent reduction of a-helix was sharper than thedecrease in random coil conformation from annealing, indi-cating that conformation transition to b-sheets was dominantin the helical segments of the protein chains. This agreeswell with Peng et al.47 who reported a faster conformationaltransition from a-helix into b-sheets when compared withrandom coil into b-sheets for spider silk protein. Uponannealing, barring W films, b-turn contents were alsoenhanced. The side-chain component (peaks 1595–1615cm�1) in the amide I regions for all films was less than 1.5%.

The ability of ethanol to induce b-sheets in W-basedfilms of B. mori silk fibroin has been demonstrated previ-ously.45,46 This transition was rapid for ethanol concentra-tions between 70 and 80%.44 Similar concentrations (75%)were used in this study, hence substantial conformationaltransition as shown after ethanol treatment is expected.Previously, Ha et al.25 revealed the presence of b-sheets inFA-based films, and a combination of random coil andb-sheets in films cast from fibroin dissolved in TFA. Simi-larly, Um et al.22 reported that FA-based films developedhigh-b-sheet fractions before alcohol treatment and that nofurther enhancement was found based on the amide IIImode of FTIR measurement after alcohol annealing. How-ever, in contrast to these previous studies, the present studyindicates that b-sheet fraction in TFA-based films was quitesimilar to W-based films. The FA plot (Fig. 3) is quite

FIGURE 2. FTIR scans of silk films. [Color figure can be viewed in the

online issue, which is available at wileyonlinelibrary.com.]

FIGURE 3. Second derivatives of amide I FTIR spectra of silk films.

[Color figure can be viewed in the online issue, which is available at

wileyonlinelibrary.com.]

40 RAJKHOWA ET AL. STRUCTURE AND PROPERTIES OF BIOMEDICAL FILMS

similar to the patterns formed during the initial phase(<0.1 min) of ethanol induced conformational transition ofwater-based films as reported by Chen et al.46 Our studyconfirms that the crystalline structure in FA-based films ismetastable, and crystallinity can be substantially enhanced bysubsequent ethanol annealing. The type and amount of crys-tal domains (b-sheet and b-turns) depend on the mechanismof hydration of hydrophobic domains during dissolution bysolvents before crystal formation.29 In addition, the rate atwhich polypeptide segments self assemble during the dehy-dration process also determines crystal structure.31,45,48

Differences in solvation and/or drying kinetics may beresponsible for increased b-sheet content in FA films.

The presence of b-sheets before ethanol treatment inFA-based films enhanced their water stability and allowedthe films to be immersed in water. Table II also shows that

following water immersion of FA films for 12 h, the b-sheetand b-turn contents increased from 34.3 to 50.9% and 7.5to 14.4%, respectively. The FTIR spectra of FA-E and FA-Ware very similar (Fig. 2). However, the b-sheet contentderived from deconvoluted spectra is marginally lower inFA-W compared to FA-E films (50.9 vs. 55.8%). The differ-ence is statistically significant (p < 0.05, Student’s t test).The ability of water to induce b-sheets in regeneratedfibroin film has been reported previously.4,5,34,49 However,no previous studies have shown such high (�50%) b-sheetcontent through water annealing.

Tensile properties of silk filmsDry regenerated silk products after induction of b-sheetsbecome brittle and have poor extensibility. For example, thepercentage breaking strain of silk films in its dry state was

FIGURE 4. FSD spectra of amide regions of silk films.

TABLE II. Percentage Contents of Different Conformations in the Silk Films Measured from FSD Spectra of the Amide I Region

(Derived From Area Normalized Data)

Silk Film Side Chain b-Sheet Random Coil a-Helix b-Turn

W 0.61 6 0.15 23.75 6 0.87 24.80 6 0.51 37.54 6 0.74 13.29 6 0.53W-E 0.53 6 0.22 50.97 6 0.60 26.36 6 2.16 8.96 6 2.39 13.18 6 1.41TFA 1.49 6 0.34 26.93 6 0.20 41.37 6 .20 21.22 6 0.28 8.98 6 0.46TFA-E 0.14 6 0.14 53.15 6 0.52 22.00 6 0.68 8.31 6 1.49 16.44 6 0.77FA 0.58 6 0.00 34.32 6 0.01 39.38 6 0.33 18.16 6 0.32 7.57 6 0.02FA-E 0.55 6 0.34 55.87 6 1.11 22.13 6 0.95 8.79 6 1.83 12.66 6 1.18FA-W 0.00 6 0.00 50.91 6 0.29 25.94 6 0.64 8.72 6 0.76 14.44 6 0.48

ORIGINAL ARTICLE

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | APR 2011 VOL 97A, ISSUE 1 41

�5% or less.50 Mechanical properties in wet conditions aremore relevant to biomedical applications. Therefore, in thisstudy, we have focused on tabulating the wet tensile testresults (Table III). For comparison, Table III includes drytest data for FA-E and data from literature.

The results in Table III show that there is no significantdifference in tensile properties (p > 0.05, two-tailedStudent’s t-test) between FA-E and W-E, despite of theirdifference in b-sheet content (Table II). FA-E had a highermean percentage breaking strain than W-E (p < 0.05). FAdid not adversely affect tensile properties of FA-E film com-pared to aqueous processing (W-E), which was also sup-ported by SDS–PAGE analysis. Conversely, strength, modulus,and percentage strain dropped notably in TFA-E films com-pared to W-E and FA-E films. This confirms the occurrenceof protein degradation of TFA films during TFA treatment, aresult supported by SDS-PAGE analysis. FA-E films in thedry state were found to have almost three times higher ten-sile strength and nearly 10 times higher Young’s modulusthan in the wet state, but their strain dropped significantlyfrom 136% to 6.4%. Data from the literature, presented inTable III for dry films, indicated even lower extensibility,although these films were unannealed.24,51

Unlike films, the mechanical properties of native silkfibers do not show large differences between dry and wetconditions. For example, we found that the tensile strengthof degummed B. mori fibers in the dry state was 745.4 MPa.This was reduced to 665.7 MPa when tested under wet con-dition (based on calculations from earlier reportedresults52). It was reported that the breaking strain of nativesilk fibers remains at �15–25% in both dry and wet condi-tions.53,54 These results may explain the pronounced effectof water on the tensile properties of silk films. Water caneasily alter the regenerated silk fibroin structure leading tolarge changes in the stress–strain profiles of silk films.Results are consistent with the general perception thatdespite the development of b-sheets in regenerated silkproducts, their mechanical properties are inferior to the silkfibers.33,55–58 However, a few studies have shown that thereis scope to significantly improve the mechanical propertiesof regenerated silk products through manipulation of struc-ture during regeneration.57,59,60 Further work is warrantedon silk structure manipulation during regeneration, so thatfilms with appropriate mechanical properties suitable fordifferent applications can be obtained.

Dissolution of silk filmsAqueous and acid-based films after ethanol annealing arewater resistant due to the dominant silk II (b-sheet) hydro-phobic structure. However, differences in the secondarystructure of films before annealing, as reflected in the FTIRstudies (Figs. 2–4 and Table II), resulted in variations intheir water solubility. W-based films lost 62.5% mass in12 h at 37�C (Table IV) and as a result disintegrated. Com-paratively, TFA films showed better water resistance,whereas FA films were virtually insoluble in water. Thepresence of higher contents of b-sheets in TFA-based films(Fig. 4 and Table II) prevented easy hydration of hydropho-bic domains of fibroin chains and thus induced some stabil-ity against water. In the case of FA films, the presence of34.3% b-sheets, although less than annealed films (51 and56%), was sufficient to provide the required stabilityagainst dissolution. Insignificant losses in mass of FA-basedfilms further proved that the reduction in random coil anda-helix and simultaneous increase in b-sheets and b-turns(Fig. 4 and Table II) during water treatment was due toconformational changes to silk II arising from water anneal-ing, rather than the solubilization of hydrophilic domains.

Enzymatic degradationA predictable degradation is highly desirable for applyingbiomaterials to wound healing and tissue engineering.In vitro enzymatic hydrolysis provides a general idea of thebiodegradability of a material. Previous studies revealedthat protease enzymes biodegrade silk faster than a-chymo-trypsin, collagenase IA,61,62 or trypsin.24 Accordingly, prote-ase XIV was selected for the investigation in this study.

Figure 5 shows that all films exhibited progressive lossof mass by enzymatic hydrolysis over time. The differencesbetween TFA-E and FA-E were not significant (p > 0.05,two-tailed Student’s t-test) at any of the time points.Conversely, W-E films lost significantly higher mass at allfour time points. As FA films before ethanol treatment were

TABLE III. Tensile Properties of Silk Films (mean 6 SD); n ¼ 15

Silk Film Tensile Strength (MPa) Young’s Modulus (MPa) Strain (%) Reference

W-E 15.8 6 3.6 311.2 6 40.6 105.7 6 34.5 This studyTFA-E 7.8 6 1.6 214.2 6 44.8 55.6 6 38.3 This studyFA-E 14.8 6 4.8 280.2 6 53.4 136.6 6 35.0 This studyFA-Ea 43.6 6 6.9 2911.0 6 501.9 6.4 6 1.4 This studyWa 23.7 6 7.3 3411 6 479 1.7 6 0.1 22FAa 23.6 6 7.7 2893 6 387 1.4 6 0.1 22Wa 58.8 6 16.7 – 2.1 6 0.4 48W 3.5 6 0.7 – 127.8 6 69.4 48

a Specimens tested in dry state.

TABLE IV. Solubility of Untreated Silk Films at 37�C for 12 h

(mean 6 SD); n ¼ 4

Silk Film % Loss in Mass

W 62.5 6 2.6TFA 19.6 6 1.6FA 2.8 6 0.9

42 RAJKHOWA ET AL. STRUCTURE AND PROPERTIES OF BIOMEDICAL FILMS

resistant to dissolution (Table IV), such films treated withwater (FA-W) were also used for enzymatic degradationstudy. These demonstrated a higher degradation rate

compared to TFA-E and FA-E at all time points. Mass loss ofFA-W was 66 and 74% at days 16 and 22 compared to 40and 58%, respectively, for the W-E films. In contrast, TFA-Eand FA-E lost only about 30% mass at day 22. All specimenstreated with the control solution of 0.1M PBS (pH 7.4) with-out enzymes did not show any perceptible change in massduring this period, thereby confirming that the mass losswas due solely to enzymatic hydrolysis.

It was observed that in the course of enzyme treatment,film surfaces were etched out gradually without disintegrat-ing films. Etched particles were clearly visible on the filmsurfaces. The AFM topography results (Fig. 6) clearly dem-onstrate that the roughness of films increased with enzymetreatment and the subsequent change in FA-W films werepronounced. It is important to note that there was no differ-ence in the topography of FA films with or without ethanoltreatment before enzyme treatment. Similar surface con-tours were observed for W-E and TFA-E films (results notshown). Hence, the increased degradation was unlikely dueto any difference in the surface roughness of the films atthe start of the enzyme treatment.

Protease from S. griseus used in this study is known tocleave aggressively and nonspecifically the amorphous

FIGURE 5. Silk film mass loss due to enzymatic degradation. [Color

figure can be viewed in the online issue, which is available at

wileyonlinelibrary.com.]

FIGURE 6. AFM images of FA-based silk films before and after protease XIV treatment. [Color figure can be viewed in the online issue, which is

available at wileyonlinelibrary.com.]

ORIGINAL ARTICLE

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | APR 2011 VOL 97A, ISSUE 1 43

domains of B. mori silk fibroin.62 The higher percentage ofrandom coil in W-E and FA-W (around 26%) compared toFA-E and TFA-E (around 22%) probably plays a role in deg-radation behavior. The resistance to enzymatic degradation(Fig. 5) also followed the trend of b-sheet content in Table II,that is, FA-E > TFA-E > W-E > FA-W. However, differences inthe scale of degradation are much larger than their differen-ces in b-sheet and random coil contents. A previous study byJoon et al. and Le et al. showed higher enzymatic degradationfor water-based water annealed films and slowly dried water-based films, respectively, compared to methanol annealing.4,48

However, their study did not show a sharp b-sheet peak asreflected in this study for FA-W. Further studies will berequired to characterize more precisely the change in molecu-lar and secondary structure due to solvent treatment toaccount for such large mass losses during enzymatic degrada-tion from films having more than 50% b-sheet contents.

Specifications of mechanical properties and biodegrada-tion of grafts and scaffold materials depend on applications.This study has demonstrated that such biomechanical prop-erties of silk fibroin films can be altered depending uponneeds through manipulation of secondary structure usingdifferent solvents. For example, the higher rate of degrada-tion of water annealed FA films in enzymes (Fig. 5), theirinsolubility in water (Table IV), and the lack of fibroin degra-dation during FA treatment (Figs. 5 and 6) may be importantfor applications where quick biodegradation and high me-chanical properties are important. We have recently pub-lished our preliminary results of FA-based films to supportthe growth of human TM keratinocytes.21 Further studiesinvolving cell viability and cytotoxicity are in progress.Despite higher safety requirements in the use of FA, previousstudies also demonstrated that silk products treated with FAsupported tissue regeneration in vitro63 and in vivo.64

CONCLUSION

Degradation of silk fibroin during TFA processing wasevident from SDS–PAGE studies. It resulted in significantreduction in tensile properties of the silk films. Conversely,the molecular weight distribution and tensile properties ofFA-based films were similar to water-based silk films. More-over, higher b-sheet content in FA-based films provided ametastable form resisting dissolution in water. In addition,their b-sheet content substantially increased during watertreatment, confirming that FA as a solvent can be used with-out alcohol annealing to prepare stable silk films. Such filmscould be degraded faster by proteolytic enzymes comparedto alcohol-annealed films. The rate of degradation of filmsincreased with an increase in random coil and decrease inb-sheet contents, but differences noted in degradation weremuch higher than the changes that occurred in conforma-tions. However, differences in secondary structure did notaffect the tensile properties of the silk films.

ACKNOWLEDGMENTS

Authors acknowledge Dr. Eun Seok Gil from Department ofBiomedical Engineering, Tufts University for his advice onFTIR analysis.

REFERENCES

1. Iizuka E. Conformation of silk sericine in solution, Bombyx mori

L. Biochim Biophys Acta: Protein Struct 1969;181:477–479.

2. Altman GH, Diaz F, Jakuba C, Calabro T, Horan RL, Chen J, Lu H,

Richmond J, Kaplan DL. Silk-based biomaterials. Biomaterials

2003;24:401–416.

3. Vepari C, Kaplan DL. Silk as a biomaterial. Progr Polym Sci 2007;

32:991–1007.

4. Jin H-J, Park J, Karageorgiou V, Kim U-J, Valluzzi R, Cebe P,

Kaplan DL. Water-stable silk films with reduced b-sheet content.

Adv Funct Mater 2005;15:1241–1247.

5. Min B-M, Jeong L, Lee KY, Ho Park W. Regenerated silk fibroin

nanofibers: Water vapor-induced structural changes and their

effects on the behavior of normal human cells. Macromol Biosci

2006;6:285–292.

6. Wang Y, Kim H-J, Vunjak-Novakovic G, Kaplan DL. Stem cell-

based tissue engineering with silk biomaterials. Biomaterials

2006;27:6064–6082.

7. Hardy JG, Romer LM, Scheibel TR. Polymeric materials based on

silk proteins. Polymer 2008;49:4309–4327.

8. Omenetto F, Kaplan D. New opportunities for an ancient material.

Science 2010;329:528.

9. Minoura N, Tsukada M, Nagura M. Physico-chemical properties of

silk fibroin membrane as a biomaterial. Biomaterials 1990;11:

430–434.

10. Minoura N, Tsukada M, Nagura M. Fine structure and oxygen per-

meability of silk fibroin membrane treated with methanol. Poly-

mer 1990;31:265–269.

11. Sugihara A, Sugiura K, Morita H, Ninagawa T, Tubouchi K, Tobe

R, Izumiya M, Horio T, Abraham NG, Ikehara S. Promotive effects

of a silk film on epidermal recovery from full-thickness skin

wounds. Proc Soci Exp Biol Med 2000;225:58–64.

12. Kimura T, Yamada H, Kohzoh T, Doi K. Accelerating effects of silk

fibroin on wound healing in hairless descendants of Mexican

hairless dogs. J Appl Sci Res 2007;3:1306–1314.

13. Domachuk P, Perry H, Amsden J, Kaplan D, Omenetto F. Bioactive

‘‘self-sensing’’ optical systems. Appl Phys Lett 2009;95:253702.

14. Kundu J, Mohapatra R, Kundu SC. Silk fibroin/sodium carboxy-

methyl cellulose blended films for biotechnological applications.

J Biomater Sci Polym Ed 2011;22:519–539.

15. Gil ES, Mandal BB, Park S-H, Marchant JK, Omenetto FG, Kaplan

DL. Helicoidal multi-lamellar features of RGD-functionalized silk

biomaterials for corneal tissue engineering. Biomaterials 2010;31:

8953–8963.

16. Chirila TV, Barnard Z, Zainuddin, Harkin DG, Schwab IR, Hirst LW.

Bombyx mori silk fibroin membranes as potential substrata for

epithelial constructs used in the management of ocular surface

disorders. Tissue Eng Part A 2008;14:1203–1211.

17. Gil ES, Park SH, Marchant J, Omenetto F, Kaplan DL. Response of

human corneal fibroblasts on silk film surface patterns. Macromol

Biosci 2010;10:664–673.

18. Lawrence BD, Marchant JK, Pindrus MA, Omenetto FG, Kaplan

DL. Silk film biomaterials for cornea tissue engineering. Biomate-

rials 2009;30:1299–1308.

19. Ghassemifar R, Redmond S, Zainuddin, Chirila TV. Advancing

towards a tissue-engineered tympanic membrante: silk fibroin as

a substratum for growing human eardrum keratinocyte. J Bio-

mater Appl 2010;24:591–606.

20. Levin B, Rajkhowa R, Redmond SL, Atlas MD. Grafts in myring-

plasty: Utilizing a silk fibroin scaffold as a novel decice. Exp Rev

Med Dev 2009;6:653–664.

21. Levin B, Redmond SL, Rajkhowa R, Eikelboom RH, Marano RJ,

Atlas MD. Preliminary results of the application of a silk fibroin

scaffold to otology. Otolaryngology Head Neck Surgery 2010;142:

S33–S35.

22. Um IC, Kweon H, Park YH, Hudson S. Structural characteristics

and properties of the regenerated silk fibroin prepared from for-

mic acid. Int J Biol Macromol 2001;29:91–97.

23. Tretinnikov ON, Tamada Y. Influence of casting temperature on

the near-surface structure and wettabiligy of cast silk fibroin films.

Langmuir 2001;17:7406–7413.

44 RAJKHOWA ET AL. STRUCTURE AND PROPERTIES OF BIOMEDICAL FILMS

24. Vasconcelos A, Freddi G, Cavaco-Paulo A. Biodegradable materi-

als based on silk fibroin and keratin. Biomacromolecules 2008;9:

1299–1305.

25. Ha SW, Tonelli AE, Hudson SM. Structural studies of Bombyx

mori silk fibroin during regeneration from solutions and wet fiber

spinning. Biomacromolecules 2005;6:1722–1731.

26. Meinel L, Hofmann S, Karageorgiou V, Kirker-Head C, McCool J,

Gronowicz G, Zichner L, Langer R, Vunjak-Novakovic G, Kaplan

DL. The inflammatory responses to silk films in vitro and in vivo.

Biomaterials 2005;26:147–155.

27. Rajkhowa R, Gil ES, Kludge JA, Numata K, Wang L, Wang X,

Kaplan DL. Reinforcing silk scaffolds with silk particles. Macromol

Biosci 2010;10:599–611.

28. Gupta MK, Khokhar SK, Phillips DM, Sowards LA, Drummy LF,

Kadakia MP, Naik RR. Patterned silk films cast from ionic liquid

solubilized fibroin as scaffolds for cell growth. Langmuir 2007;23:

1315–1319.

29. Lv Q, Cao C, Zhang Y, Ma X, Zhu H. Preparation of insoluble fi-

broin films without methanol treatment. J Appl Polym Sci 2005;

96:2168–2173.

30. Magoshi J, Nakamura S. Physical properties and structure of silk.

VIII. Effect of casting temperature on conformation of wild silk fi-

broin films. J Polym Sci Polym Phys Ed 1985;23:227–229.

31. Tsukada M. Effect of the drying rate on the structure of tussah

silk (antheraea pernyi) fibroin. J Polym Sci Polym Phys Ed 1986;

24:457–460.

32. Wang X, Kim HJ, Peng X, Matsumoto A, Kaplan DL. Biomaterial

coatings by stepwise deposition of silk fibroin. Langmuir 2005;21:

11335–11341.

33. Zuo B, Liu L, Wu Z. Effect on properties of regenerated silk fibroin

fiber coagulated with aqueous methanol/ethanol. J Appl Polym

Sci 2007;106:53–59.

34. Kawahara Y, Furukawa K, Yamamoto T. Self-expansion behaviour

of silk fibroin film. Macromol Mater Eng 2006;291:458–462.

35. Tsubouchi K, Nakao H, Igarashi Y, Takasu Y, Yamada H. Bombyx

mori fibroin enhanced the proliferation of cultured human skin

fibroblast. J Insect Biotechnol Sericol 2003;72:65–69.

36. Chen X, Knight DP, Shao Z, Vollrath F. Regenerated Bombyx silk

solutions studied with rheometry and FTIR. Polymer 2001;42:

09969–09974.

37. Hu X, Kaplan D, Cebe P. Determining beta-sheet crystallinity in fi-

brous proteins by thermal analysis and infrared spectroscopy.

Macromolecules 2006;39:6161–6170.

38. Shimura K, Kikuchi A, Ohtomo K, Katagata Y, Hyodo A. Studies

on silk fibroin of Bombyx mori. I. Fractionation of fibroin pre-

pared from the posterior silk gland. J Biochem 1976;80:693–702.

39. Tanaka K, Inoue S, Mizuno S. Hydrophobic interaction of P25,

containing ASN-linked oligosaccharide chains, with the H-L com-

plex of silk fibroin produced by Bombyx mori. Insect Biochem

Mol Biol 1999;29:269–276.

40. Zhou C-Z, Confalonieri F, Jacquet M, Perasso R, Li Z-G, Janin J.

Silk fibroin: structural implications of a remarkable amino acid

sequence. Proteins: Struct Funct Genet 2001;44:119–122.

41. Zhou C-Z, Confalonieri F, Medina N, Zivanovic Y, Esnault C, Yang

T, Jacquet M, Janin J, Duguet M, Perasso R and others. Fine or-

ganization of Bombyx mori fibroin heavy chain gene. Nucl Acids

Res 2000;28:2413–2419.

42. Yamada H, Nakao H, Takasu Y, Tsubouchi K. Preparation of unde-

graded native molecular fibroin solution from silkworm cocoons.

Mater Sci Eng C 2001;14:41–46.

43. Rajkhowa R, Wang L, Kanwar J, Wang X. Molecular weight and

secondary structure change in eri silk during alkali degumming

and powdering. J Appl Polym Sci 2011;119:1339–1347.

44. Chen X, Shao Z, Knight DP, Vollrath F. Conformation transition

kinetics of Bombyx mori silk protein. Proteins Struct Funct Bioinf

2007;68:223–231.

45. Chen X, Knight DP, Shao Z. b-turn formation during the confor-

mation transition in silk fibroin. Soft matter 2009;5:2777–2781.

46. Chen X, Shao Z, Marinkovic NS, Miller LM, Zhou P, Chance MR.

Conformation transition kinetics of regenerated Bombyx mori silk

fibroin membranne monitored by time-resolved FTIR spectros-

copy. Biophys Chem 2001;89:25–34.

47. Peng X, Shao Z, Chen X, Knight DP, Wu P, Vollrath F. Further

investigation on potassium-induced conformation transition of

nephila spidroin film with two-dimensional infrared correlation

spectroscopy. Biomacromolecules 2005;6:302–308.

48. Lu Q, Hu X, Wang X, Kluge JA, Lu S, Cebe P, Kaplan DL. Water-

insoluble silk films with silk I structure. Acta Biomater 2010;6:

1380–1387.

49. Wenk E, Wandrey AJ, Merkle HP, Meinel L. Silk fibroin spheres as

a platform for controlled drug delivery. J Controll Release 2008;

132:26–34.

50. Acharya C, Ghosh SK, Kundu SC. Silk fibroin film from non-mul-

berry tropical tasar silkworms: A novel substrate for in vitro fibro-

blast culture. Acta Biomater 2009;5:429–437.

51. Lu S, Wang X, Lu Q, Zhang X, Kluge JA, Uppal N, Omenetto F,

Kaplan DL. Insoluble and flexible silk films containing glycerol.

Biomacromolecules 2010;11:143–150.

52. Rajkhowa R, Wang L, Wang X. Ultra-fine silk powder preparation

through rotary and ball milling. Powder Technol 2008;185:87–95.

53. Das S, Chattopadhya R, Gulrajani ML, Sen K. Study of property

and structural varients of mulberry and tasar silk filaments.

AUTEX Res J 2005;5:81–86.

54. Morton WE, Hearle JWS. Physical Properties of Textile Fibres,

3rd ed. Cambridge, UK: Woodhead Publishing Limited 1993.

55. Iridag Y, Kazanc M. Preparation and characterization of Bombyx

mori silk fibroin and wool keratin. J Appl Polym Sci 2006;100:

4260–4264.

56. Freddi G, Tsukada M, Bertta S. structure and physical properties

of silk fibroin/polyacrylamide blend films. J Appl Polym Sci 1999;

71:1563–1571.

57. Ha S-W, Gracz HS, Tonelli AE, Hudson SM. Structural study of

irregular amino acid sequences in the heavy chain of Bombyx

mori silk fibroin. Biomacromolecules 2005;6:2563–2569.

58. Zuo B, Dai L, Wu Z. Analysis of structure and properties of biode-

gradable regenerated silk fibroin fibers. J Mater Sci 2006;41:

3357–3361.

59. Jiang C, Wang X, Gunawidjaja R, Lin Y-H, Gupta MK, Kaplan DL,

Naik RR, Tsukruk VV. Mechanical properties of robust ultrathin

silk fibroin films. Adv Funct Mater 2007;17:2229–2237.

60. Zhou G, Shao Z, Knight DP, Yan J, Chen X. Silk fibres extruded

artificially from aqueous solutions of regenerated Bombyx mori

silk fibroin are tougher than their natural counterpart. Adv Mater

2009;21:366–370.

61. Li M, Ogiso M, Minoura N. Enzymatic degradation behavior of po-

rous silk fibroin sheets. Biomaterials 2003;24:357–365.

62. Arai T, Freddi G, Innocenti R, Tsukada M. Biodegradation of

Bombyx mori silk fibroin fibers and films. J Appl Polym Sci 2004;

91:2383–2390.

63. Min B-M, Lee G, Kim SH, Nam YS, Lee TS, Park WH. Electrospin-

ning of silk fibroin nanofibers and its effect on the adhesion and

spreading of normal human keratinocytes and fibroblasts in vitro.

Biomaterials 2004;25:1289–1297.

64. Ghanaati S, Orth C, Unger RE, Barbeck M, Webber MJ, Motta A,

Migliaresi C, Kirkpatrick CJ. Fine-tuning scaffolds for tissue regen-

eration: Effects of formic acid processing on tissue reaction to silk

fibroin. J Tissue Eng Regen Med 2010;4:464–72.

ORIGINAL ARTICLE

JOURNAL OF BIOMEDICAL MATERIALS RESEARCH A | APR 2011 VOL 97A, ISSUE 1 45

Chapter 4

Preliminary Results of the Application of a Silk Fibroin Scaffold to Otology

Levin B, Redmond SL, Rajkhowa R, Eikelboom RH, Marano RJ, Atlas MD. Preliminary Results of the Application of a Silk Fibroin Scaffold to Otology. Otolaryngology - Head and Neck Surgery 2010; 142, Issue 3, S33-S35.

SHORT SCIENTIFIC COMMUNICATION

Preliminary results of the application of a silk

fibroin scaffold to otology

Brett Levin, BMedSci, MBBS,Sharon Leanne Redmond, Assoc Dip Appl Sci (Biol),Rangam Rajkhowa, M Tech, Robert Henry Eikelboom, B Eng, PhD,Robert Jeffery Marano, BSc, PhD, and Marcus David Atlas, MBBS, FRACS,Perth, Western Australia, and Geelong, Victoria, Australia

Sponsorships or competing interests that may be relevant to con-tent are disclosed at the end of this article.

ABSTRACT

The surgical treatment to repair chronic tympanic membrane per-forations is myringoplasty. Although multiple autologous grafts,allografts, and synthetic graft materials have been used over theyears, no single graft material is superior for repairing all perfo-ration types. Recently, the remarkable properties of silk fibroinprotein have been studied, with biomedical and tissue engineeringapplications in mind, across a number of medical and surgicaldisciplines. The present study examines the use of silk fibroin forits potential suitability as an alternative graft in myringoplastysurgery by investigating the growth and proliferation of humantympanic membrane keratinocytes on a silk fibroin scaffold invitro. Light microscopy, immunofluorescent staining, and confocalimaging all reveal promising preliminary results. The biocompat-ibility, transparency, stability, high tensile strength, and biodegrad-ability of fibroin make this biomaterial an attractive option to studyfor this utility.

© 2010 American Academy of Otolaryngology–Head and NeckSurgery Foundation. All rights reserved.

Chronic tympanic membrane (TM) perforations are acommon source of morbidity in the community, with

surgical treatment involving a myringoplasty. Temporalisfascia is universally the most common graft material usedfor all perforation types; however, other materials currentlyused include perichondrium, cartilage, fat, paper, skin, andalloderm.1 Reported success rates vary in the current liter-ature, with closure rates between 60 to 99 percent in adultsand 35 to 94 percent in children.2

Recently, the biomedical applications of fibroin (a struc-tural protein produced by silkworms) have been studiedbecause of its biocompatibility, biodegradability, mechani-cal properties, and diverse morphologies. Useful biochem-ical properties of silk include extensive hydrogen bonding

and significant crystallinity providing fiber stability, insol-ubility in most solvents, high tensile strength, and excellentelasticity. Multiple studies have shown that adverse bio-compatibility problems or hypersensitivity to silk are due toits glue-like protein, sericin.3 Once sericin is removed, thebiological responses to silk fibroin (SF) appear comparableto that of other commonly used biomaterials. Specifically,fibroin has been shown to support cell adhesion, prolifera-tion, and differentiation, as well as promote tissue repair.Because of the ability to manipulate the morphology andmechanical properties of scaffolds made from SF, they havebeen studied with tissue engineering applications relating tothe musculoskeletal system, skin, nerves, and vasculature.SF can be processed to form membranes, foams, fibers,mats, hydrogels, and meshes, depending on clinical require-ments.3

To our knowledge, our group is the first to study SF andits potential applications in the field of otology. Specifically,we report the growth of human tympanic membrane (hTM)keratinocytes on SF membranes, bearing in mind the poten-tial future use as a graft in myringoplasty. Factors that makethis application attractive are the biocompatibility of fibroin,its transparency, its biodegradability over time, and the lackof a requirement to harvest an autologous graft duringsurgery.

Methods

SF was purified from fibers of three silkworm varieties(Bombyx mori, Philosamia cynthia ricini, and Antheraeaassama). Fibroin was reconstituted and cast into thin mem-branes. These were tested for their suitability as scaffoldsfor hTM keratinocyte growth. Full ethics approval wasobtained from St. John of God Health Care Ethics Commit-tee (#193) to collect excess TM in patients undergoingrelevant otological procedures at St. John of God Hospital,Subiaco.

Received April 28, 2009; revised June 23, 2009; accepted June 30, 2009.

Otolaryngology–Head and Neck Surgery (2010) 142, S33-S35

0194-5998/$36.00 © 2010 American Academy of Otolaryngology–Head and Neck Surgery Foundation. All rights reserved.doi:10.1016/j.otohns.2009.06.746

Primary human keratinocytes were cultured from hTMexplants in 6-well plates in Dulbecco’s modified Eagle’smedium until confluent, and subsequently passaged in T-75culture flasks. Cells were seeded onto prepared 6-mm di-ameter SF membranes and incubated for 15 days at 37°Cwith five percent carbon dioxide. Media was changed every48 hours. Subsequently, immunofluorescent staining was per-formed on SF membranes containing hTM cells to demon-strate cell adhesion, proliferation, and epithelial-specific differ-entiation. Antibodies included occludin, zona occludens-1(ZO-1), mindbomb homolog 1 (MIB-1), epithelial-specificETS-1 (ESE-1), E-cadherin, and keratinocyte growth factor(KGF). DAPI (4= 9-6-diamidino-2-phenylindole) was used fornuclear counterstaining. Images were captured sequentiallywith a laser scanning confocal microscope.

Results and Discussion

All three silk scaffolds successfully supported the growth ofhTM cells. Differences were found among the three variet-ies of silk, with B mori most successfully supporting theproliferation and adhesion of TM cells. Figure 1 shows alight microscopy image of confluent hTM keratinocytegrowth on a B mori SF membrane (day 8).

Figure 2 demonstrates occludin and ZO-1 immunofluo-rescent staining of hTM keratinocyte growth on a B mori SFmembrane. Occludin and ZO-1 are integral cell membraneproteins located at tight junctions, and function in epithelialcell adhesion. Strong cell membrane expression of these twoproteins suggests that the SF membrane allows for prolif-eration of TM cells as well as excellent intercellular adhe-sion. ESE-1 stains specifically for terminally differentiatedkeratinocytes and verifies the cell lineage of hTM cells onthe SF membranes as epithelial. The cells stained positivelyfor MIB-1, a marker for proliferation. E-cadherins are trans-membrane proteins expressed by epithelial cells, and func-tion in cell adhesion. These proteins were strongly ex-

pressed, suggesting the keratinocytes maintained their cellorigin.

The lack of a robust and reliable chronic TM perforationmodel in animals that readily allows in vivo testing of thescaffolds is a study limitation.4 However, the biologicalsafety of these scaffolds can still be assessed. Futureplanned in vivo research includes animal studies using atraumatic TM perforation model with SF scaffold repair andhistological assessment of the TMs, focusing on the im-mune response to the scaffold and its biodegradability.Furthermore, the trilaminar morphology of the TM is es-sential for its function, and studies need to be undertaken toinvestigate whether the fibrous middle and inner mucosallayers will develop after SF scaffold implantation.

Initial results of the present study show that the SFscaffold successfully supports the proliferation and adhe-sion of hTM keratinocytes. The biocompatibility and bio-degradability of fibroin, allowing the simultaneous forma-tion of neotissue, are major advantages of a silk-basedtissue-engineered scaffold.5 Silk fibers have been shown tolose the majority of their tensile strength within one year invivo and are no longer recognized at the wound site aftertwo years.3 With specific reference to myringoplasty, an SFgraft may allow for the overgrowth of keratinocytes over aperiod of time, while slowly biodegrading. This approachcould potentially result in the formation of a neotympanumwith no residual SF and, importantly, no other autologousmaterial, such as fascia or cartilage. From the patient’sperspective, this may result in better surgical and long-termhearing outcomes.

Figure 1 hTM keratinocytes on SF membrane (day 8) viewedunder the light microscope. (Original magnification: �4.) Thecells are confluent on the membrane. Scale bar 20 �m.

Figure 2 hTM keratinocytes grown on SF membrane immun-ofluorescently stained with occludin (green) and ZO-1 (red). Cellnuclei are counterstained with DAPI (blue). Image taken on alaser-scanning confocal microscope. Scale bar 20 �m.

S34 Otolaryngology–Head and Neck Surgery, Vol 142, No 3S1, March 2010

Acknowledgments

Garnett Passe and Rodney Williams Memorial Foundation: funding ofresearch. Centre for Material and Fibre Innovation, Deakin University,Geelong, Victoria: manufacture and provision of silk fibroin membranesused in the study. Dr. Paul Rigby: assistance with confocal microscopycarried out using facilities at the Centre for Microscopy, Characterisationand Analysis, The University of Western Australia, which are supported byuniversity, state, and federal government funding.

Author Information

From the Ear Science Institute of Australia and Ear Sciences Centre,School of Surgery, The University of Western Australia (Drs. Levin,Eikelboom, and Marano, Ms. Redmond, and Mr. Atlas), Sir Charles Gaird-ner Hospital (Dr. Levin and Mr. Atlas), St. John of God Hospital (Mr.Atlas), Perth, Western Australia, Australia; and the Centre for Material andFibre Innovation, Deakin University (Mr. Rajkhowa), Geelong, Victoria,Australia.

Corresponding author: Dr. Brett Levin, Ear Sciences Centre, School ofSurgery, 2nd Floor M-Block, The University of Western Australia, QueenElizabeth II Medical Centre, Nedlands, WA Australia, 6009.

E-mail address: blevin@med.usyd.edu.au.

Data have been accepted for presentation at The Australian Society ofOtolaryngology Head and Neck Surgery (ASOHNS) Annual ScientificMeeting, Gold Coast, Australia, May 2009.

Author Contributions

Brett Levin, first author, experiment design, data collection, data imagingusing light and confocal microscopy, data analysis, manuscript revision;Sharon Leanne Redmond, assistant for experimental design, coauthor ofMethods section, culturing of keratinocytes, immunofluorescence staining,use of university imaging facilities for data collection; Rangam Rajk-howa, manufacture of fibroin membranes for use in experimentation,provision of suggestions on experiment conduct and validation; revision ofIntroduction and Methods sections from a materials point of view; RobertHenry Eikelboom, co-coordinating supervisor, assistance with initial de-sign and outline of project; provision of advice for drafting the manuscript,potential conflicts of interest, financial relationships, or intellectual content,approval of final draft; Robert Jeffery Marano, assistance with initial

design and outline of project, provision of scientific advice during themanuscript drafting, particularly with respect to experimental design, ap-proval of final draft; Marcus David Atlas, assistance with project design,provision of clinical advice for manuscript, provision of tympanic mem-brane tissue used in experiments, approval of final draft.

Disclosures

Competing interests: Brett Levin, Ear Science Institute Australia (ESIA):scholarship, research, employment contract has an arrangement in place fora share in any proceeds of work undertaken at ESIA in the event of futurecommercialization of the work; Sharon Leanne Redmond, Robert HenryEikelboom, Robert Jeffery Marano, ESIA: salary, employment contracthas an arrangement in place for a share in any proceeds of work undertakenat ESIA in the event of future commercialization of the work; RangamRajkhowa, Deakin University: International Post Graduate ResearchScholarship from Australian government, with an arrangement in place fora share in any proceeds of work undertaken at Deakin University in theevent of future commercialization of the work; Marcus David Atlas, ESIAand University of Western Australia: salary, employment contract has anarrangement in place for a share in any proceeds of work undertaken atESIA in the event of future commercialization of the work.

Sponsorships: The Garnett Passe and Rodney Williams Memorial Foun-dation funds Dr. Levin’s Master of Medical Science in OtolaryngologyHead and Neck Surgery; research and manuscript form part of the higherdegree; funds allow for conducting of the entire study, including datacollection and analysis.

References

1. Aggarwal R, Saeed SR, Green KJ. Myringoplasty. J Laryngol Otol2006;120:429–32.

2. Inwood JL, Wallace HC, Clarke SE. Endaural or postaural incision formyringoplasty: does it make a difference to the patient? Clin Otolaryn-gol 2003;28:396–8.

3. Altman GH, Diaz F, Jakuba C, et al. Silk-based biomaterials. Bioma-terials 2003;24:401–16.

4. Santa Maria PL, Atlas MD, Ghassemifar R. Chronic tympanic mem-brane perforation: a better animal model is needed. Wound Repair Reg2007;15:450–8.

5. Vepari C, Kaplan DL. Silk as a biomaterial. Prog Polym Sci 2007;32:991–1007.

S35Levin et al Preliminary results of the application of . . .

Chapter 5

Utilising Silk Fibroin Membranes as Scaffolds for the Growth of Tympanic Membrane Keratinocytes and their Application to

Myringoplasty Surgery

Levin B, Redmond SL, Rajkhowa R, Eikelboom RH, Marano RJ, Atlas MD. Utilising Silk Fibroin Membranes as Scaffolds for the Growth of Tympanic Membrane Keratinocytes and their Application to Myringoplasty Surgery. Under review by Biomedical Materials.

Biomedical Materials - Title Page

Submission Title

Utilising Silk Fibroin Membranes as Scaffolds for the Growth of Tympanic

Membrane Keratinocytes and their Application to Myringoplasty Surgery

Authors

1. Dr Brett Levin, B Med Sci, MBBS (Hons), Ear Science Institute of Australia;

Ear Sciences Centre, School of Surgery, The University of Western

Australia; Sir Charles Gairdner Hospital, Perth, Western Australia

2. Ms Sharon Leanne Redmond, Assoc Dip Appl Sci (Biol), Ear Science

Institute of Australia and Ear Sciences Centre, School of Surgery, The

University of Western Australia, Perth, Western Australia

3. Mr Rangam Rajkhowa, M.Tech, Centre for Material and Fibre Innovation

(CMFI), Deakin University, Geelong, Victoria, Australia

4. Adj Prof Robert Henry Eikelboom, B Eng, PhD, Ear Science Institute of

Australia and Ear Sciences Centre, School of Surgery, The University of

Western Australia, Perth, Western Australia

5. Dr Robert Jeffery Marano, BSc, PhD, Ear Science Institute of Australia and

Ear Sciences Centre, School of Surgery, The University of Western

Australia, Perth, Western Australia

6. Winthrop Prof Marcus David Atlas, MBBS, FRACS, Ear Science Institute of

Australia; Ear Sciences Centre, School of Surgery, The University of

Western Australia; St John of God Hospital; Sir Charles Gairdner Hospital,

Perth, Western Australia

Corresponding Author

Name: Dr Brett Levin

Address: Ear Sciences Centre, School of Surgery, 2nd

Floor M-Block, The

University of Western Australia, Queen Elizabeth II Medical Centre,

Nedlands, WA, 6009.

Email: DrBrettLevin@gmail.com

Ph No: (08) 9346 4153

Short Title

Silk Fibroin Scaffolds for Tympanic Membrane Keratinocyte Growth

UTILISING SILK FIBROIN MEMBRANES AS SCAFFOLDS FOR THE

GROWTH OF TYMPANIC MEMBRANE KERATINOCYTES AND THEIR

APPLICATION TO MYRINGOPLASTY SURGERY

Abstract

Chronic tympanic membrane perforations can cause significant morbidity.

Myringoplasty describes the operation that uses a graft to close such perforations. A

variety of graft materials are available, with temporalis fascia being the most

frequently used. All current grafts have limitations, and few studies report hearing

outcomes after myringoplasty surgery. In recent years, silk fibroin, has been studied

with biomedical applications in mind. Its biocompatibility, biodegradability and its

ability to act as a scaffold to support cell growth have prompted us to investigate its

interaction with human tympanic membrane keratinocytes. Silk fibroin membranes

were prepared and human tympanic membrane keratinocytes were cultured. These

keratinocytes were seeded onto the membranes and immunostained for a number of

relevant protein markers relating to cell proliferation, adhesion and epithelial specific

differentiation. The scaffold successfully supported the growth and adhesion of

keratinocytes, whilst also maintaining their cell lineage. The results from this study

and the properties of silk fibroin make it an attractive option to pursue as a potential

alternative graft in myringoplasty.

Introduction

Otitis media is defined as infection of the middle ear cleft and is the most common

reason for children to seek paediatric care (1). It is the most common otological

condition during childhood which compromises sound conduction in the middle ear

and can adversely affect long-term hearing (2). Chronic suppuratives otitis media

(CSOM) describes a stage of ear disease in which there is chronic infection of the

middle ear cleft and in which a non-intact tympanic membrane (perforation) and

discharge are present (3).

Despite advances in medical and surgical treatment, as well as increased public

health awareness, CSOM is still highly prevalent worldwide (4) and may be the most

important cause of hearing impairment in some developing countries (3). Sequelae

that may arise include speech and language deficits as well as developmental

problems in children (5). The magnitude of hearing loss is proportional to the size of

the perforation and the degree of damage caused to middle ear structures (2, 6).

Although hearing loss is the most common complication of CSOM, chronic

tympanic membrane (TM) perforations are also associated with multiple other

extracranial and intracranial complications (4) and have a significant mortality rate

(7). The most frequent extracranial complications include mastoiditis, facial

paralysis, subperiosteal abscess formation and labyrinthitis; while the most frequent

intracranial complications include meningitis, cerebral abscess formation, lateral

sinus thrombosis, extradural abscess formation, otic hydrocephalus and encephalitis

(4, 7). It is of particular concern that although antibiotic therapy has reduced the

incidence of CSOM, its complication rate has not altered in the past 10 years (8).

Myringoplasty (also known as type 1 tympanoplasty) is the term used to describe the

surgical repair of a perforated tympanic membrane by utilising a graft to close the

perforation. It is one of the most common otological procedures performed in adults

and children (9), and generally has a very high success rate with respect to

perforation closure, although this is somewhat dependent on surgical experience

(10).

Myringoplasty does, however, have a number of limitations. Few studies report

hearing outcomes after surgery, and documented success rates often refer solely to

perforation closure and the cessation of aural discharge (9). Some studies that have

examined postoperative hearing reported considerable discrepancy between pure

tone audiogram results and the subjective hearing benefit to the patient, reporting

only a 55% improvement in subjective hearing after surgery (11). Whilst many

papers refer to the high success of myringoplasty in terms of closure rates of

perforations, there is evidence to suggest that the hearing outcomes after

myringoplasty may be improved but not restored (12-14).

Multiple factors, including the type of graft used, impact on the success of

myringoplasty (15). Although multiple autologous grafts, allografts and synthetic

grafts are currently used in myringoplasty, all have documented limitations (9). The

current materials used as grafts by different surgeons include temporalis fascia,

cartilage, perichondrium, fat, paper and alloderm (15-18).

Ideally, a material used to surgically repair a perforated TM would satisfy all of the

following criteria: safety, easily obtainable, available in unlimited quantities,

inexpensive, resist negative ear pressures and cause no interference with hearing

(19). A transparent material would also be preferable as this would more closely

mimic the native TM and allow otoscopic visualisation of middle ear structures

during patient follow up.

To date, none of the above frequently used materials meet all of these requirements.

Moreover, Kaftan et al noted that a high proportion of patients found to have a

chronic perforation did not want surgical treatment, further suggesting a need for

novel therapeutic materials or procedures (20).

Silks are naturally occurring polymers that can be extruded from insects or worms

(21). They are composed of a filament core protein (fibroin) and a glue-like coating

protein (sericin). Silk from the silkworm B Mori has been used clinically for

centuries in the form of sutures. Over the last 25 years, some biocompatibility

problems initially affected its utilisation, however contamination from sericin was

thought to be responsible for this, and the isolation of silk fibroin (SF) has largely

eliminated these concerns (22).

In recent years, SF has been studied with biomedical applications in mind due its

remarkable mechanical properties, including biocompatibility and biodegradability.

It has been shown to support cell adhesion, proliferation and differentiation in vitro

and promote tissue repair in vivo. Moreover, its versatile processing and surface

modification options have expanded the clinical utility of SF and it can now be made

readily available in various forms such as films, fibres, nets, meshes, membranes,

yarns and sponges (21).

Designing scaffolds for tissue engineering purposes has in recent times become

particularly relevant (22). Studies have shown that the behaviour of cells on three-

dimensional SF scaffolds can be related to the structural differences resulting from

different scaffold preparation processes (23). Investigators can therefore be guided to

use processing scenarios that match the tissue-specific and clinical applications that

are required.

In order to assess the application of SF scaffolds to myringoplasty surgery, a chronic

TM perforation model in animals is ideally required. This is because the majority of

acute perforations will heal spontaneously (24). An inexpensive and valid animal

model would provide insights into the mechanisms of TM wound healing as well as

allow various treatment options to be evaluated (25). Although multiple attempts

have been made, the ideal animal chronic TM perforation model is yet to be created

(26).

Amoils et al had some success in developing a model in chinchillas that was

permanent, well-epithelialised and free from infection (24). Chronic perforations

were achieved in a significant percentage of animals, but persistent infection or TM

regeneration despite reperforation was problematic. Almost 20 years ago, Amoils et

al envisioned that if such a model could be successfully achieved it could be used to

assess various biomembrane scaffolds (impregnated with various growth-promoting

substances) in the regeneration of perforated TMs. They predicted that a

membranous disc with biorecombinant growth factors could provide a simple office

technique for the repair of chronic TM perforations, alleviating the morbidity and

time associated with myringoplasty surgery (24).

Although the lack of a chronic TM perforation model in animals has somewhat

hindered in vivo research, investigation of such membranous scaffolds and the

interaction with TM keratinocytes can be undertaken in vitro (27, 28).

It is important to note that the TM has a very specific multi-layered microstructure

that is responsible for its unique function (29). It is a unique membrane suspended

entirely in air. From lateral to medial, it comprises five layers: an epidermal layer of

keratinised squamous epithelium, a thin connective tissue layer, a dense connective

tissue layer, a thin layer of connective tissue and an inner single layer of epithelial

cells (30). This microstructure is vastly different from the various grafts that are

commonly used to repair TM perforations (9).

The specific distribution of various types of collagen in both healthy and perforated

TMs has also been studied (31). Interestingly, the pars tensa and pars flaccida of the

TM are composed of different types of collagen, reflecting the different

physiological properties of these tissues. Collagen types I and III are present in the

acute phase after perforation and this collagen content is modified during the

inflammatory and healing processes (31).

It is clear that the TM is anatomically and functionally a very complex membrane. It

is doubtful that any of the available autologous grafts, homografts or synthetic grafts

will be able to mimic both its structure and function. Perforations may be

successfully closed, but the difficulty remains in achieving premorbid hearing.

Perhaps this is only feasible with a scaffold that supports the overgrowth of native

TM cells and biodegrades with time. This could potentially result in the formation of

healed TM with no foreign graft material and improved hearing outcomes.

To our knowledge, our group is the first to apply the use of SF biomaterials to the

field of otology, with the aim of alleviating some of the limitations of the various

grafts currently used in myringoplasty (9, 27, 28). We report the growth and

proliferation of human TM keratinocytes on SF scaffolds, bearing in mind their

potential future use in the repair of TM perforations.

Materials and Methods

Preparation of silk fibroin membranes

Cocoon pieces from Bombyx Mori (BM) silkworms were boiled in 2 g/L sodium

carbonate and 0.6 g/L sodium dodecyl sulphate solution. The resultant fibrous

material was washed thoroughly in hot and cold water. After drying, the fibres were

dissolved in a concentrated solution (9.3M) of lithium bromide. The resultant

solution was dialysed for four days at 4°C using dialysis sacks (molecular weight cut

off 12 kDa) against Milli-Q water, which was changed every six to twelve hours.

The concentration of silk solution was adjusted to 3% (w/v) by diluting with Milli-Q

water. A portion of silk solution was retained for the preparation of water-based

films, while the remaining solution was frozen at –80oC overnight and lyophilised.

The lyophilised silk solids were dissolved separately in 98% formic acid and 98%

trifluoroacetic acid to obtain 3% (w/v) TFA and FA fibroin solutions. Silk solutions

were poured into polyethylene discs and dried inside fume hoods to obtain dry films

(approximately 33 μm in thickness). The films were stored inside the fume hood for

a further twenty-four hours to evaporate residual solvents (if any) completely. For

structural stability (i.e. conversion into the ‘silk II’ conformer which is insoluble in

water), the membranes were treated with 75% aqueous solution of EtOH for six

hours. After hydration to equilibrium and extensive rinsing in water exchanges, the

Bombyx mori silk fibroin (BMSF) membranes were stored at 4oC. Prior to

experimentation, discs were cut from membranes using an 8mm biopsy punch

(Stiefel Laboratory GmbH, Germany). Figure 1 shows the raw SF membranes after

preparation.

Primary culture of human tympanic membrane keratinocyte (hTMk) cells

Ethics approval was obtained from the St. John of God Health Care Ethics

Committee (#193) to collect excess TM in patients undergoing relevant otological

procedures at St. John of God Hospital, Subiaco. Small pieces of hTM tissue

(approximately 0.25mm2) were placed into BD Falcon 6-well culture plates (BD

Biosciences, Australia). DMEM 4500mg/L D-glucose was supplemented with

100U/mL penicillin, 100 g/mL streptomycin (Invitrogen, Australia), and 10% FBS

(Invitrogen). Plates were incubated in a humidified cell culture incubator at 37oC

with 5% carbon dioxide (CO2).

Seeding of hTMk cells on BMSF membranes

Discs of silk membrane, 8mm in diameter, were placed into 6-well culture plates

(Falcon, BD Australia). Discs were disinfected by immersion in 4 ml of 70% EtOH

for two hours at RT followed by 6 x washes in 4 ml sterile PBS over a 24 hour

period. Prior to use the discs were brief rinsed in high-glucose DMEM,

supplemented with 10% FBS and 100U/mL penicillin. hTMk cells were seeded into

each well containing membranes at 0.3 x 106 cells/mL. Plates were left for 12 days at

37oC with 5% CO2. Culture media was changed every 48 hours.

Antibodies

Primary antibodies used included rabbit polyclonal anti-human ESE-1

(NovusBiologicals, USA), mouse monoclonal anti-human E-cadherin (CDH1)

(Zymed, USA), and mouse monoclonal anti-human MIB-1 (DAKO, Denmark).

Secondary antibodies used included Alexa Fluor-488 conjugated goat anti-mouse

(1:500) and Alexa Fluor-546 conjugated goat anti-rabbit (1:500) (Molecular Probes,

USA). Nuclei were counterstained 4’,6-diamidino-2-phenylindole (DAPI),

(Molecular Probes, USA) at 1:500, diluted in 1x PBS.

Immunofluorescent cell staining

Cultured cells on SF membranes were washed briefly in PBS, pH7.2 (Invitrogen),

and then incubated in methanol at -20oC for 10 minutes, followed by three washes

with PBS, pH7.2, containing 0.1% Tween 20 (PBS-Tw). Incubation with primary

antibodies was performed at RT for 60 minutes, followed by three washes with PBS-

Tw. Primary antibodies were diluted in PBS containing 1% bovine serum albumin

(BSA). Secondary antibody incubations were performed at RT for 60 minutes

followed by three washes in PBS-Tw. Cell nuclei were counterstained with DAPI for

20 minutes, washed with PBS-Tw, cover slipped in anti-fade mounting medium and

sealed with nail varnish.

Confocal laser-scanning microscopy

Images were captured sequentially using a Nikon A1Si confocal laser-scanning

microscope (Nikon, Japan) on a Nikon TiE (Nikon, Japan) with an oil 40x objective

(Nikon Pan Fluor, NA 1.30). The pinhole was set to 1.0 Airy unit (AU). A 488nm

blue laser was used to excite AF-488 secondary antibodies and fluorescence

emission detected through a 525/50nm bandpass filter. A 561nm green laser was

used to excite AF-546 secondary antibodies and fluorescence emission detected

through a 585/50nm bandpass emission filter. A violet laser (405nm) was used for

the excitation of DAPI and fluorescence emission detected through a 450/50nm

bandpass filter. Images were saved as 12-bit greyscale images and were subsequently

merged using with ImageJ (1.42e), (Wayne Rasband, http://rsb.info.nih.gov/ij/). All

images were saved in TIFF format.

Results and Discussion

Light microscopy images confirmed that the SF scaffolds successfully supported the

growth of hTM cells. In order for these scaffolds to be considered as grafts in

myringoplasty surgery, various cellular characteristics, interactions and functions on

the membranes require investigation. This was achieved using immunofluorescent

cell staining and imaging via confocal microscopy.

Figure 2 represents a light microscopy image of keratinocytes spreading

circumferentially outwards from the TM explant in culture media (x4 magnification,

day 14). The typical, cobblestone appearance of keratinocytes is visible.

Figure 3A shows hTM keratinocytes growing and proliferating on the SF scaffold

(90 percent confluent) (day 8 on membrane, x4 magnification). Figure 3B shows the

cells 14 days after seeding onto the membrane (x20 magnification). In figure 3C, the

interface between the cells on the SF membrane and the culture media is clearly

visible (day 18 on membrane, x4 magnification). The cells are confluent on the SF

membrane. Figure 4 is a scanning electron micrograph of the same cells at day 5

(Figure 4A) and 15 (Figure 4B) on the SF membranes.

It was important to verify the cell lineage of keratinocytes as epithelial and confirm

that these cells did not transform on the SF membranes (32). ESE-1 is a highly

tissue-specific member of the ets transcription factor/oncogene family, expressed

exclusively in epithelial cells. It is induced during terminal differentiation of the

epidermis and in primary human keratinocyte differentiation (33). Figure 5 shows

the hTM keratinocytes on the SF scaffold strongly expressing the protein ESE-1.

Occludin is an integral plasma membrane protein located specifically at tight

junctions whilst ZO-1 encodes a protein located on the cytoplasmic membrane

surface of intercellular tight junctions. It is thought to be involved in signal

transduction between cells at these junctions. Occludin and ZO-1 are proteins that

play roles in cell adhesion (34, 35). Figures 6 and 7 show hTM cells immunostained

for occludin and ZO-1 respectively. The important function of these proteins is

maintained on the SF membranes.

Figure 8 shows the keratinocytes immunofluorescently stained with antibodies to E-

cadherin. E-cadherins are transmembrane proteins that also function in intercellular

adhesion between epithelial cells (36, 37). Strong uptake and staining for these

important proteins is evident.

The serial light microscopy images above show the keratinocytes proliferating on the

membranes (Figure 3). MIB-1 is a commonly used monoclonal antibody that detects

the Ki-67 antigen. This nuclear antigen is associated with cell proliferation and thus

MIB-1 is used clinically as a cell proliferation marker (38). Figure 9 demonstrates

strong protein expression for MIB-1.

Conclusion

The present study investigated the growth of hTM keratinocytes on a scaffold

composed of SF. The results showed that the scaffold supported the growth and

proliferation of hTM cells. Immunofluorescent antibody staining confirmed the

keratinocytes maintained their cell lineage on the scaffold, and integral cellular

proteins that function in proliferation, differentiation and adhesion were also

maintained.

The biocompatibility, biodegradability and transparency of SF are major advantages

of a tissue engineered scaffold. The information obtained in the present study is

important for further investigation of such a scaffold and its application to

myringoplasty surgery. To date, no scaffold that allows the overgrowth of hTM cells

and also biodegrades has been used clinically in myringoplasty. SF may provide this

solution and enable otological surgeons to successfully close chronic TM

perforations, with the additional advantages of inspecting middle ear structures

through a transparent membrane and significantly improve hearing outcomes.

Figure Legends

Figure 1 – Silk fibroin membranes prior to seeding keratinocytes.

Figure 2 - light microscopy image of human tympanic membrane cells spreading

circumferentially from the explant in culture media (x4 magnification, day 14 in

culture, scale bar = 20um).

Figure 3 – A: Human tympanic membrane keratinocytes on the SF scaffold after 8

days (x4 magnification); B: Human tympanic membrane keratinocytes on the SF

scaffold after 14 days (x20 magnification); C: Interface between the cells on the SF

membrane and the culture media (day 18 on membrane, x4 magnification). Scale bar

= 20um in all images.

Figure 4 – A: Scanning electron micrograph of keratinocytes at day 5 on the SF

membrane; B: Keratinocytes after 15 days on the SF membrane.

Figure 5 – Human tympanic membrane keratinocytes on the SF scaffold expressing

ESE-1. Scale bar = 20um.

Figure 6 – Human tympanic membrane keratinocytes on the SF scaffold

immunostained for occludin. Scale bar = 20um.

Figure 7 – Human tympanic membrane keratinocytes on the SF scaffold

immunostained for ZO-1. Scale bar = 20um.

Figure 8 – Human tympanic membrane keratinocytes on the SF scaffold expressing

E-cadherin. Scale bar = 20um.

Figure 9 – Human tympanic membrane keratinocytes on the SF scaffold expressing

MIB-1. Scale bar = 20um.

References

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

  12

CONCLUSION

Chronic perforations of the tympanic membrane continue to be a significant

source of worldwide morbidity with potentially devastating sequelae.

Although myringoplasty is a highly successful operation with respect to

perforation closure, the current grafts all have limitations, and post-operative

hearing results are often not taken into account when reporting the success

rates of myringoplasty.

The diverse biomechanical properties of fibroin including its

biocompatibility, biodegradability, tensile strength, transparency and its

wound healing abilities make it a worthwhile protein to study with

application to myringoplasty. Factors affecting the tensile strength and

biodegradation of silk fibroin films were studied. These two properties are

particularly important when considering the use of these films in a clinical

context. The ability to control biodegradation and tensile strength makes silk

fibroin a particularly valuable material to study for our purposes.

In this research, light microscopy, immunofluorescent staining, confocal

microscopy and scanning electron microscopy all confirm that the fibroin

scaffolds successfully supported the growth, proliferation and adhesion of

  13

cultured human tympanic membrane keratinocytes, whilst maintaining their

cell lineage. Integral cellular proteins that function in proliferation,

differentiation and adhesion were also maintained.

Future in vivo research using animal models with traumatic tympanic

membrane perforations needs to establish the safety of silk fibroin in the

middle ear. Fibroin membranes that degrade at rates that match the growth of

new tympanic membrane keratinocytes need to be investigated. Ideally,

strategies will also be required to manufacture silk fibroin membranes that

guide multilayered tympanic membrane regeneration.